Estimating socioeconomic status for health equity surveillance in Cameroon: an expert opinion survey
Description
Background: Despite increasing awareness of socioeconomic status’s (SES) association with health outcomes, there is no widely accepted and rapidly implementable estimation of SES measures in resource-limited settings. An exception is the Demographic and Health Surveys (DHS)’s wealth quintile index constructed from household ownership assets. To facilitate health equity surveillance, method of individual SES estimation requiring fewer number of household assets is needed. The objective of this study was to identify the DHS assets most relevant for measuring SES in Cameroon. Methods: Participants interviewed with a structured questionnaire included stakeholders involved in the design and implementation of DHS in Cameroon for many years. Using a 5-point Likert scale, experts graded DHS assets’ likelihood to measure SES. The questionnaire was strongly reliable (Cronbach’s alpha: 0.943, 95% CI: 0.920 - 0.961, p<0.001) for using the 29 items retained to measure SES. Results: The probabilities of agreeing that an asset can be a useful measure of SES varied from 0.016 to 0.047. The 12 DHS assets most likely to measure SES included having Refrigerator(85.3%), Television(83.8%), Laptop(79.4%), Mixer(77.9%), Computer(77.9%), Agricultural land(77.9%), Cable/Satellite(76.5%), Cell phone(76.5%), Modem/Internet key(73.5%), Water pump(72.1%), Car/truck(72.1%) and Gas stove(72.1%) with a respective probability (prior) of 0.047, 0.046, 0.044, 0.043, 0.043, 0.043, 0.042, 0.042, 0.041, 0.040, 0.040 and 0.040. Conclusions: This research underscores the importance of integrating local expert insights to refine the measurement of SES, promoting improved health outcomes in populations, particularly in Cameroon. Future research should explore the application of this expert-opinion-driven framework in various contexts to create more comprehensive, robust, and reliable SES indicators. Keywords: Prior Probability, Expert Opinion, Demographics and Health Survey, Socioeconomic Status, Health equity
Files
Steps to reproduce
This study used a cross-sectional design to assess expert opinion regarding the indicators used to measure SES. Data was collected between December 2023 to April 2024 in Yaoundé, Cameroon targeting experts in the field of statistics, economy, demography, social sciences, and health. The study collected data on the measurement of SES. Data was collected in a comprehensive manner using a face-to-face digital questionnaire on tablets. Study population and sampling We targeted experts in the fields of statistics, economy, demography, social sciences, and health who have been involved in DHS in Cameroon over the years. These experts were identified from DHS reports published online. An exhaustive sampling technique was used, and ensured that the experts were of diverse background Data collection procedure The expert were identified and invited to participate in the study. An Open Data Kit (ODK) was used to administer the questionnaire. The questionnaire was divided in three main section: Sociodemographic information This include age, Sex, marital status, basic training received, level of education, year of professional experience, type of Institution, researcher, grade) Participant exposure to DHS This include exposure to, access to DHS data, training on DHS use, Involvement in the design of the DHS, Involvement in the analyzes DHS data, Involved in a DHS survey Perception on socio economic indicators This include the expert opinion on the 29 indicators for measuring socio-economic level based on indicators Data analysis The first step was to present the socio-demographic characteristics of the experts surveyed. This was done using absolute and relative frequencies. The reliability of the questionnaire was assessed using the Cronbach's alpha coefficient to ensure consistency among indicators. Next, descriptive statistics on the proximity of experts to the DHS were presented in the form of frequency tables. Using a 5-point Likert scale (strongly agree, agree, neither agree nor disagree, disagree or strongly disagree), we assessed the main indicators identified by the experts as being those that best measure the well-being of the individuals. The 29 variables in the item were grouped each item into a binary variable where 1 corresponds to agree or strongly agree responses, and 0 corresponds to the other modalities and then, probability to choose an item were calculated. Data were collected, entered into ODK and analysed with R software version 25.