what is multiple case studies

Want to create or adapt books like this? Learn more about how Pressbooks supports open publishing practices.

Research Guides

Multiple Case Studies

Nadia Alqahtani and Pengtong Qu

Description

The case study approach is popular across disciplines in education, anthropology, sociology, psychology, medicine, law, and political science (Creswell, 2013). It is both a research method and a strategy (Creswell, 2013; Yin, 2017). In this type of research design, a case can be an individual, an event, or an entity, as determined by the research questions. There are two variants of the case study: the single-case study and the multiple-case study. The former design can be used to study and understand an unusual case, a critical case, a longitudinal case, or a revelatory case. On the other hand, a multiple-case study includes two or more cases or replications across the cases to investigate the same phenomena (Lewis-Beck, Bryman & Liao, 2003; Yin, 2017). …a multiple-case study includes two or more cases or replications across the cases to investigate the same phenomena

The difference between the single- and multiple-case study is the research design; however, they are within the same methodological framework (Yin, 2017). Multiple cases are selected so that “individual case studies either (a) predict similar results (a literal replication) or (b) predict contrasting results but for anticipatable reasons (a theoretical replication)” (p. 55). When the purpose of the study is to compare and replicate the findings, the multiple-case study produces more compelling evidence so that the study is considered more robust than the single-case study (Yin, 2017).

To write a multiple-case study, a summary of individual cases should be reported, and researchers need to draw cross-case conclusions and form a cross-case report (Yin, 2017). With evidence from multiple cases, researchers may have generalizable findings and develop theories (Lewis-Beck, Bryman & Liao, 2003).

Creswell, J. W. (2013). Qualitative inquiry and research design: Choosing among five approaches (3rd ed.). Los Angeles, CA: Sage.

Lewis-Beck, M., Bryman, A. E., & Liao, T. F. (2003). The Sage encyclopedia of social science research methods . Los Angeles, CA: Sage.

Yin, R. K. (2017). Case study research and applications: Design and methods . Los Angeles, CA: Sage.

Key Research Books and Articles on Multiple Case Study Methodology

Yin discusses how to decide if a case study should be used in research. Novice researchers can learn about research design, data collection, and data analysis of different types of case studies, as well as writing a case study report.

Chapter 2 introduces four major types of research design in case studies: holistic single-case design, embedded single-case design, holistic multiple-case design, and embedded multiple-case design. Novice researchers will learn about the definitions and characteristics of different designs. This chapter also teaches researchers how to examine and discuss the reliability and validity of the designs.

Creswell, J. W., & Poth, C. N. (2017). Qualitative inquiry and research design: Choosing among five approaches . Los Angeles, CA: Sage.

This book compares five different qualitative research designs: narrative research, phenomenology, grounded theory, ethnography, and case study. It compares the characteristics, data collection, data analysis and representation, validity, and writing-up procedures among five inquiry approaches using texts with tables. For each approach, the author introduced the definition, features, types, and procedures and contextualized these components in a study, which was conducted through the same method. Each chapter ends with a list of relevant readings of each inquiry approach.

This book invites readers to compare these five qualitative methods and see the value of each approach. Readers can consider which approach would serve for their research contexts and questions, as well as how to design their research and conduct the data analysis based on their choice of research method.

Günes, E., & Bahçivan, E. (2016). A multiple case study of preservice science teachers’ TPACK: Embedded in a comprehensive belief system. International Journal of Environmental and Science Education, 11 (15), 8040-8054.

In this article, the researchers showed the importance of using technological opportunities in improving the education process and how they enhanced the students’ learning in science education. The study examined the connection between “Technological Pedagogical Content Knowledge” (TPACK) and belief system in a science teaching context. The researchers used the multiple-case study to explore the effect of TPACK on the preservice science teachers’ (PST) beliefs on their TPACK level. The participants were three teachers with the low, medium, and high level of TPACK confidence. Content analysis was utilized to analyze the data, which were collected by individual semi-structured interviews with the participants about their lesson plans. The study first discussed each case, then compared features and relations across cases. The researchers found that there was a positive relationship between PST’s TPACK confidence and TPACK level; when PST had higher TPACK confidence, the participant had a higher competent TPACK level and vice versa.

Recent Dissertations Using Multiple Case Study Methodology

Milholland, E. S. (2015). A multiple case study of instructors utilizing Classroom Response Systems (CRS) to achieve pedagogical goals . Retrieved from ProQuest Dissertations & Theses Global. (Order Number 3706380)

The researcher of this study critiques the use of Classroom Responses Systems by five instructors who employed this program five years ago in their classrooms. The researcher conducted the multiple-case study methodology and categorized themes. He interviewed each instructor with questions about their initial pedagogical goals, the changes in pedagogy during teaching, and the teaching techniques individuals used while practicing the CRS. The researcher used the multiple-case study with five instructors. He found that all instructors changed their goals during employing CRS; they decided to reduce the time of lecturing and to spend more time engaging students in interactive activities. This study also demonstrated that CRS was useful for the instructors to achieve multiple learning goals; all the instructors provided examples of the positive aspect of implementing CRS in their classrooms.

Li, C. L. (2010). The emergence of fairy tale literacy: A multiple case study on promoting critical literacy of children through a juxtaposed reading of classic fairy tales and their contemporary disruptive variants . Retrieved from ProQuest Dissertations & Theses Global. (Order Number 3572104)

To explore how children’s development of critical literacy can be impacted by their reactions to fairy tales, the author conducted a multiple-case study with 4 cases, in which each child was a unit of analysis. Two Chinese immigrant children (a boy and a girl) and two American children (a boy and a girl) at the second or third grade were recruited in the study. The data were collected through interviews, discussions on fairy tales, and drawing pictures. The analysis was conducted within both individual cases and cross cases. Across four cases, the researcher found that the young children’s’ knowledge of traditional fairy tales was built upon mass-media based adaptations. The children believed that the representations on mass-media were the original stories, even though fairy tales are included in the elementary school curriculum. The author also found that introducing classic versions of fairy tales increased children’s knowledge in the genre’s origin, which would benefit their understanding of the genre. She argued that introducing fairy tales can be the first step to promote children’s development of critical literacy.

Asher, K. C. (2014). Mediating occupational socialization and occupational individuation in teacher education: A multiple case study of five elementary pre-service student teachers . Retrieved from ProQuest Dissertations & Theses Global. (Order Number 3671989)

This study portrayed five pre-service teachers’ teaching experience in their student teaching phase and explored how pre-service teachers mediate their occupational socialization with occupational individuation. The study used the multiple-case study design and recruited five pre-service teachers from a Midwestern university as five cases. Qualitative data were collected through interviews, classroom observations, and field notes. The author implemented the case study analysis and found five strategies that the participants used to mediate occupational socialization with occupational individuation. These strategies were: 1) hindering from practicing their beliefs, 2) mimicking the styles of supervising teachers, 3) teaching in the ways in alignment with school’s existing practice, 4) enacting their own ideas, and 5) integrating and balancing occupational socialization and occupational individuation. The study also provided recommendations and implications to policymakers and educators in teacher education so that pre-service teachers can be better supported.

Multiple Case Studies Copyright © 2019 by Nadia Alqahtani and Pengtong Qu is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License , except where otherwise noted.

Share This Book

• Privacy Policy

Home » Case Study – Methods, Examples and Guide

Case Study – Methods, Examples and Guide

Table of Contents

A case study is a research method that involves an in-depth examination and analysis of a particular phenomenon or case, such as an individual, organization, community, event, or situation.

It is a qualitative research approach that aims to provide a detailed and comprehensive understanding of the case being studied. Case studies typically involve multiple sources of data, including interviews, observations, documents, and artifacts, which are analyzed using various techniques, such as content analysis, thematic analysis, and grounded theory. The findings of a case study are often used to develop theories, inform policy or practice, or generate new research questions.

Types of Case Study

Types and Methods of Case Study are as follows:

Single-Case Study

A single-case study is an in-depth analysis of a single case. This type of case study is useful when the researcher wants to understand a specific phenomenon in detail.

For Example , A researcher might conduct a single-case study on a particular individual to understand their experiences with a particular health condition or a specific organization to explore their management practices. The researcher collects data from multiple sources, such as interviews, observations, and documents, and uses various techniques to analyze the data, such as content analysis or thematic analysis. The findings of a single-case study are often used to generate new research questions, develop theories, or inform policy or practice.

Multiple-Case Study

A multiple-case study involves the analysis of several cases that are similar in nature. This type of case study is useful when the researcher wants to identify similarities and differences between the cases.

For Example, a researcher might conduct a multiple-case study on several companies to explore the factors that contribute to their success or failure. The researcher collects data from each case, compares and contrasts the findings, and uses various techniques to analyze the data, such as comparative analysis or pattern-matching. The findings of a multiple-case study can be used to develop theories, inform policy or practice, or generate new research questions.

Exploratory Case Study

An exploratory case study is used to explore a new or understudied phenomenon. This type of case study is useful when the researcher wants to generate hypotheses or theories about the phenomenon.

For Example, a researcher might conduct an exploratory case study on a new technology to understand its potential impact on society. The researcher collects data from multiple sources, such as interviews, observations, and documents, and uses various techniques to analyze the data, such as grounded theory or content analysis. The findings of an exploratory case study can be used to generate new research questions, develop theories, or inform policy or practice.

Descriptive Case Study

A descriptive case study is used to describe a particular phenomenon in detail. This type of case study is useful when the researcher wants to provide a comprehensive account of the phenomenon.

For Example, a researcher might conduct a descriptive case study on a particular community to understand its social and economic characteristics. The researcher collects data from multiple sources, such as interviews, observations, and documents, and uses various techniques to analyze the data, such as content analysis or thematic analysis. The findings of a descriptive case study can be used to inform policy or practice or generate new research questions.

Instrumental Case Study

An instrumental case study is used to understand a particular phenomenon that is instrumental in achieving a particular goal. This type of case study is useful when the researcher wants to understand the role of the phenomenon in achieving the goal.

For Example, a researcher might conduct an instrumental case study on a particular policy to understand its impact on achieving a particular goal, such as reducing poverty. The researcher collects data from multiple sources, such as interviews, observations, and documents, and uses various techniques to analyze the data, such as content analysis or thematic analysis. The findings of an instrumental case study can be used to inform policy or practice or generate new research questions.

Case Study Data Collection Methods

Here are some common data collection methods for case studies:

Interviews involve asking questions to individuals who have knowledge or experience relevant to the case study. Interviews can be structured (where the same questions are asked to all participants) or unstructured (where the interviewer follows up on the responses with further questions). Interviews can be conducted in person, over the phone, or through video conferencing.

Observations

Observations involve watching and recording the behavior and activities of individuals or groups relevant to the case study. Observations can be participant (where the researcher actively participates in the activities) or non-participant (where the researcher observes from a distance). Observations can be recorded using notes, audio or video recordings, or photographs.

Documents can be used as a source of information for case studies. Documents can include reports, memos, emails, letters, and other written materials related to the case study. Documents can be collected from the case study participants or from public sources.

Surveys involve asking a set of questions to a sample of individuals relevant to the case study. Surveys can be administered in person, over the phone, through mail or email, or online. Surveys can be used to gather information on attitudes, opinions, or behaviors related to the case study.

Artifacts are physical objects relevant to the case study. Artifacts can include tools, equipment, products, or other objects that provide insights into the case study phenomenon.

How to conduct Case Study Research

Conducting a case study research involves several steps that need to be followed to ensure the quality and rigor of the study. Here are the steps to conduct case study research:

• Define the research questions: The first step in conducting a case study research is to define the research questions. The research questions should be specific, measurable, and relevant to the case study phenomenon under investigation.
• Select the case: The next step is to select the case or cases to be studied. The case should be relevant to the research questions and should provide rich and diverse data that can be used to answer the research questions.
• Collect data: Data can be collected using various methods, such as interviews, observations, documents, surveys, and artifacts. The data collection method should be selected based on the research questions and the nature of the case study phenomenon.
• Analyze the data: The data collected from the case study should be analyzed using various techniques, such as content analysis, thematic analysis, or grounded theory. The analysis should be guided by the research questions and should aim to provide insights and conclusions relevant to the research questions.
• Draw conclusions: The conclusions drawn from the case study should be based on the data analysis and should be relevant to the research questions. The conclusions should be supported by evidence and should be clearly stated.
• Validate the findings: The findings of the case study should be validated by reviewing the data and the analysis with participants or other experts in the field. This helps to ensure the validity and reliability of the findings.
• Write the report: The final step is to write the report of the case study research. The report should provide a clear description of the case study phenomenon, the research questions, the data collection methods, the data analysis, the findings, and the conclusions. The report should be written in a clear and concise manner and should follow the guidelines for academic writing.

Examples of Case Study

Here are some examples of case study research:

• The Hawthorne Studies : Conducted between 1924 and 1932, the Hawthorne Studies were a series of case studies conducted by Elton Mayo and his colleagues to examine the impact of work environment on employee productivity. The studies were conducted at the Hawthorne Works plant of the Western Electric Company in Chicago and included interviews, observations, and experiments.
• The Stanford Prison Experiment: Conducted in 1971, the Stanford Prison Experiment was a case study conducted by Philip Zimbardo to examine the psychological effects of power and authority. The study involved simulating a prison environment and assigning participants to the role of guards or prisoners. The study was controversial due to the ethical issues it raised.
• The Challenger Disaster: The Challenger Disaster was a case study conducted to examine the causes of the Space Shuttle Challenger explosion in 1986. The study included interviews, observations, and analysis of data to identify the technical, organizational, and cultural factors that contributed to the disaster.
• The Enron Scandal: The Enron Scandal was a case study conducted to examine the causes of the Enron Corporation’s bankruptcy in 2001. The study included interviews, analysis of financial data, and review of documents to identify the accounting practices, corporate culture, and ethical issues that led to the company’s downfall.
• The Fukushima Nuclear Disaster : The Fukushima Nuclear Disaster was a case study conducted to examine the causes of the nuclear accident that occurred at the Fukushima Daiichi Nuclear Power Plant in Japan in 2011. The study included interviews, analysis of data, and review of documents to identify the technical, organizational, and cultural factors that contributed to the disaster.

Application of Case Study

Case studies have a wide range of applications across various fields and industries. Here are some examples:

Business and Management

Case studies are widely used in business and management to examine real-life situations and develop problem-solving skills. Case studies can help students and professionals to develop a deep understanding of business concepts, theories, and best practices.

Case studies are used in healthcare to examine patient care, treatment options, and outcomes. Case studies can help healthcare professionals to develop critical thinking skills, diagnose complex medical conditions, and develop effective treatment plans.

Case studies are used in education to examine teaching and learning practices. Case studies can help educators to develop effective teaching strategies, evaluate student progress, and identify areas for improvement.

Social Sciences

Case studies are widely used in social sciences to examine human behavior, social phenomena, and cultural practices. Case studies can help researchers to develop theories, test hypotheses, and gain insights into complex social issues.

Law and Ethics

Case studies are used in law and ethics to examine legal and ethical dilemmas. Case studies can help lawyers, policymakers, and ethical professionals to develop critical thinking skills, analyze complex cases, and make informed decisions.

Purpose of Case Study

The purpose of a case study is to provide a detailed analysis of a specific phenomenon, issue, or problem in its real-life context. A case study is a qualitative research method that involves the in-depth exploration and analysis of a particular case, which can be an individual, group, organization, event, or community.

The primary purpose of a case study is to generate a comprehensive and nuanced understanding of the case, including its history, context, and dynamics. Case studies can help researchers to identify and examine the underlying factors, processes, and mechanisms that contribute to the case and its outcomes. This can help to develop a more accurate and detailed understanding of the case, which can inform future research, practice, or policy.

Case studies can also serve other purposes, including:

• Illustrating a theory or concept: Case studies can be used to illustrate and explain theoretical concepts and frameworks, providing concrete examples of how they can be applied in real-life situations.
• Developing hypotheses: Case studies can help to generate hypotheses about the causal relationships between different factors and outcomes, which can be tested through further research.
• Providing insight into complex issues: Case studies can provide insights into complex and multifaceted issues, which may be difficult to understand through other research methods.
• Informing practice or policy: Case studies can be used to inform practice or policy by identifying best practices, lessons learned, or areas for improvement.

Advantages of Case Study Research

There are several advantages of case study research, including:

• In-depth exploration: Case study research allows for a detailed exploration and analysis of a specific phenomenon, issue, or problem in its real-life context. This can provide a comprehensive understanding of the case and its dynamics, which may not be possible through other research methods.
• Rich data: Case study research can generate rich and detailed data, including qualitative data such as interviews, observations, and documents. This can provide a nuanced understanding of the case and its complexity.
• Holistic perspective: Case study research allows for a holistic perspective of the case, taking into account the various factors, processes, and mechanisms that contribute to the case and its outcomes. This can help to develop a more accurate and comprehensive understanding of the case.
• Theory development: Case study research can help to develop and refine theories and concepts by providing empirical evidence and concrete examples of how they can be applied in real-life situations.
• Practical application: Case study research can inform practice or policy by identifying best practices, lessons learned, or areas for improvement.
• Contextualization: Case study research takes into account the specific context in which the case is situated, which can help to understand how the case is influenced by the social, cultural, and historical factors of its environment.

Limitations of Case Study Research

There are several limitations of case study research, including:

• Limited generalizability : Case studies are typically focused on a single case or a small number of cases, which limits the generalizability of the findings. The unique characteristics of the case may not be applicable to other contexts or populations, which may limit the external validity of the research.
• Biased sampling: Case studies may rely on purposive or convenience sampling, which can introduce bias into the sample selection process. This may limit the representativeness of the sample and the generalizability of the findings.
• Subjectivity: Case studies rely on the interpretation of the researcher, which can introduce subjectivity into the analysis. The researcher’s own biases, assumptions, and perspectives may influence the findings, which may limit the objectivity of the research.
• Limited control: Case studies are typically conducted in naturalistic settings, which limits the control that the researcher has over the environment and the variables being studied. This may limit the ability to establish causal relationships between variables.
• Time-consuming: Case studies can be time-consuming to conduct, as they typically involve a detailed exploration and analysis of a specific case. This may limit the feasibility of conducting multiple case studies or conducting case studies in a timely manner.
• Resource-intensive: Case studies may require significant resources, including time, funding, and expertise. This may limit the ability of researchers to conduct case studies in resource-constrained settings.

About the author

Muhammad Hassan

Researcher, Academic Writer, Web developer

You may also like

Questionnaire – Definition, Types, and Examples

Observational Research – Methods and Guide

Quantitative Research – Methods, Types and...

Qualitative Research Methods

Explanatory Research – Types, Methods, Guide

Survey Research – Types, Methods, Examples

Have a language expert improve your writing

Run a free plagiarism check in 10 minutes, automatically generate references for free.

• Knowledge Base
• Methodology
• Case Study | Definition, Examples & Methods

Case Study | Definition, Examples & Methods

Published on 5 May 2022 by Shona McCombes . Revised on 30 January 2023.

A case study is a detailed study of a specific subject, such as a person, group, place, event, organisation, or phenomenon. Case studies are commonly used in social, educational, clinical, and business research.

A case study research design usually involves qualitative methods , but quantitative methods are sometimes also used. Case studies are good for describing , comparing, evaluating, and understanding different aspects of a research problem .

Table of contents

When to do a case study, step 1: select a case, step 2: build a theoretical framework, step 3: collect your data, step 4: describe and analyse the case.

A case study is an appropriate research design when you want to gain concrete, contextual, in-depth knowledge about a specific real-world subject. It allows you to explore the key characteristics, meanings, and implications of the case.

Case studies are often a good choice in a thesis or dissertation . They keep your project focused and manageable when you don’t have the time or resources to do large-scale research.

You might use just one complex case study where you explore a single subject in depth, or conduct multiple case studies to compare and illuminate different aspects of your research problem.

Prevent plagiarism, run a free check.

Once you have developed your problem statement and research questions , you should be ready to choose the specific case that you want to focus on. A good case study should have the potential to:

• Provide new or unexpected insights into the subject
• Challenge or complicate existing assumptions and theories
• Propose practical courses of action to resolve a problem
• Open up new directions for future research

Unlike quantitative or experimental research, a strong case study does not require a random or representative sample. In fact, case studies often deliberately focus on unusual, neglected, or outlying cases which may shed new light on the research problem.

If you find yourself aiming to simultaneously investigate and solve an issue, consider conducting action research . As its name suggests, action research conducts research and takes action at the same time, and is highly iterative and flexible. 

However, you can also choose a more common or representative case to exemplify a particular category, experience, or phenomenon.

While case studies focus more on concrete details than general theories, they should usually have some connection with theory in the field. This way the case study is not just an isolated description, but is integrated into existing knowledge about the topic. It might aim to:

• Exemplify a theory by showing how it explains the case under investigation
• Expand on a theory by uncovering new concepts and ideas that need to be incorporated
• Challenge a theory by exploring an outlier case that doesn’t fit with established assumptions

To ensure that your analysis of the case has a solid academic grounding, you should conduct a literature review of sources related to the topic and develop a theoretical framework . This means identifying key concepts and theories to guide your analysis and interpretation.

There are many different research methods you can use to collect data on your subject. Case studies tend to focus on qualitative data using methods such as interviews, observations, and analysis of primary and secondary sources (e.g., newspaper articles, photographs, official records). Sometimes a case study will also collect quantitative data .

The aim is to gain as thorough an understanding as possible of the case and its context.

In writing up the case study, you need to bring together all the relevant aspects to give as complete a picture as possible of the subject.

How you report your findings depends on the type of research you are doing. Some case studies are structured like a standard scientific paper or thesis, with separate sections or chapters for the methods , results , and discussion .

Others are written in a more narrative style, aiming to explore the case from various angles and analyse its meanings and implications (for example, by using textual analysis or discourse analysis ).

In all cases, though, make sure to give contextual details about the case, connect it back to the literature and theory, and discuss how it fits into wider patterns or debates.

Cite this Scribbr article

If you want to cite this source, you can copy and paste the citation or click the ‘Cite this Scribbr article’ button to automatically add the citation to our free Reference Generator.

McCombes, S. (2023, January 30). Case Study | Definition, Examples & Methods. Scribbr. Retrieved 14 May 2024, from https://www.scribbr.co.uk/research-methods/case-studies/

Is this article helpful?

Shona McCombes

Other students also liked, correlational research | guide, design & examples, a quick guide to experimental design | 5 steps & examples, descriptive research design | definition, methods & examples.

The Ultimate Guide to Qualitative Research - Part 1: The Basics

• Introduction and overview
• What is qualitative research?
• What is qualitative data?
• Examples of qualitative data
• Qualitative vs. quantitative research
• Mixed methods
• Qualitative research preparation
• Theoretical perspective
• Theoretical framework
• Literature reviews

Research question

• Conceptual framework
• Conceptual vs. theoretical framework

Data collection

• Qualitative research methods
• Focus groups
• Observational research

What is a case study?

Applications for case study research, what is a good case study, process of case study design, benefits and limitations of case studies.

• Ethnographical research
• Ethical considerations
• Confidentiality and privacy
• Power dynamics
• Reflexivity

Case studies

Case studies are essential to qualitative research , offering a lens through which researchers can investigate complex phenomena within their real-life contexts. This chapter explores the concept, purpose, applications, examples, and types of case studies and provides guidance on how to conduct case study research effectively.

Whereas quantitative methods look at phenomena at scale, case study research looks at a concept or phenomenon in considerable detail. While analyzing a single case can help understand one perspective regarding the object of research inquiry, analyzing multiple cases can help obtain a more holistic sense of the topic or issue. Let's provide a basic definition of a case study, then explore its characteristics and role in the qualitative research process.

Definition of a case study

A case study in qualitative research is a strategy of inquiry that involves an in-depth investigation of a phenomenon within its real-world context. It provides researchers with the opportunity to acquire an in-depth understanding of intricate details that might not be as apparent or accessible through other methods of research. The specific case or cases being studied can be a single person, group, or organization – demarcating what constitutes a relevant case worth studying depends on the researcher and their research question .

