Knowledge, Attitudes, Practices and Risk factors about Malaria in High Endemic Rural Eastern Tanzania

Background: Understanding area-specic knowledge, attitudes, practices and demographic risk factors towards malaria is becoming critical for designing appropriate awareness and control strategies. Methods: A cohort of respondents from 243 randomly selected households across 3 villages in eastern Tanzania was assessed using a standard questionnaire and direct observation. Results: Fifty-three percent of respondents cited Anopheles mosquitoes that were responsible for transmitting malaria, and 85% of those explicitly cited female Anopheles mosquitoes. The rest (47%) knew that malaria is transmitted by a mosquito but could not name it. Commonly cited mosquito breeding sites were water bodies (75%) and rubbish (24%). Majority (61%) knew that the under-ves were most vulnerable to malaria. ITNs were mostly (64%) cited as a means of malaria control. Commonly cited malaria mosquito entry routes were windows (46%), eave opening (10%) and opening on the walls (4%). Majority (70%) conrmed malaria at the health facility. The rest (30%) relied on symptoms. Majority (96%) preferred modern antimalarials. Seventy-nine percent of the respondents owned at least one ITN. Cooking was done predominantly (77%) outdoors, nished cooking and went to bed after 20:00h. Most (58%) respondents still attended night-time outdoor communal gatherings without malaria protection. Conclusions: The current study indicate that majority of participants were aware that Anopheles mosquitoes were responsible for transmitting malaria, they bred in water, under-ve children were most vulnerable to malaria, and ITNs constituted the most deployed control measure. Most respondents preferred conrming malaria at the health facility and treatment with modern antimalarials. Yet, the rate of self-diagnosis and self-medication, as well as inappropriate use of bednets requires great improvement. Demographic risk factors for early- and/or outdoor-transmission included night-time outdoor cooking and attending communal gatherings without malaria protection. The ndings warrant a subsequent study to show a causal link between night-time outdoor activities, vector biting patterns and malaria prevalence.


Introduction
Malaria remains a major cause of morbidity and mortality in tropical and subtropical regions of the world. In 2018, an estimated 228 million cases and 405 thousand deaths occurred worldwide. World Health Organization (WHO) African countries accounted for 93% of the cases and 94% of the deaths [1].
Tanzania is among the top ten countries with high malaria transmission and the population at risk in eastern and southern Africa [1,2,3]. Malaria is endemic in almost throughout the country, with over 95% of the population at risk of infection. The disease causes approximately 7.7 million con rmed and clinical cases in the country annually [2,4].
In Tanzania like other malaria-endemic countries, disease control efforts consist of timely diagnosis and treatment as well as the deployment of mosquito vector control interventions. The most frequently and widely used vector control interventions are long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS). LLINs and IRS in combination with timely diagnosis and appropriate treatment have halved malaria burden in the country [4]. Despite the progress made so far, the current control strategies in Tanzania and elsewhere in Africa are increasingly constrained by several challenges, the most important of which is the development of resistance to virtually all classes of insecticides used against malaria vectors [5][6][7][8][9]. Furthermore, the level of knowledge, attitudes and practices concerning the use of malaria control interventions varies greatly from one place to the other; and so is the disease prevalence.
Several studies have shown that malaria vector distribution, transmission rates and incidence can vary widely over short distances, between neighboring villages and even within a single settlement, as a result of small area variations in risk factors [10][11][12][13][14]. As such, well-targeted efforts that embrace area-speci c situations, at least in the remaining high malaria intensity foci, are needed to preserve the achieved health gains and advance towards elimination. Retrospective analysis of malaria cases in eastern Tanzania revealed two potentially high malaria endemic foci (Mkuyuni and Kiroka ward) in Morogoro Rural District, eastern Tanzania, with a prevalence of up to 61% [15]. Because of the aforesaid, this study assessed knowledge, attitudes, practices and demographic risk factors towards malaria amongst rural communities in a potentially high endemic area of eastern Tanzania to appropriately inform the future course of awareness campaigns, research as well as surveillance and control strategies.

Description of the study area
The study was conducted in Mkuyuni (latitude 6.57 o south and longitude 37.48 o east) and Kiroka (latitude 6.83 o south and longitude 37.78 o east) (Fig. 1). These wards are next to each other and are part of Morogoro Rural District, eastern Tanzania. Mkuyuni covers 97.4km 2 with a population of 17,935 people [16]. Kiroka covers 212km 2 with a population of 21,853 people [16]. Agriculture is the major economic activity across the two wards; and the main crops are rice, maize, banana, and coconuts. The long rain season runs from March to August and the short season runs from September to mid-December. The dry season runs from January to end of February. An estimated 96% of the houses in the area are made of mud-brick walls and iron roofs. The landscape is bestowed with temporal, semipermanent and permanent mosquito larval habitats, particularly in and around the agricultural elds. The population of malaria vectors largely composed of Anopheles gambiae sensu-lato and Anopheles funestus sensu-lato (Mnyone and Aikambe, Unpublished data). Malaria is experienced almost all year round with a peak in April and July [15].

