Objective: To determine whether clinical audit improved the performance of diabetic clinical processes in the health district in which it was implemented.
Design: Patient folders were systematically sampled annually for review.
Setting: Primary health-care facilities in the Metro health district of the Western Cape Province in South Africa.
Participants: Health-care workers involved in diabetes management.
Intervention: Clinical audit and feedback.
Main outcome measure: The Skillings-Mack test was applied to median values of pooled audit results for nine diabetic clinical processes to measure whether there were statistically significant differences between annual audits performed in 2005, 2007, 2008 and 2009. Descriptive statistics were used to illustrate the order of values per process.
Results: A total of 40 community health centres participated in the baseline audit of 2005 that decreased to 30 in 2009. Except for two routine processes, baseline medians for six out of nine processes were below 50%. Pooled audit results showed statistically significant improvements in seven out of nine clinical processes.
Conclusions: The findings indicate an association between the application of clinical audit and quality improvement in resource-limited settings. Co-interventions introduced after the baseline audit are likely to have contributed to improved outcomes. In addition, support from the relevant government health programmes and commitment of managers and frontline staff contributed to the audit's success.