Among qualitative research methods , a case study relies on multiple sources of evidence, such as documents, artifacts, interviews , or observations , to present a complete and nuanced understanding of the phenomenon under investigation. The objective is to illuminate the readers' understanding of the phenomenon beyond its abstract statistical or theoretical explanations.

Characteristics of case studies

Case studies typically possess a number of distinct characteristics that set them apart from other research methods. These characteristics include a focus on holistic description and explanation, flexibility in the design and data collection methods, reliance on multiple sources of evidence, and emphasis on the context in which the phenomenon occurs.

Furthermore, case studies can often involve a longitudinal examination of the case, meaning they study the case over a period of time. These characteristics allow case studies to yield comprehensive, in-depth, and richly contextualized insights about the phenomenon of interest.

The role of case studies in research

Case studies hold a unique position in the broader landscape of research methods aimed at theory development. They are instrumental when the primary research interest is to gain an intensive, detailed understanding of a phenomenon in its real-life context.

In addition, case studies can serve different purposes within research - they can be used for exploratory, descriptive, or explanatory purposes, depending on the research question and objectives. This flexibility and depth make case studies a valuable tool in the toolkit of qualitative researchers.

Remember, a well-conducted case study can offer a rich, insightful contribution to both academic and practical knowledge through theory development or theory verification, thus enhancing our understanding of complex phenomena in their real-world contexts.

What is the purpose of a case study?

Case study research aims for a more comprehensive understanding of phenomena, requiring various research methods to gather information for qualitative analysis . Ultimately, a case study can allow the researcher to gain insight into a particular object of inquiry and develop a theoretical framework relevant to the research inquiry.

Why use case studies in qualitative research?

Using case studies as a research strategy depends mainly on the nature of the research question and the researcher's access to the data.

Conducting case study research provides a level of detail and contextual richness that other research methods might not offer. They are beneficial when there's a need to understand complex social phenomena within their natural contexts.

The explanatory, exploratory, and descriptive roles of case studies

Case studies can take on various roles depending on the research objectives. They can be exploratory when the research aims to discover new phenomena or define new research questions; they are descriptive when the objective is to depict a phenomenon within its context in a detailed manner; and they can be explanatory if the goal is to understand specific relationships within the studied context. Thus, the versatility of case studies allows researchers to approach their topic from different angles, offering multiple ways to uncover and interpret the data .

The impact of case studies on knowledge development

Case studies play a significant role in knowledge development across various disciplines. Analysis of cases provides an avenue for researchers to explore phenomena within their context based on the collected data.

This can result in the production of rich, practical insights that can be instrumental in both theory-building and practice. Case studies allow researchers to delve into the intricacies and complexities of real-life situations, uncovering insights that might otherwise remain hidden.

Types of case studies

In qualitative research , a case study is not a one-size-fits-all approach. Depending on the nature of the research question and the specific objectives of the study, researchers might choose to use different types of case studies. These types differ in their focus, methodology, and the level of detail they provide about the phenomenon under investigation.

Understanding these types is crucial for selecting the most appropriate approach for your research project and effectively achieving your research goals. Let's briefly look at the main types of case studies.

Exploratory case studies

Exploratory case studies are typically conducted to develop a theory or framework around an understudied phenomenon. They can also serve as a precursor to a larger-scale research project. Exploratory case studies are useful when a researcher wants to identify the key issues or questions which can spur more extensive study or be used to develop propositions for further research. These case studies are characterized by flexibility, allowing researchers to explore various aspects of a phenomenon as they emerge, which can also form the foundation for subsequent studies.

Descriptive case studies

Descriptive case studies aim to provide a complete and accurate representation of a phenomenon or event within its context. These case studies are often based on an established theoretical framework, which guides how data is collected and analyzed. The researcher is concerned with describing the phenomenon in detail, as it occurs naturally, without trying to influence or manipulate it.

Explanatory case studies

Explanatory case studies are focused on explanation - they seek to clarify how or why certain phenomena occur. Often used in complex, real-life situations, they can be particularly valuable in clarifying causal relationships among concepts and understanding the interplay between different factors within a specific context.

Intrinsic, instrumental, and collective case studies

These three categories of case studies focus on the nature and purpose of the study. An intrinsic case study is conducted when a researcher has an inherent interest in the case itself. Instrumental case studies are employed when the case is used to provide insight into a particular issue or phenomenon. A collective case study, on the other hand, involves studying multiple cases simultaneously to investigate some general phenomena.

Each type of case study serves a different purpose and has its own strengths and challenges. The selection of the type should be guided by the research question and objectives, as well as the context and constraints of the research.

The flexibility, depth, and contextual richness offered by case studies make this approach an excellent research method for various fields of study. They enable researchers to investigate real-world phenomena within their specific contexts, capturing nuances that other research methods might miss. Across numerous fields, case studies provide valuable insights into complex issues.

Critical information systems research

Case studies provide a detailed understanding of the role and impact of information systems in different contexts. They offer a platform to explore how information systems are designed, implemented, and used and how they interact with various social, economic, and political factors. Case studies in this field often focus on examining the intricate relationship between technology, organizational processes, and user behavior, helping to uncover insights that can inform better system design and implementation.

Health research

Health research is another field where case studies are highly valuable. They offer a way to explore patient experiences, healthcare delivery processes, and the impact of various interventions in a real-world context.

Case studies can provide a deep understanding of a patient's journey, giving insights into the intricacies of disease progression, treatment effects, and the psychosocial aspects of health and illness.

Asthma research studies

Specifically within medical research, studies on asthma often employ case studies to explore the individual and environmental factors that influence asthma development, management, and outcomes. A case study can provide rich, detailed data about individual patients' experiences, from the triggers and symptoms they experience to the effectiveness of various management strategies. This can be crucial for developing patient-centered asthma care approaches.

Other fields

Apart from the fields mentioned, case studies are also extensively used in business and management research, education research, and political sciences, among many others. They provide an opportunity to delve into the intricacies of real-world situations, allowing for a comprehensive understanding of various phenomena.

Case studies, with their depth and contextual focus, offer unique insights across these varied fields. They allow researchers to illuminate the complexities of real-life situations, contributing to both theory and practice.

Whatever field you're in, ATLAS.ti puts your data to work for you

Download a free trial of ATLAS.ti to turn your data into insights.

Understanding the key elements of case study design is crucial for conducting rigorous and impactful case study research. A well-structured design guides the researcher through the process, ensuring that the study is methodologically sound and its findings are reliable and valid. The main elements of case study design include the research question , propositions, units of analysis, and the logic linking the data to the propositions.

The research question is the foundation of any research study. A good research question guides the direction of the study and informs the selection of the case, the methods of collecting data, and the analysis techniques. A well-formulated research question in case study research is typically clear, focused, and complex enough to merit further detailed examination of the relevant case(s).

Propositions

Propositions, though not necessary in every case study, provide a direction by stating what we might expect to find in the data collected. They guide how data is collected and analyzed by helping researchers focus on specific aspects of the case. They are particularly important in explanatory case studies, which seek to understand the relationships among concepts within the studied phenomenon.

Units of analysis

The unit of analysis refers to the case, or the main entity or entities that are being analyzed in the study. In case study research, the unit of analysis can be an individual, a group, an organization, a decision, an event, or even a time period. It's crucial to clearly define the unit of analysis, as it shapes the qualitative data analysis process by allowing the researcher to analyze a particular case and synthesize analysis across multiple case studies to draw conclusions.

Argumentation

This refers to the inferential model that allows researchers to draw conclusions from the data. The researcher needs to ensure that there is a clear link between the data, the propositions (if any), and the conclusions drawn. This argumentation is what enables the researcher to make valid and credible inferences about the phenomenon under study.

Understanding and carefully considering these elements in the design phase of a case study can significantly enhance the quality of the research. It can help ensure that the study is methodologically sound and its findings contribute meaningful insights about the case.

Ready to jumpstart your research with ATLAS.ti?

Conceptualize your research project with our intuitive data analysis interface. Download a free trial today.

Conducting a case study involves several steps, from defining the research question and selecting the case to collecting and analyzing data . This section outlines these key stages, providing a practical guide on how to conduct case study research.

Defining the research question

The first step in case study research is defining a clear, focused research question. This question should guide the entire research process, from case selection to analysis. It's crucial to ensure that the research question is suitable for a case study approach. Typically, such questions are exploratory or descriptive in nature and focus on understanding a phenomenon within its real-life context.

Selecting and defining the case

The selection of the case should be based on the research question and the objectives of the study. It involves choosing a unique example or a set of examples that provide rich, in-depth data about the phenomenon under investigation. After selecting the case, it's crucial to define it clearly, setting the boundaries of the case, including the time period and the specific context.

Previous research can help guide the case study design. When considering a case study, an example of a case could be taken from previous case study research and used to define cases in a new research inquiry. Considering recently published examples can help understand how to select and define cases effectively.

Developing a detailed case study protocol

A case study protocol outlines the procedures and general rules to be followed during the case study. This includes the data collection methods to be used, the sources of data, and the procedures for analysis. Having a detailed case study protocol ensures consistency and reliability in the study.

The protocol should also consider how to work with the people involved in the research context to grant the research team access to collecting data. As mentioned in previous sections of this guide, establishing rapport is an essential component of qualitative research as it shapes the overall potential for collecting and analyzing data.

Collecting data

Gathering data in case study research often involves multiple sources of evidence, including documents, archival records, interviews, observations, and physical artifacts. This allows for a comprehensive understanding of the case. The process for gathering data should be systematic and carefully documented to ensure the reliability and validity of the study.

Analyzing and interpreting data

The next step is analyzing the data. This involves organizing the data , categorizing it into themes or patterns , and interpreting these patterns to answer the research question. The analysis might also involve comparing the findings with prior research or theoretical propositions.

Writing the case study report

The final step is writing the case study report . This should provide a detailed description of the case, the data, the analysis process, and the findings. The report should be clear, organized, and carefully written to ensure that the reader can understand the case and the conclusions drawn from it.

Each of these steps is crucial in ensuring that the case study research is rigorous, reliable, and provides valuable insights about the case.

The type, depth, and quality of data in your study can significantly influence the validity and utility of the study. In case study research, data is usually collected from multiple sources to provide a comprehensive and nuanced understanding of the case. This section will outline the various methods of collecting data used in case study research and discuss considerations for ensuring the quality of the data.

Interviews are a common method of gathering data in case study research. They can provide rich, in-depth data about the perspectives, experiences, and interpretations of the individuals involved in the case. Interviews can be structured , semi-structured , or unstructured , depending on the research question and the degree of flexibility needed.

Observations

Observations involve the researcher observing the case in its natural setting, providing first-hand information about the case and its context. Observations can provide data that might not be revealed in interviews or documents, such as non-verbal cues or contextual information.

Documents and artifacts

Documents and archival records provide a valuable source of data in case study research. They can include reports, letters, memos, meeting minutes, email correspondence, and various public and private documents related to the case.

These records can provide historical context, corroborate evidence from other sources, and offer insights into the case that might not be apparent from interviews or observations.

Physical artifacts refer to any physical evidence related to the case, such as tools, products, or physical environments. These artifacts can provide tangible insights into the case, complementing the data gathered from other sources.

Ensuring the quality of data collection

Determining the quality of data in case study research requires careful planning and execution. It's crucial to ensure that the data is reliable, accurate, and relevant to the research question. This involves selecting appropriate methods of collecting data, properly training interviewers or observers, and systematically recording and storing the data. It also includes considering ethical issues related to collecting and handling data, such as obtaining informed consent and ensuring the privacy and confidentiality of the participants.

Data analysis

Analyzing case study research involves making sense of the rich, detailed data to answer the research question. This process can be challenging due to the volume and complexity of case study data. However, a systematic and rigorous approach to analysis can ensure that the findings are credible and meaningful. This section outlines the main steps and considerations in analyzing data in case study research.

Organizing the data

The first step in the analysis is organizing the data. This involves sorting the data into manageable sections, often according to the data source or the theme. This step can also involve transcribing interviews, digitizing physical artifacts, or organizing observational data.

Categorizing and coding the data

Once the data is organized, the next step is to categorize or code the data. This involves identifying common themes, patterns, or concepts in the data and assigning codes to relevant data segments. Coding can be done manually or with the help of software tools, and in either case, qualitative analysis software can greatly facilitate the entire coding process. Coding helps to reduce the data to a set of themes or categories that can be more easily analyzed.

Identifying patterns and themes

After coding the data, the researcher looks for patterns or themes in the coded data. This involves comparing and contrasting the codes and looking for relationships or patterns among them. The identified patterns and themes should help answer the research question.

Interpreting the data

Once patterns and themes have been identified, the next step is to interpret these findings. This involves explaining what the patterns or themes mean in the context of the research question and the case. This interpretation should be grounded in the data, but it can also involve drawing on theoretical concepts or prior research.

Verification of the data

The last step in the analysis is verification. This involves checking the accuracy and consistency of the analysis process and confirming that the findings are supported by the data. This can involve re-checking the original data, checking the consistency of codes, or seeking feedback from research participants or peers.

Like any research method , case study research has its strengths and limitations. Researchers must be aware of these, as they can influence the design, conduct, and interpretation of the study.

Understanding the strengths and limitations of case study research can also guide researchers in deciding whether this approach is suitable for their research question . This section outlines some of the key strengths and limitations of case study research.

Benefits include the following:

• Rich, detailed data: One of the main strengths of case study research is that it can generate rich, detailed data about the case. This can provide a deep understanding of the case and its context, which can be valuable in exploring complex phenomena.
• Flexibility: Case study research is flexible in terms of design , data collection , and analysis . A sufficient degree of flexibility allows the researcher to adapt the study according to the case and the emerging findings.
• Real-world context: Case study research involves studying the case in its real-world context, which can provide valuable insights into the interplay between the case and its context.
• Multiple sources of evidence: Case study research often involves collecting data from multiple sources , which can enhance the robustness and validity of the findings.

On the other hand, researchers should consider the following limitations:

• Generalizability: A common criticism of case study research is that its findings might not be generalizable to other cases due to the specificity and uniqueness of each case.
• Time and resource intensive: Case study research can be time and resource intensive due to the depth of the investigation and the amount of collected data.
• Complexity of analysis: The rich, detailed data generated in case study research can make analyzing the data challenging.
• Subjectivity: Given the nature of case study research, there may be a higher degree of subjectivity in interpreting the data , so researchers need to reflect on this and transparently convey to audiences how the research was conducted.

Being aware of these strengths and limitations can help researchers design and conduct case study research effectively and interpret and report the findings appropriately.

Ready to analyze your data with ATLAS.ti?

See how our intuitive software can draw key insights from your data with a free trial today.

• AI Content Shield
• AI KW Research
• AI Assistant
• SEO Optimizer
• AI KW Clustering
• Customer reviews
• The NLO Revolution
• Press Center
• Help Center
• Content Resources
• Facebook Group

How to Write a Multiple Case Study Effectively

Table of Contents

Have you ever been assigned to write a multiple case study but don’t know where to begin? Are you intimidated by the complexity and challenge it brings? Don’t worry! This article will help you learn how to write a multiple case study effectively that will make an impactful impression. So, let’s begin by defining a multiple case study.

What Is a Multiple Case Study?

A multiple case study is a research method examining several different entities. It helps researchers gain an understanding of the entities’ individual characteristics and disclose any shared patterns or insights. This type of investigation often uses both qualitative and quantitative data. These are usually collected from interviews, surveys, field observations, archival records, and other sources. This is done to analyze the relationships between each entity and its environment. The results can provide valuable insights for policymakers and decision-makers.

Why Is a Multiple Case Study Important?

A multiple case study is invaluable in providing a comprehensive view of a particular issue or phenomenon. Analyzing a range of cases allows for comparisons and contrasts to be drawn. And this can help identify broader trends, implications, and causes that might otherwise remain undetected. This method is particularly useful in developing theories and testing hypotheses. This is because the range of data collected provides more robust evidence than what could be achieved from one single case alone.

How to Write a Multiple Case Study

Below are the key steps on how to write a multiple case study :

1. Brainstorm Potential Case Studies

Before beginning your multiple case study, you should brainstorm potential cases suitable for the research project. Consider both theoretical and practical implications when deciding which cases are most appropriate. Think about how these cases can best illustrate the issue or question at hand. Make sure to consider all relevant information before making any decisions.

2. Conduct Background Research on Each Case

After selecting the individual cases for your multiple case study, the next step is to do background research for each case. Conducting extensive background research on each case will help you better understand the context of the study. This research will allow you to form an educated opinion and provide insight into the problems and challenges that each case may present.

3. Establish a Research Methodology

A successful multiple-case study requires a sound research methodology. This includes deciding on the methods of data collection and analysis and setting objectives. It also involves developing criteria for evaluating the results and determining what kind of data needs to be collected from each case. All of this must be done carefully, considering the purpose of the study and its outcomes.

4. Collect Data

Once a research method has been established, it is time to collect data from each case included in the study. Depending on the nature of the research project, this could involve interviewing participants, gathering statistics, or observing behaviors in certain settings. It is crucial to ensure that all data collected is accurate and reliable.

5. Analyze & Interpret Data

After the data has been collected, it must be analyzed to draw meaningful conclusions from it. This process involves examining patterns and trends within the data, identifying relationships between variables, and looking for commonalities among different cases. These findings must then be interpreted in light of the initial questions posed by the study.

6. Write the Report

After completing the analysis and interpretation of the data, it is finally time to write up the results of the multiple case study. This should include a summary of the key findings and an explanation of why these findings are significant. In addition, the limitations of the study should be acknowledged, along with recommendations for future research in this area.

Writing a multiple case study requires careful planning and execution. But the process becomes easier when you know the proper steps to conduct and create a multiple case study. It requires you to focus on the design of the study, including the sample chosen and the research methodology established. Conducting background research on each case and collecting data are also crucial steps in the process. To guide you through the process, this article outlines the key steps to help you easily write a well-structured multiple-case study .

Abir Ghenaiet

Abir is a data analyst and researcher. Among her interests are artificial intelligence, machine learning, and natural language processing. As a humanitarian and educator, she actively supports women in tech and promotes diversity.

Explore All Write A Case Study Articles

How to write a leadership case study (sample) .

Writing a case study isn’t as straightforward as writing essays. But it has proven to be an effective way of…

• Write A Case Study

Top 5 Online Expert Case Study Writing Services 

It’s a few hours to your deadline — and your case study college assignment is still a mystery to you.…

Examples Of Business Case Study In Research

A business case study can prevent an imminent mistake in business. How? It’s an effective teaching technique that teaches students…

Have you ever been assigned to write a multiple case study but don’t know where to begin? Are you intimidated…

How to Write a Case Study Presentation: 6 Key Steps

Case studies are an essential element of the business world. Understanding how to write a case study presentation will give…

How to Write a Case Study for Your Portfolio

Are you ready to showcase your design skills and move your career to the next level? Crafting a compelling case…

Log in using your username and password

• Search More Search for this keyword Advanced search
• Latest content
• For authors
• Browse by collection
• BMJ Journals More You are viewing from: Google Indexer

You are here

• Volume 14, Issue 5
• Exploring the influence of health system factors on adaptive capacity in diverse hospital teams in Norway: a multiple case study approach
• Article Text
• Article info
• Citation Tools
• Rapid Responses
• Article metrics

• http://orcid.org/0000-0002-4689-8376 Birte Fagerdal 1 ,
• http://orcid.org/0000-0001-7107-4224 Hilda Bø Lyng 1 ,
• http://orcid.org/0000-0002-9124-1664 Veslemøy Guise 1 ,
• Janet E Anderson 2 ,
• http://orcid.org/0000-0003-0296-4957 Jeffrey Braithwaite 3 ,
• http://orcid.org/0000-0003-0186-038X Siri Wiig 1
• 1 SHARE, Faculty of Health Sciences , University of Stavanger , Stavanger , Norway
• 2 Anaesthesiology and Perioperative Medicine , Monash University Faculty of Medicine Nursing and Health Sciences , Melbourne , Victoria , Australia
• 3 Australian Institute of Health Innovation , Macquarie University , North Ryde , New South Wales , Australia
• Correspondence to Mrs Birte Fagerdal; birte.fagerdal{at}uis.no

Objectives Understanding flexibility and adaptive capacities in complex healthcare systems is a cornerstone of resilient healthcare. Health systems provide structures in the form of standards, rules and regulation to healthcare providers in defined settings such as hospitals. There is little knowledge of how hospital teams are affected by the rules and regulations imposed by multiple governmental bodies, and how health system factors influence adaptive capacity in hospital teams. The aim of this study is to explore the extent to which health system factors enable or constrain adaptive capacity in hospital teams.

Design A qualitative multiple case study using observation and semistructured interviews was conducted between November 2020 and June 2021. Data were analysed through qualitative content analysis with a combined inductive and deductive approach.

Setting Two hospitals situated in the same health region in Norway.

Participants Members from 8 different hospital teams were observed during their workday (115 hours) and were subsequently interviewed about their work (n=30). The teams were categorised as structural, hybrid, coordinating and responsive teams.

Results Two main health system factors were found to enable adaptive capacity in the teams: (1) organisation according to regulatory requirements to ensure adaptive capacity, and (2) negotiation of various resources provided by the governing authorities to ensure adaptive capacity. Our results show that aligning to local context of these health system factors affected the team’s adaptive capacity.

Conclusions Health system factors should create conditions for careful and safe care to emerge and provide conditions that allow for teams to develop both their professional expertise and systems and guidelines that are robust yet sufficiently flexible to fit their everyday work context.

• Health & safety
• Organisation of health services
• Quality in health care
• Protocols & guidelines
• QUALITATIVE RESEARCH

Data availability statement

Data are available upon reasonable request.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:  http://creativecommons.org/licenses/by-nc/4.0/ .

https://doi.org/10.1136/bmjopen-2023-076945

Statistics from Altmetric.com

Request permissions.

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

STRENGTHS AND LIMITATIONS OF THIS STUDY

Data for this study were collected during the COVID-19 pandemic, which enabled the research team to observe how novel national policy measures affected the frontline.

The study contributes to resilient healthcare as there have been few multilevel studies looking at how macrolevel factors affect microlevel adaptive capacity.

The combination of observations and interviews provided a substantial amount of data which were then triangulated.

Data collected at the national level are limited as our study focused on the hospital team level.

Introduction

Healthcare systems provide the formal healthcare delivery structures for a defined population, whose funding, management, scope and content are defined by laws, policies and regulations. They provide services to people, aiming to contribute to their health and well-being. Services are usually delivered in defined settings, such as homes, nursing homes and hospitals. Healthcare systems are complex and adaptive and continuously responsive to multiple factors including patients’ needs, innovations, pressures, pandemics and funding structures. 1 Understanding flexibility and adaptive capacities in these complex healthcare systems is a key focus of investigators of resilient healthcare. 2 3 Resilience in healthcare can briefly be defined as ‘the capacity to adapt to challenges and changes at different system levels, to maintain high quality care’ p6. 4

To date, research on resilient healthcare has paid most attention to work as done at the sharp end of the system. Less is therefore known about how actions, strategies and practices enacted by regulatory bodies and policy-makers affect every day work at the microlevel, such as hospital teams. 5 While regulations in the form of standards, rules and protocols are known to be key drivers in the structuring of healthcare activities and in the design of healthcare organisations, the interfaces between policy-making, regulation and resilience are subtle and nuanced, and regulatory strategies to improve quality and safety are therefore complex and multifarious. 6 7 However, the relationship between governmental bodies and adaptive capacity at the sharp end of the system has received insufficient attention and is thus in need of closer examination. 2 8 9

In this study, we define macrolevel healthcare system actors as governmental bodies, regulators and national and regional bodies, who act or intend to shape, monitor, control and modify practices within organisations in order to achieve an identifiable, desirable state of affairs. 10 They aim to constrain action, optimise performance and attempt to prevent error.

