Mass drug administration to reduce malaria incidence in a low-to-moderate endemic setting: short-term impact results from a cluster randomised controlled trial in Senegal

Introduction

Malaria is a major public health concern in Africa. In regions where transmission is highly seasonal, seasonal malaria chemoprevention (SMC) has been widely adopted to prevent morbidity and mortality in children at-risk of severe malaria. SMC involves the monthly administration of sulfadoxine-pyrimethamine + amodiaquine given during the peak transmission season to treat existing parasitaemia and prevent new infections.¹ Since its initial recommendation by the World Health Organisation (WHO) in 2012, SMC has expanded to 18 African countries, covering 53 million children in 2023.² The scale-up of SMC has been successful, demonstrating reductions in childhood clinical malaria incidence by 60–88% under programmatic conditions.^1,3

Through high coverage of SMC, strong vector control, and prompt case management, countries in Sahelian and sub-Sahelian Africa have made significant strides in controlling malaria, prompting many to establish new goals for malaria elimination. However, recent progress toward elimination has stalled,⁴ necessitating enhanced coverage of proven core interventions and the consideration of new interventions to rapidly reduce transmission. One promising approach is mass drug administration (MDA), which involves the administration of antimalarials to all individuals in a defined geographic area at a frequency and duration tailored to the local malaria epidemiology and goals. For MDA to have an impact on transmission, high coverage (≥80%)^5,6 of the target population is needed, which requires an optimized delivery approach and strong community engagement.⁷ Achieving high coverage may be less challenging in countries successfully implementing SMC, as they can leverage their existing infrastructure of door-to-door delivery and community and health system acceptance of chemoprevention.^8,9

In addition to attaining high coverage, the effectiveness of MDA depends on the type of antimalarial regimen used. In Plasmodium falciparum-dominant regions, dihydroartemisinin-piperaquine is an attractive agent for MDA given its good safety profile, long prophylactic period, and relatively low-levels of artemisinin resistance in Africa.^10,11 However, dihydroartemisinin-piperaquine has no known efficacy against mature gametocytes^12,13—the parasites responsible for human-to-mosquito transmission. Single low-dose primaquine is a gametocytocidal agent shown to be safe and associated with the near complete prevention of human-to-mosquito transmission.^14,15 It is likely that the combination of these drugs may confer greater benefits than dihydroartemisinin-piperaquine alone, allowing MDA to have a greater impact on transmission.

Here, we present results from a cluster randomised controlled trial assessing the safety, coverage, and short-term impact of three cycles of MDA with dihydroartemisinin-piperaquine + single low-dose primaquine on P. falciparum incidence and prevalence during the intervention year. This study aimed to fill a critical evidence gap regarding the effectiveness of MDA to rapidly reduce malaria burden in a highly seasonal, low-to-moderate transmission setting where malaria control measures have been scaled-up and additional interventions are needed to accelerate malaria elimination.

Methods

Procedures

Community mobilisation and sensitisation

Upon village selection, the study team held meetings with administrative, health, and religious leaders of Tambacounda Health District to discuss the study aims, planned activities, and receive consent for study implementation. Community sensitisation materials were developed by the study team and implemented by local health staff. In the months prior to MDA, which coincided with the peak COVID-19 pandemic, additional social media campaigns, local community radio announcements, and television advertisements were conducted. Prior to each MDA cycle, town hall meetings and household visits were undertaken to ensure that the community was well-informed.

Interventions

In the control arm, the standard-of-care chemoprevention (which consisted of three cycles of SMC with sulfadoxine-pyrimethamine + amodiaquine) was administered to children aged 3-120 months at four-week intervals, initiated at the presumed start of the malaria transmission season. In the intervention arm, three cycles of MDA with dihydroartermisinin-piperaquine + single low-dose primaquine were administered to individuals aged ≥3 months at six-week intervals. To achieve maximal impact on clearing the infectious reservoir,^16,17 MDA was initiated one month prior to the presumed start of the transmission season (Figure 1). Prior to intervention implementation, pyrethroid-piperonyl butoxide (PBO) bednets were distributed door-to-door to all study villages and year-round PECADOM+ was established to monitor malaria incidence.

• Download figure
• Open in new tab

Figure 1.

