Performance and utility of more highly sensitive malaria rapid diagnostic tests.

BACKGROUND: A new more highly sensitive rapid diagnostic test (HS-RDT) for Plasmodium falciparum malaria (Alere™/Abbott Malaria Ag P.f RDT [05FK140], now called NxTek™ Eliminate Malaria Ag Pf) was launched in 2017. The test has already been used in many research studies in a wide range of geographies and use cases. METHODS: In this study, we collate all published and available unpublished studies that use the HS-RDT and assess its performance in (i) prevalence surveys, (ii) clinical diagnosis, (iii) screening pregnant women, and (iv) active case detection. Two individual-level data sets from asymptomatic populations are used to fit logistic regression models to estimate the probability of HS-RDT positivity based on histidine-rich protein 2 (HRP2) concentration and parasite density. The performance of the HS-RDT in prevalence surveys is estimated by calculating the sensitivity and positive proportion in comparison to polymerase chain reaction (PCR) and conventional malaria RDTs. RESULTS: We find that across 18 studies, in prevalence surveys, the mean sensitivity of the HS-RDT is estimated to be 56.1% (95% confidence interval [CI] 46.9-65.4%) compared to 44.3% (95% CI 32.6-56.0%) for a conventional RDT (co-RDT) when using nucleic acid amplification techniques as the reference standard. In studies where prevalence was estimated using both the HS-RDT and a co-RDT, we found that prevalence was on average 46% higher using a HS-RDT compared to a co-RDT. For use in clinical diagnosis and screening pregnant women, the HS-RDT was not significantly more sensitive than a co-RDT. CONCLUSIONS: Overall, the evidence presented here suggests that the HS-RDT is more sensitive in asymptomatic populations and could provide a marginal improvement in clinical diagnosis and screening pregnant women. Although the HS-RDT has limited temperature stability and shelf-life claims compared to co-RDTs, there is no evidence to suggest, given this test has the same cost as current RDTs, it would have any negative impacts in terms of malaria misdiagnosis if it were widely used in all four population groups explored here.

'I could not join because I had to work for pay.': A qualitative evaluation of falciparum malaria pro-active case detection in three rural Cambodian villages.

BACKGROUND: Pro-active case detection (Pro-ACD), in the form of voluntary screening and treatment (VSAT) following community mobilisation about 'asymptomatic malaria', is currently being evaluated as a tool for Plasmodium falciparum elimination in Preah Vihear Province, Cambodia. METHODS: A qualitative study was conducted to explore community understanding, perceptions, expectations and acceptability of the Pro-ACD intervention in order to identify aspects that could be improved in future Pro-ACD activities. This was ancillary to a three-round VSAT campaign, carried out in three villages between December 2015 and March 2016. Qualitative data collection began shortly after the end of the three rounds of screening. Purposive sampling was used to select participants. Nine focus group discussions with participants (n = 46) and non-participants (n = 40) in the Pro-ACD screening were conducted, in addition to in-depth interviews with key village figures (n = 9). RESULTS: Health promotion messages were well delivered and received, but it was difficult for many villagers to understand the messages around 'asymptomatic malaria'. Overall, villagers and village leaders had a positive opinion about the VSAT intervention. Acceptability was high, as a direct consequence of favourable perceptions towards the screening activity: the Pro-ACD intervention was seen by the local population as an effective, inexpensive, reliable and readily available tool to protect individuals and the community from the insurgence of malaria. Physical absence and lack of time (both linked to work-related activities) were the main reasons for non-participation. CONCLUSIONS: Although VSAT was generally well perceived and accepted, the 'time factor' related to the need to satisfy essential daily subsistence requirements played a significant role in determining participation in the screening. More well-adapted and meaningful Pro-ACD approaches could be implemented by improving the timing of the testing activites, and strengthening community participation and engagement to increase acceptability.

Closing in on the Reservoir: Proactive Case Detection in High-Risk Groups as a Strategy to Detect Plasmodium falciparum Asymptomatic Carriers in Cambodia.

