The average number of drugs prescribed by staff at CBC facilities was 2.8 drugs per prescription, with considerable variation between facilities (range 2.0-3.7 drugs). Factors found to correlate with a lower average number of drugs per prescription included supervision and prescribers' level of education, training, and years of experience. Specifically, prescribers at facilities with on-site supervision and/or bimonthly supervision by doctors prescribed significantly fewer drugs per prescription. Prescribers at facilities where the majority had completed at least a six-week training course, in comparison to either four-days or one-week's training, or no training at all, prescribed fewer drugs per prescription. Facilities where the mean years of prescribing experience was greater than nine were also found to have a lower average number of drugs per prescription, as were facilities where the majority of prescribers had university-level education.

The average antibiotic prescription rate at CBC facilities was 33.8%. There was significant variation between facilities (range 26.1-50.6%). Prescribers at facilities that received at least bimonthly supervision by doctors prescribed significantly fewer antibiotics. So did prescribers at facilities where the majority had participated in at least a six-week training course.

The total number of patients seen at a facility and number of personnel, indicators not previously examined, did not correlate with either drug or patient care indicators. Busy facilities did not prescribe a significantly different number of drugs or antibiotics. Busy facilities did not have significantly different consulting, dispensary waiting or dispensing times. Neither did the average number of drugs per prescription affect consulting, dispensary waiting or dispensing times. The only variable that correlated to significantly reduced patients' dispensary waiting time was off-site supervision. Time spent dispensing drugs was longer in facilities where the majority of dispensary personnel had at least secondary schooling. However, neither patient knowledge nor the amount of education, training or supervision of dispensing personnel correlated with the length of time spent dispensing drugs.

The average cost per prescription, also a previously untested complementary indicator, was CFAF1968 (US$3.94) for CBC facilities. Prescription cost did not correlate with any of the variables examined, including average number of drugs per prescription and antibiotic prescription rate. Neither did the drug indicators correlate with laboratory referral rates.

In addition, standard treatment guidelines (STGs) were found to have little influence on prescribing practices. Only 11.3% of the treatments prescribed in case scenarios administered to prescribers were in accordance with the STGs. This was further corroborated by chloroquine, quinine and Fansidar prescription rates, which suggested that quinine, not the Fansidar regime recommended in the STGs, was being prescribed for chloroquine-resistant Plasmodium.