Background
Cost-of-illness (COI) studies can help identify, measure, and value the economic burden an illness or disease can impose on society (Reference Rice1). It is a useful decision-making tool as their estimates can be used as a foundation for projecting disease expenses and a framework to address a certain health problem, among others (Reference Costa, Derumeaux and Rapp2;Reference Tarricone3). COI studies are a commonly used tool to provide researchers and policy/decision makers with relevant information regarding the different cost components and cost categories (or sectors) associated with an illness or disease and can describe healthcare spending as well as costs beyond healthcare (e.g., intersectoral costs) (Reference Tarricone3).
In order to allow for COI studies to optimally inform researchers and policy/decision makers, these studies need to be methodologically sound (Reference Jo4;Reference Rice5). Various checklists and guidance tools exist for full economic evaluations including, for instance, the Drummond Methods for the Economic Evaluation of Health Care Programmes (Reference Drummond, Sculpher, Claxton, Stoddart and Torrance6), the Consensus on Health Economic Criteria checklist (CHEC-list) (Reference Evers, Goossens, de Vet, van Tulder and Ament7), and others. These checklists and guidelines play an important role in assessing the (methodological or reporting) quality of economic evaluations and are widely used. In comparison, there is an evident lack of guidance for COI studies and poor consensus on how to review and assess those studies and what tool(s) to use for critical appraisal (Reference Costa, Derumeaux and Rapp2;Reference Kleine‐Budde, Touil and Moock8–Reference Zhu, Tam and Li12).
To the best of our knowledge, there are only two tools that are specifically designed to assist in developing and assessing COI studies (Reference Larg and Moss13;Reference Müller, Stock and Dintsios14). Both tools require a deeper level of technical and methodological detail and are extensive in length. The issue of length is critical because a checklist is often expected to be rigorous but also practical to use. The objective of one of the two tools, the Checklist for the Development and Assessment of Cost-of-Illness Studies by Mueller et al., was to develop a checklist in German and specifically for the German context (Reference Müller, Stock and Dintsios14). The objective of the second tool, a Guide to Critical Evaluation by Larg and Moss, was to develop a guide for understanding and evaluating COI studies (Reference Larg and Moss13). However, it is unclear whether this guide was developed based on consensus and expert opinion.
Methodological approaches for COI studies can differ in a variety of aspects (e.g., objectives, study perspective, costs included, time horizon), giving rise to considerable methodological heterogeneity (Reference Céilleachair, Hanly and Skally15;Reference Molinier, Bauvin and Combescure16). This makes comparability across COI studies difficult and the assessment of the generalizability or transferability of study results almost impossible. Because of the lack of available tools to review and assess existing COI studies, researchers often develop their own one-off list of questions as part of their work (e.g., literature reviews).
An internationally applicable, standardized checklist is needed in English to review and critically appraise the methodological approaches taken and reported in a COI study, to assess a study’s comprehensiveness, transparency and consistency, to reflect on a study’s strengths and weaknesses, and to potentially increase comparability across COI studies.
Aims and Objectives
The aim of this paper was to develop a consensus-based checklist that can be used as a minimum standard to appraise the comprehensiveness, transparency, and consistency of COI studies. This is important when, for instance, reviewing and assessing COI studies for example as part of a systematic review or when building an economic model.
Methods
The development process of the consensus-based checklist involved six sequential steps, as presented in Figure 1. These steps were based on previous approaches to the development of other relevant checklists and guidelines in health economics and related areas (Reference Evers, Goossens, de Vet, van Tulder and Ament7;Reference Müller, Stock and Dintsios14;Reference Deidda, Geue, Kreif, Dundas and McIntosh17).
Scoping Review
A targeted scoping review of systematic reviews of COI studies was conducted in MEDLINE (Ovid) to explore the different checklists or other tools used. This was complemented by hand searching and searches in Google Scholar, checking the reference lists of included articles, and reviewing articles, studies, checklists, and guidelines suggested by experts working in health economics and with COI studies. A search strategy was developed in MEDLINE using keywords and terminology relating to cost-of-illness, burden of illness, economic burden, systematic review(s), and checklist(s), focusing on papers published in 2010 (Supplementary Table 1). The search strategy combined search terms using Boolean operators “AND” and “OR” and searched for these keywords and terms in a paper’s title, abstract, and beyond. This date was selected as a previous study had considered papers prior to this date, which were reviewed (Reference Müller, Stock and Dintsios14). Studies were included in the scoping review that reported on applying at least one checklist or a similar tool for the quality or critical appraisal of COI studies. The aim was to identify all checklists and tools used by researchers for the quality or critical appraisal of COI studies, even if these checklists were not specifically designed for COI studies.
