Economic costs of fever to households in the middle belt of Ghana

Kintampo districts cover an area of approximately bout 7162 km² a resident population size of about 150,000 inhabitants in 29,438 households [19]. Kintampo Health Research Center (KHRC) located in the Kintampo districts maintains the Kintampo Health and Demographic Surveillance System (KHDSS), which was used for sampling households for this study. The KHDSS routinely collects demographic information on age, sex, pregnancies, births, deaths, and migrations among others and does verbal autopsies to determine the cause of deaths [20]. The site also collects data on household assets and possessions that enable the assessment of socio-economic status of the population. There are two hospitals, twenty health centres, three private clinics and five public Community Health Planning Services (CHPS) compound in the study area. At the time of this study about 50 % [21] of inhabitants in the study area had valid National Health Insurance cards to access health care in public and private accredited by the National Health Insurance Authority (NHIA). Currently, health insurance coverage is about 66 % [22]. The study area is mainly rural with few urban settlements. Farming is the main economic activity in the area. Malaria accounts for more than half of all out-patient visits in the study area [23].

This was a cost analysis study based on patient’s perspective. A cross-sectional household survey was conducted between October 2009 and September 2011 among community members in the Kintampo area located in the middle part of Ghana.

Assuming 6 % of the entire inhabitants will have had fever in the 2 weeks prior to the day of interview and that 50 % of them will seek care and have physical access to an authorized point of provision of ACT: then with 95 % confidence and allowing for 10 % drop-out, the estimated sample size of 4226 households will be achieved. Households were randomly selected from the 4226 households using the KHDSS household database and interviewed with a structured questionnaire. Household members who reported of fever within the last 2 weeks prior to their interview were included in the survey. However, only those who sought treatment outside their homes and incurred costs in seeking treatment were included in the analysis. Data was obtained from household members on socio-demographic characteristics, treatment-seeking patterns as well as direct and indirect costs of seeking treatment outside home.

Data was double-entered into Epidata 3.1 (THE EPIDATA ASSOCIATION, ODENSE M, DENMARK, EUROPE) and transferred to Stata 11.0© (STATA CORP. TX) for analysis. The total economic cost of fever per household was estimated by summing the direct and indirect costs incurred per household. The estimated mean and median costs with their respective 25th and 75th percentiles (iqr) and standard deviations (SD) are presented. Direct costs were defined as out-of-pocket expenditure incurred by the households as a result of seeking treatment for their fever outside their homes. It included cost of self-treatment, medical costs and non-medical costs. Cost of self-treatment comprised the cost of medicines purchased from pharmacies, chemical and other drug sellers without prescription from a doctor or medical staff at a health facility. Medical costs included consultation, laboratory and prescription fees incurred at a health facility. Non-medical cost included cost of transportation to and from the facility where care was sought and other non-medical costs incurred as a result of seeking health care for the treatment of fever.

Indirect cost was defined as the cost of productivity losses to households as a result of their fever and inability to work. It was calculated as a product of the number of days respondents were unable to work and their daily wage. Daily wage was estimated in several ways depending on the type of economic activity the respondent was involved. The daily wage for economically active adults (defined as persons aged 18 years and above; and engaged in informal income generating activities, such as farming) who hired others to do their work during the course of the fever was calculated as the cost paid out to others to get the work done. Among patients who earned monthly income, daily wage was estimated as their total monthly income divided by 22 days with the assumption that they worked 5 days per week for a month of 4 weeks. Students were asked to report on the number of school days missed due to fever. When adult patients were accompanied by caretakers, the daily wage of their caretaker was estimated in a similar manner for the patients. Seeking care outside home was defined as seeking care from the public or private health facilities, chemist or pharmacist and herbalist or drug peddler.

Wealth quintiles for each household were calculated using Principal Component Analysis (PCA) and included durable households’ possessions such as motorbike, car, bed, radio and sewing machine. Information on household architecture of material for walls, roofing material, source of drinking water, cooking utensils, toilet facility and cooking fuel [24‐26] were also obtained. Exchange rate (as at 30th September 2011): GHS 1.52 is equivalent to $1 [27].

Written informed consent was obtained from all adult participants and from care-takers of children. Additionally, assent was sought from participants who were between 12–18 years. The Kintampo Institutional Ethics Committee and the Ghana Health Service Ethical Review Committee granted ethical approval for the study.

Of the 4226 households selected, 947 (22.4 %) households with 1222 household members reported of having fever within the 2 weeks prior to their interview. Ninety-two percent (1127/1222) sought treatment outside their homes and, therefore, met the criteria for inclusion in the analysis. More females reported of fever than males (Table 1). Fifty-eight percent of the patients were below 18 years of age. Fifty-four percent and 15.1 % of those who reported of fever were farmers and unemployed, respectively. About 17.4 % of the respondents were in the highest wealth quintile whilst 21.2 % were in the lowest wealth quintile (Table 1).

Close to 32 % (356/1127) of patients who sought care outside their homes did so from chemical shops, 22.7 % (256/1127) from hospital, 17.9 % (202/1127) from private clinics and less than 5 % from Community based Health Planning and Services (CHPS). Majority of respondents in the different wealth quintiles sought care from the chemical seller shops. Most of the respondents in the rich group sought care from the private clinics, as 39 % of patients in the highest wealth quintile sought care from private clinics compared with only eight percent in the middle class who sought care from private clinics. Majority of respondents, who sought care from the CHPs compounds were from the poorest of the population. Respondents who sought care from drug peddlers were mostly from the lowest socioeconomic groups (Table 2).

Households used different means to pay for the cost of fever (Fig. 1). About 48 % of respondents (536/1132) had their cost of fever covered by the National Health Insurance Scheme (NHIS); 564 (49.8 %) respondents paid cash for the care of fever; a minority of respondents paid for their fever through donations, borrowing and sale of their assets.

A total of 2020 days with an average of 4.2 days per patient was lost due to the reported fever cases (Table 3). Sixty-five percent (252/386) of students were not able to attend school due to fever leading to a total of 904.4 school days lost and an average of 2.4 days lost per pupil/student. For adults, the average number of days lost due to fever was highest among farmers (6.4 days) and lowest among formal sector workers (5.3 days) (Table 3). The average number of days lost per patient was highest among those who sought care from herbalist (6.9 days) and lowest among those who sought care from private clinics (3.4 days) (Table 4).

Total direct cost was found to be GHS 4708.38 (US $3097.62) [with a mean cost per fever case of GHS 4.2 (US$ 2.76)]. Total indirect cost (productivity losses) was GHS 20,293.7 (US$ 13,351.12) [with a mean cost per fever case of GHS 18.0 (US$ 11.84)]. The overall total cost of treating fever for the entire study sample was GHS 25,002.04 (US$ 16,448.71) [with mean cost per fever case of GHS 22.2 (US$ 14.61)]. The highest cost was borne by the households who visited hospitals [total cost: GHS 7702.59 (US$ 5067.49); with mean cost per fever case of GHS 34.9 (US$ 22.96)]. Even though the total economic cost was lowest for those who visited herbalists [total cost: GHS 349 (US$ 229.61)] (Table 5).

Distribution of costs according to wealth quintiles showed that the total cost as well as the mean cost per participant was lowest for those in the ‘highest’ quintile [total cost: GHS 3868.92; mean cost per fever case: GHS19.9 (CI 14.0–25.9)] and highest for those in the ‘middle’ quintiles [total cost: GHS 6686.28; mean cost per fever case: GHS 24.1]. Those in the highest quintile spent a lower proportion of their income on cost of fever care compared to those in the lowest quintile (Table 6).