Background
The global burden of malaria has fallen substantially since the beginning of the millennium (37% global decrease between 2000 and 2015), primarily due to improvements in access to diagnosis and treatment and increased coverage of insecticide-treated bednets [
1]. Many of the estimated 216 million malaria cases and 445,000 deaths in 2016 occurred in remote areas of the tropics where health services are weak or non-existent [
2]. Myanmar has the greatest burden of malaria in the Greater Mekong Subregion. Much of the burden of malaria is borne by remote and hard-to-reach communities [
2]. The health care system in these communities has been weakened by decades of conflict and under-investment. The infrastructure is poor, and most villagers have no access to trained health staff and resort to informal health care providers (known by the colloquial term “quack” in Myanmar). These providers are generally untrained villagers who sell medicines unofficially. Over the past 6 years, investment in rural health care services has increased substantially [
3], supported by both the national government and international donors, in particular from
The 3 Millennium Development Goal Fund (a consortium of bilateral donors for Myanmar) and
The Global Fund to Fight AIDS, Tuberculosis and Malaria. There has been particular support for malaria control (community health workers (CHWs) and long-lasting insecticide-treated net (LLIN) distribution) [
4,
5]) for hard-to-reach communities. As a consequence, the national malaria incidence is estimated to have decreased by 49% since 2012: from 8.1 per 1000 to 4.2 per 1000 in 2015 [
6].
Community-based malaria management using rapid diagnostic tests (RDTs) and good quality treatment, implemented by CHWs, can substantially improve malaria management in remote communities. CHWs can also distribute LLIN [
7]. This approach has been successful in reducing malaria in many settings [
8‐
11] although it has failed in other contexts where community uptake of services was not maintained [
12‐
14]. Maintaining strong community uptake of CHW services is essential to sustain blood examination rates and good malaria control [
12,
13,
15,
16] and to provide longitudinal surveillance data from case detection to estimate the true malaria incidence [
17]. However, as malaria transmission and incidence falls, the proportion of febrile cases that are malaria, and receive antimalarial treatment, declines correspondingly. Community uptake of CHW programmes that offer only malaria testing and treatment is therefore likely to reduce. Patients attend a health worker because they have fever for which they want treatment. As the probability that a febrile patient has malaria and receives treatment declines, the perceived value of the malaria-only CHW programme declines too, and so the CHWs may become victims of their own success. In order to be appreciated and therefore used by their communities, CHW programmes need to provide relevant services of perceived value and benefit. To achieve this, ‘malaria-only’ CHWs can broaden their service to provide an integrated health care package to address common health problems.
Medical Action Myanmar (MAM), a medical aid organisation, has been supporting a network of CHWs in the most remote communities in Myanmar. These initially provided malaria control activities exclusively. However, as malaria decreased rapidly in these communities, CHWs were trained to offer an extended basic health care (BHC) package in addition to the malaria services. This was initiated in 2013 and 2014 in an effort to meet health needs and to ensure continued community participation with the CHW malaria programme. This extended BHC package incorporated treatment of malnutrition, diarrhoea, and respiratory tract infections, in line with the integrated community case management statement of the World Health Organization (WHO) and United Nations International Children’s Fund (UNICEF) [
18]. It also supported referral of severely ill patients and patients suspected to have tuberculosis to the nearest hospital.
All countries in the Greater Mekong Subregion have set national targets for malaria elimination in the near future. Despite interest in integrated CHW programmes, and their proven efficacy in reducing morbidity and mortality from acute respiratory infections [
19‐
23] and diarrhoeal disease [
19,
21,
23], there is relatively little information on their efficacy in malaria control and their effect on community uptake in the context of malaria elimination [
24]. We conducted a retrospective analysis of 1335 MAM-supported CHWs operating in Myanmar between 2011 and 2016. We assessed the rates of decline in
Plasmodium falciparum and
Plasmodium vivax malaria incidence and RDT positivity with each year of CHW operation. In addition, we investigated the effects of the addition of a BHC package on the uptake of malaria services in four cohorts of CHWs which had provided malaria services only for at least 1 year.
Discussion
Malaria elimination is now firmly on the agenda in much of the tropical world [
26], and many countries, particularly in Southeast Asia, have set ambitious timelines for achieving it. In Myanmar, as in many low- and middle-income countries, CHWs are key to the delivery of malaria diagnosis and treatment in remote areas. This retrospective analysis clearly demonstrates the substantial public health benefit, in terms of reduced malaria incidence, of instituting trained CHWs in remote malaria-affected villages. The two-thirds annual reduction in incidence in these communities compares with an estimated decrease of 20% per year for the overall Myanmar yearly malaria incidence between 2012 and 2015 [
6]. Thus, with relatively little training, but reliable supplies and careful monitoring, these community members provided a popular and highly effective public health service. The most likely explanation for the CHWs’ success in reducing malaria is that provision of community-based early diagnosis and effective quality-assured treatment, coupled with LLIN distribution, significantly reduced malaria transmission.
The continued function of CHWs is essential to drive malaria to elimination, and to ensure continued monitoring of any imported malaria cases once local transmission has ceased. But having achieved large reductions in malaria, the CHW who treats only malaria becomes increasingly inactive and irrelevant as fewer and fewer patients with acute febrile illness are diagnosed with malaria and receive specific treatment. From the febrile patient’s perspective, it matters little what the cause of their fever is, as long as it is treated effectively, so the incentive to seek treatment from the “malaria only” CHW declines as malaria incidence declines. Malaria-only community health workers recognise that when malaria declines, febrile patients are less likely to consult them, and wish to be able to provide help to patients with non-malaria fever [
27]. Patients return to untrained informal health care providers, who provide inappropriate medicines and interventions. This was our concern when the numbers of patients seeking malaria RDTs in remote Myanmar villages declined.
