Progress towards malaria elimination in Sabang Municipality, Aceh, Indonesia

Malaria is a major public health problem in most tropical countries, including Indonesia. Recent models estimate that over 105 million of Indonesia’s 239 million population are at risk for malaria infection, with transmission varying widely across this most populous entirely tropical country[1]. Though formally reported deaths are less than 1,000 annually[2], recent work estimates that about 11,000 people die annually due to infections with Plasmodium falciparum[3]. The number of deaths caused by Plasmodium vivax infection is not known. The most recent WHO estimate of the number of deaths due to malaria is approximately 3,000 per year in Indonesia[2]. In 2009, the Indonesian MOH explicitly made malaria elimination a national goal, to be accomplished in a step-by-step, island-by-island fashion over the next several decades[4].

Sabang municipality in Aceh Province was once one of the most malarious areas in Indonesia[5]. Now, after successful control, with only sporadic focal transmission occurring, Sabang’s government[6], with support from provincial and central authorities, is committed to eliminate malaria transmission by the end of the year 2013[7].

This effort is in line with the recent rise of malaria elimination in the global public health agenda, signaled by the commitment of the 60th World Health Assembly in 2007 that all countries should commit to eliminate malaria by 2050[8]. Indonesia aims to eliminate malaria by 2030[4], while the target for low-endemic Aceh Province is 2015[4, 7]. In 2009, the Provincial Parliament of Aceh endorsed this plan, and budgeted significant resources towards its achievement. In addition, the Governor of Aceh has provided legal support to the plan via a new regulation promulgated in 2010[7].

If the political and legal framework for malaria elimination in Indonesia and Aceh is in place, clear technical challenges remain. Indonesia straddles Wallace’s line, and has perhaps two dozen known malaria vectors with a large variety of behaviours. Further, Indonesia has a high prevalence of both P. falciparum and P. vivax, which tremendously complicates the elimination effort[9, 10]. Though decentralization of the government in 2000 (including decentralization of the health system) has resulted in some deterioration of services, some well-governed districts have risen to the challenge and have improved health service delivery[11].

In the context of Indonesia, Sabang District is an appropriate place to pilot malaria elimination by virtue of its geographic position at the western end of the archipelago, its diverse mosquito fauna, the presence of both major malaria parasites, and its strong local government. The rate of importation of malaria into Sabang is likely low, giving hope that the municipality will be able to maintain elimination in the long run. Sabang has, in general outline, followed the recommendations of the WHO for malaria elimination[8], while recognizing the necessity to adapt these guidelines according to circumstances on the ground in Sabang.

The effort has been led by the government of Sabang, with support from the Provincial Health Office of Aceh, the National Malaria Control Program of Indonesian Ministry of Health, University of Gadjah Mada, Eijkman Institute, the World Health Organization (WHO) and the United Nation Children Fund’s (UNICEF).

This account begins with a description of what WHO terms the ‘control’ phase of malaria control[8], in which post-tsunami funding allowed rapid scale-up of IRS and LLINs, improved malaria diagnosis, and treatment with ACT. The rapidly achieved high coverage ot these interventions resulted in a quick decline in malaria cases, but these persist at a low level[6]. The reorientation of activities aimed at moving the programme towards malaria elimination is then described. These efforts included intensive mapping of both vectors and malaria cases, establishment of a comprehensive database, and most recently, aggressive active case finding coupled with surveys to estimate G6PD prevalence and malaria parasite genotypes.