Background: In the frame of elimination strategies of Plasmodium falciparum (Pf), active case detection has been recommended as complementary approach to the existing passive case detection programs. We trialed a polymerase chain reaction (PCR)-based active detection strategy targeting asymptomatic individuals, named proactive case detection (PACD), with the aim of assessing its feasibility, the extra yield of Pf infections, and the at-risk population for Pf carriage status. Methods: A pilot of PACD was conducted in 3 villages in Chey Saen district (Preah Vihear province, Cambodia), from December 2015 to March 2016. Voluntary screening and treatment, following health promotion sensitization, was used as mobilization strategy. Results: A total of 2802 persons were tested, representing 54% of the population. PACD (n = 30) and the respective reactive case detection (RACD) (n = 3) identified 33 Pf carriers, approximately twice as many as the Pf infections (n = 17) diagnosed in passive case detection and respective RACD, by health centers and village malaria workers using PCR, in the same villages/period. Final positivity rate was 1.07% (30/2802). People spending nighttime in forests and plantations were found to be at increased risk for Pf infection (odds ratio [OR], 3.4 [95% CI, 1.6-7.2], P = .002 and OR, 2.3 [95% CI, 1.1-4.9], P = .03, respectively). Conclusions: We demonstrated the usefulness of the PACD component in identifying Pf asymptomatic carriers. Social mobilization and promotion led to good attendance of specific risk groups, identified to be, in the Cambodian context, individuals spending nighttime in forest and plantations.

Efficacy of artesunate-amodiaquine, dihydroartemisinin-piperaquine and artemether-lumefantrine for the treatment of uncomplicated Plasmodium falciparum malaria in Maradi, Niger.

BACKGROUND: Malaria endemic countries need to assess efficacy of anti-malarial treatments on a regular basis. Moreover, resistance to artemisinin that is established across mainland South-East Asia represents today a major threat to global health. Monitoring the efficacy of artemisinin-based combination therapies is of paramount importance to detect as early as possible the emergence of resistance in African countries that toll the highest burden of malaria morbidity and mortality. METHODS: A WHO standard protocol was used to assess efficacy of the combinations artesunate-amodiaquine (AS-AQ Winthrop®), dihydroartemisinin-piperaquine (DHA-PPQ, Eurartesim®) and artemether-lumefantrine (AM-LM, Coartem®) taken under supervision and respecting pharmaceutical recommendations. The study enrolled for each treatment arm 212 children aged 6-59 months living in Maradi (Niger) and suffering with uncomplicated falciparum malaria. The Kaplan-Meier 42-day PCR-adjusted cure rate was the primary outcome. A standardized parasite clearance estimator was used to assess delayed parasite clearance as surrogate maker of suspected artemisinin resistance. RESULTS: No early treatment failures were found in any of the study treatment arms. The day-42 PCR-adjusted cure rate estimates were 99.5, 98.4 and 99.0% in the AS-AQ, DHA-PPQ and AM-LM arms, respectively. The reinfection rate (expressed also as Kaplan-Meier estimates) was higher in the AM-LM arm (32.4%) than in the AS-AQ (13.8%) and the DHA-PPQ arm (24.9%). The parasite clearance rate constant was 0.27, 0.26 and 0.25 per hour for AS-AQ, DHA-PPQ and AM-LM, respectively. CONCLUSIONS: All the three treatments evaluated largely meet WHO criteria (at least 95% efficacy). AS-AQ and AL-LM may continue to be used and DHA-PPQ may be also recommended as first-line treatment for uncomplicated falciparum malaria in Maradi. The parasite clearance rate were consistent with reference values indicating no suspected artemisinin resistance. Nevertheless, the monitoring of anti-malarial drug efficacy should continue. Trial registration details Registry number at ClinicalTrial.gov: NCT01755559.

Community participation during two mass anti-malarial administrations in Cambodia: lessons from a joint workshop.

Two mass drug administrations (MDA) against falciparum malaria were conducted in 2015-16, one as operational research in northern Cambodia, and the other as a clinical trial in western Cambodia. During an April 2017 workshop in Phnom Penh the field teams from Médecins Sans Frontières and the Mahidol-Oxford Tropical Medicine Research Unit discussed lessons for future MDAs.

Adapting Reactive Case Detection Strategies for falciparum Malaria in a Low-Transmission Area in Cambodia.

Reactive case detection around falciparum malaria cases in Cambodia presents a low output. We improved it by including individuals occupationally coexposed with index case patients and using polymerase chain reaction-based diagnosis. The positivity rate increased from 0.16% to 3.9%.

High burden of malaria and anemia among tribal pregnant women in a chronic conflict corridor in India.

