Study design and setting
This cross-sectional study was conducted between September 2017 to October 2018. The study was conducted in the communities of two sub-counties of Wakiso district. Wakiso is a district in Central Uganda that encircles Kampala, the capital city of Uganda. Wakiso district is currently one of three districts popularly referred to as the Greater Kampala Metropolitan District (KMD). This is due to their proximity to the Central Business District of Kampala Capital City. Wakiso is home to nearly 2 million people as of 2018, who live in a mixture of urban, suburban and rural settings. Approximately 92% of the district’s population live in the rural areas of the district. Regardless of the setting in which people live, the levels of poverty are significant throughout the district. The district has a population of more than 400,000 older people aged 60 years and above (UBOS, 2017). The two sub-counties of Nansana and Busukuma were purposively selected for their representation of both urban and rural populations.
Sampling
In this study, sampling was done in several stages. Initially, we purposively selected two sub-counties for the study owing to their large populations and area size. Participants for this study had to be 60 years and above and residents in these two sub-counties. Next, we randomly selected one parish from each sub-county and purposively selected all the villages for inclusion in the study. With the help of the local village leaders known as the local councilor one (LC1) and the existing system of Ministry of Health approved village health workers, we were able to understand the layout of the villages and the households therein. Although we were not able to get a list of all villages as a sampling frame, we learnt that the villages were within walking distance from one another. Therefore, by walking down one village road, one was likely to traverse three villages in the process. Thus, we decided to start our village road treks from a central point at the sub-county headquarters. Spanning out in different directions, the research assistants were able to identify and interview study respondents who were 60 years and older from their homesteads lying along the chosen village roads. Research assistants were; a psychiatric nurse, a social worker and a community psychologist who were trained prior to conducting the study trained. The Principal Investigator (PI) also participated in interviewing respondents. The research team exhausted all households along a particular road before returning to the center to pick a new road to follow. The village health team members moved alongside each researcher and acted as the community gatekeepers that introduced the research team, as well as guiding the team on where to stop. We harnessed the intimate knowledge of the villages that the VHTs had such as, knowing specific households where an elderly person resided. We excluded all elderly people who had not lived in that village for more than 12 months and those who could not answer our questions because they were too ill. In each identified household, we interviewed one elderly person who met the study inclusion criteria. We made one call back visit to ensure that we got all potential respondents who may have not been home at the time of our visit.
Sample size calculation
A sample size of 380 respondents from the two sub-counties was calculated using the Cochran method (Cochran 1963) and the following parameters: an elected alpha of 1.96; an estimated portion of older people in Uganda of 4.6% (UBOS, 2012); q of (1- p); and an estimated margin of error d at 0.05.
Data collection tools
The data collection was conducted by three research assistants who were well-trained on the protocol and research subject protection, and the PI, using structured questionnaires. The questionnaires were first translated into the local language of Luganda by a process of forward, backward and forward translation by two teams of mental health professionals working independently of each other. Any disparities were resolved through consensus.
The study questionnaire had three parts, namely:
1) Social Demographic factors - sex, age, marital status, education level, and religion;
2) A culturally adapted version of the 35 item older persons’ QoL questionnaire (OPQOL) (Bowling & Stenner, 2011). The OPQOL has been previously validated for studies in Iran (Nikkhah, Heravi-Karimooi, Montazeri, Rejeh, & Nia, 2018), and the Czech Republic (Mares, Cigler, & Vachkova, 2016). The OPQOL is a 5-point Likert scale (strongly disagree to strongly disagree), representing questions about life overall, health, social relationships and participation, independence, control over life, freedom, area (home and neighborhood), psychological and emotional well-being, financial circumstances, religion/culture. For this study, 4 questions of the original version were dropped during the adaptation phase. The tool was then categorized into 3 subcategories of QoL, namely: Health (HE), Social Economic QoL (SE) and the Psychosocial QoL (PS). The HE QoL had 10 questions, SE QoL was represented by 9 questions and the PS QoL represented by 14 questions. 2 questions were droped because they were covered in the social demographic questions.
3) A culturally adapted version of the 26- item Brief Ageing Perceptions Questionnaire (B-APQ) (Sexton, King-Kallimanis, Morgan, & McGee, 2014). The B-APQ is a 5 point Likert scale form that assesses emotional regulation, perceptions about control of social life and perceptions about physical age. The B-APQ questionnaire has been previously validated for studies in Malaysia (Jaafar, Villiers-Tuthill, Sim, Lim, & Morgan, 2020). It was adapted for the local Ugandan context for this study.
Statistical analysis
Data were double-entered into Epidata 4.2, cleaned and exported into STATA 15.1/MP for analysis. Data was analyzed using STATA software (StataCorp LP, 4905 Lakeway Drive, College Station, TX, USA). Summary statistics for continuous variables were presented as mean ± standard deviation while categorical variables were presented as frequencies and proportions with their 95% confidence intervals in tables. A Pearson’s Chi-square test was then used to determine the relationship between the perceptions and quality of life. As this was one of the first studies conducted related to QoL in Uganda, a p-value of <0.05 was considered to be significant.
Ethical consideration
The study was approved by the Makerere School of Health Sciences Review Board (#SHSREC REF: 2017-094) and the Uganda National Council for Science and Technology (HS195ES). Study approval from the participating district was also secured. Informed consent was obtained from the respondents prior to subjecting them to any study tools. Written consent was obtained from respondents who could read and write, while those who could not read and write, we used audio recorded verbal consent. In addition to verbal consent, they were also required to append their thumb prints on the consent paper. Study respondents found to be in significant physical distress were excused and referred to the health center for medical assistance, escorted by one of the Village Health Team members. Respondents were assured of confidentiality before the start of each interview. All data collected was kept under key and lock, and only the research team had access to them. Personal identifiers were not used at data entry.