In this study, interventions were described which have dramatically reduced the burden of malaria in Sabang over the past seven years. The effect on malaria incidence has been dramatic, with a nearly 30-fold decline in incidence from 3.83 to 0.13 per thousand population from 2008 to 2011[33]. These now-standard interventions include improved malaria diagnosis, introduction of ACT for malaria treatment, and scale-up of coverage of IRS and LLINs. The backbone of the Sabang programme has been strengthening of its diagnostic system. In 2009, twenty microscopists active in the programme were tested by an expert team according to WHO guidelines; four of the Sabang microscopists tested as expert (level 1), three at reference (level 2), eight at advanced (level 3), and two at basic (level 4). These test results show that Sabang has succeeded in training a cadre of highly proficient microscopists. By 2009, blood slides from 96% of suspected malaria cases were examined by this capable team – up from 66% in 2004. The small number of cases not confirmed via microscopy was tested using RDTs. Near universal coverage of microscopic diagnosis was coupled with introduction of use of ACT for treatment of all forms of malaria (including P. vivax) in 2005. By the end of 2009, 94% of laboratory-confirmed cases of malaria in Sabang were treated with ACT[17].

After the prohibition and cessation of DDT spraying in Indonesia in 1992, only occasional IRS was carried out in Sabang until the post-tsunami period when, from 2005 – 2007, the government of Sabang, in collaboration with the Mentor Initiative (an international NGO), carried out three-times yearly spraying with alphacypermethrin in 18 villages, with reported coverage of 50-60%[34]. LLINs were introduced to Sabang after the tsunami, with more than 15,000 nets distributed since 2005 by the government of Sabang in collaboration with CARE International and UNICEF[35]. While coverage of LLINs in the six focal villages remains high, at over 75%, data from previous surveys are mixed[23]. In July 2005, Samaritan’s Purse and Mentor reported that 98% of households owned at least one LLIN[34]. A survey by community volunteers conducted in February 2006 showed that 88% of households owned an LLIN, with reported usage rates by children under five years of age of 74%[35]. In contrast, Demographic Health Survey (DHS) results for 2007 showed that only 22% of households owned an LLIN. The likely reason for this discrepancy is that the DHS results included urban parts of Sabang, which were not targeted by the LLIN campaigns, while the previous surveys used as their sampling frame only rural areas targeted for LLIN distribution[36]. In addition to interventions targeting adult mosquitoes, focal application of the larval growth regulator pyriproxyfen was carried out in five villages from 2005 – 2007[6].

Sabang, in accordance with official policy of the MOH of Indonesia, has taken the bold step of explicitly aiming to eliminate malaria transmission by the end of 2013[17]. The government of Sabang is pursuing this policy aggressively, and recognizes that the key to elimination is rapid reporting and accurate information. As described above, the routine diagnosis capability with the MOH has been steadily strengthened over the past seven years. Since mid-2009, private practitioners are asked to send blood smears to government clinics for malaria diagnosis. In May 2010, the MOH set up a system of village volunteers for active case detection. One volunteer has been recruited for each subvillage. To date, 78 volunteers have been recruited, trained, and certified by the government. In the 14 villages classified as ‘active’ or ‘residual active’ (foci A and B in Figure6) home visits are conducted twice monthly, while in non-active foci monthly visits are carried. Blood smears are taken from anyone with fever or recent history of fever and sent to the nearest community health centre. Confirmed cases are treated by health centre staff at the patient’s home, using directly observed treatment. The volunteer takes follow-up smears on days 3, 7, 14, 28 and (for P. vivax only) 90[17].

Since August 2010, all cases of malaria detected through passive or active surveillance, via either government or private practitioners, are investigated by staff from the community health centre and the municipal health office. The investigation is adapted from the procedure used in Zanzibar[37], with four steps: 1) Interview to determine travel history, blood transfusion history, or contact with malaria patients; 2) Contact survey – taking of blood smears from family members and all neighbors within 500 meters of the index case; 3) Rapid assessment of anopheles larval habitats and biting activity by the district’s entomology team; 4) Input of information into the malaria database and classification of the case as indigenous or imported. Since 2009, every case, whether detected passively or actively, has been included in the database, with information on age, sex, occupation and other aspects that may influence a person’s malaria risk); history of disease (travel history, fever/malaria history, contact history); all information related to malaria diagnosis; treatment; notification of case and epidemiology investigation (indigenous, import, induction, introduction, relapsed), source of cases (passive case detection, active case detection, survey), location (including GPS coordinates), and nearby Anopheles larval habitats[17].

