Progress and challenges to control malaria in a remote area of Chittagong hill tracts, Bangladesh

Malaria is estimated to be responsible globally for a million deaths every year, and even though 90% of mortality occurs in Africa, it remains a major health threat in South-Asian countries, including Bangladesh. There, it is endemic in 13 eastern districts where the overall infection prevalence is 3.97%. The infections are predominantly Plasmodium falciparum (90.12%) and Plasmodium vivax (5.3%), with 4.5% mixed infections. The overall prevalence in the Chittagong hill tracts (three south-eastern districts) is 11.7%, reaching 36% in a single sub-district, Rajasthali [1].

The malaria control programme in Bangladesh faces formidable challenges, including access to quality healthcare services, inadequately trained personnel, difficulty in travel, a lack of resources and education for the population at risk, and life styles depending on subsistence activities. Health facilities to manage severe malaria also are limited, surveillance is inadequate, and interventions are insufficient [2].

Many studies have described socioeconomic, demographic and environmental risk factors as part of malaria-related knowledge, attitude, and practices studies [3‐9] and while bed nets are usually considered a protective measure, they may be used insufficiently [10] or reported as ineffective [11‐15]. However, recent studies indicate that insecticide-treated nets (ITN) reduced 48-50% of malaria episodes [16] and are considered one of the most cost-effective health interventions against malaria [11, 12].

Baseline information for the implementation of ITN has shown that people were aware of malaria infection, transmission, anti-malarial drugs and malaria control [13]. WHO has recommended full coverage of all people at risk in areas targeted for malaria prevention with long-lasting insecticide-treated nets (LLINs). Especially for endemic areas, LLINs should be delivered to all people and should initially focus on priority target areas [14].

Little is known about treatment-seeking behaviour among indigenous people infected with malaria in Bangladesh. In a baseline survey, treatment-seeking at hospitals was rare, self-treatment was common and people commonly took drugs without consulting a qualified doctor. In five south-eastern districts, 32.3% people preferred to get treatment from drug vendors [3]. The choice of treatment source was related to distance from hospital, disease type, patient's gender and parent's education level. People also preferred to receive malaria treatment from the nearest health workers [17]. These results were similar to previous studies [18] indicative of suboptimal treatment regimes.

In 2006, Bangladesh received US $ 36.9 million from the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM) to support a national malaria control programme that would integrate rapid diagnosis tests (RDTs), new drug regimens (artemisinin-based combination therapy (ACT)), expanded distribution of LLIN, enhanced surveillance, vector surveillance and better documentation of activities. Bangladesh adopted artemether-lumefantrine (AL)(Coartem^®) as a first-line treatment of P. falciparum malaria and has sought to provide early diagnosis and prompt treatment to 80% of malaria patients. Other intervention objectives included effective malaria prevention to 80% of the population at risk and to strengthen the malaria epidemiological surveillance system. It was estimated that 40% of the households in high risk areas had nets, 10% of which were insecticide treated. Through this grant, it was expected that 80% of households (1.7 million) would be covered with LLINs. Nets currently present in households would be treated and re-treated twice a year with insecticide [19].

BRAC and the Ministry of Health implemented the national malaria control programme under GFATM and BRAC would be responsible for supplying LLIN to 80% household, as well as deploying health workers in every union to provide RDT and AL at the grass root level. This study was undertaken in Rajasthali, where the highest prevalence of infection had been detected to clarify factors associated with treatment seeking behaviours of malaria, distribution of LLIN, and re-treatment of ITN in remote area of a CHT district of Bangladesh two years after implementation of national control programme.

The population of Rajasthali was 24,097 with a slight preponderance of males (50.3%). There were six different ethnic communities in Rajasthali - Marma, Tripura, Tonchonga, Khiang, Chakma and Bengali. Marma was the predominant and Bengali was the second largest ethnic group. Almost half of the heads of household (46.90%) were illiterate, and only 5.70% of household heads had completed more than 10 classes of education. The majority (50.50%) of household heads were farmers (Table 1).

LLIN were provided to 90.09% of households by BRAC and 76.40% households had more than two bed nets. BRAC had treated bed nets in 40.98% of households in the six months prior to survey. It was found that 81.22% of people preferred to sleep under a bed net.

There were diverse treatment-seeking behaviours though nearly two thirds of people (66.60%) preferred to consult with trained health workers or doctors in a government hospital (control programme). A large minority of people (48.80%) also reported they received treatment from drug vendors. A small fraction (1.70%) preferred other choices (taking no measures, home-made medicine, kabiraj (who provide traditional herbal treatment in rural areas in Bangladesh), homeopathy and village doctors.

