Data were coded, captured
and analysed using Statistical Package for the Social Sciences® 20.
Frequencies and percentages were calculated to describe the categorical
variables. Pearson’s chi-square test was used to determine statistical
significance with respect to the observed associations.
the designated units for the survey. This was a targeted 20% stratified
sample designed to reflect the composition of the workforce. To achieve
this, an a priori sampling frame was created using hospital employment
data from the The Personal and Salary System® (PERSAL®) database to
determine the number of HCWs from each occupational category who
would need to be surveyed in order to achieve a 20% sample of HCWs.
A target sample size of 20% of HCWs was selected, based on what
the primary investigators determined would be achievable in terms of
the recruitment process described herein. The target sample size was
calculated to be more than large enough to serve as a valid representation
of the various outcomes of interest in the study population. In order to
achieve a representative 20% sample, almost double the number of
required questionnaires were distributed to potential respondents. The
self-administered questionnaire comprised open- and closed-ended
questions. It was pilot tested and the final version was made available in
English and Sesotho, the language spoken predominantly in Free State
province.