BACKGROUND: With more than 200 million cases a year, malaria is an important global health concern, especially among pregnant women. The forested tribal areas of Andhra Pradesh, Telangana and Chhattisgarh in India are affected by malaria and by an on-going chronic conflict which seriously limits access to health care. The burden of malaria and anemia among pregnant women in these areas is unknown; moreover there are no specific recommendations for pregnant women in the Indian national malaria policy. The aim of this study is to measure the burden of malaria and anemia among pregnant women presenting in mobile clinics for antenatal care in a conflict-affected corridor in India. METHODS: This is a descriptive study of routine programme data of women presenting at first visit for antenatal care in Médecins sans Frontières mobile clinics during 1 year (2015). Burden of malaria and anemia were estimated using rapid diagnostic tests (SD BIOLINE® and HemoCue® respectively). RESULTS: Among 575 pregnant women (median age: 26 years, interquartile range: 25-30) 29% and 22% were in their first and second pregnancies respectively. Mid-Upper Arm Circumference (MUAC) was below 230 mm in 74% of them. The prevalence of anemia was 92.4% (95% Confidence Intervals (CI): 89.9-94.3), while severe anemia was identified in 6.9% of the patients. The prevalence of malaria was 29.3% (95%CI: 25.7-33.2) with 64% caused by isolated P. falciparum, 35% by either P. falciparum or mixed malaria and 1% by either P. vivax, or P.malariae or P. ovale. Malaria test was positive in 20.8% of asymptomatic cases. Malaria was associated with severe anemia (prevalence ratio: 2.56, 95%CI: 1.40-4.66, p < 0.01). CONCLUSIONS: Systematic screening for malaria and anemia should be integrated into maternal and child health services for conflict affected populations in highly endemic tribal areas. Interventions should include the use of rapid diagnostic test for all pregnant women at every visit, regardless of symptoms. Further studies should evaluate the impact of this intervention alone or in combination with intermittent malaria preventive treatment.

Efficacy of artemether-lumefantrine in relation to drug exposure in children with and without severe acute malnutrition: an open comparative intervention study in Mali and Niger.

BACKGROUND: Severe acute malnutrition (SAM) affects almost all organs and has been associated with reduced intestinal absorption of medicines. However, very limited information is available on the pharmacokinetic properties of antimalarial drugs in this vulnerable population. We assessed artemether-lumefantrine (AL) clinical efficacy in children with SAM compared to those without. METHODS: Children under 5 years of age with uncomplicated P. falciparum malaria were enrolled between November 2013 and January 2015 in Mali and Niger, one third with uncomplicated SAM and two thirds without. AL was administered under direct observation with a fat intake consisting of ready-to-use therapeutic food (RUTF - Plumpy'Nut®) in SAM children, twice daily during 3 days. Children were followed for 42 days, with PCR-corrected adequate clinical and parasitological response (ACPR) at day 28 as the primary outcome. Lumefantrine concentrations were assessed in a subset of participants at different time points, including systematic measurements on day 7. RESULTS: A total of 399 children (360 in Mali and 39 in Niger) were enrolled. Children with SAM were younger than their non-SAM counterparts (mean 17 vs. 28 months, P < 0.0001). PCR-corrected ACPR was 100 % (95 % CI, 96.8-100 %) in SAM at both day 28 and 42, versus 98.8 % (96.4-99.7 %) at day 28 and 98.3 % (95.6-99.4 %) at day 42 in non-SAM (P = 0.236 and 0.168, respectively). Compared to younger children, children older than 21 months experienced more reinfections and SAM was associated with a greater risk of reinfection until day 28 (adjusted hazard ratio = 2.10 (1.04-4.22), P = 0.038). Day 7 lumefantrine concentrations were significantly lower in SAM than non-SAM (median 251 vs. 365 ng/mL, P = 0.049). CONCLUSIONS: This study shows comparable therapeutic efficacy of AL in children without SAM and in those with SAM when given in combination with RUTF, but a higher risk of reinfection in older children suffering from SAM. This could be associated with poorer exposure to the antimalarials as documented by a lower lumefantrine concentration on day 7. TRIAL REGISTRATION: ClinicalTrials.gov: NCT01958905 , registration date: October 7, 2013.

Performance and time to become negative after treatment of three malaria rapid diagnostic tests in low and high malaria transmission settings.

