Improving Primary Care Pharmaceutical Use
The results regarding whether simply participating in a norms development process results in positive change are ambiguous. Agyepong and colleagues (1996) in Ghana showed that involving health staff in the process of developing treatment norms for malaria, ARI, diarrhea, and hypertension resulted in only minor changes in their subsequent practice. However, Hadiyono (1996) showed that in Indonesia a two-hour group discussion between health workers and community members about perceptions and misconceptions regarding injections caused immediate and sustained reductions in injection prescribing of 18.7%. The reasons for the differences in these findings remain to be explored, e.g., culture, institutional context, target problem, the focus of the discussion, etc.
The only strategies which consistently to achieve moderate or large improvements were the group of interventions that depended on simple distribution of clinical guidelines or unbiased prescribing information to health workers. Although educational in intent, distribution of such materials is often carried out as a simple administrative process without supportive training or other reinforcing strategies. The median improvement in target prescribing in the four RCTs which tested this strategy was only 4.7%; all four interventions resulted in non-significant changes of 10% or less.(Agunawela 1991, Chowdhury 1995, Chowdhury 1996, Kafuko 1996) In fact, this result was anticipated; all of these studies were set up to contrast the efficacy of distributing printed materials in comparison to a training or administrative intervention that the researchers felt would be more powerful.
Once the simple yet consistently ineffective strategy of simply disseminating printed information has been discarded, the remaining administrative/managerial approaches are among the most effective, resulting in a median improvement of 27.5% in the outcomes targeted. These interventions have involved implementing two types of systems: supervision or performance monitoring using indicators; and prescription record audit and feedback of data to prescribers. Some interventions involved a mixture of both approaches.
Public sector supervisory systems exist in principle in most countries. Four of the administrative interventions involved training existing district- or provincial-level supervisory personnel how to function more effectively, and especially how to collect performance data rapidly using simple tools based on the WHO drug use indicators (Kafle 1995a, 1995b, Kafuko 1996) or integrated checklists (Loevinsohn 1995). The Uganda Essential