Articles excluded due to language restriction and inaccessible detailed results were analyzed further to avoid exclusion of relevant studies. Discrepancies between the two reviewers were resolved through a discussion. When the discrepancies could not be resolved, a third reviewer was involved. The quality assessment results were analyzed qualitatively. The assessment items that were associated with the structural approaches of the model were highlighted.
Data extraction was performed on items related to the following topics: 1. TB progression modeling. The data extraction was performed by one reviewer TIAP. However, the results were discussed with other authors, mainly authors who performed quality assessment on the studies AMH or GWF.
It also included the settings in which the analysis was performed. A pilot study performed following the development of the data extraction forms, tested the applicability of the forms. Its result was discussed among authors and necessary changes, as well as additional fields were incorporated into the forms. The extracted data were analyzed qualitatively.
The cost-effectiveness conclusions and main cost-effectiveness outcomes of similar studies i. The methodological approaches of the studies were assessed to confirm inconsistencies found in previous reviews. Their applicability in various settings and potential influence on the cost-effectiveness result were also analyzed. After removing duplications, the search in five electronic databases identified articles.
Of these, only 27 fulfilled the inclusion criteria and were included in the review. In total, studies were excluded. The number of studies excluded for each exclusion criterion is given in Table 1.
Sub-Saharan Africa. The included studies evaluated the cost-effectiveness of various active pulmonary TB diagnosis tools and strategies. Two studies evaluated the negative impact of serology based testing for active disease diagnosis in high burden settings. Almost all studies concluded that novel diagnosis tools or strategies were cost effective compared to the best available diagnosis, as depicted in Fig 2.
Exceptions were found in three studies assessing NAAT e. Xpert and Mycobacterium Tuberculosis Direct Test or MTD in low-burden settings, [ 33 , 39 , 45 ] two studies evaluating serology based testing in high burden setting, [ 36 , 42 ] and one study investigating TB diagnosis in private and public health sector. NAAT was cost effective when applied selectively on smear positive cases due to its high sensitivity and the high rate of true positive in this group.
This can be seen in Figs 3 and 4 , which depicts the ICER of various diagnosis strategies from studies utilizing generic health outcomes measurement, i. The adjusted cost-effectiveness outcome for all included studies can be found in S3 Table.
Two studies which evaluated the cots-effectiveness of implementing Xpert as an initial test for all TB cases in South Africa exemplified the variations of cost-effectiveness outcome.
As detailed in Table 2 , the included studies utilized various methodological approaches. Around half of the studies failed to report structural assumptions and their justification.
A complete quality assessment of the studies can be found in S1 Table. The static model framework, i. As detailed in Fig 5 , studies generally included important setting characteristics, i. The inclusion of these characteristics reflected the disease burden in the study settings.
India and South Africa. Studies using a dynamic model framework showed few structural assumptions disparities. In contrast, the 24 static models showed extensive structural assumption disparities.
These included different assumptions regarding treatment outcomes, clinical diagnosis and empirical treatment, inpatient discharge decision, and re-diagnosis of false negative patients. The disparities are detailed in Table 3. However, it could also have been influenced by modeling practice inconsistencies.
As reported in one study, modeling practice inconsistencies, such as utilization of different methodological and structural approaches, were strongly related to the wide range of results. The included studies were also afflicted by quality issues pertaining to structural approaches identified by a previous review, such as failure to address structural uncertainty. As in previous reviews, [ 6 , 10 ] a static model framework, which omitted TB transmission process, was the preferred options among the included studies.
This is understandable, since static models development is relatively straightforward and can be easily understood by non-modelers, including decision makers. In contrast, dynamic models require complicated mathematical computation. Furthermore, their development is challenged by the difficulties in modeling the transmission process itself [ 73 ] especially when data and knowledge regarding the transmission process is limited. However, the use of static models to address infectious disease intervention strategies may potentially underestimate the indirect effect of the strategies in preventing secondary cases.
The argument should be considered carefully, since the underestimation may not only affect health benefit but also the demand of intervention and its related cost. The use of a dynamic model framework is recommended for investigating an infectious disease management strategy that influences the disease transmission process. They may also influence transmission process due to their operational aspect, such as a rapid process to obtain diagnosis result.
The rapid process causes shorter delay in starting treatment, which consequently causes shorter period of infectiousness. The ongoing TB transmission process in high burden TB settings is an important factor that might hinder countries progressing toward TB control goals. A dynamic model framework may also be useful in facilitating long term analysis. Studies utilizing a dynamic model framework in a setting with a high prevalence of TB and HIV comorbidity could identify budget increase in the long term 10 to 20 years , not only for TB management, but also for Anti-Retroviral Treatment ART , due to the higher survival rate of patients.
HIV comorbidity and MDR TB were included in the model governed by underlying assumptions driven by currently limited knowledge and data. Furthermore, these characteristics were excluded in several relevant settings. This exclusion potentially influences study result, e.