BACKGROUND: The performance of different malaria rapid diagnostic tests (RDT) may be influenced by transmission intensity and by the length of time each test requires to become negative after treatment and patient's recovery. METHODS: Results of three RDTs (two HRP2 and one pLDH antigen-based tests) were compared to blood smear microscopy (the gold standard method) in children under 5 years of age living in a high versus low malaria intensity setting in southwestern Uganda. In each setting, 212 children, who tested positive by at least one RDT and by microscopy, were treated with artemether-lumefantrine. RDTs and microscopy were then repeated at fixed intervals to estimate each test's time to negativity after treatment and patient recovery. RESULTS: In the two settings, sensitivities ranged from 98.4 to 99.2 % for the HRP2 tests and 94.7 to 96.1 % for the pLDH test. Specificities were 98.9 and 98.8 % for the HRP2 tests and 99.7 % for the pLDH test in the low-transmission setting and 79.7, 80.7 and 93.9 %, respectively, in the high-transmission setting. Median time to become negative was 35-42 or more days for the HRP2 tests and 2 days for the pLDH test. CONCLUSIONS: High transmission contexts and a long time to become negative resulted in considerably reduced specificities for the HRP2 tests. Choice of RDT for low- versus high-transmission settings should balance risks and benefits of over-treatment versus missing malaria cases. TRIAL REGISTRATION: Registry number at ClinicalTrial.gov: NCT01325974.

Assessing the asymptomatic reservoir and dihydroartemisinin-piperaquine effectiveness in a low transmission setting threatened by artemisinin resistant Plasmodium falciparum.

BACKGROUND: In Cambodia, elimination of artemisinin resistance through direct elimination of the Plasmodium falciparum parasite may be the only strategy. Prevalence and incidence at district and village levels were assessed in Chey Saen district, Preah Vihear province, North of Cambodia. Molecular and clinical indicators for artemisinin resistance were documented. METHODS: A cross sectional prevalence survey was conducted at village level in the district of Chey Saen from September to October 2014. Plasmodium spp. was assessed with high volume quantitative real-time polymerase chain reaction (qPCR). Plasmodium falciparum-positive samples were screened for mutations in the k13-propeller domain gene. Treatment effectiveness was established after 28 days (D28) using the same qPCR technique. Data from the provincial surveillance system targeting symptomatic cases, supported by Médecins Sans Frontières (MSF), were used to assess incidence. RESULTS: District P. falciparum prevalence was of 0.74 % [0.41; 1.21]; village prevalence ranged from 0 to 4.6 % [1.4; 10.5]. The annual incidence of P. falciparum was 16.8 cases per 1000 inhabitants in the district; village incidence ranged from 1.3 to 54.9 for 1000 inhabitants. Two geographical clusters with high number of cases were identified by both approaches. The marker for artemisinin resistance was found in six samples out of the 11 tested (55 %). 34.9 % of qPCR blood analysis of symptomatic patients were still positive at D28. CONCLUSIONS: The overall low prevalence of P. falciparum was confirmed in Chey Saen district in Cambodia, while there were important variations between villages. Symptomatic cases had a different pattern and were likely acquired outside the villages. It illustrates the importance of prevalence surveys in targeting interventions for elimination. Mutations in the k13-propeller domain gene (C580Y), conferring artemisinin resistance, were highly prevalent in both symptomatic and asymptomatic cases (realizing the absolute figures remain low). Asymptomatic individuals could be an additional reservoir for artemisinin resistance. The low effectiveness of dihydroartemisinin-piperaquine (DHA-PPQ) for symptomatic cases indicates that PPQ is no longer able to complement the reduced potency of DHA to treat falciparum malaria and highlights the need for an alternative first-line treatment.

Plasmodium Parasitemia Associated With Increased Survival in Ebola Virus-Infected Patients.