Timeline of study activities

MDA and SMC were delivered door-to-door by DSDOMs via directly observed therapy for all three doses. Study drugs were administered using an age-based dosing strategy per manufacturer’s instructions (Appendix 2). For each MDA cycle, dihydroartemisinin-piperaquine (Eurartesim™, Sigma-Tau, Italy) was given for three consecutive days and single low-dose primaquine (Remedica Ltd, Limassol, Cyprus) was given with the first dose of dihydroartemisinin-piperaquine. For each SMC cycle, amodiaquine (Fosun Pharma, Shanghai, China) was given for three consecutive days and sulfadoxine-pyrimethamine (Fosun Pharma, Shanghai, China) was given with the first dose of amodiaquine. For participants who were unable to swallow tablets (e.g., young children), tablets were crushed and mixed in water. If the participant vomited within 30 minutes of administration, the full dose was re-administered. If the re-administered dose was vomited within 30–60 minutes, half the dose was administered.

During drug administration, individuals with symptomatic malaria confirmed by a histidine-rich protein 2-based rapid diagnostic test (RDT) (ParaHIT®-f, ARKRAY Healthcare Pvt Ltd, Surat, India) were treated with artemether-lumefantrine and did not receive study drugs until the subsequent cycle.

Malaria surveillance

Malaria cases were captured through health facility and PECADOM+ registries. Suspected cases (i.e., presentation of fever or history of fever in the past 48 hours) were confirmed by RDT. To ensure high-quality capture of incident cases, PECADOM+ was expanded year-round implementation in all study villages and fully scaled-up by March 1, 2021. Data were collected on paper-based registries and abstracted onto electronic databases. Duplicates between registries were removed. Village-level population size was estimated by averaging estimates from two censuses conducted before-and-after intervention implementation (Figure 1).

Pharmacovigilance

Passive and active pharmacovigilance systems were used to monitor the safety of MDA. For passive surveillance, adverse events (AEs) following drug intake were recorded by study staff or DSDOMs into standardised case report forms. Participants were encouraged to inform local health or study staff if they experienced an AE within one month after drug intake. AEs were graded by a study clinician (e.g., mild, moderate, or severe) and managed free of charge. For active surveillance monitoring, the study staff surveyed 220 random households per study arm on the day after the last dose of study drugs. Households were sampled from every village, proportionate to village population size, such that five households were sampled from villages with <300 residents and ten households from villages with ≥300 residents. In intervention villages, the study team randomly sampled three members per household according to their age group: <5, 5–15, and ≥15 years. In control villages, three household members <10 years of age were randomly sampled. Survey participants were asked if they experienced an AE and to describe the event, including type, onset date, and duration. Severity was graded by a study clinician.

Cross-sectional surveys

To determine parasite prevalence, cross-sectional surveys were conducted at the end of the transmission season before-and-after intervention implementation (December 10–20, 2020 and December 9–21, 2021). A two-stage cluster sampling strategy was undertaken to randomly select households and household members from all villages. Participants were asked about their demographic characteristics, malaria prevention measures, and history of fever. Suspected malaria cases were confirmed by RDT and those with a positive test were treated with artemether-lumefantrine. A fingerprick blood sample was taken for microscopy and for dried blood spots (DBS) to confirm parasitaemia by PCR and to genotype drug resistance markers.

Laboratory analysis

Microscopy slides and DBS from surveys were transported to Université Iba Der Thiam de Thiès. Slides were stained with 6% Giemsa for 20 min and read by two microscopists. A third reviewer settled discrepant findings. Parasite DNA was extracted from DBS using the Chelex-100 extraction method¹⁸ and tested for parasitemia by real-time PCR using species-specific primers based on 18s rRNA gene as previously described.¹⁹ PCR-positive samples were genotyped to assess the presence of point mutations in the pfK13, pfdhps, pfdhfr, pfcrt, and pfmdr1 genes using high-resolution melting analysis as previously described.²⁰

Results

Between September 1, 2020 and October 25, 2020, 523 villages in the study area were geolocated and screened for eligibility, and 111 met the inclusion criteria (Figure 2). Of these, 60 villages were randomly selected allowing for a ≥2.5 km distance between village centroids and randomised to intervention or control. Village-level factors included in the constrained randomisation were balanced across arms (Table 1). Overall, coverage of pyrethroid-PBO bednets was high (98%) and similar between arms. In the pre-intervention year, 81% of children <10 years of age reported receiving the most recent cycle of SMC and 71% reported receiving all three cycles. Twenty percent reported sleeping away from their home in the past 15 days.

• Download figure
• Open in new tab

Figure 2.