Assessment and Comparison of the Different Checklists and Their Questions
The different checklists identified in the scoping review in Step 1 were listed, compared, and critically reviewed to determine whether they had been specifically designed for COI studies or were based on other existing health economic guidelines. This involved developing a matrix, charting all the questions and subquestions included in the identified checklists, to allow discussion and comparison by all authors. Due to the fact that not all checklists identified in Step 1 were specifically designed for COI studies (for example they may be for full economic evaluations), the questions (or criteria) included in these different checklists were carefully and critically reviewed in terms of their applicability and relevance for COI studies to identify and synthesize a set of key questions for assessing these studies. This meant that questions that were listed in existing checklists but were only applicable to full economic evaluations were excluded.
Development of a Checklist for COI Studies
This step involved synthesizing the output of the scoping review (Step 1) and the results of the critical assessment (Step 2) to determine key areas that would need to be included in a checklist for COI studies. This was further refined to develop a list of the key questions that would be relevant and applicable and that could be used as a minimum standard for the critical appraisal of COI studies. From here on this will be referred to as the ‘preliminary checklist’. This provided an initial outline for discussion with the experts engaged in COI studies in the next step.
Expert Interviews
Semi-structured, open-ended interviews were conducted with health economists and other experts from different countries working with COI studies to seek their expert opinion on the preliminary checklist and potentially identify questions to be added, removed, or revised. This process is fully described in a separate paper https://doi.org/10.1017/S0266462323000181. In this study, we use the term ‘experts’ to refer to individuals that are knowledgeable in a particular area, in this case in health economics/COI studies, and are/were actively involved in doing research around COI studies, including professors, assistant or associate professors, research fellows, among other. Experts were selected purposively based on their knowledge and expertise in relation to COI studies using network and snowball sampling. Interviews were audio-recorded, with the participant’s consent, and anonymized. A Framework approach was applied for the thematic analysis of the interviews, following systematic steps (Reference Gale, Heath, Cameron, Rashid and Redwood18): interview recordings were transcribed verbatim by one author (L.S.); transcripts were entered and coded in NVivo, identifying themes and subthemes (L.S.); a set of transcripts and the coding framework were cross-checked by another author (L.J.); both authors familiarized themselves with the transcripts and agreed on a final framework listing relevant themes and sub-themes; the framework and findings were discussed among the author team (A.P., L.J., L.S., S.E., T.R.); findings were reported narratively. More detailed information on the methodology, conduct, and analysis of the interviews is provided in a separate paper https://doi.org/10.1017/S0266462323000181.
Finalization of the Checklist
Experts’ feedback, suggestions, and recommendations on the preliminary checklist were carefully considered. The checklist was modified based on the experts’ feedback, removing certain questions, adding relevant questions, and rephrasing other questions, where applicable. The checklist was also presented at internal seminars in the Health Economics Unit at the University of Birmingham and at international health economics conferences, including the lolaHESG 2021 (The lowlands Health Economists’ Study Group) and the iHEA Conference 2021 (International Health Economics Association) to seek further feedback from experts in health economics. This step also involved the development of a list and description of answer categories suggested for use when answering the questions in the checklist, based on discussions with experts during the interviews and at the international conferences.
Development of Guidance Statements
Guidance statements were developed for each question listed in the checklist to provide further information on the purpose and meaning behind each question and to give an example of a best practice. These guidance statements were based on existing health economic guidelines and best practices, to align the language and terminology in the checklist with the existing economic literature (Reference Tarricone3;Reference Jo4;Reference Evers, Goossens, de Vet, van Tulder and Ament7;Reference Drummond and Jefferson19–Reference Chisholm, Stanciole, Edejer and Evans22).
Ethical Approval
Ethical approval to conduct this study was obtained from the University of Birmingham (ERN_20-1240).
Results
The result was a consensus-based checklist for the critical appraisal of COI studies covering relevant questions in relation to study characteristics, methodology and cost analysis, and results and reporting (Table 1). Guidance statements explaining the questions and suggested answer categories were also established (Table 2). This study further generated relevant interview findings that are summarized and presented in a separate paper https://doi.org/10.1017/S0266462323000181.
a Suggested answer categories: Yes, No, Partially, Not Applicable (NA), and Unclear. See Table 2 for further detail and guidance.
a Other international literature might also refer to the components of resource use as “resources” or “costs.”