BACKGROUND: The ongoing Ebola outbreak in West Africa has resulted in 28 646 suspected, probable, and confirmed Ebola virus infections. Nevertheless, malaria remains a large public health burden in the region affected by the outbreak. A joint Centers for Disease Control and Prevention/National Institutes of Health diagnostic laboratory was established in Monrovia, Liberia, in August 2014, to provide laboratory diagnostics for Ebola virus. METHODS: All blood samples from suspected Ebola virus-infected patients admitted to the Médecins Sans Frontières ELWA3 Ebola treatment unit in Monrovia were tested by quantitative real-time polymerase chain reaction for the presence of Ebola virus and Plasmodium species RNA. Clinical outcome in laboratory-confirmed Ebola virus-infected patients was analyzed as a function of age, sex, Ebola viremia, and Plasmodium species parasitemia. RESULTS: The case fatality rate of 1182 patients with laboratory-confirmed Ebola virus infections was 52%. The probability of surviving decreased with increasing age and decreased with increasing Ebola viral load. Ebola virus-infected patients were 20% more likely to survive when Plasmodium species parasitemia was detected, even after controlling for Ebola viral load and age; those with the highest levels of parasitemia had a survival rate of 83%. This effect was independent of treatment with antimalarials, as this was provided to all patients. Moreover, treatment with antimalarials did not affect survival in the Ebola virus mouse model. CONCLUSIONS: Plasmodium species parasitemia is associated with an increase in the probability of surviving Ebola virus infection. More research is needed to understand the molecular mechanism underlying this remarkable phenomenon and translate it into treatment options for Ebola virus infection.

The Merits of Malaria Diagnostics during an Ebola Virus Disease Outbreak.

Malaria is a major public health concern in the countries affected by the Ebola virus disease epidemic in West Africa. We determined the feasibility of using molecular malaria diagnostics during an Ebola virus disease outbreak and report the incidence of Plasmodium spp. parasitemia in persons with suspected Ebola virus infection.

Effect of Artesunate-Amodiaquine on Mortality Related to Ebola Virus Disease.

BACKGROUND: Malaria treatment is recommended for patients with suspected Ebola virus disease (EVD) in West Africa, whether systeomatically or based on confirmed malaria diagnosis. At the Ebola treatment center in Foya, Lofa County, Liberia, the supply of artemether-lumefantrine, a first-line antimalarial combination drug, ran out for a 12-day period in August 2014. During this time, patients received the combination drug artesunate-amodiaquine; amodiaquine is a compound with anti-Ebola virus activity in vitro. No other obvious change in the care of patients occurred during this period. METHODS: We fit unadjusted and adjusted regression models to standardized patient-level data to estimate the risk ratio for death among patients with confirmed EVD who were prescribed artesunate-amodiaquine (artesunate-amodiaquine group), as compared with those who were prescribed artemether-lumefantrine (artemether-lumefantrine group) and those who were not prescribed any antimalarial drug (no-antimalarial group). RESULTS: Between June 5 and October 24, 2014, a total of 382 patients with confirmed EVD were admitted to the Ebola treatment center in Foya. At admission, 194 patients were prescribed artemether-lumefantrine and 71 were prescribed artesunate-amodiaquine. The characteristics of the patients in the artesunate-amodiaquine group were similar to those in the artemether-lumefantrine group and those in the no-antimalarial group. A total of 125 of the 194 patients in the artemether-lumefantrine group (64.4%) died, as compared with 36 of the 71 patients in the artesunate-amodiaquine group (50.7%). In adjusted analyses, the artesunate-amodiaquine group had a 31% lower risk of death than the artemether-lumefantrine group (risk ratio, 0.69; 95% confidence interval, 0.54 to 0.89), with a stronger effect observed among patients without malaria. CONCLUSIONS: Patients who were prescribed artesunate-amodiaquine had a lower risk of death from EVD than did patients who were prescribed artemether-lumefantrine. However, our analyses cannot exclude the possibility that artemether-lumefantrine is associated with an increased risk of death or that the use of artesunate-amodiaquine was associated with unmeasured patient characteristics that directly altered the risk of death.

Malaria PCR detection in Cambodian low-transmission settings: dried blood spots versus venous blood samples.

In the context of malaria elimination, novel strategies for detecting very low malaria parasite densities in asymptomatic individuals are needed. One of the major limitations of the malaria parasite detection methods is the volume of blood samples being analyzed. The objective of the study was to compare the diagnostic accuracy of a malaria polymerase chain reaction assay, from dried blood spots (DBS, 5 µL) and different volumes of venous blood (50 µL, 200 µL, and 1 mL). The limit of detection of the polymerase chain reaction assay, using calibrated Plasmodium falciparum blood dilutions, showed that venous blood samples (50 µL, 200 µL, 1 mL) combined with Qiagen extraction methods gave a similar threshold of 100 parasites/mL, ∼100-fold lower than 5 µL DBS/Instagene method. On a set of 521 field samples, collected in two different transmission areas in northern Cambodia, no significant difference in the proportion of parasite carriers, regardless of the methods used was found. The 5 µL DBS method missed 27% of the samples detected by the 1 mL venous blood method, but most of the missed parasites carriers were infected by Plasmodium vivax (84%). The remaining missed P. falciparum parasite carriers (N = 3) were only detected in high-transmission areas.