Trial profile

In the intervention arm, 8931, 9571, and 9703 residents were screened for the first, second, and third cycle of the MDA campaign. Distribution coverage of single low-dose primaquine (where the denominator excluded ineligible residents) was 79% (6286/7992), 82% (6949/8462), and 84% (7199/8575) across the three cycles. Distribution coverage of ≥1 dose of dihydroartemisinin-piperaquine among the eligible individuals was 79% (6505/8229), 82% (7125/8673), and 83% (7250/8690) in the first, second, and third cycles. Distribution coverage of all three doses of dihydroartemisinin-piperaquine was 74%, 79%, and 81% across cycles.

Distribution coverage was higher in those <10 years compared to ≥10 years (distribution coverage of ≥1 dose was 85%, 86%, and 87% in <10 years across cycles, and 75%, 80%, and 81% in ≥10 years across cycles). Crude coverage of ≥1 dose of dihydroartemisinin-piperaquine was 73%, 74%, and 75% across cycles. By village, distribution coverage ranged from 58%–97%; 50%, 67%, and 70% of intervention villages reached the WHO target coverage of ≥80% in the first, second, and third cycles, respectively (Appendix 3). The major reasons for non-participation were absence (range: 14%–21%) and illness (range: 5%–7%) (Appendix 4).

Absences were similar between males and females (1⋅12:1 ratio) and in age to those who received MDA (16 years [interquartile range (IQR): 7–26] versus 13 years [IQR: 6–27]). Refusals were rare (1%–2%) and mostly among males (70%) with a median age of 22 years [IQR: 15–30].

In the control arm, 3492, 3489, and 3487 children aged 3–120 months were screened for the first, second, and third SMC cycle. Distribution coverage of ≥1 dose of SMC drugs was 93% (3202/3457), 92% (3174/3442), and 92% (3146/3434) across cycles. Distribution coverage for all three SMC doses was 92%, 92%, and 91% across cycles. Crude coverage of ≥1 dose was 92%, 91%, and 90% across cycles. By village, distribution coverage ranged from 62%–100%; 93% of SMC villages reached ≥80% coverage across all cycles (Appendix 3). Major reasons for non-receipt of SMC were absence (7%) and illness (1%–2%) (Appendix 4). Refusals of SMC were low across cycles (≤1%).

During the pre-intervention transmission season (July-December 2020), malaria incidence was 181 and 204 cases/1000 population in the intervention and control arms, respectively (Table 2). In the transmission season of the intervention year (July-December 2021), malaria incidence reduced to 93 cases/1000 population in the intervention arm and to 173 cases/1000 population in the control arm (Table 2; Figure 3). The unadjusted intervention effect of MDA was 52% [95% CI: 21%, 71%]. The adjusted intervention effect, which accounted for variables included in the constrained randomisation was 55% [95% CI: 28%, 72%].

• Download figure
• Open in new tab

Figure 3.

Graphical representation of village-level monthly incidence of symptomatic malaria during the intervention implementation year by intervention group (A) and stratified by age group (B), and baseline transmission intensity (C).

Pre-specified subgroup analyses showed evidence of an interaction by age group; the adjusted intervention effect was 58% [95% CI: 33%, 74%] in the ≥10 years age group and 46% [95% CI: 10%, 67%] in the <10 years age group (p_interaction=0·016) (Table 2; Figure 3). The impact of MDA on malaria incidence was similar in low and moderate transmission settings defined as microscopy parasite prevalence of <10% versus ≥10% (adjusted intervention effect=55% [31%, 71%] versus 52% [-19%, 81%]; p_interaction=0·88) (Table 2; Figure 3).

By microscopy, parasite prevalence decreased between 2020 and 2021, from 6·1% to 1·8% in the intervention arm and from 6·7% to 4·7% in the control arm (adjusted intervention effect=62% [95% CI: 22%, 80%]) (Table 3). The impact of MDA differed between the <10 and ≥10 years age group (adjusted intervention effect=76% [95% CI: 42%, 90%] versus 51% [95%% CI: -14%, 79%]), but this finding did not reach statistical significance (p_interaction=0·17). By PCR, parasite prevalence decreased between 2020 and 2021, from 17·9% to 4·5% in the intervention arm and 19·9% to 8·3% in the control arm (adjusted intervention effect=47% [95% CI: 3%, 71%]). The effect of MDA differed between the <10 and ≥10 years age groups: adjusted intervention effect=71% [95% CI: 35%, 87%] and 33% [95% CI: -27%, 65%], respectively (p_interaction=0·050). The impact of MDA on microscopy- or PCR-confirmed parasite prevalence did not significantly differ between low and moderate transmission settings (p_interaction for microscopy-confirmed prevalence=0·73, p_interaction for PCR-confirmed prevalence=0·73) (Table 3).