In complex systems like hospitals, much work is performed in teams. 11–13 Understanding the nature of teams and team performance is important to promote team effectiveness. The few studies that have been undertaken are limited in scope as they have not considered how teams are defined and structured, what their functions are or differences across healthcare teams. 11 14 Most research on teams in healthcare has focused on the dynamic domains in healthcare, such as emergency medicine or operating rooms, and teams that are similar to the teams in other industries, for instance in aviation. 15 16 However, not all teams in hospitals operate in an emergency setting. Teams in hospitals differ depending on their goals, tasks, structure, membership and situation, affecting how they adapt to a multitude of contingencies that are encountered in everyday work. 17 Hence, their requirements for support could differ depending on these attributes but this question has not been addressed sufficiently in previous research. Knowledge of these differences may enable optimisation of support and better function for the different teams. This study will address these knowledge gaps.

Aim and research question

This study aims to explore whether and how health system factors enable adaptive capacity in different types of hospital teams in Norway. We asked: What kind of health system factors enable adaptive capacity in hospital teams, and how do these factors affect adaptive capacity?

Design and setting

A qualitative exploratory methodology was chosen, using a multiple-embedded case study design. 11 18 A case was defined as one hospital containing four different types of teams. Two case hospitals were recruited to the study, featuring a total of eight teams. The study’s design was in line with that of an international comparative study, involving six countries (The Netherlands, Japan, Australia, England, Switzerland and Norway), where this article reports partial findings from the Norwegian case (see protocol of Anderson et al ). 11 The two Norwegian hospitals and the four team types were recruited in line with the study protocol. Findings from each of the countries will be written up as country case reports following an agreed on template. Furthermore, an international cross-case comparative analysis will be performed using the Qualitative Comparative Analysis method 19 with the aim of exploring how multilevel system factors interact to support or hinder adaptive capacity in different types of hospital teams in different countries, and how this leads to performance variability. This international comparative analysis is currently in progress. This article stands alone and uses Norwegian data only.

Recruitment and study context

The Norwegian health system is a semidecentralised system with the Norwegian Parliament as its highest decision-making body. The municipalities are responsible for providing primary care for their citizens, mainly through nursing homes, homecare, general practitioners and rehabilitation services. The hospitals are mainly state owned and administered by four Regional Health Authorities. The Norwegian Board of Health Supervision is a national regulatory body, organised under the Ministry of Health and Care Services. County Governors at the regional level oversee services within primary and specialised healthcare. Norway has a comprehensive set of legislation governing the health services, including requirements for the quality of services, regulations for authorised healthcare personnel and service users’ rights. These legislated requirements are subject to supervision and investigation by the Norwegian Board of Health Supervision and the County Governors. 20 21

The two hospitals in this study were selected and recruited based on their size and role in teaching provision. 11 Both hospitals are situated in the same health region in Norway. Hospital 1 is a large teaching hospital and hospital 2 is a middle-sized local hospital which is also responsible for educating healthcare professionals. The four different team types were structural, hybrid, responsive and coordinating, and are displayed in table 1 . See Fagerdal et al 22 for further descriptions of the teams.

• View inline

Descriptions of the four different teams studied in each hospital

Data were collected through observation, interviews and document analysis, all undertaken between December 2020 and June 2021. Researcher BF and HBL conducted the observations, which entailed following one or more team members for two workdays using an observation guide. Both researchers wrote their own individual fields notes which were both included in the data material. Using the observation guide enabled a structuring of the text in line with the central concepts used in resilience literature. 23 During observations, we looked for various types of demands from the different levels of the organisations, the teams’ capacities to meet the demands and types of adaptations that were performed by the teams and team members. The observed teams differed in how they work together and consequently our undertaking of the observations had to align with those differences. The structural and hybrid teams were observed during two shifts, including evening and dayshifts. With the responsive teams, we followed one team member during their workday and their response to acute alarms. The coordinating teams meet for 10 min every weekday, and the researchers attended all their meetings during a 14-day period. Due to the COVID-19 pandemic, one of the coordinating teams held their meetings digitally, which we also attended. The observations totalled 115 hours (see table 2 ).

Overview of data collection methods and data material according to team types and case sites

All interviews were undertaken post observation by researcher BF using a semistructured interview guide based on content from the Concepts for Applying Resilience Engineering (CARE) model, that is, demand, capacity, misalignments and adaptations, 24 and the four potentials of resilience; monitoring, anticipating, responding and learning. 23 Team members and one leader from each team were interviewed, resulting in 30 interviews (see table 3 ). Participants comprised 27 females and 3 males and their ages ranged between 24 and 56. The interview length varied from 40 to 90 min with a median of 55 min. All participants signed a written consent form and were given the opportunity to withdraw without any negative implications; all invited participants accepted the invitation to interview.

Overview of the interviewed participants in the study

Patient and public involvement statement

A coresearcher employed in the overall Resilience in Healthcare project, of which this study is a part, 11 collaborated in the planning and design of the study, and access to teams at hospital 1. In hospital 2, we used a local coordinator to help identify and facilitate access to the different teams.

All interviews were audio recorded and transcribed verbatim by researcher BF. Observation notes were included in the analysis, and all notes and interview transcripts were grouped according to hospital and team types to streamline the analysis work. We conducted a within-case analysis of each hospital and a cross-case analysis to identify patterns and themes in our overall material. 18 The data material was first read through in full by all the researchers to get a sense of the whole. The analysis was then done using a combined deductive and inductive approach. 25 The CARE model 24 was used as a framework to assist the deductive part of the analysis as visualised in figure 1 .

• Download figure
• Open in new tab
• Download powerpoint

Concepts for Applying Resilience Engineering model after Anderson et al 24 visualising the study’s focus on team adaptation.

Data were organised using three of the four key concepts in the CARE model matrix: capacities, misalignments and adaptations. The capacities were defined as health system factors in this analysis and represent the factors that influence teams’ ability to adapt. All data were in addition coded for team type and hospital which allowed for a cross hospital and cross-team analysis. After the data material had been divided into three parts of text, to enable further analysis, we proceeded with an inductive content analysis approach. 25 The categories were inductively reviewed and recoded and further developed into latent themes across the four teams. This process resulted in overarching themes representing health system factors, that influence teams’ adaptive capacity (see table 4 ).

Inductive coding structure

The national and regional health authorities set the scene for how the hospitals prioritises and arrange their work. System-level decisions filter down through the organisation and influence the team’s everyday work. Our analysis shows that the effect of system factors on teams’ everyday work and adaptive capacity can be divided into two main themes, each with associated subthemes: (1) organisation according to regulatory requirements to ensure adaptive capacity and (2) negotiation of various resources provided by the governing authorities to ensure adaptive capacity. In table 4 , we present the themes along with their subthemes, codes and examples of quotes from the participants or description from the observation.

Organising according to regulatory requirements to ensure adaptive capacity

National and regional guidelines, financial governance and regulatory inspections by the health supervision authorities all shaped the organisation of the hospitals.

Context and organisational structure

The organisational context was important. It affected how the teams enacted and performed patient care. For instance, the smaller hospital 2 had restrictions and limitations regarding both the types of diagnoses and the number of patients they were able to treat due to regional regulations. These regulations had a large impact on the smaller hospital and their teams in how they organised their work, their competence requirements and what kind of learning opportunities were available to the team members. For instance, since hospital provided an acute function for surgical patients, it could continue to be an educational institution for healthcare personnel, which also meant that healthcare professionals in the structural and hybrid teams could maintain and develop their skills in acute care. In addition, it also impacted the hybrid and structural teams in how they arranged their work by always being prepared for the admission of acute surgical patients during their workday. Furthermore, the regional health authority maintained overall flexibility in acute care provision by having this function in both hospitals.

Both the coordinating teams in our study had been established by the hospitals in response to a government policy of preventing corridor beds in hospitals as a means of improving patients’ safety. The teams were set up to include all ward managers cooperating to manage patient flow, and with a goal of evening out the overall strain across the hospital. These teams’ main assignment was to allocate patients to free beds within the hospital. In addition, a positive consequence of having these teams was that the team members got a better mutual understanding of the overall situation within the hospitals and an improved understanding of each other’s challenges across the hospital. This provided them with a greater range of solutions to use when making adaptations to avoid patients in the corridors. The coordinating team in hospital 2 also functioned as an arena for the team members to exchange advice and suggest solutions to other challenges in their work. This was to a certain extent also valid for the team in hospital 1, but due to the comparatively larger size of the team there, it was more difficult for the team members to get well acquainted. In addition to better patient flow and avoiding corridor patients, the hospitals aimed for the teams to focus on building a culture of helping each other across their respective hospitals and to foster a feeling of joint responsibility for the betterment of the hospital overall (see table 4 ). Similar to the responsive teams, the coordinating teams had been enabled to make quick decisions spanning hospital units, allowing for a wider range of alternative solutions to the problems encountered than if they were to make decisions on their own. Also, team members felt more of a responsibility to help each other and found that it was more difficult to say no to requests for free beds when meeting face to face with colleagues. Both the individual team members and the hospital organisation as a whole were thus found to have widened their adaptive capacities after establishing these teams.

Aligning with national and regional guidelines

The use of clinical guidelines provided teams with direction in the different treatment courses offered to patients. National guidelines were translated and aligned to work practices within the organisation to fit the current work in the teams. This gave the team a standard to maintain, a structure for their work and also brought them a sense of safety in knowing their boundaries and priorities for adaptation. For instance, the national guideline for sepsis treatment recommends starting antibiotics treatment within 1 hour of the start of symptoms and also lists early important diagnostic signs to look for in patients who are deteriorating. Early intervention and treatment improve the overall survival of these patients and both hospitals needed to ensure proper alignment to these standards (see table 4 ). The hybrid and structural teams were well aware of this, due to guidelines and information campaigns. The teams thus adapted their work to meet the national demands imposed here, prioritising this work over what were considered other less important tasks, such as helping patients with personal hygiene.

Another example of how guidelines shaped the organisation of hospital teams and how teams acted was seen in the work of both the responsive teams in the study. The two hospitals had to comply with the national requirements of diagnostic and treatment guidelines for cerebral infarction, and both hospitals had created responsive stroke teams to allow for quick diagnostics and treatment. Tailoring the responsive teams to fit the requirements of the national guidelines, reduced the ‘door to needle time’ in both hospitals significantly. This was accomplished by providing and designing equipment, procedures, role descriptions and facilities along with the right competent personnel. The responsive teams frequently made adaptations to the clinical procedure to fit with the patient’s condition, the proximity of the competent team members and the tailored equipment and location enabled for quick decision-making within the team, instead of encountering communication via phones or waiting for each other to finish other tasks.

Negotiating various resources provided by the governing authorities to ensure adaptive capacity

Financial incentives.

Incentives like the national funding model which generates income for the hospitals impacted both what kind of and how the hospitals prioritised treatment. Governing authorities use financial incentives to orient the hospitals towards planned direction. Budget cuts and other financial restraints imposed on hospitals demanded that both hospitals adapt their priorities, which consequently affected the teams’ delivery of treatment and care in the sharp end of the system. The government requirements for increased efficiency in the healthcare system, such as financial incentives for reducing beds, increased the pace of work and often required development of new work practices to cope with these demands. For instance, in both hospitals, there had been a decrease in hospital beds, and a shift towards outpatient treatment due to governing authorities funding schemes. To cope with this, both the hybrid and structural teams in both hospitals treated patients for a shorter amount of time. For example, the structural teams no longer admitted patients overnight preoperatively and discharged patients earlier postoperatively to primary healthcare service or the home. The teams coped with this by planning the discharge of the patient already at admittance to facilitate a safe and good-quality discharge. However, they often adapted their plans by not discharging patients due to either lack of capacity in primary care services, or disagreement and concern with the level of care offered in the municipalities. This example shows that the teams in practice negotiated the consequences of government funding restrictions to suit the patients’ needs.

In addition, they could to some extent handle some demands by determining how they could change procedures to fit certain requirements. For instance, one of the changes the structural team in hospital 2 made to manage earlier discharge was to have the nightshift staff remove the postoperative urine catheter from patients. The clinical procedure stated that for the patient to be discharged, they had to be able to urinate spontaneously after catheter removal. Catheter removal later in the day regularly meant that the patient had to stay an extra night, so by changing the timing of its removal staff still managed to provide care within the frame of guidelines given.

Physical surroundings

Both the hybrid and responsive teams in both hospitals had been placed in new premises designed specifically to accommodate their way of working, with well-designed spaces to facilitate their workday with proximity to necessary equipment, and a nearness to each other that enabled team members to easily assist if needed. Similarly, the structural team in hospital 2 had new premises, with a uniform design across the new hospital building making it easy for personnel to change teams and wards since their premises were already familiar to them. This uniformity in building design improved the teams’ overall adaptive capacity in peak situations, or when there was an absence of key personnel across wards and teams. Staff could easily assist personnel from other wards as they knew where equipment was stored and how the different facilities in the ward functioned (patient rooms, nurses’ stations, etc). The structural team in hospital 1, however, worked in old premises with narrow hallways and few physical meeting arenas for the team members, which hampered their workflow in that they had to spend time looking for each other, and otherwise had few opportunities to engage in direct communication with each other during their workday. The physical surroundings of the two coordinating teams differed. Due to the size of the team in hospital 1, the team there used digital software to manage the overall patient flow in the hospital. The smaller team in hospital 2 managed the same using a paper form that each member completed. However, both of the teams used the meeting to elaborate on their numbers with additional information as the numbers alone did not provide a sufficient representation of the overall situation on the wards.

Training and development resources

Training and development resources were crucial for a team’s adaptive capacity. The national attention on patient safety in recent decades has led to improved treatment courses and changed the focus on how healthcare personnel can learn from adverse events to avoid similar incidents in the future. Consequently, this has led to innovative solutions in how hospital managers organise learning activities for their employees. In accordance with a growing focus on simulation-based training and learning from regulatory bodies and policy-makers, all the teams in the study apart from the coordinating teams increasingly used simulation training (see table 4 ). Often, the teams would make simulation scenario cases based on adverse events or incidents that had happened on their ward and used them in their training. For the responsive teams, this type of training was mandatory and part of regulatory requirements for the teams. Also, for these teams that only worked together for limited episodes and had changing membership and different professional cultures, these simulation trainings were their only chance to practice and improve their team communication. During the period of our observation, they developed new cases with COVID-19 themes and used them to train and learn before they received actual COVID-19 patients. This improved their performance, as they had found several shortcomings in their COVID-19 procedure and thus changed it accordingly. For instance, they made efforts to prevent unnecessary contamination of team members and had detected a lack in the procedure of personal protective equipment. This shows that these types of prescribed training exercises enable teams to adapt procedures to fit their everyday work conditions.

Quality improvement resources

Quality improvement resources outside the hospital organisation supported team’s adaptive capacity. The national and regional healthcare authorities arrange various conferences and campaigns for hospitals and other healthcare institutions. Here, policy-makers, leaders and healthcare professionals meet and create reflexive spaces. As part of such efforts, the best practices are displayed and workshops are provided to encourage and translate quality and safety improvement into practice in different ways, alongside guidelines, learning tools and other materials for the different organisations to use and implement in their quality improvement work. Having this competence base within the health regions and at the national level to support teams added knowledge and increased adaptive capacity as it required knowledge transfer and new ideas anchored in research and practice. Moreover, the patient safety focus within the wards and teams like the safe care screening programme and safety huddles, launched by the Norwegian Directorate of Health and implemented through the regional health authorities, increased the team’s awareness of patient safety culture. The increased amount of quality measures the clinicians had to undertake and report on in their daily work were generally seen as good quality measures from both the organisations and the team’s point of view. However, it sometimes felt counterproductive constantly having to cope with balancing patients’ needs with the requirements of screening procedures, especially if staff felt they had little room for autonomous clinical assessment. For instance, the safe care screening programme where every patient over the age of 18 had to be screened for their risk of falling, bedsores and possible malnutrition within 24 hours was questioned. Screening young patients for this felt unnecessary and if there were other more pressing tasks that were seen as more important, they adapted the way they prioritised.

This study investigated the relationship between health system factors and adaptive capacity in hospital teams. Our results have shown that health system-level factors influence adaptive capacity in the teams through the provision of guidelines and resources, and how the teams align these to their current demands and capacity situation. Their effects on different teams are not uniform; some are advantageous to one team but disadvantageous to another. 5 6 We argue that it is the team’s opportunity to align these factors to context that are key for enabling adaptive capacity, as illustrated in figure 2 .

Illustrating the teams aligning system-levels factors to context for adaptive capacity.

All levels of a health system can influence each other, especially in an integrated and tightly coupled system. Higher system levels can affect lower levels through, for example, explicit instructions, by the provision or limitation of resources, or by establishing incentive systems. 26–28 On the other hand, lower system levels may use discretion when they interpret and implement directives from higher levels, and they may control the information flow to higher levels. 26 Our results show that decisions made at one level of the system can support or hinder adaptive capacity at other lower hierarchical levels of the system. 29–31 Accordingly, the system-level governing factors affect adaptive capacity at the sharp end by setting the framework and boundaries within which activity can take place. Regulatory bodies have system-wide responsibilities and must respond to system-wide disturbances, without detailed knowledge of how work is done in practice at the sharp end. Consequently, the sharp end must adapt to respond appropriately to disturbances within its own field of responsibility. 32

This study has operationalised adaptation using the CARE model 24 to see how different teams at the sharp end work in practice to negotiate system-level factors, such as regulations and guidelines. The findings show that factors at the macrolevel required different forms of adaptations within different team types to managing everyday work. Enabling adaptation at the team level by taking action at the macrolevel to attempt to reconcile work as imagined with work as done ( figure 1 ). The system-level factors also represent long-term planning and transformation of practices rather than short-term adaptations or adjustments in the system. 33 They envisage setting up the processes that design, produce and circulate resources that underpin safety, and prevent errors through standardisation, regulation and training. 32 How the teams negotiate these long-term transformations to their everyday work determines their adaptive capacity as our results have shown. Adaptation and adjustments to local context are inevitable in healthcare. 9 11 34 35 However, the vast number of protocols, policies, checklists, standards, guidelines, pathways and other regulatory requirements may lead those working at the sharp end to feel overwhelmed. 6 If not aligned with goals, tasks and current challenges, these governing factors may end up being counterproductive. 5 The teams studied talked about their everyday work and their primary focus on patient care along with their willingness to act in the best interest of the patients. 36 They talked about feeling a compound pressure in order to align system-level demands with their context and patients’ wishes and needs. 37 38 Taking the perspective of the patient into account was important to the teams. 39 40 Consequently, different teams had to align system-level demands differentially to ensure quality care for patients.

Our study showed that teams must balance continuous efficiency with thoroughness assessments 32 41–43 in everyday work (eg, making the nightshift prepare discharge adding more work to reduce corridor patients). Ways that the teams in our study continuously adapted regulatory requirements to their work context illuminated how resilient systems must have robust yet flexible structures to assist the system to deal with both everyday work and unexpected events. 8 30 44 45 System-level factors must therefore provide flexibility to fit different situations and types of teams, as teams differ in how they cooperate and function in everyday work. To ensure alignment of perspectives between macrolevel and microlevel actors, common arenas and structures for mutual feedback and reflections between stakeholders are crucial. 7 Furthermore, system factors need to entail robustness in the directions they provide to practice and the implementation of improvement efforts. 33

The findings show that for the responsive and coordinating teams the size of the hospital played a significant role in their ability to adapt. These two team types operated in part at the mesolevel of the hospital organisation, spanning hospital departments. Their work was characteristically ad hoc, dynamically changing team memberships and members who work primarily in other teams. The large size of hospital 1 hampered development of relationships between the team members in both the responsive and the coordinating team, whereas in the smaller hospital 2 it was easier to develop close relationships between colleagues. This implies that ad hoc teams, and especially large ones, need to have structure and guidelines in place that direct their work, and support to adapt their work based on the team members understanding of the tasks and their roles. The structural and hybrid teams were colocated and this seemed to allow for the development of long-term collegial relationships, better cooperation between team members, more flexible adaptation of their work and also seemed to allow for working with greater levels of independence and a larger room for self-organisation. Their work is influenced by system-level demands, but the size of the organisation does not affect their day-to-day work to the same degree as for the coordination and responsive teams.

Strengths and limitations

A strength of the study is that by combining observation and interviews we have gathered in-depth data of the team’s everyday work.

Data collection during COVID-19 pandemic could hamper everyday work practice; however, we collaborated closely with the sites to avoid any problems for the involved teams and units. Only two hospitals contributed to the data collection and including additional hospitals could add more than we have from two hospitals. However, the inclusion of eight teams, the total amount of data gave rich information to analyse our research questions.

Interview data from the macrolevel could have added additional perspectives from the regulators and policy-makers. We suggest further studies to integrate this in their activities to uncover the role of system factors seen from the policy-makers’ and regulators’ perspectives.

Conclusions and implications

This study illuminated how teams negotiate the health system factors that shape their work to provide as much adaptive capacity as possible and attempt to align system-level regulation and guidelines with everyday work demands. The results show that the size of both the organisation and team had an effect on adaptive capacity. Our findings imply that healthcare systems need to facilitate conditions that allow for teams to develop their professional expertise and develop systems that are robust and flexible to fit the context. Teams should be enabled to adapt to the functions and structure of the health system to carry out their everyday work in a changing environment.

Ethics statements

Patient consent for publication.

Not applicable.

Ethics approval

This study involves human participants and was approved by Regional committe for Medical and Health Research Ethics, ref.nr. 166280. Participants gave informed consent to participate in the study before taking part.

Acknowledgments

The authors would like to thank all participating teams and their leaders at the two hospitals who shared their valuable knowledge and reflections.

• Hollnagel E ,
• Braithwaite J
• Anderson JE ,
• Macrae C , et al
• Billett S , et al
• Smaggus A ,
• Ellis LA , et al
• Bal R , et al
• Christian JS ,
• Christian MS ,
• Pearsall MJ , et al
• Roberts APJ ,
• Webster LV ,
• Salmon PM , et al
• Lavelle M ,
• Cross S , et al
• Ballangrud R ,
• Husebø SE ,
• Aase K , et al
• Salas E , et al
• Saunes IS ,
• Karanikolos M ,
• Saunes, I.S., M. Karanikolos, and A. Sagan, Norway
• Fagerdal B ,
• Guise V , et al
• Hollnagel E
• Back J , et al
• Rasmussen J
• Bergström J ,
• Blanchet K ,
• Ramalingam B , et al
• Schubert CC
• Johannessen T ,
• Krystallidou D ,
• Deveugele M , et al
• Dillon EC ,
• Tai-Seale M ,
• Meehan A , et al
• Hofmeyer A ,
• Svensson I ,
• von Knorring M ,
• Hagerman H , et al
• Churruca K ,
• Clay-Williams R , et al
• Glette MK ,

X @fagerbirte

Contributors The study design was developed in collaboration with the whole research team. BF and HBL conducted the data collection. BF conducted and transcribed all the interviews. The analysis and interpretation of data were conducted in close collaboration between BF, HBL, VG, JEA and SW. SW is the guarantor of this study. All authors contributed with writing, critical revision and approval of the final version.

Funding This project is part of the Resilience in Healthcare Research program which has received funding from the Research Council of Norway from the FRIPRO TOPPFORSK program, grant agreement no. 275367. The University of Stavanger, Norway, NTNU Gjøvik, Norway supports the study with kind funding.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

Read the full text or download the PDF:

Artificial intelligence  is being used in healthcare for everything from answering patient questions to assisting with surgeries and developing new pharmaceuticals.

According to  Statista , the artificial intelligence (AI) healthcare market, which is valued at $11 billion in 2021, is projected to be worth $187 billion in 2030. That massive increase means we will likely continue to see considerable changes in how medical providers, hospitals, pharmaceutical and biotechnology companies, and others in the healthcare industry operate.