The emphasis on mapping larval habitats near transmission is appropriate in the context of Indonesia, as some, but not all, malaria vectors in the country are amenable to larval control via environmental management. In the malaria database, information on coverage of LLINs and IRS is linked to foci of transmission[23], so that interventions targeting larvae may be carried out should transmission continue even though coverage of interventions targeting adults is high. Effective larval control might also allow the possibility for scaling back in IRS and LLIN coverage, as continuous high coverage is difficult to support politically as malaria incidence declines to near zero levels[38].

In Sabang, as throughout Indonesia and much of southeast and south Asia, both P. falciparum and P. vivax are common[9, 10, 39]. The steps taken after successful control of malaria in Sabang – improved reporting, coordination with the private sector, active case detection, mapping of cases and mosquito habitats – may be sufficient to eliminate malaria on the island. However, results of our mass screening of nearly half of the population of Sabang raise doubts on the likelihood of success, especially for P. vivax. Of note, only three of the 19 cases detected in the survey were symptomatic, and these were all detected with microscopy. Of 16 asymptomatic cases, five were detected with microscopy, while the balance of 11 cases was detected with PCR. Of these 11 asymptomatic cases, 10 were pure P. vivax infections, and all of these were detected via PCR. Thus, for both species of parasite, but more so for P. vivax, subpatent, asymptomatic infections were found via PCR that were not detectable with microscopy[24]. Indeed, by extrapolation from this survey, about 165 asymptomatic infections of malaria - mostly P. vivax -- were present in Sabang at the time of the survey, all undetectable by the microscopy. Even though Sabang’s programme is based upon rapid detection, reporting, and intervention with diagnosis via conventional microscopy, it may still be possible to eliminate malaria on the island without resorting to mass treatment or expensive screening via PCR. Whether elimination is possible will depend upon the balance between the efficiency of the programme, and the rate at which subpatent, asymptomatic infections recrudesce – allowing detection and attack via medication – or spontaneously disappear[39‐41]. Both processes are known to occur. An important third unknown parameter is the probability of transmission of subpatent infections to mosquitos[42]. Though presumably low, this probability can be further minimized via vector control interventions targeting both adults and larvae. These three parameters presumably vary between the two parasite species. The higher prevalence of asymptomatic P. vivax infections, coupled with the operational difficulty of radical treatment of hypnozoites, makes is likely that this parasite will be more difficult to eliminate than P. falciparum[8, 39, 43, 44].

Aggressive efforts to treat P. vivax carry risks associated with G6PD deficiency[45, 46]. Though reported prevalence in Sabang is low, based upon results of our survey, it is possible that other parts of Indonesia[29‐31, 47] or other parts of Asia[30, 48], may have higher levels of this risk factor. Minimization of this risk will require availability of cheap and effective rapid screening tests for G6PD deficiency or better anti-hypnozoite drugs, but neither of these is available now[39, 49‐51]. Further, the current anti-hypnozoite drug, primaquine, must be taken over 14 days by patients that already feel recovered, raising serious problems with patient compliance[50, 51].

Further complicating malaria elimination efforts in Sabang is the high mobility of Indonesians, and the fact that many Indonesians may seek employment in parts of Aceh, Kalimantan or eastern Indonesia, where risk of malaria infection is substantial. A recent imported case of malaria to Sabang in a gold miner illustrates this point. Sabang will need to establish good community based systems for screening and detection of malaria in residents returning from malaria endemic areas. Such a system has been established in Wonosobo, central Java[52], which sends many temporary workers to malaria-endemic Kalimantan. Sabang has, as yet, not set up any systems to specifically detect imported cases. Perhaps of greater importance are efforts on the part of the MOH to reduce malaria incidence throughout the country. Malaria incidence has declined in eastern Indonesia thanks to scale up of many of the same interventions rolled out in Sabang, but many pockets of high transmission remain, and these are all potential sources of imported malaria to Sabang.

All surveillance activities described in this report were consistent with and conform to policies set by the Ministry of Health of the Republic of Indonesia.

These studies were funded by UNICEF Indonesia.