To improve access to health care and to remedy the declining consultation rate, the remit of the CHWs was broadened to include other common febrile illnesses and referral of severely ill patients to the nearest hospital. This initiative was welcomed by the communities as non-malaria febrile illnesses now received a specific treatment as well, and for malaria control activities, it had the major advantage of sustaining quality. In this retrospective assessment, there was a decrease in health seeking behaviour and malaria RDT testing after malaria transmission had gone down, but this trend reversed after broadening the health care package of the CHWs. A limitation of this study was that it was observational, not experimental. However, the consistency and magnitude of the estimates observed, across four separate cohorts, suggest that the improvements in malaria RDT testing rates are real, accurate, and representative of what could be expected in future implementations of integrated care. Of note, we observed different magnitudes of effect across different cohorts, suggesting that the context of a given community will be an important factor in service uptake. Cohort 4 had the smallest magnitude step-increase in blood examination rate, though in sensitivity analyses this estimate was higher when seasonality was not accounted for, or was accounted for using first-order autoregression (Additional file
2). Cohort 4 also had the highest malaria RDT positivity rate prior to BHC package introduction; a small step-increase was consistent with an expectation that this intervention will be most effective in communities that feel malaria is no longer a primary concern. Interrupted time series analyses are well suited to the evaluation of intervention effects in real-world settings [
28], and as it seems unlikely that the incidence of febrile illnesses changed coincidentally with the change in CHW practice, a causal relationship is likely.
Most health professionals recognise that CHWs can have an important positive effect on malaria diagnosis and treatment in the community and that the success of community case management can be replicated for other common diseases. WHO and UNICEF outlined their support for integrated community case management in 2012 [
18], noting that appropriately trained, supervised, and supported CHWs can identify and correctly treat most childhood respiratory tract infections and diarrhoea [
29]. It has been estimated that community management of childhood pneumonia could reduce mortality from pneumonia in children less than 5 years old by 70% [
22]. Oral rehydration salts and zinc reduce the mortality of diarrhoeal disease in community settings; community promotion of oral rehydration salts was estimated to reduce the number of deaths due to diarrhoea by 69% (95% CI 51–80%) [
30], and zinc supplementation is estimated to decrease diarrhoea mortality by 23% (95% CI 15–31%) [
31]. The approach of integrating disease-specific programmes with other health services (the “diagonal approach”) is likely to be appropriate for other narrow disease-specific programmes, which can also see their sustainability threatened by their own success, and is critical for effective health systems strengthening [
32,
33].
During the period of this study, financial incentive schemes for CHWs in Myanmar varied across different organisations; the value of incentives offered by MAM was neither the highest nor the lowest, among organisations supporting Myanmar CHWs [
34]. There are concerns that per test incentives that are too high can have a negative influence on CHW diagnostic practices. Conversely, when incentives are too low or absent, the CHW may stop testing altogether and retention may become difficult. More research is needed to identify the correct balance of monthly incentive and test incentive in low-income areas. Ultimately, frequent monitoring and supervision are essential under any incentive scheme to maintain high standards of care and to check for aberrant practices.
Some policy makers worry that CHWs are not capable of providing the correct diagnosis for patients with non-malaria fevers. In several countries, CHWs are allowed to prescribe antimalarials but not antibiotics. Health professionals fear that CHWs may become the new generation of informal health care providers, overprescribing unnecessary antibiotics without proper diagnosis. Instead, they suggest that it is better to refer patients with non-malaria febrile illnesses to the nearest government health service. However, in remote areas with poor infrastructure, numerous barriers of geographical access, availability, affordability, and acceptability hamper access to government health services [
35,
36], and such a strategy is simply not feasible. Even when trusted CHWs refer patients to the nearest hospital, many will not go [
37,
38]. Smaller villages are generally more remote and face more barriers to health service access so typically have a higher uptake of CHW services. It is in these villages that integration of community-based health services is most important. Integrated CHWs can be carefully regulated, which could diminish irrational antibiotic use, prevent antimicrobial resistance, and substantially reduce morbidity and mortality in a cost-effective manner.
This retrospective analysis includes over half a million malaria RDT results from a large number of communities (1335) over 5 years. The downward trends of incidence for both P. falciparum and P. vivax malaria after the introduction of community-based malaria management are substantial and convincing. By contrast, the analysis of the impact of the introduction of a BHC package was performed only in a relatively small number of cohorts (four) including 154 communities. This was unavoidable as the number of CHWs with enough data before and after introduction of the BHC package was limited, but the results are likely to be relevant to other remote communities.
Acknowledgements
We thank all MAM field staff and community health workers in remote communities who were essential for the success of the programme, the National Malaria Control Programme, the Karen Department of Health and Welfare, Community and Health Development Network and staff from United Nations Office for Project Services in Myanmar, the 3 Millennium Development Goal Fund, and The Global Fund to Fight AIDS, Tuberculosis and Malaria. We thank Julie Simpson for her valuable feedback and comments. MOCRU is part of the Mahidol Oxford Research Unit, supported by the Wellcome Trust of Great Britain.