Abstract
Background: The biological diagnosis of imported malaria cases in nonendemic areasis an infrequent challenge that requires efficient methods, trained staff and high-quality proficiency. Microscopy, rapid diagnosis tests and molecular tests are widely available and provide excellent results. However, there is a continuous flow of recently developed methods, either at a preliminary step or commercially available. Among the latter, flow cytometry using hematology analysers has gained more attention in recent years and is expected to be used in endemic and nonendemic areas. However, the real cost of using these methods, from historic microscopy to more recent molecular or cytometry methods, is frequently approximate. In the context of limited resources for medical care, a complete cost-effectiveness analysis of the different scenarios of biological methods used in a nonendemic area should aid in the decision-making process for the most appropriate scenario. Therefore, the aim of this study was to provide an extensive cost-effectiveness analysis and a comparison between different scenarios available in France.
Methods: The full cost-effectiveness of each malaria diagnosis method relative to the clinical benefits of the outcome was measured in terms of monetary and nonmonetary values. The study was conducted in agreement with the CHEERS 2022 checklist and recommendations from the B&M Gates Foundation. The study population was a cohort of patients who were receiving health care at Lyon University Hospital for fever and suspected malaria during 2023. Age, Plasmodium species, hospitalization levels (ICU, non-ICU), and positive or negative outcomes were documented for the included patients. Four scenarios were tested among the most likelytreatments: 1) microscopy, 2) RDT + microscopy, 3) LAMP + microscopy, and 4) Haematology analyser XN-31 + microscopy. The direct costs of the intervention and control tests were calculated on the basis of prices paid in France in 2023 for one dedicated machine with a specific depreciation rate and maintenance, quality controls and all consumables needed to perform malaria diagnosis for one sample among 1000 tests per year. The indirect cost of technical training, supervision and quality proficiency was calculated based on the hourly salary of the laboratory technician and junior and senior doctorsaccording to the time needed for each scenario.
Results: A decision tree was developed to compare the intervention to the three comparator scenarios, and an incremental cost-effectiveness ratio was used to compare the intervention and controls. The obtained cost-effectiveness plane clearly demonstrated that the intervention (XN-31+microscopy) was the most cost-effective scenario, as it was more effective and less expensive than scenarios 2 and 3 (RDT+microscopy and LAMP+microscopy). Microscopy was also dominated by the intervention because of the significantly greater cost of training and quality proficiency.
Conclusion: Thisstudy is based on data available in France and should not be directly translated to other countries or other health care systems. However, this approach provides a global approach for determining the cost-effectiveness of the most frequent methods for diagnosing malaria. This allows us to compare those methods and will help final decision makers to select the most appropriate scenario depending on local constraints. The cost-effectiveness results clearly demonstrated that the intervention (XN-31 + microscopy) was dominant (most effective and least costly) to the comparators. Intervention also stochastically dominates (first order) microscopy, while the direct cost of one XN-31 test is greater than that of microscopy. Indeed, the better performance of the XN-31 in terms of sensitivity and specificity and thereduced time needed for training and operational execution of the test were the basis for themajor impact on the cost-effectiveness ratio.