Improving Primary Care Pharmaceutical Use
improving pharmaceutical use in these settings not published in international journals, or that were reported in languages inaccessible to us. Our review is biased toward including published studies carried out by academics and interventions sponsored by international agencies. Because of this bias, our results probably overstate the ability of interventions as usually conducted to improve practice.
It is clear that many efforts to improve drug use are not evaluated. Because of this, the impacts of countless activities such as training courses, in-service seminars, formularies, drug supply improvements, drug bulletins, or standard treatment guidelines designed to improve drug use are simply not known. Furthermore, when an attempt is made to evaluate impacts, study designs are frequently invalid; the evidence from over one-third of the studies identified for this review was not considered because of methodological inadequacy.
There are fundamental problems in comparing interventions which have a diversity of objectives, measurement methods, and outcomes. Many of the included interventions are multifactorial, and difficult to characterize in a simple classification. Training, the most common intervention, can be conducted in many different ways with a broad range of objectives. Categories like “community case management” or “essential drugs program” blur large distinctions in the way these approaches have been implemented. In addition, our comparison strategy of calculating a single “effect size” to compare across different outcomes and interventions is a technically unsatisfactory and conceptually limiting way to gauge either the magnitude or the importance of impacts.
Nevertheless, within these limitations, our review uncovers some clear lessons about what we know - and what we do not yet know - regarding how to improve use of medicines in primary health care in developing countries.
What do we know?
Despite the substantial number of interventions to improve drug use in primary care, the breadth of our experience is still quite limited. Most reported interventions have been directed at acute illnesses in children, specifically ARI and diarrhea. From the standpoint of prescribing practices, much of the effort has been directed at influencing the decision to use or not use antibiotics, with some attention to reducing the prescribing of multiple