A multiphase program for malaria elimination in southern Mozambique (the Magude project): A before-after study

Abstract

Background: Malaria eradication remains the long-term vision of the World Health Organization (WHO). However, whether malaria elimination is feasible in areas of stable transmission in sub-Saharan Africa with currently available tools remains a subject of debate. This study aimed to evaluate a multiphased malaria elimination project to interrupt Plasmodium falciparum malaria transmission in a rural district of southern Mozambique.

Methods and findings: A before-after study was conducted between 2015 and 2018 in the district of Magude, with 48,448 residents living in 10,965 households. Building on an enhanced surveillance system, two rounds of mass drug administrations (MDAs) per year over two years (phase I, August 2015-2017), followed by one year of reactive focal mass drug administrations (rfMDAs) (phase II, September 2017-June 2018) were deployed with annual indoor residual spraying (IRS), programmatically distributed long-lasting insecticidal nets (LLINs), and standard case management. The four MDA rounds covered 58%-72% of the population, and annual IRS reported coverage was >70%. Yearly parasite surveys and routine surveillance data were used to monitor the primary outcomes of the study-malaria prevalence and incidence-at baseline and annually since the onset of the project. Parasite prevalence by rapid diagnostic test (RDT) declined from 9.1% (95% confidence interval [CI] 7.0-11.8) in May 2015 to 2.6% (95% CI 2.0-3.4), representing a 71.3% (95% CI 71.1-71.4, p < 0.001) reduction after phase I, and to 1.4% (95% CI 0.9-2.2) after phase II. This represented an 84.7% (95% CI 81.4-87.4, p < 0.001) overall reduction in all-age prevalence. Case incidence fell from 195 to 75 cases per 1,000 during phase I (61.5% reduction) and to 67 per 1,000 during phase II (65.6% overall reduction). Interrupted time series (ITS) analysis was used to estimate the level and trend change in malaria cases associated with the set of project interventions and the number of cases averted. Phase I interventions were associated with a significant immediate reduction in cases of 69.1% (95% CI 57.5-77.6, p < 0.001). Phase II interventions were not associated with a level or trend change. An estimated 76.7% of expected cases were averted throughout the project (38,369 cases averted of 50,005 expected). One malaria-associated inpatient death was observed during the study period. There were 277 mild adverse events (AEs) recorded through the passive pharmacovigilance system during the four MDA rounds. One serious adverse event (SAE) that resulted in death was potentially related to the drug. The study was limited by the incomplete coverage of interventions, the quality of the routine and cross-sectional data collected, and the restricted accuracy of ITS analysis with a short pre-intervention period.

Conclusion: In this study, we observed that the interventions deployed during the Magude project fell short of interrupting P. falciparum transmission with the coverages achieved. While new tools and strategies may be required to eventually achieve malaria elimination in stable transmission areas of sub-Saharan Africa, this project showed that innovative mixes of interventions can achieve large reductions in disease burden, a necessary step in the pathway towards elimination.

Trial registration: ClinicalTrials.gov NCT02914145.

Conflict of interest statement

The authors have declared that no competing interests exist. CM is a member of the Editorial Board of PLOS Medicine.

Figures

Fig 1. Project interventions and outcome evaluation timeline.

Timing of the Magude project interventions and of the different data collection approaches used to measure and compare the primary outcomes of the project, and to estimate impact. IPTp, intermittent preventive treatment for pregnant women; IRS, indoor residual spraying; LLIN, long-lasting insecticidal net; MDA, mass drug administration; NMCP, National Malaria Control Program; rfMDA, reactive focal mass drug administration.

Fig 2. All-age and age-specific malaria infection prevalence estimates measured by RDT before (May 2015) and during the Magude project (2016–2018).

CI, confidence interval; MDA, mass drug administration; RDT, rapid diagnostic test.

Fig 3. Routine outpatient data collected through the RRS of Magude.

(A) Weekly number of outpatients who attended any HF of Magude (dotted line), the diagnostics performed (dashed line) and cases confirmed (solid line). (B) Weekly number of cases in children under 5 years and individuals five years old and older. (C) All-age,

Fig 4. ITS regression model estimates of the weekly number of passively detected malaria cases in Magude before (October 2013 to July 2015) and after the intervention during phase I (August 2015–2017) and phase II (September 2018 to June 2018), and the estimated cases that would have happened since the first phase of the Magude project, had this not taken place (counterfactual).

ITS, interrupted time series.