Plasmodium prevalence and artemisinin-resistant falciparum malaria in Preah Vihear Province, Cambodia: a cross-sectional population-based study.

BACKGROUND: Intensified efforts are urgently needed to contain and eliminate artemisinin-resistant Plasmodium falciparum in the Greater Mekong subregion. Médecins Sans Frontières plans to support the Ministry of Health in eliminating P. falciparum in an area with artemisinin resistance in the north-east of Cambodia. As a first step, the prevalence of Plasmodium spp. and the presence of mutations associated with artemisinin resistance were evaluated in two districts of Preah Vihear Province. METHODS: A cross-sectional population-based study using a two-stage cluster sampling was conducted in the rural districts of Chhaeb and Chey Saen, from September to October 2013. In each district, 30 clusters of 10 households were randomly selected. In total, blood samples were collected for 1,275 participants in Chhaeb and 1,224 in Chey Saen. Prevalence of Plasmodium spp. was assessed by PCR on dried blood spots. Plasmodium falciparum positive samples were screened for mutations in the K13-propeller domain gene (PF3D7_1343700). RESULT: The prevalence of Plasmodium spp. was estimated at 1.49% (95% CI 0.71-3.11%) in Chhaeb and 2.61% (95% CI 1.45-4.66%) in Chey Saen. Twenty-seven samples were positive for P. falciparum, giving a prevalence of 0.16% (95% CI 0.04-0.65) in Chhaeb and 2.04% (95% CI 1.04-3.99%) in Chey Saen. Only 4.0% of the participants testing positive presented with fever or history of fever. K13-propeller domain mutant type alleles (C580Y and Y493H) were found, only in Chey Saen district, in seven out of 11 P. falciparum positive samples with enough genetic material to allow testing. CONCLUSION: The overall prevalence of P. falciparum was low in both districts but parasites presenting mutations in the K13-propeller domain gene, strongly associated with artemisinin-resistance, are circulating in Chey Saen.The prevalence might be underestimated because of the absentees - mainly forest workers - and the workers of private companies who were not included in the study. These results confirm the need to urgently develop and implement targeted interventions to contain and eliminate P. falciparum malaria in this district before it spreads to other areas.

Low referral completion of rapid diagnostic test-negative patients in community-based treatment of malaria in Sierra Leone.

BACKGROUND: Malaria is hyper-endemic and a major public health problem in Sierra Leone. To provide malaria treatment closer to the community, Médecins Sans Frontières (MSF) launched a community-based project where Community Malaria Volunteers (CMVs) tested and treated febrile children and pregnant women for malaria using rapid diagnostic tests (RDTs). RDT-negative patients and severely ill patients were referred to health facilities. This study sought to determine the referral rate and compliance of patients referred by the CMVs. METHODS: In MSF's operational area in Bo and Pujehun districts, Sierra Leone, a retrospective analysis of referral records was carried out for a period of three months. All referral records from CMVs and referral health structures were reviewed, compared and matched for personal data. The eligible study population included febrile children between three and 59 months and pregnant women in their second or third trimester with fever who were noted as having received a referral advice in the CMV recording form. RESULTS: The study results showed a total referral rate of almost 15%. During the study period 36 out of 2,459 (1.5%) referred patients completed their referral. There was a significant difference in referral compliance between patients with fever but a negative RDT and patients with signs of severe malaria. Less than 1% (21/2,442) of the RDT-negative patients with fever completed their referral compared to 88.2% (15/17) of the patients with severe malaria (RR = 0.010 95% CI 0.006 - 0.015). CONCLUSIONS: In this community-based malaria programme, RDT-negative patients with fever were referred to a health structure for further diagnosis and care with a disappointingly low rate of referral completion. This raises concerns whether use of CMVs, with referral as backup in RDT-negative cases, provides adequate care for febrile children and pregnant women. To improve the referral completion in MSF's community-based malaria programme in Sierra Leone, and in similar community-based programmes, a suitable strategy needs to be defined.