Study location
The study was conducted among primary health workers in Local Government Health Authorities (LGHAs) in Oyo State. Oyo state is situated in the South-western part of the country. The State has 33 local government areas (LGAs) with varying incidence of COVID-19 (Fig 1). The state has a projected 2019 population of 8, 635, 793 using a growth rate of 3.4% and 2006 population figure as the baseline [14]. Oyo State is predominantly populated by the Yorubas with a sparse presence of other ethnic nationalities. The inhabitants are majorly engaged in farming, trading and some of them are artisans and civil servants. Rural-urban migration is a pronounced phenomenon in the state and this was found to result from the predilection for living in high-density urban centres which can rapidly aid the transmission of any disease outbreak. There are 351 political wards in the State and these wards housed 733 primary health care facilities, primary health centers and health posts. In these health facilities, there are 19 Medical Officers of Health (MOH), 224 nurses/midwives, 200 community health officers (CHOs), 589 senior community extension workers (CHEWs), 131 junior community extension workers (JCHEWs), 43 medical laboratory scientists (MLS), 46 science laboratory scientists (SLTs), 13 pharmacy technicians, 104 medical record officers, 890 health assistants, 122 health attendants and 141 other staff [15].
Fig 1: Map of Oyo State showing the LGAs with the number of COVID-19 cases as at 21/07/2020
Study design and population
This was an online questionnaire-based, cross-sectional survey carried out among PHC workers in the local government health authorities (LGHAs) of Oyo State between March and April, 2020. The study population included all the PHC workers working in the 33 LGHAs in Oyo State at the time of study. All consenting PHC workers who had been in the service for at least one year were recruited for the study. Medically unfit personnel, administrative staff and casual workers engaged by LGHAs were exempted from the study.
Sample size and sampling technique
The minimum sample size was calculated using the sample size formula for estimating single proportion. Since there was no previous prevalence study of COVID-19 among health workers, a prevalence of 0.5% was used and the margin of error was set at 5%. A non-response rate of 10% was envisaged among our respondents and adjustment for this was made to arrive at a minimum sample size of 422. A two-stage sampling technique was used. Stage one: Five LGHAs were selected from each of the lists of the LGHAs in each of the three senatorial districts by ballot making 15 LGAs in all. Stage two: Three WhatsApp platforms were selected from the list of all WhatsApp platforms of the different cadres in each of the 15 selected LGHAs. The number of PHC workers interviewed was proportionately allocated to the LGHAs.
Research instruments and data collection methods
Consequent upon the restriction of movement and lockdown order issued out by the Government of Nigeria, we decided to design an online data collection tool. This collection method was executed using Google Forms (docs.google.com/forms). A total of 460 PHC workers who were members in the several WhatsApp groups were approached to participate in the study. An online data collection tool was designed using Google Forms (via docs.google.com/forms). The Google Form linked to the questionnaire was sent to the participants via the identified WhatsApp groups. A semi-structured questionnaire was developed from previous studies [16,17] to collect data on socio-demographic characteristics, knowledge about COVID-19 and concern of health workers which include; self-satisfaction related concern, social-status related concern, work-related concern and government-related concern. The prepared questionnaire was linked to Google Form and sent to the participants via the already identified WhatsApp groups. The administrators of each of the identified WhatsApp groups were contacted on phones before the final dissemination of the questionnaire to the respondents.
Validation and Pretest of the instrument
The validity and reliability of the questionnaire were done before the final collection of data. Three Nigerian experts in the field of epidemiology and medical statistics in a Nigerian university evaluated the extent to which the variables in the questionnaires were relevant to the objectives of the study. Thereafter, questionnaire was pretested among PHC workers in LGHAs different from the ones used for the main study. The pretest helped to assess the relevance of the questions in producing responses from the participants. Ambiguous questions were either removed or rephrased in line with study objectives. Cronbach’s alpha internal consistency reliability of 0.84 was achieved for the analyzed variables.
Measurement of main outcome variables
Knowledge of PHC workers on COVID-19 was assessed using 13 questions adapted from the previous studies [18,19]. Respondents were scored according to their responses and each correct response was scored one point while each wrong response was scored zero. The total obtainable score was 13. Composite scores were calculated and respondents having ≥ 78% of the total obtainable score categorized as good knowledge. This scoring system agrees with a study on knowledge of COVID-19 among Chinese residents [20].
The concern of PHC workers about COVID-19 was assessed using an adapted questions from a previous study [18] and experts opinion. The Likert scale with five items ranging from strongly disagreed (1) to strongly agreed (5) was used. Responses were scored 1,2,3,4 and 5 in that order to obtain the level of concern score. The sums of the scores for individual respondent were calculated and the mean of all the scores was determined. The mean level of concern’s score was 72.4, while the minimum and maximum satisfaction score were 33.0 and 110.0 respectively. The respondents who scored up to or above the mean were categorized as those with high level of concern about COVID-19.
Data Analysis
At the expiration of the time set for the responses, all the filled questionnaires were extracted from Google Forms and transferred to a Microsoft Excel 2016 for data cleaning and coding. The cleaned data was exported to Statistical Package for Social Sciences (SPSS) version 25 (SPSS Inc, Chicago, IL, IBM Version) for entry and analysis. Univariate analyses were done using frequency distribution tables and charts. In the bivariate analysis, association between level of concern about COVID-19 and socio-demographic variables was determined using Chi Square. For every cell with an expected value less than 5, Fisher’s Exact Test was used to determine the statistical significance. In the multivariate analysis, binary logistic regression analysis was carried out to determine the demographic predictors of concern of PHC workers about COVID-19. Variables imputed into the logistic model were selected based on their level of significance during bi-variate analysis (p value was set at <0.08). Adjusted odds ratio and 95% confidence interval were obtained to identify determinants of PHC workers’ concern about COVID-19 at p value of < 0.05.