Data collection
Simple random sampling technique was used to select study villages. The study households in each of the study villages were selected using a systematic random sampling technique. The relative contribution in terms of sample size was predetermined based on the size and number of households in the three study villages. A total of 243 households (Changa = 103, Mkuyuni = 72 and Mfumbwe = 68) were nally selected and their respective heads or any other responsible adult interviewed after a brief, but thorough introduction to the study. A standardized questionnaire with structured and semi-structured questions was used as the main data collection tool. The questionnaire was written in English and translated into Kiswahili, a national language which was understood by all of the respondents in the study villages. The structured questionnaire and direct observations were done by a team of trained eld workers and young graduates. The questions were mostly meant to seek information on, knowledge, attitudes and practices about malaria transmission, diagnosis, treatment and control. Also, there were questions meant to identify physical and socio-demographic features with potential for increasing the transmission risk of malaria in the study villages. Only one person per household was interviewed. The questionnaire was supplemented by direct observation to ascertain certain responses with immediate and/or readily available evidence including among other, ownership of bed net, other preventive measures, open eaves and other openings on the houses.

Data analysis and presentation
The collected data were summarized and analyzed by using Statistical Package for Social Sciences (SPSS) and Excel and presented in the form of tables. The collected data were computed in SPSS to obtain appropriate proportions and frequencies.

Socio-demographic characteristics
A total of 243 respondents were interviewed during the period between February and April 2020 (Table 1). They comprised of males (47%, 113/243) and females (53%, 130/243) females. Most of them were over thirty years old (167%, 162/243) and had primary education (60%, 146/243). The next majority was below the primary level of education (33%, 81/243). The most common occupation reported by respondents was crop farming (87%, 211/243). Most of the study households had children below 5 years (63%, 153/243) and/or 5-15 years old (74%, 181/243). screening their windows would protect them from mosquitoes and malaria thereof. Furthermore, they a rmed that the risk of contracting malaria was much higher during the wet season. The results are summarized in Table 2.

Attitudes towards malaria interventions
When asked about the performance of bednets, 47% (113/243) of the respondents doubted that bednets can still reduce malaria. However, with an exception of 26% (29/113) of the respondents who said that mosquitoes can penetrate even intact nets, the rest (64%, 84/113) were not sure of the reason(s).
Dry season 10 (10) 18 (27) Table 3. . When asked about gatherings and/or ceremonies taking place outdoors during night-time, most respondents (58%, 141/243) reported that they still take place and that people may spend more than three nights outdoors (51%, 121/243) oftentimes without any protection against malaria (61%, 89/243). About 39% (55/243) claimed to protect themselves against malaria during those events, but they could not substantiate how they achieved that. The results are summarized in Table 4.