Scoping Review
The scoping review of systematic reviews of COI studies published between 2010 and 2020 identified twenty-six studies that reported to have used a checklist or similar tool to assess COI studies. Six different checklists and guidelines were identified: the BMJ Checklist, the CHEC-list, the CHEERS checklist, the Drummond 10-point checklist, and the Drummond Methods for the Economic Evaluation of Health Care Programmes, and the Guide to Critical Evaluation by Larg and Moss (Reference Drummond, Sculpher, Claxton, Stoddart and Torrance6;Reference Evers, Goossens, de Vet, van Tulder and Ament7;Reference Larg and Moss13;Reference Drummond and Jefferson19;Reference Husereau, Drummond and Petrou23). (The Drummond 10-point checklist is adapted from the Drummond Methods, but we followed the study authors’ ways of reporting). A seventh tool was identified through handsearching, the Checklist for the Development and Assessment of Cost-of-Illness Studies by Mueller et al. (Supplementary Table 2) (Reference Müller, Stock and Dintsios14). This scoping review revealed that most of the studies predominantly applied quality or critical appraisal tools that are intended for the assessment of full economic evaluations to assess the quality of COI studies. For example, eight studies identified in the scoping review reported to have used (part of) the BMJ Checklist, five studies used the CHEERS Checklist, four studies used the Drummond Methods, three studies used the Drummond 10-point Checklist, and another three studies used the CHEC-list. Some studies reported to have used more than one checklist or other tool/source; in part to develop their own checklist based on existing tools or guidelines. Where checklists and guidelines for full economic evaluations were applied, many studies only adopted a subset of the questions included in the checklists or guidelines for full economic evaluations.
Only two of the identified tools are designed for the assessment of COI studies: the Guide to Critical Evaluation by Larg and Moss, and the Development and Assessment of Cost-of-Illness Studies by Mueller et al. Where these tools were applied, studies mostly modified the original checklist by removing some questions or changing the wording of questions.
Some of the studies reported to have developed their own ad hoc checklist for their study or systematic reviews (which simply had a one-off purpose) based on existing guidelines, previous studies, and/or health economic guidelines.
Assessment and Comparison of the Different Checklists and Their Questions
The matrix analysis charting all the questions and subquestions included in the identified checklists showed similarities in terms of key areas (dimensions) covered in the checklists such as study characteristics, the detailed methods that were used in the cost analysis, and how the study had been reported. There were some key differences between the checklists for full economic evaluations and the two tools specifically designed for COI studies. The latter two are more extensive and require the user/researcher to look at COI studies in more (technical) detail. The number of questions (criteria) in each checklist was recorded: the BMJ Checklist (n = 35), the CHEC-list (n = 19), the CHEERS checklist (n = 24, or n = 27 when including subquestions), the Drummond 10-point checklist (n = 10), and the Drummond Methods for the Economic Evaluation of Health Care Programmes (n = NA), the Guide to Critical Evaluation by Larg and Moss (n = 37), and the Checklist for the Development and Assessment of Cost-of-Illness Studies by Mueller et al. (n = 35) (Supplementary Table 2).
Development of a Checklist for COI Studies
Following the assessment of the questions and subquestions in Step 2, a list of key questions relevant and applicable to COI studies that would need to be included in a checklist was developed, and the CHEC-list was used as a foundation for further development. The CHEC-list was chosen as a foundation because a rigorous process had been followed to build the checklist. This process included literature searches, taking into consideration existing health economic checklists and criteria, and building consensus using Delphi methods involving a panel of international experts. In addition, the CHEC-list is concise and comprehensive in its format as well as manageable with a total of nineteen questions. This was considered an advantage as the aim of this study was to develop a checklist for COI studies that is concise but comprehensive and can be expanded, where needed. Due to the fact that the CHEC-list is intended for full economic evaluations, the author team reviewed all nineteen questions in terms of their applicability and relevance to COI studies. We excluded those questions that were only relevant for full economic evaluations (e.g., a description of competing alternatives; an identification, measurement, and valuation of relevant outcomes for each alternative; information on an incremental analysis of costs and outcomes). This resulted in a preliminary list of fourteen questions applicable to COI studies (Supplementary Table 3). Findings from a previous study comparing the original CHEC-list to two other checklists (the BMJ checklist and the Quality of Health Economic Studies (QHES) checklist) suggested that the original CHEC-list is missing a question assessing whether study limitations are specified (Reference Gerkens, Crott and Cleemput24). Hence, the author team added a question on study limitations to the preliminary checklist for COI studies, resulting in a total of fifteen questions for the preliminary checklist. The order and wording of the fourteen questions were kept almost identical to the original CHEC-list as health economists and other experts working with COI studies that were to be interviewed were likely to be familiar with the questions and terminology, and this was considered helpful for the interviews. The questions could (preliminarily) be divided into the following three dimensions: study characteristics; methodology and cost analysis; and results and reporting.
