Cost-effectiveness analysis of introducing malaria diagnostic testing in drug shops: A cluster-randomised trial in Uganda

Research output: Contribution to journalJournal articlepeer-review

Background Private sector drug shops are an important source of malaria treatment in Africa, yet diagnosis without parasitological testing is common among these providers. Accurate rapid diagnostic tests for malaria (mRDTs) require limited training and present an opportunity to increase access to correct diagnosis. The present study was a cost-effectiveness analysis of the introduction of mRDTs in Ugandan drug shops. Methods Drug shop vendors were trained to perform and sell subsidised mRDTs and artemisinin-based combination therapies (ACTs) in the intervention arm while vendors offered ACTs following presumptive diagnosis of malaria in the control arm. The effect on the proportion of customers with fever ‘appropriately treated of malaria with ACT’ was captured during a randomised trial in drug shops in Mukono District, Uganda. Health sector costs included: training of drug shop vendors, community sensitisation, supervision and provision of mRDTs and ACTs to drug shops. Household costs of treatment-seeking were captured in a representative sample of drug shop customers. Findings The introduction of mRDTs in drug shops was associated with a large improvement of diagnosis and treatment of malaria, resulting in low incremental costs for the health sector at US $0.55 per patient appropriately treated of malaria. High expenditure on non-ACT drugs by households contributed to higher incremental societal costs of US$3.83. Sensitivity analysis showed that mRDTs would become less cost-effective compared to presumptive diagnosis with increasing malaria prevalence and lower adherence to negative mRDT results. Conclusion mRDTs in drug shops improved the targeting of ACTs to malaria patients and are likely to be considered cost-effective compared to presumptive diagnosis, although the increased costs borne by households when the test result is negative are a concern.

Original languageEnglish
Article numbere0189758
JournalPLOS ONE
Volume12
Issue number12
Number of pages18
ISSN1932-6203
DOIs
Publication statusPublished - Dec 2017

Number of downloads are based on statistics from Google Scholar and www.ku.dk


No data available

ID: 187288603