An evidence of this influence was shown by a recent multi-model study. Other important setting characteristics include various key drivers of TB epidemics and health system characteristics, such as domination of sub-standard private sector TB care. One example of this limitation is the uncertainty regarding the impact of shifting patients from private sector to the high quality public sector care. Other disparities found concerned the structural assumptions, which were more prevalent among the static models.
Assuming a successful treatment outcome for all correct diagnosis was applied by several studies. This assumption would not apply in settings with significant numbers of unsuccessful treatment outcome such as treatment loss-to-follow-up. Assuming only successful treatment outcome in such settings may lead to an overestimation of the cost-effectiveness.
This was proven by an empirical study in South Africa which showed that extensive loss to follow up mitigated the benefit of an accurate diagnosis by a novel tool. Another structural assumption disparity concerned the practice of clinical diagnosis, consisting of additional diagnosis, such as Chest X-ray, followed by an empirical TB treatment.
It is usually performed in highly suspected TB cases which obtain negative results for the main diagnosis tests such as smear microscopy. As argued by one study, the effectiveness of novel diagnosis tools could have been overestimated by modeling studies when clinical diagnosis practice was underestimated.
When a rigorous clinical diagnosis practice is applied, as found in several high burden settings, [ 82 , 83 ] most TB cases will be diagnosed and treated under the existing diagnosis practice. Consequently, this will minimize the number of additional cases diagnosed by the novel diagnosis tools and lower their effectiveness.
Another study argued that assuming clinical diagnosis practice to be unchanged by the availability of novel diagnostic tools with better accuracy could underestimate the effectiveness of the novel tools. Unfortunately, the influences of novel diagnosis tools on clinical diagnosis practice have not been fully understood. In settings where TB is managed in an inpatients-setting, the assumption regarding decision to discharge from hospital or isolation can also be influential.
Disparities were also found in assumptions around the false negative re-diagnosis process. In high burden settings, this may be challenged by several circumstances such as long distances between home and the health center. Despite its important findings, this review could have overlooked some relevant studies, such as those published in languages other than English and those with inaccessible detailed result.
However, additional analysis performed to studies excluded based on these two criteria did not result in additional inclusion of studies. The complete description of the additional analysis can be found in S1 File.
A meta-analysis could be performed on the cost and health benefit estimates of the reviewed studies to produce a single estimate of ICER. However, the single estimate might be futile since it is considered non-transferable to any setting.
This transferability issue has resulted in the absence of a recommended meta-analysis method for cost-effectiveness analyses results. Bias could be introduced during the screening process, which was done by only one author. However, this risk of bias was reduced by performing validation. Bias could also be introduced during data extraction, since it was also conducted by one author.
It was, however, reduced through extensive discussions among the co-authors; especially with those who also analyzed the studies for quality assessment. This review included studies conducted in multiple settings and covered a variety of diagnostic tools and strategies.
It also managed to confirm past systematic review findings. Thus, inclusion of more studies in this review would most likely substantiate current findings. Indication of bias was shown by the extensive disparities of cost-effectiveness result among studies with similar diagnostic strategies and settings.
This bias could be caused by methodological inconsistencies, including structural approach disparities. Another source of bias, which is not discussed in this review, is input parameters. Input parameters may have reliability issue. For example, several settings in Southeast Asia showed a high burden of TB, but population screening found that actual TB prevalence was significantly higher than what had been estimated through TB program case notification.
The structural approaches of the models were often reported inadequately. This practice could cause difficulties for decision makers when assessing the suitability of an approach to their setting, which emphasizes the need to report structural approaches and their justification transparently.
This can be done following a recent guidance on model validation reporting. Modeling practice inconsistencies, found in the past and in the current review, highlight the need for a clear standard for study conduct and reporting, especially in high burden TB settings, which mostly consists of Low-Middle Income Countries.
Moreover, the standard should be enforced by important stakeholders, such as shown by The Bill and Melinda Gates Foundation, which developed a case-reference containing general methodologic specifications and reporting standards. This systematic review identified extensive structural approach disparities in model-based cost-effectiveness analyses addressing TB diagnosis strategies.
It shows that several structural approaches could be inapplicable in certain settings. Furthermore, certain approaches could potentially contribute to under- or overestimation of the cost-effectiveness of a diagnosis tool or strategy. Eventually, they will lead to ambiguities and difficulties when interpreting the study result. We would like to thank Prof. Mirjam Kretzschmar for the comments and input on a draft of this article, as well as Linda McPhee for the input on the redaction of the article.
Browse Subject Areas? Click through the PLOS taxonomy to find articles in your field. Abstract Background Structural approach disparities were minimally addressed in past systematic reviews of model-based cost-effectiveness analyses addressing Tuberculosis management strategies.