Better  machine learning (ML)  algorithms, more access to data, cheaper hardware, and the availability of 5G have contributed to the increasing application of AI in the healthcare industry, accelerating the pace of change. AI and ML technologies can sift through enormous volumes of health data—from health records and clinical studies to genetic information—and analyze it much faster than humans.

Healthcare organizations are using AI to improve the efficiency of all kinds of processes, from back-office tasks to patient care. The following are some examples of how AI might be used to benefit staff and patients:

• Administrative workflow:  Healthcare workers spend a lot of time doing paperwork and other administrative tasks. AI and automation can help perform many of those mundane tasks, freeing up employee time for other activities and giving them more face-to-face time with patients. For example, generative AI can help clinicians with note-taking and content summarization that can help keep medical records as thoroughly as possible. AI might also help with accurate coding and sharing of information between departments and billing.
• Virtual nursing assistants:  One study found that  64% of patients  are comfortable with the use of AI for around-the-clock access to answers that support nurses provide. AI virtual nurse assistants—which are AI-powered chatbots, apps, or other interfaces—can be used to help answer questions about medications, forward reports to doctors or surgeons and help patients schedule a visit with a physician. These sorts of routine tasks can help take work off the hands of clinical staff, who can then spend more time directly on patient care, where human judgment and interaction matter most.
• Dosage error reduction:  AI can be used to help identify errors in how a patient self-administers medication. One example comes from a study in  Nature Medicine , which found that up to 70% of patients don’t take insulin as prescribed. An AI-powered tool that sits in the patient’s background (much like a wifi router) might be used to flag errors in how the patient administers an insulin pen or inhaler.
• Less invasive surgeries:  AI-enabled robots might be used to work around sensitive organs and tissues to help reduce blood loss, infection risk and post-surgery pain.
• Fraud prevention:  Fraud in the healthcare industry is enormous, at $380 billion/year, and raises the cost of consumers’ medical premiums and out-of-pocket expenses. Implementing AI can help recognize unusual or suspicious patterns in insurance claims, such as billing for costly services or procedures that are not performed, unbundling (which is billing for the individual steps of a procedure as though they were separate procedures), and performing unnecessary tests to take advantage of insurance payments.

A recent study found that  83% of patients  report poor communication as the worst part of their experience, demonstrating a strong need for clearer communication between patients and providers. AI technologies like  natural language processing  (NLP), predictive analytics, and  speech recognition  might help healthcare providers have more effective communication with patients. AI might, for instance, deliver more specific information about a patient’s treatment options, allowing the healthcare provider to have more meaningful conversations with the patient for shared decision-making.

According to  Harvard’s School of Public Health , although it’s early days for this use, using AI to make diagnoses may reduce treatment costs by up to 50% and improve health outcomes by 40%.

One use case example is out of the  University of Hawaii , where a research team found that deploying  deep learning  AI technology can improve breast cancer risk prediction. More research is needed, but the lead researcher pointed out that an AI algorithm can be trained on a much larger set of images than a radiologist—as many as a million or more radiology images. Also, that algorithm can be replicated at no cost except for hardware.

An  MIT group  developed an ML algorithm to determine when a human expert is needed. In some instances, such as identifying cardiomegaly in chest X-rays, they found that a hybrid human-AI model produced the best results.

Another  published study  found that AI recognized skin cancer better than experienced doctors.  US, German and French researchers used deep learning on more than 100,000 images to identify skin cancer. Comparing the results of AI to those of 58 international dermatologists, they found AI did better.

As health and fitness monitors become more popular and more people use apps that track and analyze details about their health. They can share these real-time data sets with their doctors to monitor health issues and provide alerts in case of problems.

AI solutions—such as big data applications, machine learning algorithms and deep learning algorithms—might also be used to help humans analyze large data sets to help clinical and other decision-making. AI might also be used to help detect and track infectious diseases, such as COVID-19, tuberculosis, and malaria.

One benefit the use of AI brings to health systems is making gathering and sharing information easier. AI can help providers keep track of patient data more efficiently.

One example is diabetes. According to the  Centers for Disease Control and Prevention , 10% of the US population has diabetes. Patients can now use wearable and other monitoring devices that provide feedback about their glucose levels to themselves and their medical team. AI can help providers gather that information, store, and analyze it, and provide data-driven insights from vast numbers of people. Using this information can help healthcare professionals determine how to better treat and manage diseases.

Organizations are also starting to use AI to help improve drug safety. The company SELTA SQUARE, for example, is  innovating the pharmacovigilance (PV) process , a legally mandated discipline for detecting and reporting adverse effects from drugs, then assessing, understanding, and preventing those effects. PV demands significant effort and diligence from pharma producers because it’s performed from the clinical trials phase all the way through the drug’s lifetime availability. Selta Square uses a combination of AI and automation to make the PV process faster and more accurate, which helps make medicines safer for people worldwide.

Sometimes, AI might reduce the need to test potential drug compounds physically, which is an enormous cost-savings.  High-fidelity molecular simulations  can run on computers without incurring the high costs of traditional discovery methods.

AI also has the potential to help humans predict toxicity, bioactivity, and other characteristics of molecules or create previously unknown drug molecules from scratch.

As AI becomes more important in healthcare delivery and more AI medical applications are developed, ethical, and regulatory governance must be established. Issues that raise concern include the possibility of bias, lack of transparency, privacy concerns regarding data used for training AI models, and safety and liability issues.

“AI governance is necessary, especially for clinical applications of the technology,” said Laura Craft, VP Analyst at  Gartner . “However, because new AI techniques are largely new territory for most [health delivery organizations], there is a lack of common rules, processes, and guidelines for eager entrepreneurs to follow as they design their pilots.”

The World Health Organization (WHO) spent 18 months deliberating with leading experts in ethics, digital technology, law, and human rights and various Ministries of Health members to produce a report that is called  Ethics & Governance of Artificial Intelligence for Health . This report identifies ethical challenges to using AI in healthcare, identifies risks, and outlines six  consensus principles  to ensure AI works for the public’s benefit:

• Protecting autonomy
• Promoting human safety and well-being
• Ensuring transparency
• Fostering accountability
• Ensuring equity
• Promoting tools that are responsive and sustainable

The WHO report also provides recommendations that ensure governing AI for healthcare both maximizes the technology’s promise and holds healthcare workers accountable and responsive to the communities and people they work with.

AI provides opportunities to help reduce human error, assist medical professionals and staff, and provide patient services 24/7. As AI tools continue to develop, there is potential to use AI even more in reading medical images, X-rays and scans, diagnosing medical problems and creating treatment plans.

AI applications continue to help streamline various tasks, from answering phones to analyzing population health trends (and likely, applications yet to be considered). For instance, future AI tools may automate or augment more of the work of clinicians and staff members. That will free up humans to spend more time on more effective and compassionate face-to-face professional care.

When patients need help, they don’t want to (or can’t) wait on hold. Healthcare facilities’ resources are finite, so help isn’t always available instantaneously or 24/7—and even slight delays can create frustration and feelings of isolation or cause certain conditions to worsen.

IBM® watsonx Assistant™ AI healthcare chatbots  can help providers do two things: keep their time focused where it needs to be and empower patients who call in to get quick answers to simple questions.

IBM watsonx Assistant  is built on deep learning, machine learning and natural language processing (NLP) models to understand questions, search for the best answers and complete transactions by using conversational AI.

Get email updates about AI advancements, strategies, how-tos, expert perspective and more.

See IBM watsonx Assistant in action and request a demo

Get our newsletters and topic updates that deliver the latest thought leadership and insights on emerging trends.

Your company may already be a member. View our member list to find out, or create a new account .

Forgot Password?

Content Library

You can search our content library for case studies, research, industry insights, and more.

You can search our website for events, press releases, blog posts, and more.

Why Claritin Is Planting Trees to Help Your Allergies

May 16, 2024    

Claritin is on a mission to plant female trees to counteract the pollen-producing effects of male trees. Bayer crafted a story across multiple video ad formats to showcase Claritin's commitment to sustainability, while following the principles of Ad Net Zero to curb emissions from creative production.

This Is an ANA Member Exclusive

Access to this item is reserved for ANA members only.

Already have an account? The industry's best insights and resources await:

No Account?

Use your business email address to create your free account ; if you're a member through your company, we'll know.

Members can access their benefits as soon as they sign up and log in.

Not a Member?

You can still create a free account to access the latest from our online publication, ANA Magazine , receive content and special event offers through our newsletters, get breaking industry updates, and so much more.

The content you're trying to see is available to:

• ANA Client-Side Marketer Tier Members
• Platinum Tier Members
• Gold Tier Members
• Silver Tier Members
• Individual Members

Discover everything the ANA can do to help drive growth for your organization. Connect with our membership team.

• Open access
• Published: 13 May 2024

Lipidomic studies revealing serological markers associated with the occurrence of retinopathy in type 2 diabetes

• Mingqian He 1   na1 ,
• Guixue Hou 2   na1 ,
• Mengmeng Liu 1   na1 ,
• Zhaoyi Peng 1 ,
• Hui Guo 1 ,
• Yue Wang 1 ,
• Jing Sui 3 ,
• Hui Liu 4 ,
• Xiaoming Yin 5 ,
• Meng Zhang 1 ,
• Ziyi Chen 1 ,
• Patrick C.N. Rensen 1 , 6 ,
• Liang Lin 2 , 8 ,
• Yanan Wang 1 , 7 &
• Bingyin Shi 1  

Journal of Translational Medicine volume  22 , Article number:  448 ( 2024 ) Cite this article

268 Accesses

1 Altmetric

Metrics details

The duration of type 2 diabetes mellitus (T2DM) and blood glucose levels have a significant impact on the development of T2DM complications. However, currently known risk factors are not good predictors of the onset or progression of diabetic retinopathy (DR). Therefore, we aimed to investigate the differences in the serum lipid composition in patients with T2DM, without and with DR, and search for potential serological indicators associated with the development of DR.

A total of 622 patients with T2DM hospitalized in the Department of Endocrinology of the First Affiliated Hospital of Xi’an JiaoTong University were selected as the discovery set. One-to-one case–control matching was performed according to the traditional risk factors for DR (i.e., age, duration of diabetes, HbA1c level, and hypertension). All cases with comorbid chronic kidney disease were excluded to eliminate confounding factors. A total of 42 pairs were successfully matched. T2DM patients with DR (DR group) were the case group, and T2DM patients without DR (NDR group) served as control subjects. Ultra-performance liquid chromatography–mass spectrometry (LC–MS/MS) was used for untargeted lipidomics analysis on serum, and a partial least squares discriminant analysis (PLS-DA) model was established to screen differential lipid molecules based on variable importance in the projection (VIP) > 1. An additional 531 T2DM patients were selected as the validation set. Next, 1:1 propensity score matching (PSM) was performed for the traditional risk factors for DR, and a combined 95 pairings in the NDR and DR groups were successfully matched. The screened differential lipid molecules were validated by multiple reaction monitoring (MRM) quantification based on mass spectrometry.

The discovery set showed no differences in traditional risk factors associated with the development of DR (i.e., age, disease duration, HbA1c, blood pressure, and glomerular filtration rate). In the DR group compared with the NDR group, the levels of three ceramides (Cer) and seven sphingomyelins (SM) were significantly lower, and one phosphatidylcholine (PC), two lysophosphatidylcholines (LPC), and two SMs were significantly higher. Furthermore, evaluation of these 15 differential lipid molecules in the validation sample set showed that three Cer and SM(d18:1/24:1) molecules were substantially lower in the DR group. After excluding other confounding factors (e.g., sex, BMI, lipid-lowering drug therapy, and lipid levels), multifactorial logistic regression analysis revealed that a lower abundance of two ceramides, i.e., Cer(d18:0/22:0) and Cer(d18:0/24:0), was an independent risk factor for the occurrence of DR in T2DM patients.

Disturbances in lipid metabolism are closely associated with the occurrence of DR in patients with T2DM, especially in ceramides. Our study revealed for the first time that Cer(d18:0/22:0) and Cer(d18:0/24:0) might be potential serological markers for the diagnosis of DR occurrence in T2DM patients, providing new ideas for the early diagnosis of DR.

Introduction

Type 2 diabetes mellitus (T2DM) is a common chronic disease in many countries, and its prevalence is growing as people’s lifestyles are changing [ 1 ]. Diabetes causes various complications, classified as either macrovascular complications (such as cardiovascular disease and stroke) or microvascular complications (such as kidney disease) [ 2 ]. Diabetic retinopathy (DR), a specific microvascular complication of diabetes, is the most common cause of vision loss in people of working age [ 3 , 4 ]. Poor glycemic control, hypertension, and diabetes duration are major risk factors for DR [ 5 ]. Although intensive risk factor control reduces the risk of DR progression and vision loss, many diabetic patients continue to develop DR with strict glycemic and blood pressure control [ 6 ]. Despite increasing research supporting the efficacy of routine DR screening to prevent DR and early treatment to reduce the risk of vision loss, there are no specific biomarkers for diagnosing the onset and early progression of DR. Additionally, new and more effective strategies are awaited to prevent and treat the progression of DR.

Accumulating evidence suggests that disruption in lipid metabolism is an early event in the pathogenesis of diabetes complications. Previous studies found that levels of multiple lipid species, including glycerophospholipids, sphingolipids and glycerolipids, are critical risk factors for T2DM and its complications [ 7 , 8 ]. Lysophosphatidylcholine (LPC) is a main glycerophospholipid known for its essential role in lipid and glucose metabolism, and LPC has been intensively studied in the development of metabolic diseases including T2DM [ 9 ]. Sphingolipids, including ceramides (Cer), sphingomyelins (SM) and gangliosides, have a variety of intra- and extracellular effects on glucose homeostasis and metabolic disease [ 10 ] Numerous studies suggest Cer, a crucial lipid intermediate in sphingolipid metabolism, is a major contributing factor for insulin resistance, and inhibition or depletion of enzymes driving de novo ceramide synthesis can prevent the development of diabetes in mice [ 7 , 11 , 12 ]. In contrast, a decrease in very long chain Cer is correlated with the development of macroalbuminuria in diabetes [ 13 ]. Accelerated sphingolipid catabolism’ leading to an increase in glucosylceramide or glycosphingolipids might contribute to the neuronal pathologies of DR [ 14 ]. In addition, SM produced by the transfer of a phosphocholine moiety from phosphatidylcholine to the ceramide backbone has been linked to insulin resistance [ 15 , 16 ] and is also an independent marker of cardiovascular disease [ 17 ]. Thus, dysregulated lipid metabolism is a major contributor to the pathogenesis of T2DM and its complications, and specific lipid species that are responsible for the occurrence of DR are rather obscure.

Lipidomics offers solid platforms for identifying novel lipid mediates in biochemical processes of lipid metabolism, thus providing new opportunities for disease prediction and detection [ 18 , 19 ]. Lipidome analysis is performed by liquid chromatography and electrospray ionization-tandem mass spectrometry (LC–MS/MS) for molecular lipid identification and quantification and multiple reaction monitoring (MRM) for targeted quantification of those lipid species. Lipid-based biomarkers offer unique options for precision medicine by providing sensitive diagnostic tools for disease prediction and monitoring [ 20 ]. Using a quantitative metabolomics approach, Emil et al. compared the aqueous humor and serum concentrations of metabolites in senior adults with an without diabetes who underwent cataract surgery [ 21 ]. However, the field of lipidomics studies of DR is still in its early stages, with few studies published and little replication of results [ 22 ].

In this study, we aimed to find reliable serum lipid-based biomarkers for the presence of DR in patients with T2DM by using two cohorts. To this end, serum samples of the discovery cohort was subjected to untargeted lipidomics analysis to search for differentially abundant lipids between individuals without and with DR. In the validation cohort, the observed differential lipid molecules were validated using mass spectrometry MRM targeting techniques. We hypothesized that DR has a distinctive serum lipid signature and that particular lipid species can act as biomarkers for T2DM patients with DR.

Research design and methods

Participants.

A total of 622 participants with T2DM hospitalized in the Endocrinology Department of the First Affiliated Hospital of Xi’an JiaoTong University were screened as the discovery set. Participants with chronic kidney disease [estimated glomerular filtration rate (eGFR) < 90 (mL/min/1.73 m 2 )] were excluded from the selection. We conducted pair matching according to the traditional risk factors for DR (including age, duration of diabetes, HbA1c level, and hypertension). For the discovery cohort, we selected 42 T2DM patients with DR (DR group). The control participants were 42 T2DM patients without DR (NDR group), and they were matched to patients in the DR group by age (in 5-year bands), diabetes duration (in 5-year bands), HbA1c levels (in 0.5% bands), and hypertension status.

Lipid markers of DR identified from the discovery cohort were quantified in a separate sample cohort (validation cohort). We first screened 531 T2DM patients. Individuals with chronic kidney disease [eGFR < 90 (mL/min/1.73 m 2 )] were excluded from the selection. Then, we conducted 1:1 propensity score matching (PSM) (matching tolerance = 0.02) by age, diabetes duration, HbA1c level, hypertension status, sex, BMI, systolic blood pressure (SBP), diastolic blood pressure (DBP), and eGFR. For the validation cohort, 95 T2DM patients with DR (DR group) and 95 T2DM patients without DR (NDR group) were included.

Sample collection

Fasting blood samples and clinical data were collected from the individuals. All blood samples were collected at the First Affiliated Hospital of Xi’an JiaoTong University physical examination center. Blood samples were centrifuged for 20 min at 1500 rpm and 4 °C. Then, serum was collected and stored at -80 °C until analysis. HbA1c was measured using an automatic HbA1c analyzer (TOSOH BIOSCIENCE, INC.; HLC-723G8). Total cholesterol (CHOL), triglyceride (TG), high density lipoprotein-cholesterol (HDL-c), low density lipoprotein-cholesterol (LDL-c), uric acid (UA), aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), gamma-glutamyl transpeptidase (GGT), total bilirubin (TBIL), direct bilirubin (DBIL), total protein (TP), albumin (ALB), glucose (GLU), blood urea nitrogen (BUN), creatinine (CRE) were measured using standard reagents on an automatic biochemistry analyzer (HITACHI, Inc.; LAbOSPECT, 008AS). Blood pressure was measured in triplicate using an Omron HBP-9020 digital automatic blood pressure machine (Kyoto, Japan).

Lipid extraction

The serum samples were thawed slowly at 4 °C, 100 µL of the sample was placed in a 96-well plate, 300 µL of isopropanol (prechilled at -20 °C) spiked with internal standards (SPLASH® LIPIDOMIX® Mass Spec Standard, Avanti, USA) was added, and the samples were vortexed and mixed for 1 min and then centrifuged at 4 °C for 20 min at 4000 rcf after resting overnight at -20 °C as previously reported [ 23 ]. The supernatant was injected for LC–MS/MS analysis, and 10 µL of each supernatant was mixed into quality control (QC) samples to assess the reproducibility and stability of the LC–MS analysis process.

LC–MS/MS analysis

Lipids were separated and detected by an UPLC (CSH C18 column, 1.7 μm 2.1*100 mm, Waters, USA) equipped with a Q Exactive Plus high-resolution mass spectrometer (Thermo Fisher Scientific, USA) as previously reported [ 24 ]. The following gradient was used for elution: 0–2 min, 40-43% mobile phase B (10 mM ammonia formate, 0.1% formic acid, 90% isopropyl alcohol, and 10% acetonitrile); 2–2.1 min, 43-50% liquid B; 2.1–7 min, 50-54% solution B; 7–7.1 min, 54-70% liquid B; 7.1–13 min, 70-99% liquid B with a flow rate of 0.35 mL/min. Mobile phase A was an aqueous solution containing 10 mM ammonia formate, 0.1% formic acid and 60% acetonitrile in water.

All samples were analyzed in data-dependent acquisition (DDA) mode with the following positive/negative ionization settings: spray voltage, 3.8/–3.2 kV; aux gas heater temperature, 350 °C; and capillary temperature, 320 °C. The full scan mass range was 200–2000 m/z with 70,000 mass resolution at m/z 200 and AGC set to 3e6 with a maximum ion injection time of 100 ms. The top three precursors were selected for subsequent MS fragmentation with a maximum ion injection time of 50 ms and resolution of 17,500 at m/z 200, and the AGC was 1e5. The stepped normalized collision energy was set to 15, 30, and 45 eV.

Data preprocessing and quality control

The raw data obtained from the LC–MS/MS detection were imported into LipidSearch v.4.1 (Thermo Fisher Scientific, USA) for lipid identification and quantification. The following parameters were used for lipid identification and peak extraction: the type of identification was Product, the mass deviation of the parent and daughter ions was 5 ppm, and the response threshold was set to 5.0% of the relative response deviation of the daughter ions; the quantitative parameters were set to calculate the peak areas of all identified lipids, and the peak extraction mass deviation was set to 5 ppm. For ESI + data, [M + H]+, [M + NH4]+, and [M + Na] + were selected as adducts, while for ESI- data, [M-H]-, [M-2 H]-, and [M-HCOO]- were selected as adducts. The peak alignment was performed for all identified lipids, and those not marked as “rejected” were considered for inclusion in the subsequent analysis.

For data preprocessing, raw data exported from LipidSearch were further analyzed by meta X [ 25 ]. The data preprocessing included (1) Removing lipid molecules with more than 50% missing information in QC samples and more than 80% missing information in experimental samples (i.e., LipidIon in the table); (2) Filling the missing values using the k-nearest neighbor (KNN) algorithm; (3) Correcting the batch effect using quality control-based robust LOESS signal correction (QC-RLSC); (4) Using probabilistic quotient normalization (PQN) to normalize the data to obtain the relative peak areas; and (5) Removing the lipid molecules with a coefficient of variation (CV) greater than 30% of the relative peak areas from all QC samples.

Data quality was assessed by the reproducibility of QC sample assays. The assessment included chromatogram overlap of QC samples, principal component analysis (PCA), number of extracted peaks, and differences in peak response intensity.

Data processing

A combination of multivariate statistical analysis and univariate analysis was used to screen for lipids of which the abundance differed between groups. The multivariate statistical analysis methods used were principal component analysis (PCA) and partial least squares method-discriminant analysis (PLS-DA). PCA is an unsupervised pattern recognition method, and PLS-DA is a supervised pattern recognition method. The univariate analyses were fold change (FC) and Student’s t test. The FC was obtained by fold change analysis, and the p  value pairs of the t test were corrected for the false discovery rate (FDR) to obtain a q-value. The differential lipid molecule screening conditions were as follows: (1) variable importance in the projection (VIP) ≥ 1 for the first two principal components of the PLS-DA model; (2) fold change ≥ 1.2 or ≤ 0.83; and (3) p  value < 0.05.

Targeted lipid quantification by MRM in validation samples

The identified differential lipids were further quantified by multiple reaction monitoring (MRM). For lipid extraction, the procedure was consistent with the untargeted experiment as described. The MRM transition list is shown in Table S1 . For MRM quantification, all validation samples were analyzed on a QTRAP 5500 mass spectrometer with a CSH C18 column (1.7 μm 2.1*100 mm, Waters, USA) for separation. All lipids were subjected to targeted quantification in ESI + mode with a specific transition setting.

Statistical analysis

The clinical data of samples are presented as the mean ± standard deviation (SD) for normally distributed variables or the median (interquartile range) for abnormal distribution. Comparisons between the case group and the control group were made using a two-tailed t test or Mann-Whitney U test for continuous data and the X 2 test for categorical data. The calculation of the area under the curve (AUC) in receiver operating characteristic (ROC) curve analysis was used to evaluate the discriminatory ability of the markers. Logistic regression models were applied to assess the relationship between lipid molecules and the presence of DR. The odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for the molecules with 1-SD changes. The known risk factors for DR, such as CHOL, TGs, LDL-c, and HDL-c, were added to multivariate logistic regression to calculate the adjusted odds ratios. Ordinal logistic regression models were used to assess the relationships between lipid molecules and DR stages [NDR, nonproliferative DR (NPDR) and proliferative DR (PDR)].

