Human factors affecting the quality of routinely collected data in South Africa.

Evaluations that have looked at the people aspect of the health information system in South Africa have only focused on the availability of human resources and not on competence or other behavioural factors. Using the Performance of Routine Information System Management (PRISM) tool that assumes relationships between technical, behavioural and organizational determinants of the routine information processes and performance, this paper highlights some behavioural factors affecting the quality of routinely collected data in South Africa. In the context of monitoring maternal and child health programmes, data were collected from 161 health information personnel in 58 health facilities and 2 district offices from 2 conveniently sampled health districts. A self-administered questionnaire was used to assess confidence and competence levels of routine health information system (RHIS) tasks, problem solving anddata quality checking skills, and motivation. The findings suggest that 64% of the respondents have poor numerical skills and limited statistical and data quality checking skills. While the average confidence levels at performing RHIS tasks is 69%, only 22% actually displayed competence above 50%. Personnel appear to be reasonably motivated but there is considerable deficiency in their competency to interpret and use data. This may undermine the quality and utility of the RHIS.