Study setting and design
This study utilized an analytical descriptive cross-sectional design, employing a quantitative approach. The study was conducted in Bwengu area in Mzimba District. Bwengu is one of the rural areas in the northern region of Malawi, specifically located in Mzimba North in traditional authority (TA) Jalavikuba Munthali. Its geographical coordinates are 11° 4' 0" South, 33° 55' 0" East. Bwengu is 54 km-north from Mzuzu city in the northern region, and 408.5 km from Lilongwe, Malawi’s capital city.
Study population
The study population constituted of all the elderly aged 60 years and above, in Bwengu area. According to the National Statistical Office (NSO) projection report, Bwengu has a total population of 15,077, of which 5% are 60 years and above [21]. Consequently, the total population of people aged 60 years and above in Bwengu was estimated at 754.
Sample size determination
Since the target population size was known, sample size was determined using [22], which gave us a total sample of 287 participants after factoring in a 10% non-response rate, based on a similar study by [23].
Sampling technique
Participants were identified through the use of nonprobability method of snowball sampling technique. Using exponential non-discriminative snowball sampling technique, a single initial respondent aged 60 years and above was identified in Bwengu area from whom a multiple chain of other respondents was obtained to be interviewed until the desired sample size was reached.
Inclusion criteria
The study included older adults, both male and female aged 60 years and above who gave their consent, as well as those who cared for them (caregivers), if they were unable to speak for themselves. The legal guardian of the elderly, including biological and adoptive family, who were in charge of their general wellbeing, was referred to as a caregiver. Interviews were conducted at the home of the elderly person or their caregiver following the completion of a consent form by each. In order to obtain the desired sample size, we made care to employ a successive sampling strategy, interviewing eligible research participants each time.
Data collection and management
Data was collected from 1st December to 30th December, 2023 among individuals aged 60 years and above. A newly developed closed-ended pretested questionnaire was used to collect data from all participants. The questionnaire was divided into three sections with the first section containing demographic information, second section with information on the magnitude of utilization of health services among the elderly and the last section with information on the factors influencing health seeking behaviour. As one way of managing data; codes ranging from 001 to 287 were used instead of participants names.
Statistical analysis
The collected data was entered and analysed using STATA version 17. Descriptive statistics in form of frequencies and percentages were presented using figures and tables. Pearson chi-square test of independence was conducted to establish the association between the outcome variable and the potential predictor variables. In binary logistic regression model, only those explanatory variables which were significant at 5% level of significance in chi-square test of independence were entered. The strength of association was estimated using Crude Odds Ratio (COR) with 95% confidence intervals.
Study variables
Outcome variable
In this study, health seeking behaviour was the dependent variable. It was coded as 1 if the respondent demonstrated ‘comprehensive health seeking behaviour’ and 0 if the respondent ‘did not demonstrate comprehensive health seeking behaviour’. Using three previous consecutive times when a respondent was sick, the health seeking behaviour was coded as comprehensive if the respondent went to seek health care at the health facility for at least two times. The health seeking behaviour was coded as not comprehensive if the respondent sought health care services at a health facility at most once in the three consecutive times, they were sick. Respondents’ health passports were used to provide evidence that they sought health services.
Potential predictors
Factors which are associated with the health seeking behaviour as reviewed in literature were used as predictor variables. These factors were gender (male, female), age(60-70 years, 71 – 80 years, above 80 years), marital status (never married before, married, widowed/divorced), occupation (farmer, retired/working, business), education (primary, secondary, tertiary), family type (nuclear, joint), place of seeking health services (hospital, herbalist/traditional healers, others), condition of seeking healthcare (just after starting feeling unwell, failed alternatives), medical services (free, purchased), distance to the nearest health facility (less than 3 km, equal to or more than 3 km), and satisfaction of the accessed health services (very satisfied, satisfied, not satisfied).
Ethical consideration
The study sought ethical approval from the University of Livingstonia Review Board (Ref: UNILIA-REC/UGS/16/2023). Verbal and written informed consent were obtained from each and every respondent before data collection. Clearance was obtained from the Village Headmen to access the elderly in the villages.