Characteristics of the discovery cohort

Table  1 shows the clinical characteristics of individuals selected for the discovery cohort. There were no significant differences in age and sex between the DR and NDR groups. In fact, these groups were comparable for most metabolic characteristics, such as BMI, diabetes duration, and HbA1c, and there were no significant between-group differences for hypertension status, antihypertensive agent use, hypoglycemic therapy status or NSAID use. The blood pressure and glucose of the participants were treated and controlled. Compared with control subjects with T2DM, T2DM patients with DR had higher levels of LDL-c levels, AST, TBIL, and BUN (Table  1 and Table S2 ).

Untargeted lipidome-derived biomarkers for diabetic retinopathy: results from the discovery cohort

A total of 1721 lipids were detected. The number of lipids with an RSD (CV) less than or equal to 30% in the QC samples was 1421. The ratio of the number of lipids with CV less than or equal to 30% to the number of all detected lipids in QC samples was 81%.

Fifteen candidate lipids were identified from the discovery cohort. Compared with those of the NDR group, the levels of three Cer and seven SM were significantly lower in the DR group. In contrast, two SM, two LPC and one PC were significantly higher in the DR group (Fig.  1 A and B). More specifically, compared with T2DM patients without DR, T2DM patients with DR showed lower levels of Cer(d18:0/24:0), Cer(d18:0/22:0), Cer(d42:3), SM(d22:0/16:0), SM(d18:1/24:1), SM(d42:0), SM(d40:0), SM(d39:0), SM(d38:0), and SM(d36:0), and higher levels of SM(d20:1/16:1), SM(d34:1), LPC(18:2), LPC(16:0) and PC(34:2). The heat map shows the distribution of these lipids between individuals of the NDR and DR groups (Fig.  1 C). The results of ROC analysis and the odds ratios of the lipid markers in the basic logistic regression models are shown in Table  2 . The AUC values for the 15 lipids ranged from 0.72 to 0.94. All lipids retained significant ORs after adjusted for CHOL, TG, LDL-c, and HDL-c (adjusted ORs are shown in Table  2 ). Furthermore, we used ordinal logistic regression, which estimated the odds of being in one higher category of the DR stage (from NDR to PDR) for lipid species, to test the associations between lipid species and DR stage (Table S3 ; n  = 42 in the NDR group, n  = 37 in the NPDR group, n  = 5 in the PDR group), and we analyzed the data while excluding participants with diabetic macular edema (DME) ( n  = 4 in the DR group), as before, all lipids retained significant ORs (Table S4 ).

Lipidome-derived markers identified from the discovery cohort. Lipidomic analysis identified fifteen candidate lipids of which serum levels were different between 42 T2DM patients with DR (DR group) and 42 T2DM patients without DR (NDR group) from the discovery cohort. ( A ) Mean peak intensity of lipids was analyzed after Log2 transformation of the data. ( B ) Fold change in DR/NDR was analyzed after Log2 transformation of the data. ( C ) Heatmap showing the distribution of lipid markers. Each row in the figure represents a different lipid, and each column represents a sample. Different colors indicate different intensities, and Log2 conversion was used for the data

Characteristics of the validation cohort and targeted lipidomics analysis

The 15 differential lipids found from the discovery cohort were validated in another set of samples. The clinical characteristics of individuals selected for the validation cohort are shown in Table  3 . Most metabolic and clinical features were comparable (Table S5 ), and there was no significant difference in LDL-c between the DR and NDR groups.

In the validation cohort, when compared with subjects in the NDR group, T2DM patients with DR showed lower levels of Cer(d18:0/24:0), Cer(d18:0/22:0), Cer(d42:3) and SM(d18:1/24:1) by univariate logistic regression, which was consistent with the results of the discovery cohort. However, the levels of SM(d20:1/16:1), LPC(18:2) and LPC(16:0) were lower in T2DM patients with DR from the validation cohort, opposite to the result obtained in the discovery cohort (Fig.  2 A and B). The AUC values for these lipids were higher than 0.61. The other 8 lipids did not significantly differ between the DR and NDR groups in the validation cohort (Table  4 ). Of note, compared with those in T2DM patients, the peak area (after Log2 transformation) of Cer(d18:0/24:0) (20.48 ± 0.82 vs. 20.12 ± 0.99, p  = 0.006, Fig.  2 C) and Cer(d18:0/22:0) (19.91 ± 0.75 vs. 19.64 ± 0.92, p  = 0.028, Fig.  2 C) remained significantly lower in T2DM patients with DR, and the levels of these two lipids retained significant ORs when adjusted for known risk factors (i.e., CHOL, TG, LDL-c and HDL-c). In the ordinal regression, these two lipids maintained significant ORs (Table S7 , n  = 95 in the NDR group, n  = 87 in the NPDR group; n  = 8 in the PDR group), and were also significant while excluding patients with DME (Table S6 , n  = 2 in the DR group). These findings imply that levels of Cer(d18:0/24:0) and Cer(d18:0/22:0) were independent markers for T2DM patients with DR in both the discovery cohort (Table  2 ) and validation cohort (Table  4 ).

The results of targeted lipidomics analysis in the validation cohort. For the validation cohort, the cases were 95 T2DM patients with DR (DR group), and the control subjects were 95 T2DM patients who had no DR (NDR group). ( A ) Peak area of lipids was analyzed after Log2 transformation of the data. ( B ) Fold change in DR/NDR was analyzed after Log2 transformation of the data. ( C ) The log2 conversion was used for the intensities of Cer(d18:0/24:0) and Cer(d18:0/22:0). All data are presented as the mean ± standard deviation (SD). Each symbol represents an individual participant. * p  < 0.05, ** p  < 0.01, pairwise comparisons of change scores between the groups were evaluated by t test

DR is the most common microvascular complication of diabetes and the main factor contributing to visual impairment in working-age individuals [ 3 ]. T2DM patients often develop DR despite of proper control of systemic risk factors, indicating the involvement of other pathogenic factors for DR development. To find new and more effective strategies for preventing and treating DR, it is necessary for us to identify novel biomarkers for DR screening or detection. Lipidomics will aid in understanding the mechanism of DR at various stages of the disease, early diagnosis, and the identification of new therapeutic targets. In this study, by using two clinical cohorts, we found that the serum lipidomic profiles in T2DM patients with DR showed significant differences from those in T2DM patients without DR. The differential lipid species in the DR group were linked to disturbances in sphingolipid metabolism. Compared with those in the NDR group, the levels of Cer(d18:0/24:0) and Cer(d18:0/22:0) were significantly lower in the DR group after adjusting for covariates, i.e. known risk factors in both the discovery and validation cohorts. These findings suggest that these two lipid species may be potential serological markers for the diagnosis of DR in patients with T2DM.

In this study, we found two ceramide molecules that were significantly lower in T2DM patients with DR, indicating that they may have disturbed ceramide metabolism compared to T2DM patients without DR. Ceramide is sphingolipid [ 11 ] and can be found in VLDL, LDL, and HDL. Consistent with our findings, Fort et al. found a significantly lower abundance of Cer in central retinal tissue obtained postmortem from T2DM patients with DR compared to those without DR [ 26 ]. Similarly, ceramide levels were shown to be lower and glucosylceramide levels higher in the retinas of diabetic rodents [ 27 ]. This indicates that diabetes reduces the retinal ceramide content and may suggest that dysregulated sphingolipid metabolism may cause retinal resistance to insulin action [ 27 ]. These findings imply that ceramide is diverted from the overall pools of retinal sphingolipids toward the glycosylated forms due to hyperglycemia. In contrast, Levitsky et al. found that diabetes-induced increases in mitochondrial ceramide led to impaired mitochondrial function in the retinal pigment epithelial (RPE) cells of the retina [ 28 ], and disruption of the blood-retinal barrier might be caused by diabetes-induced overexpression of acid sphingomyelinase. Additionally, inflammation is a common underlying factor in DR, and inflammation generates Cer from SM in the serum membrane. This induces death receptor ligand formation and leads to apoptosis of RPE and photoreceptor cells [ 29 ]. In addition to diabetes, circulating Cer was shown to strongly correlate with future adverse cardiovascular events. It has recently been discovered that in individuals with atherosclerotic CVD, serum levels of specific Cer species can predict the future risk of cardiovascular death. In the Corogene study, higher concentrations of Cer(d18:1/16:0), Cer(d18:1/18:0), and Cer(d18:1/24:1) and lower concentrations of Cer(d18:1/24:0) were associated with a higher risk of fatal myocardial infarction [ 30 ]. Our study found that Cer(d18:0/24:0) and Cer(d18:0/22:0) were significantly lower in T2DM patients with DR compared to those without DR, which suggests that different numbers of carbons and double bonds in ceramides might play differential roles in DR and CVD. The distinct ceramides and ceramide metabolites involved in metabolic regulation play unanticipated roles [ 31 ]. Watt et al. discovered that circulating ceramides present in LDL particles were sufficient to induce insulin resistance in vitro and in vivo [ 32 ]. However, how these two identified ceramides influence lipid metabolism in T2DM remains unclear and needs further exploration. Thus, disturbed Cer metabolism may contribute to dysfunction in DR, and therapeutic strategies to restore normal Cer metabolism might be an effective approach for treatment of DR.

In the discovery cohort, LPC(18:2) and LPC(16:0) were significantly higher in T2DM patients with DR. However, these two lipids were significantly lower in DR in the validation cohort. The previous findings point to a change in sphingolipid composition between control and T2DM [ 33 ]. LPC is an inflammatory phospholipid and an important atherogenic substance in LDL that contributes to diabetic complications [ 34 ]. Lipoprotein-associated phospholipase A2 (Lp-PLA2) plays a crucial role in diabetes-related retinal vasopermeability, a response mediated by LPC, and inhibiting Lp-PLA2 reduces diabetes-induced retinal vasopermeability [ 35 ]. LPC O-16:0, LPC P-16:0, LPC O-18:0, and LPC 18:1 were all found to be inversely related to incident T2DM [ 36 ]. The differences between the discovery and validation cohorts may be related to the populations studied, medications used, and stages of diabetic retinopathy [ 37 ].

There are some limitations of this study. First, only a Chinese ethnic group was selected, and future validation of our findings in other races or ethnic groups is warranted. Second, instead of chronic risk factors associated with the development of DR, some of the identified lipid markers might only represent temporary metabolic perturbations in this cross-sectional study. Third, the exact mechanism of DR development in patients with T2DM through which ceramide functions has not been explained. Therefore, more extensive preclinical and clinical studies are needed to clarify the mechanisms behind the potential effects of specific lipids.

Overall, the deregulation of sphingolipid metabolism in the diabetic retina appears to be a significant and seldom-studied element of DR pathophysiology. The precise mechanism underlying this disease is still unknown and requires further investigation. We showed the potential value of lipidomics research in understanding the pathophysiology of DR, and the results suggest that lipidomics profiling may be capable of identifying early-stage DR diagnostic indicators in high-risk Chinese populations. In addition, the findings from this study may help in the elucidation of new therapeutic targets for DR prevention and treatment.

Data availability

All relevant data and materials have been included in the article and its supplementary data files. Further inquiries can be directed to the corresponding authors.

Abbreviations

• Type 2 diabetes mellitus

Diabetic retinopathy

Liquid chromatography–mass spectrometry

Partial least squares discriminant analysis

Variable importance in the projection

Propensity score matching

Sphingomyelins

Phosphatidylcholine

Lysophosphatidylcholines

Multiple reaction monitoring

Systolic blood pressure

Diastolic blood pressure

Total cholesterol

Triglyceride

High density lipoprotein-cholesterol

Low density lipoprotein-cholesterol

Aspartate aminotransferase

Alanine aminotransferase

Alkaline phosphatase

Gamma-glutamyl transpeptidase

Total bilirubin

Direct bilirubin

Total protein

Blood urea nitrogen

Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87:4–14.

Article   CAS   PubMed   Google Scholar  

Wong TY, Cheung CM, Larsen M, Sharma S, Simo R. Diabetic retinopathy. Nat Rev Dis Primers. 2016;2:16012.

Article   PubMed   Google Scholar  

American Diabetes A. 10. Microvascular Complications and Foot Care: Standards of Medical Care in Diabetes-2018. Diabetes Care. 2018;41:S105-S18.

Jampol LM, Glassman AR, Sun J. Evaluation and care of patients with Diabetic Retinopathy. N Engl J Med. 2020;382:1629–37.

Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556–64.

Article   PubMed   PubMed Central   Google Scholar  

Beulens JW, Patel A, Vingerling JR, Cruickshank JK, Hughes AD, Stanton A, et al. Effects of blood pressure lowering and intensive glucose control on the incidence and progression of retinopathy in patients with type 2 diabetes mellitus: a randomised controlled trial. Diabetologia. 2009;52:2027–36.

Meikle PJ, Summers SA. Sphingolipids and phospholipids in insulin resistance and related metabolic disorders. Nat Reviews Endocrinol. 2017;13:79–91.

Article   CAS   Google Scholar  

Zeng W, Beyene HB, Kuokkanen M, Miao G, Magliano DJ, Umans JG, et al. Lipidomic profiling in the strong heart study identified American indians at risk of chronic kidney disease. Kidney Int. 2022;102:1154–66.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Knuplez E, Marsche G. An updated review of Pro- and Anti-inflammatory properties of plasma lysophosphatidylcholines in the Vascular System. Int J Mol Sci. 2020;21.

Tonks KT, Coster ACF, Christopher MJ, Chaudhuri R, Xu AM, Gagnon-Bartsch J, et al. Skeletal muscle and plasma lipidomic signatures of insulin resistance and overweight/obesity in humans. Obesity. 2016;24:908–16.

Hannun YA, Obeid LM. Sphingolipids and their metabolism in physiology and disease. Nat Rev Mol Cell Biol. 2018;19:175–91.

Chaurasia B, Tippetts TS, Mayoral Monibas R, Liu J, Li Y, Wang L, et al. Targeting a ceramide double bond improves insulin resistance and hepatic steatosis. Volume 365. New York, NY: Science; 2019. pp. 386–92.

Google Scholar  

Klein RL, Hammad SM, Baker NL, Hunt KJ, Al Gadban MM, Cleary PA, et al. Decreased plasma levels of select very long chain ceramide species are associated with the development of nephropathy in type 1 diabetes. Metab Clin Exp. 2014;63:1287–95.

Fox TE, Han X, Kelly S, Merrill AH 2nd, Martin RE, Anderson RE, et al. Diabetes alters sphingolipid metabolism in the retina: a potential mechanism of cell death in diabetic retinopathy. Diabetes. 2006;55:3573–80.

Coen PM, Goodpaster BH. Role of intramyocelluar lipids in human health. Trends Endocrin Met. 2012;23:391–8.

Hannun YA, Obeid LM. Many ceramides. J Biol Chem. 2011;286:27855–62.

Jiang XC, Paultre F, Pearson TA, Reed RG, Francis CK, Lin M et al. Plasma sphingomyelin level as a risk factor for coronary artery disease. Arteriosclerosis, thrombosis, and vascular biology. 2000;20:2614–8.

Hyotylainen T, Oresic M. Analytical Lipidomics in metabolic and clinical research. Trends Endocrin Met. 2015;26:671–3.

Xuan Q, Ouyang Y, Wang Y, Wu L, Li H, Luo Y, et al. Multiplatform Metabolomics reveals novel serum metabolite biomarkers in Diabetic retinopathy subjects. Adv Sci (Weinh). 2020;7:2001714.

Hyotylainen T, Oresic M. Optimizing the lipidomics workflow for clinical studies-practical considerations. Anal Bioanal Chem. 2015;407:4973–93.

Grochowski ET, Pietrowska K, Godlewski A, Gosk W, Buczynska A, Wojnar M et al. Simultaneous comparison of aqueous humor and serum metabolic profiles of Diabetic and nondiabetic patients undergoing cataract Surgery—A targeted and quantitative Metabolomics Study. Int J Mol Sci. 2023;24.

Liew G, Lei Z, Tan G, Joachim N, Ho IV, Wong TY, et al. Metabolomics of Diabetic Retinopathy. Curr Diab Rep. 2017;17:102.

Sarafian MH, Gaudin M, Lewis MR, Martin FP, Holmes E, Nicholson JK, et al. Objective set of criteria for optimization of sample preparation procedures for ultra-high throughput untargeted blood plasma lipid profiling by ultra performance liquid chromatography-mass spectrometry. Anal Chem. 2014;86:5766–74.

Li J, Li L, Liu R, Zhu L, Zhou B, Xiao Y, et al. Integrative lipidomic features identify plasma lipid signatures in chronic urticaria. Front Immunol. 2022;13:933312.

Wen B, Mei Z, Zeng C, Liu S. metaX: a flexible and comprehensive software for processing metabolomics data. BMC Bioinformatics. 2017;18:183.

Fort PE, Rajendiran TM, Soni T, Byun J, Shan Y, Looker HC et al. Diminished retinal complex lipid synthesis and impaired fatty acid β-oxidation associated with human diabetic retinopathy. Jci Insight. 2021;6.

Fox TE, Han XL, Kelly S, Merrill AH, Martin RE, Anderson RE, et al. Diabetes alters sphingolipid metabolism in the retina - A potential mechanism of cell death in diabetic retinopathy. Diabetes. 2006;55:3573–80.

Levitsky Y, Hammer SS, Fisher KP, Huang C, Gentles TL, Pegouske DJ et al. Mitochondrial Ceramide effects on the retinal pigment epithelium in diabetes. Int J Mol Sci. 2020;21.

Mondal K, Mandal N. Role of bioactive sphingolipids in inflammation and Eye diseases. Adv Exp Med Biol. 2019;1161:149–67.

Laaksonen R, Ekroos K, Sysi-Aho M, Hilvo M, Vihervaara T, Kauhanen D, et al. Plasma ceramides predict cardiovascular death in patients with stable coronary artery disease and acute coronary syndromes beyond LDL-cholesterol. Eur Heart J. 2016;37:1967–76.

Summers SA, Chaurasia B, Holland WL. Metabolic messengers: ceramides. Nat Metabolism. 2019;1:1051–8.

Article   Google Scholar  

Zhang QJ, Holland WL, Wilson L, Tanner JM, Kearns D, Cahoon JM, et al. Ceramide mediates vascular dysfunction in Diet-Induced obesity by PP2A-Mediated dephosphorylation of the eNOS-Akt complex. Diabetes. 2012;61:1848–59.

Wilmott LA, Grambergs RC, Allegood JC, Lyons TJ, Mandal N. Analysis of sphingolipid composition in human vitreous from control and diabetic individuals. J Diabetes Complications. 2019;33:195–201.

Shi AH, Yoshinari M, Wakisaka M, Iwase M, Fujishima M. Lysophosphatidylcholine molecular species in low density lipoprotein of type 2 diabetes. Horm Metab Res. 1999;31:283–6.

Canning P, Kenny BA, Prise V, Glenn J, Sarker MH, Hudson N, et al. Lipoprotein-associated phospholipase A2 (Lp-PLA2) as a therapeutic target to prevent retinal vasopermeability during diabetes. Proc Natl Acad Sci U S A. 2016;113:7213–8.

Zhong J, Cheung CYY, Su X, Lee CH, Ru Y, Fong CHY, et al. Specific triacylglycerol, diacylglycerol, and lyso-phosphatidylcholine species for the prediction of type 2 diabetes: a  ∼  16-year prospective study in Chinese. Cardiovasc Diabetol. 2022;21:234.

Krischer JP, Liu X, Lernmark A, Hagopian WA, Rewers MJ, She JX, et al. The influence of type 1 diabetes genetic susceptibility regions, Age, Sex, and Family History on the Progression from multiple autoantibodies to type 1 diabetes: a TEDDY Study Report. Diabetes. 2017;66:3122–9.

Download references

Acknowledgements

Y.N.W. is supported by the China “Thousand Talents Plan” (Young Talents), Shaanxi province “Thousand Talents Plan” (Young Talents) and Foundation of Xi’an Jiaotong University (Plan A).

This study was supported by grants from The Natural Science Foundation Program of Shaanxi (2024JC-YBQN-0828) and National Key R&D Program of China (No. 2018YFC1311501).

Author information

Mingqian He, Guixue Hou and Mengmeng Liu contributed equally to this work.

Authors and Affiliations

Department of Endocrinology, the First Affiliated Hospital of Xi’an JiaoTong University, No.277, West Yanta Road, Xi’an, Shaanxi, 710061, P.R. China

Mingqian He, Mengmeng Liu, Zhaoyi Peng, Hui Guo, Yue Wang, Meng Zhang, Ziyi Chen, Patrick C.N. Rensen, Yanan Wang & Bingyin Shi

BGI-SHENZHEN, No. 21 Hongan 3rd Street, Yantian District, Shenzhen, Guangdong, 518083, P.R. China

Guixue Hou & Liang Lin

Department of Endocrinology and International Medical Center, the First Affiliated Hospital of Xi’an JiaoTong University, No.277, West Yanta Road, Xi’an, Shaanxi, 710061, P.R. China

Biobank, The First Affiliated Hospital of Xi’an JiaoTong University, Xi’an, Shaanxi, 710061, China

Chengdu HuiXin Life Technology, Chengdu, Sichuan, 610091, P.R. China

Xiaoming Yin

Department of Medicine, Division of Endocrinology, Leiden University Medical Center, P.O. Box 9600, Leiden, 2300 RA, The Netherlands

Patrick C.N. Rensen

Med-X institute, Center for Immunological and Metabolic Diseases, the First Affiliated Hospital of Xi’an JiaoTong University, Xi’an JiaoTong university, Xi’an, Shaanxi, 710061, P.R. China

Building NO.7, BGI Park, No. 21 Hongan 3rd Street, Yantian District, Shenzhen, Guangdong, 518083, P.R. China

You can also search for this author in PubMed   Google Scholar

Contributions

B.S., Y.W., and L.L. conceived this review and critically revised the manuscript. M.H., G.H., and M.L. drafted the manuscript. Z.P., H.G., Y.W., and J.S. drew the figures and collected the related references. H.L., X.Y., M.Z., Z.C. and P.C.N. supervised and revised the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Liang Lin , Yanan Wang or Bingyin Shi .

Ethics declarations

Ethics approval and consent to participate.

This study was conducted with approval from the Institutional Review Board at the First Affiliated Hospital of Xi’an Jiaotong University, Shaanxi, China (approval number: XJTU1AF2018LSK-055). Written informed consent was obtained from all participants.

Consent for publication

All authors critically reviewed and approved the final manuscript.

Competing interests

All authors declared no conflicts of interest.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

He, M., Hou, G., Liu, M. et al. Lipidomic studies revealing serological markers associated with the occurrence of retinopathy in type 2 diabetes. J Transl Med 22 , 448 (2024). https://doi.org/10.1186/s12967-024-05274-9

Download citation

Received : 21 August 2023

Accepted : 04 May 2024

Published : 13 May 2024

DOI : https://doi.org/10.1186/s12967-024-05274-9

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Retinopathy
• Serological markers

Journal of Translational Medicine

ISSN: 1479-5876

• Submission enquiries: Access here and click Contact Us
• General enquiries: [email protected]

Short and sweet: multiple mini case studies as a form of rigorous case study research

• Original Article
• Open access
• Published: 15 May 2024

Cite this article

You have full access to this open access article

• Sebastian Käss   ORCID: orcid.org/0000-0002-0640-3500 1 ,
• Christoph Brosig   ORCID: orcid.org/0000-0001-7809-0796 1 ,
• Markus Westner   ORCID: orcid.org/0000-0002-6623-880X 2 &
• Susanne Strahringer   ORCID: orcid.org/0000-0002-9465-9679 1  

135 Accesses

Explore all metrics

Case study research is one of the most widely used research methods in Information Systems (IS). In recent years, an increasing number of publications have used case studies with few sources of evidence, such as single interviews per case. While there is much methodological guidance on rigorously conducting multiple case studies, it remains unclear how researchers can achieve an acceptable level of rigour for this emerging type of multiple case study with few sources of evidence, i.e., multiple mini case studies. In this context, we synthesise methodological guidance for multiple case study research from a cross-disciplinary perspective to develop an analytical framework. Furthermore, we calibrate this analytical framework to multiple mini case studies by reviewing previous IS publications that use multiple mini case studies to provide guidelines to conduct multiple mini case studies rigorously. We also offer a conceptual definition of multiple mini case studies, distinguish them from other research approaches, and position multiple mini case studies as a pragmatic and rigorous approach to research emerging and innovative phenomena in IS.

