Treatment-seeking behaviour for febrile illness in an area of seasonal malaria transmission in rural Ethiopia

Malaria remains the major cause of morbidity and mortality particularly in sub-Saharan Africa [1, 2]. In areas with low endemicity, malaria is characterized by frequent and often large-scale epidemics associated with high case fatality rates [3, 4]. Although children are at the greatest risk in these areas, all age groups of the population are at risk of severe malaria and death due to lack of protective immunity. Early diagnosis and prompt treatment has been a cornerstone of malaria control [5, 6]. However, effective case management partly depends on early recognition of the disease and the subsequent treatment-seeking behaviour from proper care providers.

In the past, treatment-seeking behaviours have been extensively studied in areas of high endemicity [7, 8]. Studies conducted in these areas have shown that response to malaria mainly depends on access to a health facility, perceived severity of the illness, perceived quality of care, cultural and traditional beliefs, and knowledge of the symptoms [9]. It is also recognized that treatment of malaria usually starts at home and households seek care outside if home treatment was not successful [10‐13], resulting in delays of seeking treatment from a proper care provider.

Despite the extensive work on treatment-seeking behaviours for malaria particularly among children under 5 years of age in areas of intense transmission in sub-Saharan Africa, little is known about the management of malaria and treatment-seeking patterns among children and adults in areas of seasonal malaria transmission [7, 14‐16]. Due to the severe effects of malaria epidemics, however, increased attention has recently been given to malaria perceptions and practices of the community in areas of low but seasonal transmission [4, 7, 15‐17]. This is of particular relevance to improve early diagnosis and prompt treatment of malaria in epidemic prone areas. The present study assessed the treatment-seeking behaviour of reported malaria in all age groups in an area of seasonal transmission in Ethiopia.

In all age groups of patients, the most commonly reported symptoms were fever, shivering and chills, loss of appetite, headache and high body temperature (Table 1). Reported shivering and chills among <5 children and adults was higher than that reported among children 5–14 (P = 0.013). However, vomiting was more reported among <5 children compared to children between 5–14 year-old and adults ≥15 (19%, 14% Vs. 12%; p=0.018).

The two-week prevalence of reported malaria across the three age group categories indicated that adults' prevalence was higher than that reported for children (Table 2). At the time of the survey, about 77% (n = 1348) of those reported with malaria sought any form of treatment from various sources. The most frequently reported first response was visiting CHWs, followed by public health facilities and private clinics. A considerable proportion (16%) of patients took a second or third action to the same episode of illness since the previous treatment(s) did not provide a cure. Private clinics and public health facilities were the main source of antimalarial treatment by the majority of the patients during the second (41%) and third (35%) visits, respectively. Although symptomatic diagnosis for malaria was reported, 23% of the individuals were neither taken to health facilities nor received any form of home treatment, the main reasons being mild illness (41%), financial constraint (37%), distant health facility (18%), shortage of time due to work overload (3.5%), and the perception that antimalarial drugs were expensive (0.5%).

Table 3 shows the type of diagnosis for the perceived malaria illness two weeks prior to the survey. Family or self-diagnosis was the most common type of diagnosis within the first 24 hours of the onset of illness. Clinical diagnosis by health workers including CHWs and laboratory diagnosis was infrequently mentioned within the 24 hours of illness onset. The vast majority (76%) did not get any type of treatment or intervention within the first 24 hours of the onset of illness. When respondents were asked about the general diagnosis for this particular reported malaria episode, clinical and laboratory diagnoses increased to 49% and 16%, respectively.

Among those who sought any type of treatment, only 13% had sought treatment within the first 24 hours of onset of illness (Table 4). Most waited until the second day of fever onset before seeking treatment. There were no significant differences in the timings of the first treatment by age. As shown in Table 4, seeking treatment within 24 hours of illness onset was more common among those who practiced home treatment, followed by those who sought treatment from CHWs and private clinics. Treatments from public health facilities were sought lately compared to other sources of treatment.

This paper has presented evidence on the treatment-seeking behaviour for reported malaria illness in an area of seasonal transmission in rural Ethiopia. A number of malaria-related symptoms particularly fever, shivering, chills, loss of appetite and headache were presented, which taken together may approximate a clinical diagnosis of malaria although using these symptoms, particularly fever, as a proxy for malaria appears to be neither sensitive nor specific when compared to parasitologically confirmed diagnosis [24, 25]. In the present study area, laboratory tests were unavailable at public health facilities such as health posts and health stations. At health center and hospital levels, malaria diagnosis through blood smear examination is usually recommended, but there are occasions that these facilities might not provide blood testing services due to patient overload or shortage of supplies.

