In countries where malaria is endemic, the use of rapid diagnostic tests(RDTs) has become routine, especially in rural settings. Such regions are characterised by often having other co-endemic infectious diseases, at high levels of prevalence. To illustrate the potential added-value of "sentinel" screening for patients presenting for a routine diagnostic test for malaria, at healthcare facilities in Uganda. We developed an economic model by combining two decision trees, one for malaria and a second for the co-endemic disease schistosomiasis. The integrated model was designed to inform policy strategies for the co-endemic disease in addition to malaria (i.e., whether to test opportunistically for schistosomiasis or use mass drug administration(MDA) as per usual practice).We performed the analysis on three comparators varying testing accuracy and costs. Sentinel screening can provide added value to the testing of patients compared with the status quo: when schistosomiasis prevalence is high then MDA is preferential; if low prevalence, treating no one is preferred. If the disease has average levels of prevalence, then a strategy involving testing is preferred. Prevalence thresholds driving the dominant strategy are dependent upon the model parameters, which are highly context specific. At average levels of prevalence for schistosomiasis and malaria for Uganda, adding a sentinel screening was cost-effective when the accuracy of test was higher than current diagnostics and when economies of scope were generated(Expected value clinical Information = 0.65$ per DALY averted, 137.91$ per correct diagnoses).Protocols using diagnostics with current accuracy levels were preferred only for levels of MDA coverage below 75%. The importance of the epidemiological setting is crucial in determining the best cost-effective strategy for detecting endemic disease. Economies of scope can make sentinel screenings cost-effective strategies in specific contexts. Blanket thresholds recommended for MDA may not always be the preferred option for endemic diseases.
Read full abstract