Study settings and participants
This cross-sectional study was carried out in the Khulna City Corporation (KCC) area of Khulna district; a southwestern divisional headquarter in Bangladesh. Located between 24°45’ and 24°54’ north latitudes and between 89°28’ and 89°35’ east longitudes, KCC covered an area of 45.7 Km2 and consists of 31 wards, and it is the fifth-largest city corporation in Bangladesh in terms of concentration of slums (1,134) and slum dwellers (79,827 people in 40,015 households) [18-20]. Moreover, a significant percentage of older people (7.34%) live in these urban slums [21]. therefore, certain attributes were specified to identify the households with older people in the slums of KCC to facilitate a face-to-face interview: i.e., (a) a household with at least one resident who must be an aged people in the age bracket of 60≥ (above) years; (b) s/he must be living in the selected slums of KCC area; (c) for at least five consecutive years; (d) with or without a spouse, but married at least once; (e) staying with their family or their own. Considering the criteria, the participants were identified using a door-to-door census following a two-stage area probability sampling approach. At the initial stage, three Thana – an administrative unit in the local government system in Bangladesh – i.e., Khulna Sadar, Khalishpur, and Sonadanga, were selected considering the concentration of slums; in the second stage, four slums were selected based on the number of slum dwellers. After a week-long census by a group of ten data enumerators, 1,104 older people were identified from 2,167 households within the selected areas. Later, 636 older people were interviewed by administering an interview schedule, proportionate to the geographical location as well as the concentration of the population.
Ethical issues
The research was performed in accordance with the Declaration of Helsinki, and the ethical clearance committee of Khulna University, Bangladesh, approved this study (Reference No. KUECC – 2022/06/02). In this study, informed consent was obtained from all the participants, i.e., the elderly slum dwellers, and they were notified about the purpose of the study, and they were assured by the data enumerators about anonymity and confidentiality of the information. The participation was voluntary, and there was no incentive for the participants. Moreover, the participants had the right to revoke their participation and shared information without prior justification.
Procedures
A semi-structured interview schedule (IS) was developed after carefully reviewing relevant literature considering the research objectives. The IS was divided into three mutually inclusive sections, e.g., the first section focused on socio-demographic information, including age, sex, religion, occupation, income, beneficiary/non-beneficiary status in cooperatives, recipient/non-recipient of social assistance, the second section comprised information regarding the non-monetary wealth (NMW) of the households [22], the ability to manage activities of daily living (ADL) [23] and satisfaction with domain of life (SDL) [24], whereas the second section highlighted the health status of participants, including ailment, healthcare status, care-seeking behavior and so on. Following the development of IS, it was verified by a pre-test on 20 elderly slum dwellers to make sure of its adequacy to extract relevant information from the participants, minimize redundancy and non-response rate, as well as provide first-hand experience for the data enumerators to curb the timing of the interview [25]. It is important to note that the researchers extensively trained the data enumerators for a week through classroom-based lectures, role-playing, and practice sessions on the content of the IS and the techniques to establish rapport and extract information. Data were collected for three months, starting in July, and ending in September. Later, to ensure standardized data collection, twenty households were re-surveyed randomly by the researchers to identify inconsistencies and re-visited families with the data enumerators to confirm the highest data quality.
Measures
Socio-demographic information
Some specific socio-demographic factors, including age, sex, education, occupation, income, marital status, were considered as the predictors of health status, e.g., ailment, of the older people living in slums of KCC areas.
Indices
An index of non-monetary wealth (NWM) was measured considering different dimensions, including ‘elements of comfort’ – television, refrigerator, ceiling fan, stand fan; ‘communication and comfort’ – mobile, bicycle, rickshaw, van, easy bike; ‘consumption of water’ – potable, protected and unprotected – for primary, secondary and tertiary us; ‘housing structure’ – floor, walls, ceiling; ‘energy consumption’ – electricity, solar power and kerosene/wood/leaf – for power and cooking; ‘sanitation facility’ – unitary or common, modern or tradition [22]. The summation of the response for each item was added and an index of non-monetary wealth was developed. Likewise, an index of activities of daily living (ADL) was used to assess the physical capacity of older people to execute certain daily activities, including bathing, personal hygiene, medication, chores with and outside of the household or their dependence on others [23], and the summation of each item led to the ADL index. Finally, the satisfaction of domains of life (SDL) assessed the satisfaction of older people regarding their physical health, economic status, relationship with spouse and children as well as their overall life [24]. SDL was measured in a five-point Likert scale, and summation of each item resulted in the SDL index.
Health status (ailment)
Health status was measured by a dichotomous response – ‘No = 0’ and ‘Yes = 1’ – for the question – ‘did you suffer any kind of ailment or physical problem in the last one month?’ This dichotomous response was considered as the dependent variable for this study.
Analysis
Data were analyzed in two consecutive staged using IBM SPSS Statistics (version 26). At first, descriptive statistics, including frequency and percentage analysis was used followed by Pearson’s Chi-square (χ2) and Yate’s continuity correction (χ2Yate’s) to explore the association between health status (ailment) and the socio-demographic factors. Finally, multivariable logistic regression was executed considering the variables found statistically significant in Pearson’s Chi-square and Yate’s continuity correction test. Findings were shown using the crude odds ratio (COR) and the adjusted odds ratio (AOR) with 95% confidence interval at a 10% level of significance.