Improving Primary Care Pharmaceutical Use
drugs, lowering the use of injections, and improving compliance with guidelines. Our conclusions about impact are based on this limited range of topics, although these conclusions may prove to be more widely generalizable as a greater variety of problems are addressed.
The quality of study designs used in these interventions has been improving over time, with a greater emphasis on RCTs and an increasing number of time series studies. However, despite gains in design, the quality of statistical analysis in these interventions is often weak. Some work has been done in describing appropriate design and analysis strategies when the unit of randomization is actually the group (e.g., health centers, hospitals, or districts), as if often is in public sector interventions.(Diwan 1992). However, much more needs to be done in developing standardized approaches for the analysis of drug use interventions that are methodologically and statistically correct, yet still accessible to researchers and managers who are not academics.
A variety of educational and administrative intervention approaches have been tested. It is encouraging that the people and problems targeted have been responsive to improvement; more than three-fourths of the reviewed interventions have had moderate to large impacts on at least one targeted outcome. Based on these experiences, we can draw the following conclusions about effectiveness:
Well-designed training interventions, whether conducted in large or small groups, can successfully improve targeted prescribing outcomes by an average of 15% or more. The impact of training seems to be increased by: employing multiple training modalities (lectures, group problem-solving, role playing, opportunity to practice skills); repeated sessions; focus on one clinical problem at a time; training at the work site; and using opinion leaders or district-level staff as trainers.
The dissemination of printed clinical guidelines or unbiased drug information alone is not sufficient to cause measurable improvements in behavior. However, active dissemination of guidelines through staff training, peer group re-invention of guidelines, or audit and feedback has resulted in improved compliance with the guidelines.
The evidence is clear that community case management through CHWs can reduce child mortality from ARI and diarrhea. However, we know little about the impacts of pharmaceutical treatment by CHWs on