Similar content being viewed by others

The theory contribution of case study research designs

Case Study Research

Avoid common mistakes on your manuscript.

1 Introduction

Case study research has become a widely used research method in Information Systems (IS) research (Palvia et al. 2015 ) that allows for a comprehensive analysis of a contemporary phenomenon in its real-world context (Dubé and Paré, 2003 ). This research method is particularly useful due to its flexibility in covering complex phenomena with multiple contextual variables, different types of evidence, and a wide range of analytical options (Voss et al. 2002 ; Yin 2018 ). Although case study research is particularly useful for studying contemporary phenomena, some researchers feel that it lacks rigour, particularly in terms of the validity of findings (Lee and Hubona 2009 ). In response to these criticisms, Yin ( 2018 ) provides comprehensive methodological steps to conduct case studies rigorously. In addition, many other publications with a partly discipline-specific view on case study research, offer guidelines for achieving rigour in case study research, e.g., Benbasat et al. ( 1987 ), Dubé and Paré ( 2003 ), Pan and Tan ( 2011 ), or Voss et al. ( 2002 ). Most publications on case study methodology converge on four criteria for ensuring rigour in case study research: (1) construct validity, (2) internal validity, (3) external validity, and (4) reliability (Gibbert et al. 2008 ; Voss et al. 2002 ; Yin 2018 ).

A key element of rigour in case study research is to look at the unit of analysis of a case from multiple perspectives in order to draw informed conclusions (Dubois and Gadde 2002 ). Case study researchers refer to this as triangulation, for example, by using multiple sources of evidence per case to support findings (Benbasat et al. 1987 ; Yin 2018 ). However, in our own research experience, we have come across numerous IS publications with a limited number of sources of evidence per case, such as a single interview per case. Some researchers refer to these studies as mini case studies (e.g., McBride 2009 ; Weill and Olson 1989 ), while others refer to them as multiple mini cases (e.g., Eisenhardt 1989 ). We were unable to find a definition or conceptualisation of this type of case study. Therefore, we will refer to this type of case study as a multiple mini case study (MMCS). Interestingly, many researchers use these MMCSs to study emerging and innovative phenomena.

From a methodological perspective, multiple case study publications with limited sources of evidence, also known as MMCSs, may face criticism for their lack of rigour (Dubé and Paré 2003 ). Alternatively, they may be referred to as “marginal case studies” (Piekkari et al. 2009 , p. 575) if they fail to establish a connection between theory and empirical evidence, provide only limited context, or merely offer illustrative aspects (Piekkari et al. 2009 ). IS scholars advocate conducting case study research in a mindful manner by balancing methodological blueprints and justified design choices (Keutel et al. 2014 ). Consequently, we propose MMCSs as a mindful approach with the potential for rigour, distinguishing them from marginal case studies. The following research question guides our study:

RQ: How can researchers rigorously conduct MMCSs in the IS discipline?

As shown in Fig.  1 , we develop an analytical framework by synthesising methodological guidance on how to rigorously conduct multiple case study research. We then address three aspects of our research question: For aspect (1), we analyse published MMCSs in the IS discipline to derive a "Research in Practice" definition of MMCSs and research situations for MMCSs. For aspect (2), we use the analytical framework to analyse how researchers in the IS discipline ensure that existing MMCSs follow a rigorous methodology. For aspect (3), we discuss the methodological findings about rigorous MMCSs in order to derive methodological guidelines for MMCSs that researchers in the IS discipline can follow.

Overview of the research approach

We approach these aspects by introducing the conceptual foundation for case study research in Sect.  2 . We define commonly accepted criteria for ensuring validity in case study research, introduce the concept of MMCSs, and distinguish them from other types of case studies. Furthermore, as a basis for analysis, we present an analytical framework of methodological steps and options for the rigorous conduct of multiple case study research. Section  3 presents our methodological approach to identifying published MMCSs in the IS discipline. In Sect.  4 , we first define MMCSs from a research in practice perspective (Sect.  4.1 ). Second, we present an overview of methodological options for rigorous MMCSs based on our analytical framework (Sect.  4.2 ). In Sect.  5 , we differentiate MMCSs from other research approaches, identify research situations of MMCSs (i.e., to study emerging and innovative phenomena), and provide guidance on how to ensure rigour in MMCSs. In our conclusion, we clarify the limitations of our study and provide an outlook for future research with MMCSs.

2 Conceptual foundation

2.1 case study research.

Case study research is about understanding phenomena by studying one or multiple cases in their context. Creswell and Poth ( 2016 ) define it as an “approach in which the investigator explores a bounded system (a case) or multiple bounded systems (cases) over time, through detailed, in-depth data collection” (p. 73). Therefore, it is suitable for complex topics with little available knowledge, needing an in-depth investigation, or where the research subject is inseparable from its context (Paré 2004 ). Additionally, Yin ( 2018 ) states that case study research is useful if the research focuses on contemporary events where no control of behavioural events is required. Typically, this type of research is most suitable for how and why research questions (Yin 2018 ). Eventually, the inferences from case study research are based on analytic or logical generalisation (Yin 2018 ). Instead of drawing conclusions from a representative statistical sample towards the population, case study research builds on analytical findings from the observed cases (Dubois and Gadde 2002 ; Eisenhardt and Graebner 2007 ). Case studies can be descriptive, exploratory, or explanatory (Dubé and Paré 2003 ).

The contribution of research to theory can be divided into the steps of theory building , development and testing , which is a continuum (Ridder 2017 ; Welch et al. 2011 ), and case studies are useful at all stages (Ridder 2017 ). In theory building, there is no theory to explain a phenomenon, and the researcher identifies new concepts, constructs, and relationships based on the data (Ridder 2017 ). In theory development, a tentative theory already exists that is extended or refined (e.g., by adding new antecedents, moderators, mediators, and outcomes) (Ridder 2017 ). In theory testing, an existing theory is challenged through empirical investigation (Ridder 2017 ).

In case study research, there are different paradigms for obtaining research results, either positivist or interpretivist (Dubé and Paré 2003 ; Orlikowski and Baroudi 1991 ). The positivist paradigm assumes that a set of variables and relationships can be objectively identified by the researcher (Orlikowski and Baroudi 1991 ). In contrast, the interpretivist paradigm assumes that the results are inherently rooted in the researcher’s worldview (Orlikowski and Baroudi 1991 ). Nowadays, researchers find that there are similar numbers of positivist and interpretivist case studies in the IS discipline compared to almost 20 years ago when positivist research was perceived as dominant (Keutel et al. 2014 ; Klein and Myers 1999 ). As we aim to understand how to conduct MMCSs rigorously, we focus on methodological guidance for positivist case study research.

The literature proposes a four-phased approach to conducting a case study: (1) the definition of the research design, (2) the data collection, (3) the data analysis, and (4) the composition (Yin 2018 ). Table 1 provides an overview and explanation of the four phases.

Case studies can be classified based on their depth and breadth, as shown in Fig.  2 . We can distinguish five types of case studies: in-depth single case studies , marginal case studies , multiple case studies , MMCSs , and extensive in-depth multiple case studies . Each type has distinct characteristics, yet the boundaries between the different types of case studies is blurred. Except for the marginal case studies, the italic references in Fig.  2 are well-established publications that define the respective type and provide methodological guidance. The shading is to visualise the different types of case studies. The italic references in Fig.  2 for marginal case studies refer to publications that conceptualise them.

Simplistic conceptualisation of MMCS

In-depth single case studies focus on a single bounded system as a case (Creswell and Poth 2016 ; Paré 2004 ; Yin 2018 ). According to the literature, a single case study should only be used if a case meets one or more of the following five characteristics: it is a critical, unusual, common, revelatory, or longitudinal case (Benbasat et al. 1987 ; Yin 2018 ). Single case studies are more often used for descriptive research (Dubé and Paré 2003 ).

A second type of case studies are marginal case studies , which generally have low depth (Keutel et al. 2014 ; Piekkari et al. 2009 ). Marginal case studies lack a clear link between theory and empirical evidence, a clear contextualisation of the case, and are often used for illustration purposes (Keutel et al. 2014 ; Piekkari et al. 2009 ). Therefore, marginal case studies provide only marginal insights with a lack of generalisability.

In contrast, multiple case studies employ multiple cases to obtain a broader picture of the researched phenomenon from different perspectives (Creswell and Poth 2016 ; Paré 2004 ; Yin 2018 ). These multiple case studies are often considered to provide more robust results due to the multiplicity of their insights (Eisenhardt and Graebner 2007 ). However, often discussed criticisms of multiple case studies are high costs, difficult access to multiple sources of evidence for each case, and long duration (Dubé and Paré 2003 ; Meredith 1998 ; Voss et al. 2002 ). Eisenhardt ( 1989 ) considers four to ten in-depth cases as a suitable number of cases for multiple case study research. With fewer than four cases, the empirical grounding is less convincing, and with more than ten cases, researchers quickly get overwhelmed by the complexity and volume of data (Eisenhardt 1989 ). Therefore, methodological literature views extensive in-depth multiple case studies as almost infeasible due to their high complexity and resource demands, which can easily overwhelm the research team and the readers (Stake 2013 ). Hence, we could not find a methodological publication outlining the approach for this case study type.

To solve the complexity and resource issues for multiple case studies, a new phenomenon has emerged: MMCS . An MMCS is a special type of multiple case study that focuses on an investigation's breadth by using a relatively high number of cases while having a somewhat limited depth per case. We characterise breadth not only by the number of cases but also by the variety of the cases. Even though there is no formal conceptualisation of the term, we understand MMCSs as a type of multiple case study research with few sources of evidence per case. Due to the limited depth per case, one can overcome the resource and complexity issues of classical multiple case studies. However, having only some sources of evidence per case may be considered a threat to rigour. Therefore, in this publication, we provide suggestions on how to address these threats.

2.2 Rigour in case study research

Rigour is essential for case study research (Dubé and Paré 2003 ; Yin 2018 ) and, in the early 2000s, researchers criticised case study research for inadequate rigour (e.g., Dubé and Paré 2003 ; Gibbert et al. 2008 ). Based on this, various methodological publications provide guidance for rigorous case study research (e.g., Dubé and Paré 2003 ; Gibbert et al. 2008 ).

Methodological literature proposes four criteria to ensure rigour in case study research: Construct validity , internal validity , external validity , and reliability (Dubé and Paré 2003 ; Gibbert et al. 2008 ; Yin 2018 ). Table 2 outlines these criteria and states in which research phase they should be addressed (Yin 2018 ). Methodological literature agrees that all four criteria must be met for rigorous case study research (Dubé and Paré 2003 ).

The methodological literature discusses multiple options for achieving rigour in case study research (e.g., Benbasat et al. 1987 ; Dubé and Paré 2003 ; Eisenhardt 1989 ; Yin 2018 ). We aggregated guidance from multiple sources by conducting a cross-disciplinary literature review to build our analytical foundation (cf. Fig. 1 ). This literature review aims to identify the most relevant multiple case study methodology publications from a cross-disciplinary and IS-specific perspective. We focus on the most cited methodology publications, while being aware that this may over-represent disciplines with a higher number of case study publications. However, this approach helps to capture an implicit consensus among case study researchers on how to conduct multiple case studies rigorously. The literature review produced an analytical framework of methodological steps and options for conducting multiple case studies rigorously. Appendix A Footnote 1 provides a detailed documentation of the literature review process. The analytical framework derived from the set of methodological publications is presented in Table  3 . We identified required and optional steps for each research stage. The analytical framework is the basis for the further analysis of MMCS and an explanation of all methodological steps is provided in Appendix B. Footnote 2

3 Research methodology

For our research, we analysed published MMCSs in the IS discipline with the goal of understanding how these publications ensured rigour. This section outlines the methodology of how we identified our MMCS publications.

First, we searched bibliographic databases and citation indexing services (Vom Brocke et al. 2009 ; Vom Brocke et al. 2015 ) to retrieve IS-specific MMCSs (Hanelt et al. 2015 ). As shown in Fig.  3 , we used two sets of keywords, the first set focusing on multiple case studies and the second set explicitly on mini case studies. We decided to follow this approach as many MMCSs are positioned as multiple case studies, avoiding the connotation “mini” or “short”. We restricted our search to completed research publications written in English from litbaskets.io size “S”, a set of 29 highly ranked IS journals (Boell and Wang 2019 ) Footnote 3 and leading IS conference proceedings from AMCIS, ECIS, HICSS, ICIS, and PACIS (published until end of June 2023). We focused on these outlets, as they can be taken as a representative sample of high quality IS research (Gogan et al. 2014 ; Sørensen and Landau 2015 ).

The search process for published MMCSs in the IS discipline

Second, we screened the obtained set of IS publications to identify MMCSs. We only included publications with positivist multiple cases where the majority of cases was captured with only one primary source of evidence. Further, we excluded all publications which were interview studies rather than case studies (i.e., they do not have a clearly defined case). In some cases, it was unclear from the full text whether a publication fulfils this requirement. Therefore, we contacted the authors and clarified the research methodology with them. Eventually, our final set contained 50 publications using MMCSs.

For qualitative data analysis, we employed axial coding (Recker 2012 ) based on the pre-defined analytical framework shown in Table  3 . For the coding, we followed the explanations of the authors in the manuscripts. The coding was conducted and reviewed by two of the authors. We coded the first five publications of the set of IS MMCS publications together and discussed our decisions. After the initial coding was completed, we checked the reliability and validity by re-coding a sample of the other author’s set. In this sample, we achieved inter-coder reliability of 91% as a percent agreement in the decisions made (Nili et al. 2020 ). Hence, we consider our coding as highly consistent.

In the results section, we illustrate the chosen methodological steps for each MMCS type (descriptive, exploratory, and explanatory). For this purpose, we selected three publications based on two criteria: only journal publications, as they have more details about their methodological steps and publications which applied most of the analytical framework’s methodology steps. This led to three exemplary IS MMCS publications: (1) McBride ( 2009 ) for descriptive MMCSs, (2) Baker and Niederman ( 2014 ) for exploratory MMCSs, and (3) van de Weerd et al. ( 2016 ) for explanatory MMCSs.

4.1 MMCS from a “Research in Practice" perspective

In this section, we explain MMCSs from a "Research in Practice" perspective and identify different types based on our sample of 50 MMCS publications. As outlined in Sect.  2.1 , an MMCS is a special type of a multiple case study, which focuses on an investigation’s breadth by using a relatively high number of cases while having a limited depth per case. In the most extreme scenario, an MMCS only has one source of evidence per case. Moreover, breadth is not only characterised by the number of cases, but also by the variety of the cases. MMCSs have been used widely but hardly labelled as such, i.e., only 10 of our analysed 50 MMCS publications explicitly use the terms mini or short case in the manuscript . Multiple case study research distinguishes between descriptive, exploratory, and explanatory case studies (Dubé and Paré 2003 ). The MMCSs in our sample follow the same classification with three descriptive, 40 exploratory, and seven explanatory MMCSs. Descriptive and exploratory MMCSs are used in the early stages of research , and exploratory and explanatory MMCSs are used to corroborate findings .

Descriptive MMCSs provide little information on the methodological steps for the design, data collection, analysis, and presentation of results. They are used to illustrate novel phenomena and create research questions, not solutions, and can be useful for developing research agendas (e.g., McBride 2009 ; Weill and Olson 1989 ). The descriptive MMCS publications analysed contained between four to six cases, with an average of 4.6 cases per publication. Of the descriptive MMCSs analysed, one did not state research questions, one answered a how question and the third answered how and what questions. Descriptive MMCSs are illustrative and have a low depth per case, resulting in the highest risk of being considered a marginal case study.

Exploratory MMCSs are used to explore new phenomena quickly, generate first research results, and corroborate findings. Most of the analysed exploratory MMCSs answer what and how questions or combinations. However, six publications do not explicitly state a research question, and some MMCSs use why, which, or whether research questions. The analysed exploratory MMCSs have three to 27 cases, with an average of 10.2 cases per publication. An example of an exploratory MMCS is the study by Baker and Niederman ( 2014 ), who explore the impacts of strategic alignment during merger and acquisition (M&A) processes. They argue that previous research with multiple case studies (mostly with  three cases) shows some commonalities, but much remains unclear due to the low number of cases. Moreover, they justify the limited depth of their research with the “proprietary and sensitive nature of the questions” (Baker and Niederman 2014 , p. 123).

Explanatory MMCSs use an a priori framework with a relatively high number of cases to find groups of cases that share similar characteristics. Most explanatory MMCSs answer how questions, yet some publications answer what, why, or combinations of the three questions. The analysed explanatory MMCSs have three to 18 cases, with an average of 7.2 cases per publication. An example of an explanatory MMCS publication is van de Weerd et al. ( 2016 ), who researched the influence of organisational factors on the adoption of Software as a Service (SaaS) in Indonesia.

4.2 Applied MMCS methodology in IS publications

4.2.1 overarching.

In the following sections, we present the results of our analysis. For this purpose, we mapped our 50 IS MMCS publications to the methodological options (Table  3 ) and present one example per MMCS type. We extended some methodological steps with options from methodology-in-use. A full coding table can be found in Appendix D Footnote 4 . Tables 4 , 5 , 6 and 7 summarise the absolute and percentual occurrences of each methodological option in descriptive, exploratory, and explanatory IS MMCS publications. All tables are structured in the same way and show the number of absolute and, in parentheses, the percentual occurrences of each methodological option. The percentages may not add up to 100% due to rounding. The bold numbers show the most common methodological option for each MMCS type and step. Most publications were classified in previously identified options. Some IS MMCS publications lacked detail on methodological steps, so we classified them as "step not evident". Only 16% (8 out of 50) explained how they addressed validity and reliability threats.

4.2.2 Research design phase

There are six methodological steps in the research design phase, as shown in Table  4 . Descriptive MMCSs usually define the research question (2 out of 3, 67%), clarify the unit of analysis (2 out of 3, 67%), bound the case (2 out of 3, 67%), or specify an a priori theoretical framework (2 out of 3, 67%). The case replication logic is mostly not evident (2 out of 3, 67%). Descriptive MMCS use a criterion-based selection (1 out of 3, 33%), a maximum variation selection (1 out of 3, 33%), or do not specify the selection logic (1 out of 3, 33%). Descriptive MMCSs have a high risk of becoming a marginal case study due to their illustrative nature–our chosen example is not different. McBride ( 2009 ) does not define the research question, does not have a priori theoretical framework, nor does he justify the case replication and the case selection logic. However, he clarifies the unit of analysis and extensively bounds each case with significant context about the case organisation and its setup.

The majority of exploratory MMCSs define the research question (34 out of 40, 85%) clarify the unit of analysis (35 out of 40, 88%), and specify an a priori theoretical framework (33 out of 40, 83%). However, only a minority (6 out of 40, 15%) follow the instructions of bounding the case or justify the case replication logic (13 out of 40, 33%). The most used case selection logic is the criterion-based selection (23 out of 40, 58%), followed by step not evident (5 out of 40, 13%), other selection approaches (3 of 40, 13%), maximum variation selection (3 out of 40, 13%), a combination of approaches (2 out of 40, 5%), snowball selection (2 out of 40, 5%), typical case selection (1 out of 40, 3%), and convenience-based selection (1 out of 40, 3%). Baker and Niederman ( 2014 ) build their exploratory MMCS on previous multiple case studies with three cases that showed ambiguous results. Hence, Baker and Niederman ( 2014 ) formulate three research objectives instead of defining a research question. They clearly define the unit of analysis (i.e., the integration of the IS function after M&A) but lack the bounding of the case. The authors use a rather complex a priori framework, leading to a high number of required cases. This a priori framework is also used for the “theoretical replication logic [to choose] conforming and disconfirming cases” (Baker and Niederman 2014 , p. 116). A combination of maximum variation and snowball selection is used to select the cases (Baker and Niederman 2014 ). The maximum variation is chosen to get evidence for all elements of their rather complex a priori framework (i.e., the breadth), and the snowball sampling is chosen to get more details for each framework element.

All explanatory MMCS s define the research question, clarify the unit of analysis, and specify an a priori theoretical framework. However, only one (14%) bounds the case. The case replication logic is mostly a mixture of theoretical and literal replication (3 out of 7, 43%) and one (14%) MMCS does a literal replication. For 43% (3 out of 7) of the publications, the step is not evident. Most explanatory MMCSs use criterion-based selection (4 out of 7, 57%), followed by maximum variation selection (2 out of 7, 29%) and snowball selection (1 out of 7, 14%). In their publication, van de Weerd et al. ( 2016 ) define the research question and clarify the unit of analysis (i.e., the influence of organisational factors on SaaS adoption in Indonesian SMEs). Further, they specify an a priori framework (i.e., based on organisational size, organisational readiness, and top management support) to target the research (van de Weerd et al. 2016 ). A combination of theoretical (between the groups of cases) and literal (within the groups of cases) replication was used. To strengthen the findings, van de Weerd et al. ( 2016 ) find at least one other literally replicated case for each theoretically replicated case.

To summarize this phase, we see that in all three types of MMCSs, the majority of publications define the research question, clarify the unit of analysis, and specify an a priori theoretical framework. Moreover, descriptive MMCSs are more likely to bound the case than exploratory and explanatory MMCSs. However, only a minority across all MMCSs justify the case replication logic, whereas the majority does not. Most MMCSs justify the case selection logic, with criterion-based case selection being the most often applied methodological option.

4.2.3 Data collection phase

In the data collection phase, there are four methodological steps, as summarised in Table  5 .

One descriptive MMCS applies triangulation via multiple sources, whereas for the majority (2 out of 3, 67%), the step is not evident. One (33%) of the analysed descriptive MMCSs creates a full chain of evidence, none creates a case study database, and one (33%) uses a case study protocol. McBride ( 2009 ) applies triangulation via multiple sources, as he followed “up practitioner talks delivered at several UK annual conferences” (McBride 2009 , p. 237). Therefore, we view the follow-up interviews as the primary source of evidence per case, as dedicated questions to the unit of analysis can be asked per case. Triangulation via multiple sources was then conducted by combining practitioner talks and documents with follow-up interviews. McBride ( 2009 ) does not create a full chain of evidence, a case study database, nor a case study protocol. This design decision might be rooted in the objective of a descriptive MMCS to illustrate and open up new questions rather than find clear solutions (McBride 2009 ).

Most exploratory MMCSs triangulate via multiple sources (20 out of 40, 50%) or via multiple investigators (4 out of 40, 10%). Eight (20%) exploratory MMCSs apply multiple triangulation types and for eight (20%), no triangulation is evident. At first glance, a triangulation via multiple sources may seem contradictory to the definition of MMCSs–yet it is not. MMCSs that triangulate via multiple sources have one source per case as the primary, detailed evidence (e.g., an interview), which is combined with easily available supplementary sources of evidence (e.g., public reports and documents (Baker and Niederman 2014 ), press articles (Hahn et al. 2015 ), or online data (Kunduru and Bandi 2019 )). As this leads to multiple sources of evidence, we understand this as a triangulation via multiple sources; however, on a different level than triangulating via multiple in-depth interviews per case. Only a minority of exploratory MMCSs create a full chain of evidence (14 out of 40, 35%), and a majority (23 out of 40, 58%) use a case study database or a case study protocol (20 out of 40, 50%). Baker and Niederman ( 2014 ) triangulate with multiple sources (i.e., financial reports as supplementary sources) to increase the validity of their research. Further, the authors create a full chain of evidence from their research question through an identical interview protocol to the case study’s results. For every case, an individual case report is created and stored in the case study database (Baker and Niederman 2014 ).