Expert Interviews
Between October 2020 and April 2021, 21 professionals (eleven male, ten female) from eleven different countries and with expertise in health economics (n = 17), economics, (n = 1), health policy (n = 2), and psychology (n = 1) participated in the interviews and provided feedback on the checklist. Experts were affiliated with academia, international policy organizations, governmental organizations, and consulting firms. More detailed information on the interview sample is provided in a separate paper https://doi.org/10.1017/S0266462323000181.
This study reached data saturation and consensus after those 21 interviews, finding similarities across those interview findings with little to no new findings emerging. Overall, experts were in favor of the checklist and expressed the urgent need for a checklist for COI studies. They suggested to remove, add, or rephrase some of the questions. Their feedback was considered and discussed carefully to finalize the checklist (before further presenting this checklist to experts at international health economic conferences). A more detailed analysis of the interview findings including relevant quotations is provided in a separate paper https://doi.org/10.1017/S0266462323000181.
Finalization of the Checklist
Following expert feedback and discussions with experts at international health economic conferences, the final version of the checklist was agreed upon. The final checklist comprised 17 main questions (and some additional subquestions) across three domains: study characteristics; methodology and cost analysis; and results and reporting (Table 1). These domains are briefly described below.
Domain 1 – Study characteristics: This dimension aims to assist the user of this checklist in assessing whether a COI study formulated an objective (Item 1.), described the characteristics of the study population (Item 2.), and is explicit about the perspective chosen for the cost analysis (Item 3.).
Domain 2 – Methodology and cost analysis: This dimension aims to assist the user of this checklist in assessing whether a COI study reported their choice for their epidemiological approach (Item 4.), costing approach (Item 5.), and data collection approach (Item 6.) as well as whether it stated which resources their study identified (Item 7.), measured (Item 8.), and valued (Item 9.). It also guides the user in assessing whether a COI study stated their time horizon for analysis (Item 10.), reported whether they discounted future costs (Item 11.) and conducted sensitivity analysis/analyses (Item 12.).
Domain 3 – Results and reporting: This dimension aims to assist the user of this checklist in assessing whether a COI study presented their study results by cost category/sector (depending on their study perspective) (Item 13.) and discussed the generalizability of study results (Item 14.), study limitations (Item 15.), and ethical and/or distributional issues(Item 16.). It also asks whether the study reported any conflict of interest (Item 17.).
Application of the Checklist
Based on the consensus built throughout the interviews and at international conferences, the following answer categories are suggested to be applied when answering the questions in the checklist: Yes, Partially, No, Not Applicable (NA), or Unclear. The checklist contains one column listing all questions and subquestions, one column to note down the answer, and one column to add Supporting Information. Users are advised to extract relevant information from COI studies when answering the questions to support or justify their answer narratively and to increase accountability. Further detail on the above answer categories is given in the guidance (Table 2), and the reasons for choosing intermediate categories are summarized and published in a separate paper https://doi.org/10.1017/S0266462323000181.
It is considered sufficient if one reviewer completes the checklist, assuming that they use the data extraction column to add information that justifies their answer. It is recommended to seek out to a second reviewer where information is not clear, and discrepancies need to be discussed. This checklist does not require the user to add scores to their answers or produce a total score for each study and a ranking of studies by score. Where desired or needed, it is however open to and possible for users to add scores to their answers (e.g., yes = 1, partially = 0.5, no = 0). When answering the questions as suggested in the guidelines, users will be able to identify the number of yeses or nos, which could give them an idea of the comprehensiveness of the study and an opportunity to reflect upon the reasons for a certain indication.
Where needed, the checklist can be modified and/or expanded, but it is suggested to clearly report any modifications or expansions to maintain consistent use.
Development of Guidance Statements
Additional guidance was developed describing the purpose and meaning behind each question and listing examples of best practice, see Table 2.
Discussion
The aim of this study was to develop a consensus-based checklist that can be used as a minimum standard to appraise the comprehensiveness, transparency, and consistency of COI studies. This study is the first to establish a checklist in English to review and assess the methodological approaches taken and reported in COI studies. The checklist was developed with the engagement of international experts from relevant backgrounds such as health economics, health policy, and psychology who had conducted or provided guidance for COI studies. The checklist is based on existing checklists and guidelines for health economic studies, expert qualitative interviews, and feedback from stakeholders at international health economic conferences. It is a pragmatic, generic, concise, and comprehensive tool that can be applied in several scenarios and can be considered a minimum standard, for instance, when reviewing and assessing COI studies for example as part of a systematic review or when building an economic model.