The present data is based on self-reported treatment-seeking patterns based on perceived malaria. As a result, all cases reported as malaria may not be malaria, or those not reported as malaria may be truly malaria. The study was conducted during a high malaria transmission season in the area, and its sensitivity would be high considering the transmission season. A study conducted in the adjacent district around Butajira demonstrated 75% sensitivity and 60% specificity using a combination of fever, previous attack of malaria, or the absence of cough during high transmission season [26]. Self/family diagnosis was common among all age groups of our study population, and this pattern of diagnosis is expected because more people in a rural setting hardly access health facilities for better clinical or laboratory diagnosis.

Nevertheless, fever serves as a proxy for malaria both at household level for home management of malaria [27] and peripheral health facilities where diagnosis is made presumptively upon the presence or history of fever [6, 23]. However, a recent change of first-line treatment of malaria to artemisinin-based combination therapies (ACTs) in many countries of sub-Saharan Africa has highlighted the potential cost implications of malaria over-diagnosis based upon clinical signs and symptoms [28]. Use of a rapid diagnostic test would help in identifying malaria parasites especially in areas with seasonal malaria transmission where presumptive diagnosis of the disease may be inaccurate but the magnitude of asymptomatic carriers is assumed to be low.

Home treatment in this area of seasonal transmission of malaria was much lower than that observed in areas of similar transmission in Kenya [15] and Uganda [16], and also in areas of high transmission [9, 11]. Seeking care from the health facility mainly comes after the failure of care at home [29], or malaria patients might directly seek treatment from health care facilities without the initiation of treatment at home [30]. In a study conducted in east Shewa zone in central Ethiopia, 87% of malaria patients diagnosed at MCLs directly sought treatment without taking anti-malarial drugs at home or prior visit to other health facilities [31]. The present study confirms previous observations in Ethiopia that use of health services for malaria is widely practiced despite the low level of home treatment.

Although public health facilities are assumed to have better institutional organization and capacity that include trained health workers and sustainable supply of drugs, about one-third of the reported malaria cases made first visit to CHWs for malaria treatment. This may not be due to the technical quality of CHWs, but due to their ease accessibility, the availability of antimalarials and their responsiveness to the demand of people. In other words, people in rural areas have little access to public health facilities, despite the inefficient services rendered by the public health-care system [32]. The reasons for the high utilization of private clinics included the perception that they provided high quality services and gave much attention compared to the public services, despite their high cost of treatment. Most private health services are concentrated in the urban centers and small towns. Here it is possible to hypothesize that people are willing to travel long distances to seek health care when quality service is demanded, despite high transportation costs. In the present study area, CHWs were found at a closer distance, but public and private health services were found at a distant for the majority of the rural households. The widespread use of CHWs and private clinics indicates the need to strengthen the community-based malaria control interventions and highlight the importance of private sector for malaria treatment.

The present study findings indicate that delays in seeking treatment for reported malaria is more common among the users of public health facilities, followed by the users of private clinics, compared to the users of CHWs and home treatment. Most malaria deaths among young children occur within 2–3 days after onset of symptoms [33], yet in our study only 13% had sought treatment within 24 hours of onset of symptoms, very much lower than a finding observed in an area of low transmission in Uganda [16] and Kenya [15]. The present study finding has strong implications for improving early diagnosis and prompt treatment with effective antimalarial drug.

Initiation of treatment at home for malaria was started earlier than those sought care from other providers, a finding which is in line with the results of studies conducted elsewhere [10‐12]. However, under conditions in which the health service coverage is poor and most facilities are not well strengthened due to lack of resources, seeking treatment from these facilities may not guarantee either prescription of an efficacious antimalarial drug or the correct dosage due to lack of appropriate drug or poor awareness of health workers about the drug policy [16].

This study has indicated that seeking antimalalria treatment from non-public health facilities such as CHWs and private providers as a first resort to malaria constituted more than half of the visits, necessitating the importance of strengthening community-based interventions and peripheral health facilities. Of special finding is the early treating-seeking pattern of those who visited CHWs and practiced home treatment, with delays most common among the users of the public health facility. Although home-based management of malaria is a promising strategy for improving provision of prompt effective treatment [27, 34], the present findings indicate very low level of home treatment, requiring the need for further research. The supply of effective antimalarial drugs such as artemether-lumefantrine to the commonly used care providers is the most important step to reduce the impact of malaria among the rural community. People should be informed and educated about the importance of early diagnosis and prompt treatment with effective antimalarials within 24 hours of symptom onset.