Access to care, as characterized by distance appeared to play a significant role in treatment preference. A large proportion of households lived within 2 km of health facilities supported by national malaria control programme (77.95%), as well as drug stores (53.70%) (Table 1). People who lived within 2 km of health facilities led by the malaria control programme preferred it compared with people who lived more than 2 km away from health facilities. As the distance increased (≥ 2 km), people were less likely to prefer the control programme's health facility (OR = 0.48; 95% CI: 0.40-0.08), but, instead, preferred drug vendors (OR = 2.02; 95% CI: 1.71-2.39). There were also ethnic differences in the utilization of health care facilities, study result confirm that people from Tonchonga ethnic community prefer most the malaria control program for treatment (OR = 7.07; 95% CI: 5.57-8.98) while people of Tripura ethnic community (OR = 3.12; 95% CI: 2.41-4.05), and Khiang ethnic community (OR = 2.46; 95% CI: 1.78-3.41) prefer the malaria control programme for treatment. Compared with other communities, Bengali people were more likely to seek treatment from drug vendors (Table 2). These heterogeneities were also associated with spatial clustering in health seeking approaches. There were two clusters where people preferred to use the malaria control programme. One was most likely cluster and another one was secondary cluster (Table 3; Figure 1). The most likely cluster was in the central-western part of the study area (RR = 2.24; P = 0.001). There were also several isolated villages in eastern part where people did not prefer treatment provide by malaria control programme. There was also a cluster favouring using drug vendors (RR = 2.97;P = 0.001) in the western part of Rajasthali. Eight secondary clusters were observed in the study area where people preferred drug vendors for treatment. Among them six were statistically significant (Table 3). All secondary clusters were in the southern and eastern part of Rajasthali (Figure 2). People who were involved in the service and business industries seemed to have a better treatment-seeking approach compared with people involved in agriculture.

This is the first independent evaluation of a GFATM supported malaria control programme in a malaria endemic remote area of Bangladesh. According to published literature another study in Africa has examined GFATM supported malaria control programmes [21]. The analysis indicates significant levels of success in the delivery of interventions to the communities most at risk. However, a number of challenges still remain for the control of malaria in this region of Bangladesh.

The target goal for the supply of LLIN was 80% of households in the malaria endemic areas [2] were surpassed as we have found that LLIN were supplied to 90.09% of households by BRAC. Additionally, during the study BRAC had treated 40.98% of household ITN/bed nets within the previous six months - substantially ahead of the two year timeline that was anticipated [2]. The final programmatic metric, based on the national control programme, was that every household should have more than two bed nets [2]. This goal has not yet been reached, with slightly more than a third of households lacking a second LLIN. These households tended to occur in more isolated regions. In hard to reach areas, a visualization approach using geographic information system (GIS) can be applied for target intervention. There were six significant clusters of households who have less than two bed nets (Table 4). Large clusters of households who have less than two bed nets were in the eastern part of the study area (Figure 3) where access is most limited by the landscape.

In sub-Saharan Africa, recent studies show that 66% people prefer to be treated for malaria in a health care facility, 19% people prefer to buy drugs from a shop and a non-significant proportion of people take medicine from traditional healers, use herbal medicines, self-treatment or have no treatment [22]. Despite the overwhelming preference for health care facilities, and free treatment in government hospitals, people continue to use private clinic/drug vendors at high rates [23]. Results obtained in south-eastern part of Bangladesh are similar to that of sub-Saharan Africa. The widespread distribution of BRAC health workers provide better coverage than other health service providers and they can provide treatment to the majority of the population. However, at a relatively short distance (2 km) individuals appear to choose alternative health care sources. Nearly half of the people preferred drug vendors who provide anti-malarial drugs for fever without diagnosis. In the 2007 malaria baseline survey in the CHT districts, it was found only 51.5% (103/200) fever cases were positive for malaria. Total subjects enrolled in CHT for that survey was 2250. This suggests a significant potential for improper treatment with ACT or other anti-malarials from drug vendors in the region.

It is still unclear why people continue to prefer to pay drug vendors for treatment when the malaria control programme provides diagnosis and treatment in the community at free of cost.

The target goal for the supply of LLIN and retreatment of ITN were surpassed. Distance from the national malarial control programme facilities may be a factor although some people preferred to receive treatment from a drug vendor rather than the malaria control programme, even if the drug vendor was more than 2 km away. This suggests that factors related to social interactions with the vendor, such as history, consistency of availability or other factors may come into play. At least some of these factors may be associated with ethnic experiences and the bases for these clustering's need further examination. Unsurprisingly educational experiences were associated with preferences for the use of malaria control programme services. A high proportion of people preferred drug vendors without having a proper diagnosis. Drug vendors are highly patronized and thus there may be the need to improve their services for public health good. Otherwise it may cause incomplete treatment, misuse of anti-malarial drugs, contribute to the risk of drug resistance and jeopardize the present malaria control efforts in Bangladesh.

The authors hereby certify that no conflict of interest of any kind occurred in the framework of this study.