In both passive and active pharmacovigilance systems, the frequency of AEs reduced with each subsequent cycle and no serious adverse events (SAEs) were detected in either arm (Appendix 5; Table 4). Through passive surveillance, 129 AEs were observed in 67/20 887 (0·003%) participants of the intervention arm and four AEs were observed in 2/9524 (0·0002%) participants of the control arm; all were mild (Appendix 5). Both the frequency and proportion of participants who experienced an AE were higher in the intervention arm compared to the control arm (p<0·0001). The most common AEs found in the intervention arm were gastrointestinal issues (45·7%), headaches (25·6%), and fever (17·1%).

Through active surveillance, more participants of the intervention arm reported an AE compared to the control arm (13·7% versus 9·4%; p<0·0001) (Table 4). Among children aged <10 years, the proportion of reported AEs were similar between arms (10·1% versus 9·4%; p=0·57). In the intervention arm, the most common AEs were gastrointestinal issues (41·2%), fever (22·7%), and headaches (11·8%). Of the 498 AEs reported in the intervention arm, 64·1% were mild, 32·9% were moderate, and 3·0% were severe. Severe AEs included fever (n=7), headache (n=4), drowsiness (n=1), vomiting (n=1), diarrhoea (n=1), and loss of appetite (n=1). All AEs appeared three hours after drug intake, resolved within 72 hours, and did not require hospitalization. No cases of anaemia were found through either passive or active surveillance.

Of the 597 PCR-positive samples collected from surveys, 433 were successfully genotyped to determine molecular markers of antimalarial resistance (Appendix 6). Mutations associated with intermediate resistance to SP (PfDHFR N51I, C59R, S108N and PfDHPS A437G) were seen at high proportions (range: 58%–100%). PfDHFR I164L was seen at proportions of 2%–12%. None of the studied mutations differed in proportions between arms or time periods. There was no evidence of PfDHPS K540E, PfDHPS A581G, or PfK13 C580Y mutations.

Discussion

We conducted a cluster randomised controlled trial evaluating the safety, coverage, and short-term effectiveness of three rounds of MDA with dihydroartemisinin-piperaquine + single low-dose primaquine in a low-to-moderate transmission setting of Senegal where malaria control interventions were scaled-up through pyrethroid-PBO bednet distribution, three monthly SMC cycles, and expansion of year-round proactive community case management of malaria. During the trial intervention year, malaria incidence and parasite prevalence reduced in both arms, likely due in part to pyrethroid-PBO bednet distribution. In intervention villages, MDA was associated with a 55% and 62% reduction in malaria incidence and microscopy-confirmed parasite prevalence. Subgroup analyses showed that MDA had a substantial impact on incidence and parasite prevalence in those <10 and ≥10 years of age and in both low and moderate transmission settings. Overall, MDA was well-tolerated; most AEs were mild or moderate and no cases of anaemia or SAEs were observed.

While the trial was not designed to separately estimate MDA’s effect on transmission versus its direct, prophylactic effect, there was some evidence to suggest that MDA indirectly affected transmission. First, subgroup analyses demonstrated that among children under ten, MDA was associated with a 46% reduction in incidence as compared to children under ten in the control arm who received SMC. We expect that this reduction is unlikely by differences in the prophylactic effects of these drugs given we expect dihydroartemisinin-piperaquine and sulfadoxine-pyrimethamine + amodiaquine to have similar protective efficacies,¹¹ MDA covered less of the transmission season, and MDA had slightly lower coverage than SMC in under tens. Moreover, MDA’s impact on parasite prevalence further supports this reasoning, given the endline survey was conducted more than two months after the last cycle of drug administration when the prophylactic effects of both MDA and SMC drugs would have waned.

Our study has several important findings/caveats. First, our monthly incidence analyses (Figure 4) revealed that SMC should have started one month later and both chemoprevention campaigns should have been extended to four cycles to cover the entire transmission season. If this had this been done, the impact of MDA in children under tens might not have been as substantial. Second, our findings may not be generalisable to most settings deploying SMC where the malaria burden is mostly moderate-to-high. While our study found MDA was associated with reductions in malaria burden and short-term transmission, evidence of its sustained benefit and cost-effectiveness is needed before considering MDA as an intermediate intervention to accelerate malaria elimination.