Methods A systematic search to identify studies published before October was performed in five electronic databases. Results A total of 27 studies were included in the review.
Conclusion In cost-effectiveness analysis studies addressing active pulmonary Tuberculosis diagnosis, models showed numerous disparities in their structural approaches. Methods A systematic literature review was performed to identify model-based cost-effectiveness analyses addressing active pulmonary TB diagnosis, which were published before October Download: PPT. Table 1. Exclusion criteria and the number of excluded studies per criteria.
Data extraction and analysis Data extraction was performed on items related to the following topics: 1. Results After removing duplications, the search in five electronic databases identified articles.
Fig 2. Cost-effectiveness conclusion of novel diagnosis strategies. Table 2. Methodological variations and quality attributes found in the included studies. Fig 5. Table 3. Variations of structural assumptions in static models. Table 4. Recommendation of approaches to manage issues related to structural disparities.
Conclusion This systematic review identified extensive structural approach disparities in model-based cost-effectiveness analyses addressing TB diagnosis strategies. Supporting information. S1 File. Details of the search and screening strategy, and quality assessment approach. S2 File.
S1 Table. Quality assessment of the included studies. S2 Table. Data extraction for general information. S3 Table. Data extraction for main outcomes.
S4 Table. Data extraction of tuberculosis progression modeling approaches. Acknowledgments We would like to thank Prof.
References 1. The WHO Global tuberculosis report—further to go. Lancet Glob Health. N Engl J Med. Johannesson M, Weinstein MC. On the decision rules of cost-effectiveness analysis. J Health Econ. Karlsson G, Johannesson M. The decision rules of cost-effectiveness analysis. Optimal allocation of resources over health care programmes: dealing with decreasing marginal utility and uncertainty. Health Econ. Article Google Scholar. Briggs A, Gray A. Using cost effectiveness information.
Weinstein M, Zeckhauser R. Critical ratios and efficient allocation. J Public Econ. Birch S, Gafni A. Health Care Manag Sci. Cost-utility analysis from a societal perspective. Health Policy.
The role of value for money in public insurance coverage decisions for drugs in Australia: a retrospective analysis — Med Decis Making. Guide to the methods of technology appraisal Changes to NICE drug appraisals: what you need to know.
NICE; Institute for Clinical and Economic Review. Overview of the ICER assessment framework and update for — Updating cost-effectiveness. Looking back and moving forward: on the application of proportional shortfall in health priority setting in the Netherlands.
Five-hundred life-saving interventions and their cost-effectiveness. Risk Anal. Review of Australian health economic evaluation— interventions: what can we say about cost effectiveness? Cost Eff Resour Alloc. Ranking 93 health interventions for low- and middle-income countries by cost-effectiveness.
Standards for economic analyses of interventions for the field of health psychology and behavioral medicine. Health Psychol. Cost-effectiveness in health and medicine. Google Scholar. Recommendations for reporting cost-effectiveness analyses.
Panel on cost-effectiveness in health and medicine. Bracco A, Krol M. Economic evaluations in European reimbursement submission guidelines: current status and comparisons. Expert Rev Pharmacoecon Outcomes Res. Framework for systematic identification of ethical aspects of healthcare technologies: the SBU approach. Neyt M, Van Brabandt H. The importance of the comparator in economic evaluations: working on the efficiency frontier. How to select the best comparator?
An international economic evaluation guidelines comparison. Value Health. Incremental cost-effectiveness analysis: the optimal strategy depends on the strategy set. J Clin Epidemiol. Luce RD, Raiffa H. Games and decisions: introduction and critical survey. Hoboken: Wiley; Keeney RL, Raiffa H. Decisions with multiple objectives: Preferences and value tradeoffs.
Cambridge: Cambridge University Press; Arrow KJ. Risk perception in psychology and economics. Econ Inq. Kahneman D, Tversky A. Choices, values, and frames. Am Psychologist. Development of WHO guidelines on generalized cost-effectiveness analysis. Culyer AJ. Cost-effectiveness thresholds in health care: a bookshelf guide to their meaning and use. Health Econ Policy Law. Bach P. New math on drug cost-effectiveness. Tversky A. Elimination by aspects: a theory of choice.
Psychol Rev. Some guidelines on the use of cost effectiveness league tables. Tversky A, Simonson I. Context-dependent preferences. Manage Sci. Download references. You can also search for this author in PubMed Google Scholar. JAS generated the initial idea and wrote the first draft of the manuscript.
All authors made relevant contributions to the work. All authors read and approved the final manuscript. JS is employee of Pfizer. The views or opinions presented in this work are solely those of the authors and do not represent those of the companies. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
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.
Reprints and Permissions. Some reflections on the use of inappropriate comparators in CEA. Cost Eff Resour Alloc 18, 29 Download citation. Received : 15 November Accepted : 18 August Published : 27 August 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. Skip to main content. Search all BMC articles Search.
0コメント