Discussion
Uptake and utilization of existing and novel malaria control interventions are often affected by knowledge, attitudes and practices amongst communities living in malaria-endemic settings. Similarly, documenting household demographic/household characteristics associated with increased mosquito biting risk is essential to inform the course of action in terms of designing behavioural change communication and other control interventions. This study assessed knowledge, attitudes, practices and demographic risk factors towards malaria amongst rural communities in a potentially high endemic area of eastern Tanzania. Findings from the study indicate an impressive level of knowledge on different aspects of malaria disease and control measures. More than 50% of the respondents knew that malaria is transmitted by Anopheles mosquitoes and that they breed in water bodies. However, rubbish was also reported by 24% of the respondents as a breeding site for Anopheles mosquitoes. Interestingly, up to half of the respondents were aware they could still get malaria if bitten before midnight including early morning and evening. Most communities in endemic areas believed malaria transmissions happen after midnight [17] most likely due to continued dissemination of outdated awareness messages [18]. Most messages date back to the 1990s and early 2000s when malaria mosquito bites were predominantly occurring indoors and late night-time. Despite the improvement in community awareness noted in this study, a considerable proportion of respondents as the case may be in other endemic areas, still rely on outdated messages. Not much advocacy, if at all, has been made to re ect the present situation on mosquito biting behaviours and malaria transmission [18]. Studies are increasingly documenting earlyand outdoor-biting in primary malaria vectors in Tanzania and elsewhere in Africa [19][20][21][22]. Therefore, the behavioural change communications need to be revised to re ect such changes to heighten awareness on the upcoming challenges in disease control. Otherwise, they will continue being inattentive to earlyand outdoor-malaria transmissions and thus constrain the global efforts to sustain the current health gains and advance towards elimination. This study documented a considerable level of awareness among respondents concerning the age group most vulnerable to malaria and control measures. More than 60% of the respondents cited the under-ve children as the most vulnerable group and ITNs as one of the most powerful control intervention currently available. Indeed, the under-ve children across all endemic countries suffer the greatest malaria burden mostly due to lack of immunity to the disease [1,[23][24]. Similarly, ITNs remain the predominant control measure of choice across all malaria-endemic countries [1]. To enhance the impact of disease control efforts, the improved knowledge needs to be re ected on daily community practices. This is not fully realized in the present study areas as the case may be in other resource-poor malaria-endemic areas.
Thus warranting continued behavioural change communications and improved access to control measures at local levels. The respondents recognized multiple entry routes for malaria mosquitoes however windows were by far the most commonly cited route. Extremely few respondents cited eaves although most of their houses had open eaves, and these have long been considered as the primary entry route into houses for malaria vectors [25][26][27][28]. This implies low awareness on the contribution of open eaves and therefore future awareness strategies should also emphasize the role of this entry route. These strategies will be enhanced by considerable improvement on literacy level and access to information in the study area and many other rural communities in Tanzania. We envision great improvement in malaria control if the advocacy is done alongside the improvement of housing styles happening in Tanzania and elsewhere in Africa. Interestingly, there are several on-going initiatives that emphasize on the house improvement as an alternative malaria control strategy. Expectedly, the bene ts of such initiatives will soon be extended to Tanzania.
Most respondents (70%) reported con rming malaria at the health facilities instead of self-diagnosis based on clinical presentations. Furthermore, 62% of the respondents reported that oftentimes they con rm the disease before taking antimalarial drugs obtained either from the health facilities or private drug dispensing shops. This complies with the government and WHO recommendation that all suspected cases should be con rmed either through microscopy or rapid diagnostic test before treatment [29]. A preference for testing before taking anti-malarial drugs has also been reported in other settings [30,31].
The documented improvement in the study area may be explained by easy accessibility and quality of malaria diagnosis services provided by the resident health facilities and wide access to behavioural change communication (BCC) messages on malaria diagnosis through radios, televisions and health workers. Yet, self-diagnosis and medication remain a paramount concern in the study area (~30%) and elsewhere in Tanzania [32][33][34][35]. The most common reasons for these concerns include long distance to health facilities, inability to pay for healthcare charges, long waiting time at the health facilities, frequent shortages of medicines and poor attitudes on health workers [34,35,[36][37]. We recommend a study to determine the contribution of these and/or other factors within the current study area. This will add great value to the current ndings in favour of designing appropriate awareness and control strategies.
Up to 96% of the respondents, self-and health facility-diagnosed reported that they preferentially treat the disease using modern remedies. Only 4% of the respondents reported that they preferentially treat the disease using traditional remedies. This implies that reported self-medication is most likely done using modern antimalarials. As such, continued encouragement for health-facility based diagnosis must go hand-in-hand with improvement on self-medication particularly in areas without or with insu cient healthcare facilities. Banning self-medication may encourage people to practice it more in hiding and result to even more devastating consequences. The sought solutions should otherwise aim to provide the necessary knowledge and ensure that self-medication is well guided. This can be achieved among other ways by disseminating clear messages about malaria as part of health education, and formulation of realistic treatment policies [38]. Such messages should also seek to raise community awareness and knowledge on non-malarial fevers. If practiced appropriately, self-medication is recognized worldwide as a major contributor to most health care systems [39].
A vast majority of respondents recognized the importance of ITNs but noted that they were not perfect. This is substantiated by the fact that up to 79% of the respondents owned at-least one ITN. High ITN ownership despite the admiration that they were not perfect agrees with the ndings of studies in other malaria endemic areas in Tanzania [17]. If appropriately and regularly used, ITNs remain reliable preventive measures against malaria [1,[40][41][42], therefore their deployment within and beyond the study area must be encouraged. All respondents admitted that they were still concerned with malaria, more so when considering deaths of their family members or close relatives in the past. Therefore, ITNs remain inevitable in preventing malaria resurgence and devastating morbidities and mortalities of the past.
The proportion of malaria mosquitoes biting people outdoors is increasing in most endemic countries within the African region [19][20][21][22]. Such exposures are greatest for rural and resource-poor communities due to extensive outdoor night-time social, cultural and economic activities [18,43]. This study documented demographic features of similar nature which potentially increase the malaria transmissions outdoors and/or early before going to bed. A vast majority of the respondents in this study were cooking outdoors between 18:00h and 22:00h. Also, majority, including those not cooking outdoors went to bed after 20:00h. Furthermore, most respondents reported night-time communal ceremonies/gatherings, which may take more than three days. Similar observations are increasingly documented in many other malaria endemic resource-poor settings [17,43,[44][45][46]. Cooking and communal ceremonies/gatherings have been reported among the popular activities that kept people outdoors in many endemic areas of Africa

Conclusions
Findings from the current study indicate that majority of participants were aware that Anopheles mosquitoes were responsible for transmitting malaria, they bred in water, under-ve children were most vulnerable to malaria, and ITNs constitute the most powerful control measure. But, knowledge gaps requiring consideration when designing behavioural change communications were also identi ed. Concerning practices towards malaria disease and control interventions, we documented promising proportion in terms of respondents who con rmed the disease at the health facilities, preference to modern rather than herbal remedies, and bednet ownership. However, the rate of self-diagnosis, selfmedication, and appropriate use of bednets require great improvement. Concerning demographic risk factors for early-and/or outdoor-malaria transmission, the majority cooked outdoors, slept late, and participated in several days of night-time communal gatherings without any protection against the disease. The ndings warrant subsequent study to show a causal link between night-time outdoor activities, vector biting patterns and malaria prevalence.
Declarations Figure 1 Map of the study area, Morogoro Rural District, Eastern Tanzania.