All explanatory MMCSs triangulate during the data collection phase, either via multiple sources (2 out of 7, 29%) or a combination of multiple investigators and sources (5 out of 7, 71%). Interestingly, only three explanatory MMCSs (43%) create a full chain of evidence. All create a case study database (7 out of 7, 100%) and the majority creates a case study protocol (6 out of 7, 86%). In their explanatory MMCS, van de Weerd et al. ( 2016 ) use semi-structured interviews as the primary data collection method. The interview data is complemented “with field notes and (online) documentation” (van de Weerd et al. 2016 , p. 919), e.g., data from corporate websites or annual reports. Moreover, a case study protocol and a case study database in NVivo are created to increase reliability.

To summarise the data collection phase, we see that most (40 out of 50, 80%) of MMCSs apply some type of triangulation. However, only 36% (18 out of 50) of the analysed MMCSs create a full chain of evidence. Moreover, descriptive MMCSs are less likely to create a case study database (0 out of 3, 0%) or a case study protocol (1 out of 3, 33%). In contrast, most exploratory and explanatory MMCS publications create a case study database and case study protocol.

4.2.4 Data analysis phase

There are three methodological steps (cf. Table 6 ) for the data analysis phase, each with multiple methodological options.

One descriptive MMCS (33%) corroborates findings through triangulation, and two do not (67%). Further, one (33%) uses a rich description of findings as other corroboration approaches, whereas for the majority (2 out of 3, 67%), the corroboration with other approaches is not evident. Descriptive MMCSs mostly do not define their within-case analysis strategy (2 out of 3, 67%). However, pre-defined patterns are used to conduct a cross-case analysis (2 out of 3, 67%). In the data analysis, McBride ( 2009 ) triangulates via multiple sources of evidence (i.e., talks at practitioner conferences and resulting follow-up interviews), but does not apply other corroboration approaches or provides methodological explanations for the within or cross-case analysis. This design decision might be rooted in the illustrative nature of his descriptive MMCS and the focus on analysing each case standalone.

Exploratory MMCSs mostly corroborate findings through a combination of triangulation via multiple investigators and sources (15 out of 40, 38%) or triangulation via multiple sources (9 out of 40, 23%). However, for ten (25%) exploratory MMCSs, this step is not evident. For the other corroboration approaches, a combination of approaches is mostly used (15 out of 40, 38%), followed by rich description of findings (11 out of 40, 28%), peer review (6 out of 40, 15%), and prolonged field visits (1 out of 40, 3%). For five (13%) publications, other corroboration approaches are not evident. Pattern matching (17 out of 40, 43%) and explanation building (5 out of 40, 13%) are the most used methodological options for the within-case analysis. To conduct a cross-case analysis, 11 (28%) MMCSs use a comparison of pairs or groups of cases, nine (23%) pre-defined patterns, and six (15%) structure their data along themes. Interestingly, for 14 (35%) exploratory MMCSs, no methodological step to conduct the cross-case analysis is evident. Baker and Niederman ( 2014 ) use a combination of triangulation via multiple investigators (“The interviews were coded by both researchers independently […], with a subsequent discussion to reach complete agreement” (Baker and Niederman 2014 , p. 117)) and sources to increase internal validity. Moreover, the authors use a rich description of the findings. An explanation-building strategy is used for the within-case analysis, and the cross-case analysis is done based on pre-defined patterns (Baker and Niederman 2014 ). This decision for the cross-case analysis is justified by a citation of Dubé and Paré ( 2003 , p. 619), who see it as “a form of pattern-matching in which the analysis of the case study is carried out by building a textual explanation of the case.”

Explanatory MMCSs corroborate findings through a triangulation via multiple sources (4 out of 7, 57%) or a combination of multiple investigators and sources (3 out of 7, 43%). For the other corroboration approaches, a rich description of findings (3 out of 7, 43%), a combination of approaches (3 out of 7, 43%), or peer review (1 out of 7, 14%) are used. To conduct a within-case analysis, pattern matching (5 out of 7, 71%) or explanation building (1 out of 7, 14%) are used. For the cross-case analysis, pre-defined patterns (3 out of 7, 43%) and a comparison of pairs or groups of cases (2 out of 7, 29%) are used; yet, for two (29%) explanatory MMCSs a cross-case analysis step is not evident. van de Weerd et al. ( 2016 ) corroborate their findings through a triangulation via multiple sources, a combination of rich description of findings and solicitation of participants’ views (“summarizing the interview results of each case company for feedback and approval” (van de Weerd et al. 2016 , p. 920)) as other corroboration approaches. Moreover, for the within-case analysis, the authors “followed an explanation-building procedure to strengthen […] [the] internal validity” (van de Weerd et al. 2016 , p. 920). For the cross-case, the researchers compare groups of cases. They refer to this approach as an informal qualitative comparative analysis.

To summarize the results of the data analysis phase, we see that some type of triangulation is used by most of the MMCSs, with source triangulation (alone or in combination with another approach) being the most often used methodological option. For the within-case analysis, pattern matching (22 of 50, 44%) is the most often used methodological option. For the cross-case analysis, pre-defined patterns are most often used (14 out of 50, 28%). However, depending on the type of MMCS, there are differences in the options used and some methodological options are never used (e.g., time-series analysis and solicitation of participants’ views).

4.2.5 Composition phase

We can find two methodological steps for the composition phase, as summarized in Table  7 .

Descriptive MMCSs do not apply triangulation in the composition phase (3 out of 3, 100%), nor do they use the methodological step to let key informants review the draft of the case study report (3 of 3, 100%). Also, the descriptive MMCS by McBride ( 2009 ) does not apply any of the methodological steps.

Exploratory MMCSs mostly use triangulation via multiple sources (25 out of 40, 63%), a combination of multiple sources and theories (2 out of 40, 5%), triangulation via multiple investigators (1 out of 40, 3%), and a combination of multiple sources and methods (1 out of 40, 3%). However, for 11 (28%) exploratory MMCS publications, no triangulation step is evident. Moreover, the majority (24 out of 40, 85%) do not let key informants review a draft of the case study report. Baker and Niederman ( 2014 ) do not use triangulation in the composition phase nor let key informants review the draft of the case study report. An example of an exploratory publication that applies both methodological steps is the publication by Kurnia et al. ( 2015 ). The authors triangulate via multiple sources and let key informants review their interview transcripts and the case study report to increase construct validity.

Explanatory MMCSs mostly use triangulation via multiple sources (5 out of 7, 71%) and for two (29%), the step is not evident. Furthermore, only two MMCS (29%) publications let key informants review the draft of the case study report, whereas the majority (5 out of 7, 71%) do not. In their publication , van de Weerd et al. ( 2016 ) use both methodological steps of the composition phase. The authors triangulate via multiple sources by presenting interview snippets from different cases for each result in the case study manuscript. Moreover, each case and the final case study report were shared with key informants for review and approval to reduce the risk of misinterpretations and increase construct validity.

To summarize, most exploratory and explanatory MMCSs use triangulation in the composition phase, whereas descriptive MMCSs do not. Moreover, only a fraction of all MMCSs let key informants review a draft of the case study report (8 out of 50, 16%).

5 Discussion

5.1 mmcs from a “research in practice" perspective, 5.1.1 delineating mmcs from other research approaches.

In this section, we delineate MMCSs from related research approaches. In the subsequent sections, we outline research situations for which MMCSs can be used and the benefits MMCSs provide.

Closely related research approaches from which we delineate MMCSs are multiple case studies , interviews, and vignettes . As shown in Fig.  2 , MMCSs differ from multiple case studies in that they focus on breadth by using a high number of cases with limited depth per case. In the most extreme situation, an MMCS only has one primary source of evidence per case. Moreover, MMCSs can also consider a greater variety of cases. In contrast, multiple case studies have a high depth per case and multiple sources of evidence per case to allow for a source triangulation (Benbasat et al. 1987 ; Yin 2018 ). Moreover, multiple case studies mainly focus on how and why research questions (Yin 2018 ), whereas MMCSs can additionally answer what, whether, and which research questions. The rationale why MMCSs are used for more types of research questions is their breadth, allowing them to also answer rather explorative research questions.

Distinguishing MMCSs from interviews is more difficult . Yet, we see two differences. First, interview studies do not have a clear unit of analysis. Interview studies may choose interviewees based on expertise (expert interviews), whereas case study researchers select informants based on the ability to inform about the case (key informants) (Yin 2018 ). Most of the 50 analysed MMCS (88%) specify their unit of analysis. Second, MMCSs can use multiple data collection methods (e.g., observations, interviews, documents), while interviews only use one (the interview) (Lamnek and Krell 2010 ). An example showing these delineation difficulties between MMCSs and interviews is the publication of Demlehner and Laumer ( 2020 ). The authors claim to take “a multiple case study approach including 39 expert interviews” (Demlehner and Laumer 2020 , p. 1). However, our criteria classify this as an interview study. Demlehner and Laumer ( 2020 ) contend that the interviewees were chosen using a “purposeful sampling strategy” (p. 5). However, case study research selects cases based on replication logic, not sampling (Yin 2018 ). Moreover, the results are not presented on a per-case basis (as usual for case studies); instead, the findings are presented on an aggregated level, similar to expert interviews. Therefore, we would not classify this publication as an MMCS but find that it is a very good example to discuss this delineation.

MMCSs differ from vignettes, which are used for (1) data collection , (2) data analysis , and (3) research communication (Klotz et al. 2022 ; Urquhart 2001 ). Researchers use vignettes for data collection as stimuli to which participants react (Klotz et al. 2022 ), i.e., a carefully constructed description of a person, object, or situation (Atzmüller and Steiner 2010 ; Hughes and Huby 2002 ). We can delineate MMCS from vignettes for data collection based on this definition. First, MMCSs are not used as a stimulus to which participants can react, as in MMCSs, data is collected without the stimulus requirement. Furthermore, vignettes for data collection are carefully constructed, which contradicts the characteristics of MMCS, that are all based on collected empirical data and not constructed descriptions.

A data analysis vignette is used as a retrospective tool (Klotz et al. 2022 ) and is very short, which makes it difficult to analyse deeper relationships between constructs. MMCSs differ from vignettes for data analysis in two ways. First, MMCSs are a complete research methodology with four steps, whereas vignettes for data analysis cover only one step (the data analysis) (e.g., Zamani and Pouloudi 2020 ). Second, vignettes are too short to conduct a thorough analysis of relationships, whereas MMCSs foster a more comprehensive analysis, allowing for a deeper analysis of relationships.

Finally, a vignette used for research communication “(1) is bounded to a short time span, a location, a special situation, or one or a few key actors, (2) provides vivid, authentic, and evocative accounts of the events with a narrative flow, (3) is rather short, and (4) is rooted in empirical data, sometimes inspired by data or constructed.” (Klotz et al. 2022 , p. 347). Based on the four elements for the vignettes’ definition, we can delineate MMCS from vignettes used for research communication. First, MMCSs are not necessarily bounded to a short time span, location, special situation, or key actors; instead, with MMCSs, a clearly defined case bounded in its context is researched. Second, the focus of MMCSs is not on the narrative flow; instead, the focus is on describing (c.f., McBride ( 2009 )), exploring (c.f., Baker and Niederman ( 2014 )), or explaining (c.f., van de Weerd et al. ( 2016 )) a phenomenon. Third, while MMCSs do not have the depth of multiple case studies, they are much more comprehensive than vignettes (e.g., the majority of analysed publications (42 of 50, 84%) specify an a priori theoretical framework). Fourth, every MMCS must be based on empirical data, i.e., all of our 50 MMCSs collect data for their study and base their results on this data. This is a key difference from vignettes, which can be completely fictitious (Klotz et al. 2022 ).

5.1.2 MMCS research situations

The decision to use an MMCS as a research method depends on the research context. MMCSs can be used in the early stages of research (descriptive and exploratory MMCS) and to corroborate findings (exploratory and explanatory MMCS). Academic literature has yet to agree on a uniform categorisation of research questions. For instance, Marshall and Rossman ( 2016 ) distinguish between descriptive, exploratory, explanatory, and emancipatory research questions. In contrast, Yin ( 2018 ) distinguishes between who , what , where , how , and why questions, where he argues that the latter two are especially suitable for explanatory case study research. MMCSs can answer more types of research questions than Yin ( 2018 ) proposed. The reason for this is rooted in the higher breadth of MMCSs, which allows MMCSs to also answer rather exploratory what , whether , or which questions, besides the how and why questions that are suggested by Yin ( 2018 ).

For descriptive MMCSs , the main goal of the how and what questions is to describe the phenomenon. However, in our sample of analysed MMCSs, the analysis stops after the description of the phenomenon. The main goal of the five types of exploratory MMCS research questions is to investigate little-known aspects of a particular phenomenon. The how and why questions analyse operational links between different constructs (e.g., “How do different types of IS assets account for synergies between business units to create business value?” (Mandrella et al. 2016 , p. 2)). Exploratory what questions can be answered by case study research and other research methods (e.g., surveys or archival analysis) (Yin 2018 ). Nevertheless, all whether and which MMCS research questions can also be re-formulated as exploratory what questions. The reason why many MMCSs answer what , whether , or which research questions lies in the breadth (i.e., higher number and variety of cases) of MMCS, that allow them to answer these rather exploratory research questions to a satisfactory level. Finally, the research questions of the explanatory MMCSs aim to analyse operational links (i.e., how or why something is happening). This is also in line with the findings of Yin ( 2018 ) for multiple case study research. However, for MMCSs, this view must be extended, as explanatory MMCSs are also able to answer what questions. We explain this with the higher breadth of MMCS.

To discuss an MMCS’s contribution to theory, we use the idea of the theory continuum proposed by Ridder ( 2017 ) (cf. Section  2.1 ). Despite being used in the early phase of research (descriptive and exploratory), we do not recommend using MMCSs to build theory . We argue that for theory building, data with “as much depth as […] feasible” (Eisenhardt 1989 , p. 539) is required on a per-case basis. However, a key characteristic of MMCSs is the limited depth per case, which conflicts with the in-depth requirements of theory building. Moreover, a criterion for theory building is that there is no theory available which explains the phenomenon (Ridder 2017 ). Nevertheless, in our analysed MMCSs, 84% (42 out of 50) have an a priori theoretical framework. Furthermore, for theory building, the recommendation is to use between four to ten cases; with more, “it quickly becomes difficult to cope with the complexity and volume of the data” (Eisenhardt 1989 , p. 545). However, a characteristic of MMCSs is to have a relatively high number of cases, i.e., the analysed MMCSs often have more than 20 cases, which is significantly above the recommendation for theory building.

The next phase in the theory continuum is theory development , where a tentative theory is extended or refined (Ridder 2017 ). MMCSs should and are used for theory development, i.e., 84% (42 out of 50) of analysed MMCS publications have an a priori theoretical framework extended and refined using the MMCS. An MMCS example for theory development is the research of Karunagaran et al. ( 2016 ), who use a combination of the diffusion of innovation theory and technology organisation environment framework as tentative theories to research the adoption of cloud computing. As Ridder ( 2017 ) outlined, for theory development, literal replication and pattern matching should be used. Both methodological steps are used by Karunagaran et al. ( 2016 ) to identify the mechanisms of cloud adoption more precisely.

The next step in the theory continuum is theory testing , where existing theory is challenged by finding anomalies that existing theory cannot explain (Ridder 2017 ). The boundaries between theory development and testing are often blurred (Ridder 2017 ). In theory testing, the phenomenon is understood, and the research strategy focuses on testing if the theory also holds under different circumstances, i.e., hypotheses can be formed and tested based on existing theory (Ridder 2017 ). In multiple case study research, theory testing uses theoretical replication with pattern matching or addressing rival explanations (Ridder 2017 ). In our MMCS publications, no publication addresses rival explanations, and only a few apply theoretical replication and pattern matching–yet not for theory testing. A few publications claim to test propositions derived from an a priori theoretical framework (e.g., Schäfferling et al. 2011 ; Spiegel and Lazic 2010 ; Wagner and Ettrich-Schmitt 2009 ). However, these publications either do not state their replication logic (e.g., Spiegel and Lazic 2010 ; Wagner and Ettrich-Schmitt 2009 ) or use a literal replication (e.g., Schäfferling et al. 2011 ), both of which weaken the value of their theory testing.

5.1.3 MMCS research benefits

MMCSs are beneficial in multiple research situations and can be an avenue to address the frequent criticism of multiple case study research of being time-consuming and costly (Voss et al. 2002 ; Yin 2018 ).

Firstly, MMCSs can be used for time-critical topics where it is beneficial to publish results quicker and discuss them instead of conducting in-depth multiple case studies (e.g., COVID-19 (e.g., dos Santos Tavares et al. 2021 ) or emergent technology adoption (e.g., Bremser 2017 )). Especially with COVID-19, research publishing saw a significantly higher speed due to special issues of journals and faster review processes. Further, due to the fast technological advancements, there is a higher risk that the results are obsolete and of less practical use when researched with time-consuming multiple in-depth case studies.

Secondly, MMCSs can be used in research situations when it is challenging to gather in-depth data from multiple sources of evidence for each case due to the limited availability of sources of evidence or limited accessibility of sources of evidence. When researching novel phenomena (e.g., the adoption of new technologies in organisations), managers and decision-makers are usually interviewed as sources of evidence. However, in most organisations, only one (or very few) decision-makers have the ability to inform and should be interviewed, limiting the potential sources of evidence per case. These decision-makers often have limited availability for multiple in-depth interviews. Furthermore, the sources of evidence are often difficult to access, as professional organisations have regulations that prevent sharing documents with researchers.

Thirdly, MMCSs can be beneficial when the research framework is complex and requires many cases for validation (e.g., Baker and Niederman ( 2014 ) validate their rather complex a priori framework with 22 cases) or when previous research has led to contradictory results . Therefore, in both situations, a higher breadth of cases is required to also research combinatorial effects (e.g., van de Weerd et al. 2016 ). However, conducting an in-depth multiple case study would take time and effort. Therefore, MMCSs can be a mindful way to collect many cases, but in the same vein, being time and cost-efficient.

5.2 MMCS research rigour

Table 8 outlines two types of methodological steps for MMCSs. The first are methodological steps, where MMCSs should follow multiple case study methodological guidance (e.g., clarify the unit of analysis ), while the second is unique to MMCSs due to its characteristics. This section focuses on the latter, exploring MMCS characteristics, problems, validity threats, and proposed solutions.

The characteristics of MMCSs of having only one primary source of evidence per case prevents MMCSs from using source triangulation, which is often used in multiple case study research (Stake 2013 ; Voss et al. 2002 ; Yin 2018 ). By only having one source of evidence, researchers can fail to develop a sufficient set of operational measures and instead rely on subjective judgements, which threatens construct validity (Yin 2018 ). The threats to construct validity must be addressed throughout the MMCS research process. To do so, we propose to use easily accessible supplementary data or other triangulation approaches to increase construct validity in a MMCS. For the other triangulation approaches, we see that the majority of publications use supplementary data (e.g., publicly available documents) as further sources of evidence, multiple investigators, multiple methods (e.g., quantitative and qualitative), multiple theories, or combinations of these (cf. Tables 5 , 6 and 7 ). Having one or, in the best case, all of them reduces the risk of reporting spurious relationships and subjective judgements of the researchers, as a phenomenon is analysed from multiple perspectives. Besides the above-mentioned types of triangulation, we propose to apply a new type of triangulation, which is specific to MMCSs and triangulates findings across similar cases combined to groups instead of multiple sources per case. We propose that all reported findings have to be found in more than one case in a group of cases. This is also in line with previous methodological guidelines, which suggest that findings should only be reported if they have at least three confirmations (Stake 2013 ). To triangulate across multiple cases in one group, researchers have to identify multiple similar cases by applying a literal case replication logic to reinforce similar results. One should also apply a theoretical replication to compare different groups of literally replicated cases (i.e., searching for contrary results). Therefore, researchers have to justify their case replication logic . However, in our sample of MMCS, the majority (32 of 50, 64%) does not justify their replication logic, whereas the remaining publications use either literal replication (8 of 50, 16%), theoretical replication (6 of 50, 12%), or a combination (4 of 50, 8%). We encourage researchers to use a combination of literal and theoretical replication because it allows triangulation across different groups of cases. An exemplary MMCS that uses this approach is the publication of van de Weerd et al. ( 2016 ), who use theoretical replication to find cases with different outcomes (e.g., adoption and non-adoption) and use literal replication to find cases with similar characteristics and form groups of them.

Two further methodological steps, which are not exclusive to MMCS but recommended for increasing the construct validity, are creating a chain of evidence and letting key informants review a draft of the case study report . Only 36% (18 out of 50) of the analysed MMCS publications establish a chain of evidence. One reason for this lower usage may be that the majority (35 out of 50, 70%) of the publications analysed are conference proceedings. While we understand that these publications face space limitations, we note that no publication offers a supplementary appendix with in-depth insights. However, we encourage researchers to create a full chain of evidence with as much transparency as possible. Therefore, online directories for supplementary appendices could be a valuable addition. As opposed to a few years ago, these repositories today are widely available and using them for such purposes could become a good research practice for qualitative research. Interestingly, only 16% (8 of 50) analysed MMCS publications let key informants review the draft of the case study report . As MMCSs only have one source of evidence per case, misinterpretations and subjective judgement by the researcher have a significantly higher impact on the results compared to multiple case study research. Therefore, MMCS researchers should let key informants review the case study report before publishing.

MMCSs only have few (one) sources of evidence per case, so the risk of focusing on spurious relationships is higher, threatening internal validity (Dubé and Paré 2003 ). This threat to internal validity must be addressed in the data analysis phase. In the context of MMCSs, researchers may aggregate fewer data points to obtain a within-case overview. Therefore, having a clear perspective of the existing data points and rigorously applying the within-case analysis methodological steps (e.g., pattern matching) is even more critical. However, due to the limited depth of data at MMCSs, the within-case analysis must be combined with an analysis across groups of cases (to allow triangulation via multiple groups of cases). For MMCSs, we propose not doing the cross-case analysis on a per-case basis. Instead, we propose to build groups of similar cases across which researchers could conduct an analysis across groups of cases. This solidifies internal validity in case study research (Eisenhardt 1989 ) by viewing and synthesising insights from multiple perspectives (Paré 2004 ; Yin 2018 ).

Another risk of MMCSs is the relatively high number of cases (i.e., we found up to 27 for exploratory MMCSs) that is higher than Eisenhardt’s ( 1989 ) recommendation of maximal ten cases in multiple case study research. With more than ten in-depth cases, researchers struggled to manage the complexity and data volume, resulting in models with low generalisability and reduced external validity (Eisenhardt 1989 ). We propose to use two methodological steps to address the threat to external validity.

First, like Yin’s ( 2018 ) recommendation to use theory for single case studies, we suggest an a priori theoretical framework for MMCSs. 84% (42 out of 50) of the analysed MMCS publications use such a framework. An a priori theoretical framework has two advantages: it simplifies research by pre-defining constructs and relationships, and it enables analytical techniques like pattern matching. Second, instead of doing the within and then cross-case analysis on a per-case basis, for MMCSs, we propose first doing the within-case analysis and then forming groups of similar cases. Then, the cross-case analysis is performed on the formed groups of cases. To form case groups, replication logic (literal and theoretical) must be chosen carefully. Cross-group analysis (with at least two cases per group) can increase the generalisability of results.