Additionally, this study addresses the inconsistency in the use of checklists and guidelines to appraise COI studies and provides further evidence that there is an urgent need for a standardized checklist to review and assess COI studies. It is the first study to fill this gap and provide a tool that could be used by users/researchers more consistently and internationally.
Comparison to Other Checklists
The checklists identified in the scoping review of this study cover similar questions and show methodological parallels. Those checklists, in particular the CHEC-list, provided a starting point to the development of a consensus-based checklist for COI studies available in English. The Guide to Critical Evaluation by Larg and Moss and the Checklist for the Development and Assessment of Cost-of-Illness Studies by Mueller et al. are designed for COI studies but may require a deeper level of technical and methodological detail and are extensive in length. It was unclear as to how the guide by Larg and Moss had been developed and whether it is based on expert opinion and consensus (Reference Aumann, Kreis, Zeidler and von der Schulenburg25). The checklist by Mueller et al. was developed using expert opinion but it was first and foremost established for the German context and is officially only published in German (Reference Müller, Stock and Dintsios14). Both tools have been taken into consideration for the development of the present consensus-based checklist for COI studies.
Strengths and Weaknesses of the Study
This study established and followed an extensive, structured, and iterative approach to the development of the checklist, involving literature searches, expert interviews, and further discussions among experts at international health economic conferences. The involvement of twenty-one international experts working in health economics, health policy, and psychology, and users as well as developers of existing quality or critical appraisal tools is one of the key strengths of this study. The checklist is a generic tool that can be applied to different disease areas and scenarios. A potential weakness of this study is that the checklist has not yet been formally pilot-tested. The checklist has, however, been applied by staff and students at Maastricht University and the University of Birmingham who have provided constructive feedback. The author team will initiate further piloting and testing of the checklist across different disease areas, and potentially refine its criteria, where relevant. This will be undertaken by students and researchers (initially by the authors and their research groups) using this tool for reviewing and assessing COI studies as part of future systematic reviews. It is planned that the checklist will be published on a university Web site (Maastricht University) alongside other quality and critical appraisal tools. This will allow us to monitor use of the checklist and to provide details on correspondence, in order to collect feedback from a wider range of users (outside of our research groups). Another limitation of this study is that due to the iterative process of development, the feedback collected during the interviews and the feedback collected at conferences were merged for analysis, making it difficult to compare the changes made based on the interviews and those made due to conference discussions. Further, the use of the CHEC-list as a starting point may be a limitation due to the CHEC-list being developed for full economic evaluations. However, other tools and checklists including the guide by Larg and Moss and the checklist by Mueller et al. were carefully considered during the development of the new checklist.
Conclusion
There is currently no standard checklist for the critical appraisal of COI studies and, as a result, the use of checklists for COI studies is inconsistent and heterogeneous. The consensus-based checklist for COI studies is a first step toward standardizing the critical appraisal of COI studies and is one that could be considered a minimum standard. The consensus-based checklist can help to improve comprehensiveness, transparency, and transparency in COI studies, to address heterogeneity and to enable better comparability of methodological approaches across international studies.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0266462323000193.
Acknowledgments
We would like to thank the following interview participants for their interest, time, and constructive feedback on the checklist: Alexander Konnopka, Brendan Kennelly, Brian Chun Fai Chan, Dan Chisholm, Diarmuid Coughlan, Dirk Mueller, Feng Xie, Isabelle Durand-Zaleski, Jared Streatfeild, Joel Segel, Luke Vale, Lynne Pezzullo, Manuela Deidda, Marcia Ward, Mark Monahan, Nichola Naylor, Rana Rizk, Rosanna Tarricone, and Sarah Byford. Two participants are not listed by name to safeguard their anonymity. We would also like to thank the students at Maastricht University and the University of Birmingham for pilot testing the consensus-based checklist and providing us with their feedback.
Author contribution
All authors made substantial contributions to the conception and design of the work. L.S. led the project, analysis, and writing of this paper. All authors helped with the selection of interview participants for Step 4. L.S. performed the data collection including the conduct of the interviews. All authors contributed to the analysis and interpretation of data and provided relevant input to the final checklist. All authors contributed to and have read and approved the submitted version of the manuscript. L.J. and L.S. jointly submitted the manuscript.
Funding statement
L.S. is supported by a PhD Studentship, which is funded jointly by the University of Birmingham and Maastricht University. This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Competing interest
The authors declare none.