The WHO currently recommends MDA for transmission reduction in very low-to-low transmission settings (i.e., parasite prevalence of <10% or incidence <250 cases/1000 population). This recommendation is based on evidence from eight cluster randomised controlled trials which found MDA can have a substantial, but short-term impact in these settings.^1,21,22 However, MDA is not recommended for transmission reduction in moderate-to-high transmission settings, based on evidence from two cluster randomised trials and two non-randomized studies which did not show significant short- or long-term impacts on prevalence or incidence. Our study provides new evidence indicating MDA may have a short-term impact on transmission in both low and moderate transmission strata. Our findings are consistent with a recent trial conducted in a moderate transmission setting of The Gambia²³ which demonstrated a 70% reduction in the odds of PCR-confirmed infection two months after the last MDA cycle. Evidence from these two studies should be considered when determining future recommendations in moderate transmission settings.

There were several strengths of the study, including our large sample size, rigorous monitoring of safety, and achievement of high coverage. By the final MDA cycle, 70% (21/30) of villages reached the WHO target for MDA coverage (≥80%)⁷, despite significant operational challenges in door-to-door drug administration amidst the peak of the global COVID-19 pandemic. The high coverage observed in our study was likely attributable to community acceptance of annual SMC campaigns, established infrastructure of community case management, support from key administrative and health authorities, and repeated community sensitisation campaigns. However, absences were common during MDA cycles (14-21%), especially among adolescents and young adults who are generally asymptomatic and known to be important drivers of transmission.²⁴ Thus, in order to sustain gains made, future MDA campaigns will need to consider additional strategies to reach these groups in order to prevent malaria importation and potential resurgence after MDA.

Our study had a few limitations. First, as explained earlier, SMC was not optimally timed, making it difficult to draw direct comparisons between SMC and MDA for burden reduction in children. Second, baseline incidence was likely non-differentially mismeasured in villages where DSDOMs were absent at the start of the study (33% of villages). While the stratified randomization scheme balanced this factor between arms, this may have biased our effect estimates toward the null (see Appendix 7 for analyses restricted to villages with prior PECADOM+). Third, given baseline incidence was likely mismeasured, subgroup analyses by baseline malaria incidence were not reported. Fourth, parasitological confirmation of malaria incidence relied on RDT which could have resulted in false positives potentially biasing our effect estimates toward the null. Fifth, MDA was conducted for only one year and additional cycles may have been needed to reach pre-elimination status (i.e., incidence <5 cases/1000 population). Finally, the current analyses only evaluated short-term impact. After discontinuation of MDA, there is potential for malaria to rebound, and analyses of longer-term impact and cost-effectiveness from an additional year of follow-up of these cohorts are forthcoming which will help to determine the sustainability of MDA.

In low-to-moderate malaria transmission settings, where transmission is seasonal and coverage of standard malaria control interventions is high, we demonstrated that three cycles of MDA with dihydroartemisinin-piperaquine + single-low dose primaquine could safely and rapidly reduce malaria burden and potentially have a short-term impact on transmission. However, future work investigating whether the impact of MDA can be sustained, and its cost-effectiveness is needed to inform countries of the potential of MDA as an intervention to accelerate toward elimination.

Role of the funding source

The study was funded under the US President’s Malaria Initiative Impact Malaria Consortium. The funders participated in the design of the study and reviewed the manuscript prior to submission. The funders had no role in data collection and data analysis. The first authors had full access to the study data and final responsibility for the decision to submit the manuscript.

Contributions

JH conceived of the study and wrote the first draft of the protocol. RG, JLN, MER, AB, KSR, JT, EBKC, FB, ST, ID, and ABG contributed to the protocol and approved of the final version. EBKC, AD, TG, AS, ST, SG, ACL, ED, ID, OGB, and JLN implemented the study. JLN, EBKC, AD, TG, ABG, ST, KSR, MER, MSH, TK, MH, EE, RG, DS, and FB provided oversight of the study. ACL conducted laboratory analysis of samples. The statistical analysis plan was developed by MSH, PM, AF, MER, AB, EBKC, and JLN and data were analysed by AF with XW, MER, AS, and EBKC. MER, EBKC, AF, XW, PM, AB, JH, KSR, MSH, and JLN interpreted the results. MER, EBKC, MSH, and JLN wrote the first draft of the manuscript. PM, JH, AF, and KSR provided additional inputs to the writing. All authors reviewed and approved the final manuscript before submission.

Declaration of interests

The study was funded by the US President’s Malaria Initiative. MER is supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health (Award Number K99HD111572). JH and ABG receive salary support from the US President’s Malaria Initiative. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institutes of Health, US Centers of Disease Control and Prevention, and the US Agency for International Development.

Data sharing

De-identified data may be available upon reasonable request and approval from Principal Investigators via a data sharing agreement.