To increase MMCS reliability, a case study database and protocol should be created, similar to multiple case studies. To ensure higher reliability, researchers should document MMCS design decisions in more detail. As outlined in the results section, the documentation on why design decisions were taken is often relatively short and should be more detailed. This call for better documentation is not exclusive to MMCSs, as Benbasat et al. ( 1987 ) and Dubé and Paré ( 2003 ) also criticised this for multiple case study research.To ensure rigour in MMCS, we suggest following the steps for multiple case study research. However, MMCSs have unique characteristics, such as an inability to source triangulate on a per-case level, a higher risk of marginal cases, and difficulty in managing a high number of cases. Therefore, for some methodological steps (cf. Table 8 ), we propose MMCS-specific methodological options. First, MMCS should include supplementary data per case (to increase construct validity). Second, instead of doing a cross-case analysis, we propose to form groups of similar cases and focus on the cross-group analysis (i.e., in each group, there must be at least two cases). Third, researchers should justify their case replication logic , i.e., a combination of theoretical replication (to form different groups) and literal replication (to find the same cases within groups) should be conducted to allow for this cross-group analysis.

6 Conclusion

Our publication contributes to case study research in the IS discipline and beyond by making four methodological contributions. First, we provide a conceptual definition of MMCSs and distinguish them from other research approaches. Second, we provide a contemporary collection of exemplary MMCS publications and their methodological choices. Third, we outline methodological guidelines for rigorous MMCS research and provide examples of good practice. Fourth, we identify research situations for which MMCSs can be used as a pragmatic and rigorous approach.

Our findings have three implications for research practice: First, we found that MMCSs can be descriptive, exploratory, or explanatory and can be considered as a type of multiple case study. Our set of IS MMCS publications shows that this pragmatic approach is advantageous in three situations. First, for time-sensitive topics, where rapid discussion of results, especially in the early stages of research, is beneficial. Second, when it is difficult to collect comprehensive data from multiple sources for each case, either because of limited availability or limited accessibility to the data source. Third, in situations where the research setting is complex, many cases are needed to validate effects (e.g., combinatorial effects) or previous research has produced conflicting results. It is important, however, that the pragmatism of the MMCS should not be misunderstood as a lack of methodological rigour.

Second, we have provided guidelines that researchers can follow to conduct MMCSs rigorously. As we observe an increasing number of MMCSs being published, we encourage their authors to clarify their methodological approach by referring to our analytical MMCS framework. Our analytical framework helps researchers to justify their approach and to distinguish it from approaches that lack methodological rigour.

Third, throughout our collection of MMCS publications, we contacted several authors to clarify their case study research methodology. In many cases, these publications lacked critical details that would be important to classify them as MMCS or marginal cases. Many researchers responded that some details were not mentioned due to space limitations. While we understand these constraints, we suggest that researchers still present these details, for example, by considering online appendices in research repositories.

Our paper has five limitations that could be addressed by future research. First, we focus exclusively on methodological guidelines for positivist multiple case study research. Therefore, we have not explicitly covered methodological approaches from other research paradigms.

Second, we aggregated methodological guidance on multiple case study research from the most relevant publications by citation count only. As a result, we did not capture evidence from publications with far fewer citations or that are relevant in specific niches. However, our design choice is still justified as the aim was to identify established and widely accepted methodological strategies to ensure rigour in case study research.

Third, the literature reviews were keyword-based. Therefore, concepts that fall within our understanding of MMCS but do not include the keywords used for the literature search could not be identified. However, due to the different search terms and versatile search approaches, our search should have captured the most relevant contributions.

Fourth, we selected publications from highly ranked IS MMCS publications and proceedings of leading IS conferences to analyse how rigour is ensured in MMCSs in the IS discipline. We therefore excluded all other research outlets. As with the limitations arising from the keyword-based search, we may have omitted IS MMCS publications that refer to short or mini case studies. However, the limitation of our search is justified as it helps us to ensure that all selected publications have undergone a substantial peer review process and qualify as a reference base in IS.

Fifth, we coded our variables based on the characteristics explicitly stated in the manuscript (i.e., if authors position their MMCS as exploratory, we coded it as exploratory). However, for some variables, researchers do not have a consistent understanding (e.g., the discussion of what constitutes exploratory research by cf., Sarker et al. ( 2018 )). Therefore, we took the risk that MMCS may have different understandings of the coded variables.

For the future, our manuscript on positivist MMCSs provides researchers with guidance for an emerging type of case study research. Based on our study, we can identify promising areas for future research. By limiting ourselves to the most established strategies for ensuring rigour, we also invite authors to enrich our methodological guidelines with other, less commonly used steps. In addition, future research could compare the use of MMCSs in IS with other disciplines in order to solidify our findings.

Data availability

Provided at https://doi.org/10.6084/m9.figshare.24916458

The information can be found in the online Appendix: https://doi.org/10.6084/m9.figshare.24916458 .

litbaskets.io is a web interface that allows searching for literature across the top 847 IS journals. It offers ranging from 2XS (Basket of Eight) to 3XL (847) essential IS journals and a full list of 29 journals which are the basis for this study can be found in Appendix C ( https://doi.org/10.6084/m9.figshare.24916458 ).

Atzmüller C, Steiner PM (2010) Experimental vignette studies in survey research. Method Eur J Res Methods Behav Soc Sci. https://doi.org/10.1027/1614-2241/a000014

Article   Google Scholar  

Baker EW, Niederman F (2014) Integrating the IS functions after mergers and acquisitions: analyzing business-IT alignment. J Strateg Inf Syst 23(2):112–127. https://doi.org/10.1016/j.jsis.2013.08.002

Benbasat I, Goldstein DK, Mead M (1987) The case research strategy in studies of information systems. MIS Q 11(3):369–386. https://doi.org/10.2307/248684

Boell S, Wang B (2019) www.litbaskets.io , an IT artifact supporting exploratory literature searches for information systems research. In: Proceedings ACIS 2019

Bremser CP, Gunther Rothlauf F (2017) Strategies and influencing factors for big data exploration. In: proceedings AMCIS 2017

Vom Brocke J, Simons A, Niehaves B, Riemer K, Plattfaut R, Cleven A (2009) Reconstructing the giant: on the importance of rigour in documenting the literature search process. In: Proceedings ECIS 2009

Creswell JW, Poth CN (2016) Qualitative inquiry and research design: choosing among five approaches, 4th edn. Sage Publications, California

Google Scholar  

Demlehner Q, Laumer S (2020) Shall we use it or not? Explaining the adoption of artificial intelligence for car manufacturing purposes. In: Proceedings ECIS 2020

Dubé L, Paré G (2003) Rigor in information systems positivist case research: current practices, trends, and recommendations. MIS Q 27(4):597–636. https://doi.org/10.2307/30036550

Dubois A, Gadde L-E (2002) Systematic combining: an abductive approach to case research. J Bus Res 55(7):553–560. https://doi.org/10.1016/S0148-2963(00)00195-8

Eisenhardt KM (1989) Building theories from case study research. Acad Manag Rev 14(4):532–550. https://doi.org/10.2307/258557

Eisenhardt KM, Graebner ME (2007) Theory building from cases: opportunities and challenges. Acad Manag J 50(1):25–32. https://doi.org/10.5465/amj.2007.24160888

Gibbert M, Ruigrok W, Wicki B (2008) What passes as a rigorous case study? Strateg Manag J 29(13):1465–1474. https://doi.org/10.1002/smj.722

Gogan JL, McLaughlin MD, Thomas D (2014) Critical incident technique in the basket. In: Proceedings ICIS 2014

Hahn C, Röher D, Zarnekow R (2015) A value proposition oriented typology of electronic marketplaces for B2B SaaS applications. In: Proceedings AMCIS 2015

Hanelt A, Hildebrandt B, Polier J (2015) Uncovering the role of IS in business model innovation: a taxonomy-driven approach to structure the field. In: Proceedings ECIS 2015

Hughes R, Huby M (2002) The application of vignettes in social and nursing research. J Adv Nurs 37(4):382–386. https://doi.org/10.1046/j.1365-2648.2002.02100.x

Karunagaran S, Mathew S, Lehner F (2016) Differential adoption of cloud technology: a multiple case study of large firms and SMEs. In: Proceedings ICIS 2016

Keutel M, Michalik B, Richter J (2014) Towards mindful case study research in IS: a critical analysis of the past ten years. Eur J Inf Syst 23(3):256–272. https://doi.org/10.1057/ejis.2013.26

Klein HK, Myers MD (1999) A set of principles for conducting and evaluating interpretive field studies in information systems. MIS Q 23(1):67–93. https://doi.org/10.2307/249410

Klotz S, Kratzer S, Westner M, Strahringer S (2022) Literary sketches in information systems research: conceptualization and guidance for using vignettes as a narrative form. Inf Syst Manag. https://doi.org/10.1080/10580530.2021.1996661

Kunduru SR, Bandi RK (2019) Fluidity of power structures underpinning public discourse on social media: a multi-case study on twitter discourse in India. In: Proceedings AMCIS 2019

Kurnia S, Karnali RJ, Rahim MM (2015) A qualitative study of business-to-business electronic commerce adoption within the indonesian grocery industry: a multi-theory perspective. Inf Manag 52(4):518–536. https://doi.org/10.1016/j.im.2015.03.003

Lamnek S, Krell C (2010) Qualitative sozialforschung: mit online-materialien, 6th edn. Beltz Verlangsgruppe, Germany

Lee AS, Hubona GS (2009) A scientific basis for rigor in information systems research. MIS Q 33(2):237–262. https://doi.org/10.2307/20650291

Mandrella M, Zander S, Trang S (2016) How different types of IS assets account for synergy-enabled value in multi-unit firms: mapping of critical success factors and key performance indicators. In: Proceedings AMCIS 2016

Marshall C, Rossman GB (2016) Designing qualitative research, 6th edn. SAGE Publications, Inc., California

McBride N (2009) Exploring service issues within the IT organisation: four mini-case studies. Int J Inf Manag 29(3):237–242. https://doi.org/10.1016/j.ijinfomgt.2008.11.010

Meredith J (1998) Building operations management theory through case and field research. J Oper Manag 16:441–454. https://doi.org/10.1016/S0272-6963(98)00023-0

Nili A, Tate M, Barros A, Johnstone D (2020) An approach for selecting and using a method of inter-coder reliability in information management research. Int J Inf Manage 54:102154. https://doi.org/10.1016/j.ijinfomgt.2020.102154

Orlikowski WJ, Baroudi JJ (1991) Studying information technology in organizations: research approaches and assumptions. Inf Syst Res 2(1):1–28

Palvia P, Daneshvar Kakhki M, Ghoshal T, Uppala V, Wang W (2015) Methodological and topic trends in information systems research: a meta-analysis of IS journals. Commun Assoc Inf Syst 37(1):30. https://doi.org/10.17705/1CAIS.03730

Pan SL, Tan B (2011) Demystifying case research: a structured–pragmatic–situational (SPS) approach to conducting case studies. Inf Organ 21(3):161–176. https://doi.org/10.1016/j.infoandorg.2011.07.001

Paré G (2004) Investigating information systems with positivist case research. Commun Assoc Inf Syst 13(1):18. https://doi.org/10.17705/1CAIS.01318

Piekkari R, Welch C, Paavilainen E (2009) The case study as disciplinary convention: evidence from international business journals. Organ Res Methods 12(3):567–589. https://doi.org/10.1177/109442810831990

Recker J (2012) Scientific research in information systems: a beginner’s guide. Springer, Berlin

Ridder H-G (2017) The theory contribution of case study research. Bus Res 10(2):281–305. https://doi.org/10.1007/s40685-017-0045-z

dos Santos Tavares AP, Fornazin M, Joia LA (2021) The good, the bad, and the ugly: digital transformation and the Covid-19 pandemic. In: Proceedings AMCIS 2021

Sarker S, Xiao X, Beaulieu T, Lee AS (2018) Learning from first-generation qualitative approaches in the IS discipline: an evolutionary view and some implications for authors and evaluators (PART 1/2). J Assoc Inf Syst 19(8):752–774. https://doi.org/10.17705/1jais.00508

Schäfferling A, Wagner H-T, Schulz M, Dum T (2011) The effect of knowledge management systems on absorptive capacity: findings from international law firms. In: Proceedings PACIS 2011

Sørensen C, Landau JS (2015) Academic agility in digital innovation research: the case of mobile ICT publications within information systems 2000–2014. J Strateg Inf Syst 24(3):158–170. https://doi.org/10.1016/j.jsis.2015.07.001

Spiegel F, Lazic M (2010) Incentive and control mechanisms for mitigating relational risk in IT outsourcing relationships. In: Proceedings AMCIS 2010

Stake RE (2013) Multiple case study analysis. The Guilford Press

Urquhart C (2001) Bridging information requirements and information needs assessment: Do scenarios and vignettes provide a link? Inf Res 6(2):6–2

van de Weerd I, Mangula IS, Brinkkemper S (2016) Adoption of software as a service in indonesia: examining the influence of organizational factors. Inf Manag 53(7):915–928. https://doi.org/10.1016/j.im.2016.05.008

Vom Brocke J, Simons A, Riemer K, Niehaves B, Plattfaut R, Cleven A (2015) Standing on the shoulders of giants: challenges and recommendations of literature search in information systems research. Commun Assoc Inf Syst 37(1):9. https://doi.org/10.17705/1CAIS.03709

Voss C, Tsikriktsis N, Frohlich M (2002) Case research in operations management. Int J Oper Prod Manag 22(2):195–219

Wagner H-T, Ettrich-Schmitt K (2009) Integrating value-adding mobile services into an emergency management system for tourist destinations. In: Proceedings ECIS 2009

Welch C, Piekkari R, Plakoyiannaki E. et al (2011) Theorising from case studies: Towards a pluralist future for international business research. J Int Bus Stud 42, 740–762. https://doi.org/10.1057/jibs.2010.55

Weill P, Olson MH (1989) Managing investment in information technology: mini case examples and implications. MIS Q 13(1):3–17. https://doi.org/10.2307/248694

Yin RK (2018) Case study research and applications: design and methods, 5th edn. Sage Publications, California

Zamani E, Pouloudi N (2020) Generative mechanisms of workarounds, discontinuance and reframing: a study of negative disconfirmation with consumerised IT. Inf Syst J 31(3):284–428. https://doi.org/10.1111/isj.12315

Download references

Open Access funding enabled and organized by Projekt DEAL. No funding was received for conducting this study.

Author information

Authors and affiliations.

Technische Universität Dresden, Dresden, Germany

Sebastian Käss, Christoph Brosig & Susanne Strahringer

OTH Regensburg, Seybothstr 2, 93053, Regensburg, Germany

Markus Westner

You can also search for this author in PubMed   Google Scholar

Contributions

All authors contributed to the study conception and design. Literature search and analyses were performed by the first two authors, and reviewed by the other two. All authors contributed to the interpretation and the discussion of the results. The first draft of the manuscript was written by the first two authors and all authors commented on the previous versions of the manuscript and critically revised the work. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Susanne Strahringer .

Ethics declarations

Conflict of interest.

The authors have no competing interests to declare that are relevant to the content of this study.

Ethical approval

Not Applicable, no human participants.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Käss, S., Brosig, C., Westner, M. et al. Short and sweet: multiple mini case studies as a form of rigorous case study research. Inf Syst E-Bus Manage (2024). https://doi.org/10.1007/s10257-024-00674-2

Download citation

Received : 24 January 2024

Accepted : 23 February 2024

Published : 15 May 2024

DOI : https://doi.org/10.1007/s10257-024-00674-2

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Case study research
• Multiple mini case study
• Short case study
• Methodological guidance
• Find a journal
• Publish with us
• Track your research

The times for multiples: Five situations when multiples need more than a second look

For more than 60 years, Bob Dylan’s “Don’t Think Twice, It’s All Right” has endured as a brilliant song and a terrible principle for valuation—especially when it comes to multiples. Managers and finance practitioners should always think twice about multiples. When multiples are used properly and the correct peer groups are selected, they can provide a quick estimation, serve as a reality check against a traditional discounted cash flow model, and slot into common shorthand. But investment or financial decisions should never be made based primarily, let alone solely, on multiples . 1 Marc Goedhart, Vartika Gupta, Peeyush Karnani, and Werner Rehm, “ The times for multiples: Why value creation always comes first ,” McKinsey, March 17, 2023.

In some cases, multiples can be particularly misleading. In this article, we’ll look at five of the most common situations where multiples can provide an incomplete or distorted picture.

1. Shocks to an industry or the broader market

It’s easiest to compare multiples when conditions hold steady over time: each 12- or 24-month earnings forecast is an iterative prediction of a company’s earnings over a very long horizon. But during short periods of turbulent change, the current or 12-month forward multiples can lead to nonsense results; the short term (informed by current, highly atypical performance) tells investors or managers little about what cash flow will be after broad turbulence has passed. Likewise, any future comparison between how a company performed during a crisis and how its consensus earnings appear after a crisis can be wildly inapt. We’d expect, for example, that some retailers would sell more at times when people were “just stocking up before the hoarders get here,” 2 Gokhan Dogan and John Sterman, “‘I’m not hoarding, I’m just stocking up before the hoarders get here.’: Behavioral causes of phantom ordering in supply chains,” Journal of Operations Management , 2015, Volumes 39–40. or that physical fitness chains would have lower earnings during pandemic lockdowns. But unless these occasions are regular and not one-offs, they tell us little to nothing about a company’s prospects to create value over the longer term.

Market reactions during the initial phase of the COVID-19 pandemic provide a case in point. Consider the multiples of one biotech company. In the months leading up to the global pandemic, the company’s forward EBITDA expectations were negative. When the worldwide health crisis became evident, however, the company shifted toward developing COVID-19 vaccines. Investors applauded, sending the company’s two-year 3 In this case, two fiscal years. enterprise value (EV) multiple to a dizzying 33-times EBITDA. Over the following months, however, as the analysts caught up and adjusted their forecasts, the company’s EV/EBITDA multiple fell and then flattened out to between four and seven times consensus earnings.

The COVID-19 pandemic’s economic effects, of course, were not limited to single companies; the crisis affected entire industries. For example, airlines’ earnings declined significantly (Exhibit 1). As a result, earnings multiples for all major airlines expanded. Why? Because EV incorporates cash flow during and well after a crisis has passed; when the pandemic first hit, the airlines’ earnings denominator shrank much more than its EV numerator. When travel resumed a little over a year after the pandemic, multiples quickly corrected to historical levels. Now, if one were to look back and consider an airline’s historical multiples during the first year or so of the pandemic, that period would stand out as an aberration. And conducting a multiples analysis during those fraught times invited massive error; minor differences in small earnings estimates led to a wide distribution of one-off multiples.

2. Periods of heavy investment

Multiples also get distorted when companies incur very large capital expenditures within a very short span. That may seem counterintuitive, since capital expenditures are not explicitly included in a typical multiple. However, consider the following example: Company A, a refinery, has just completed a major technological upgrade. Its competitor, Company B, will begin investing tomorrow and continue investing over the next two years for the same technology. Since the cost of the upgrade is known, all else equal, Company B’s observed net enterprise value (enterprise value before excess cash) will be lower by the present value of the upgrade cost. Consequently, the observed EV/EBITDA multiple will shrink, even though the two refineries will have the same capabilities to generate cash flows in the long term. The lower multiple doesn’t mean lower growth. But it would be impossible to know that unless one were to take a closer look.

The differences between current state and future state are particularly acute for green start-ups or high-tech companies. They can have very high growth expectations and mostly negative near-term earnings. Because of the challenge and expense of scaling up, multiples won’t be a useful way to compare early-stage companies with peers that are only a few years older. The best approach for assessing high-growth companies is to go back to discounted cash flow basics—“back from the potential future”—using probability-weighted scenarios to arrive at value today. 4 See Tim Koller, Marc Goedhart, and David Wessels, Valuation: Measuring and Managing the Value of Companies , Hoboken, NJ: John Wiley & Sons, 2015.

3. Cyclical companies

A typical commodity company, precisely because it is subject to cycles, will not have a stable EBITDA. Yet its enterprise value changes much less over time. That’s because sophisticated investors price in the cycles. As a result, we typically see marked shifts in forward multiples at different stages during the commodity cycle. For example, forward multiples at the top of a cycle tend to be high because of an expected downturn, even though little to no growth is expected. Consequently, a point-in-time multiple for cyclical companies can be very deceptive (Exhibit 2). To gain greater insight, we recommend using a through-cycle multiple or a two- to three-year forward multiple. Practitioners can also use multiples such as EV over installed capacity, EV per barrel (in the case of an oil and gas company), or EV over mine reserves (in the case of a mining company).

When a company announces an acquisition, its market value should reflect its own value plus any synergies, net of any premium being paid to the target—after pricing in the possibility that the deal will not close. However, even when a deal closes and uncertainty is effectively reduced to zero, 5 In rare situations, a merger or acquisition can be unwound after closing. forward-multiples will likely still not be adequately reflected in analysts’ forecasts. Many analysts do not even adjust numbers reported to typical data aggregators until after closing, which leads to a distorted “mechanical” multiple. This is especially the case for deals with a long time lag between announcement and closing.

Consider one large acquisition in the consumer-packaged-goods industry. At closing, its EV/EBITDA multiple sharply increased to approximately 20 times EBITDA. In the year before closing, the EV multiple had been 15 times EBITDA, and two years before closing, it had been 13 times. Soon after the deal closed, though, the 20-times multiple began to regress to prior levels. Analysts were now including both the initial earnings impact of the acquisition and a more informed perspective about expected synergies.

For deals in which the target’s earnings and historical financials are publicly available, we recommend always double-checking the multiple. This is relatively easy to do: combine the acquirer’s expected earnings with those of the target and make a rough forecast to start. You can then refine your analysis from there, based on changes to EBITDA that one could reasonably expect.

5. Changes in strategy or business model

Companies that announce a new growth or portfolio strategy that shifts their business mix typically see some change in market expectations. However, given the disconnect between current profitability and short-term expectations compared with where the company is headed over a longer time, enterprise value multiples may not fully represent the company’s long-term trajectory (Exhibit 3).

For example, from 2003 to 2005, one industrial company traded at an EV multiple of ten to 12 times two-year forward EBITDA. In 2006, the company began to expand into medtech and life sciences. Yet it took a full decade for the company to fully transition and its underlying multiples (16 to 20 times earnings) to reflect the profitability of its new business model. Since the market hadn’t fully grasped how the company was changing, it also failed to appreciate which companies were its peers. Indeed, any peer comparison based on its 2006 model would have been a poor predictor indeed. That’s why it is always best to use different multiples for each business segment—particularly when a corporation is changing its businesses.

A current manifestation of this challenge is in the energy sector. One traditional energy company, for example, has been aggressively transitioning its portfolio to include renewables. For the first several years of its transition, the company’s multiple was consistently in line with those of traditional energy companies, even though it determinedly pushed into alternative energy. Only after a decade of making the transition—when its renewable portfolio exceeded more than 20 percent of its business—did the market adjust the company’s multiples.

Market misses are particularly likely when a company moves more sharply away from its long-time core businesses. Because a new business model has different fundamentals, a point-in-time model (for example, EV to one-year forward earnings) won’t give a full picture of the corporation’s underlying economics for a longer term. The more a company changes its business mix, the weirder its multiples may appear—at least before disaggregating into segments. Even in the age of large databases and AI, practitioners need to look at company details, economic circumstances, and facts on the ground to correctly apply multiples. Common sense and a little legwork go a long way.

Used properly, multiples can be an effective supplemental tool. But traditional EV multiples can provide an incomplete picture or inapposite results during industry- or market-wide shocks, periods of heavy investment, single points in time over a longer commodity cycle, after a merger or acquisition has just closed, and when a company is making profound changes to its business portfolio. In those cases, it’s always best to think twice about multiples—and sometimes, more than that.

Peeyush Karnani is an associate partner in McKinsey’s New York office, and Werner Rehm is a partner in the New Jersey office.

The authors wish to thank Satvik Bansal, Marc Goedhart, and Avineet Sadani for their contributions to this article.

This article was edited by David Schwartz, an executive editor in the Tel Aviv office.

Explore a career with us

Related articles.

The times for multiples: Why value creation always comes first

Warren Buffett: An appreciation

Solving the paradox of growth and profitability in e-commerce