Improving Primary Care Pharmaceutical Use
Unfortunately, none of the well-designed ARI CCM studies report data on the appropriateness of differential diagnosis and treatment of respiratory infections by CHWs. The WHO ARI clinical case management algorithm is known to be sensitive in detecting pneumonia but may not be specific; whether widespread use of this algorithm by CHWs increases overall rates of inappropriate antibiotic use, and the effect of such increases on rates of antibiotic resistance, are important unanswered questions. The diarrhea CCM interventions were somewhat more attentive to measuring changes in drug use. Two studies report significant increases in community ORS use (Hetta 1984, Jintanganont 1988); two also report significant decreases in the use of antibiotics (Hetta 1984, Paredes Solari 1996).
Group process interventions are structured to combine learning, consideration of evidence about appropriate treatment, and discussion about actual practice by a facility-based peer group. The five interventions of this type in our review resulted in a median improvement in the targeted prescribing practice of 18.7%. In a notable series of implementations of the group process approach, Guiscafré (1988, 1995), Guttierez (1994), Perez-Cuevas (1996) and their colleagues targeted treatment of ARI and diarrhea in the Mexico MOH and Social Security health systems. Their intervention combined group workshops to develop facility-specific standard treatment norms with subsequent practice audits and a peer review panel that monitored adherence to these norms. The interventions resulted in both short-term (3 month) and long-term (18 month) improvements in use of specific indicated and non-indicated drugs, as well as in much greater rates of adherence to standard treatment norms.
The Mexico studies demonstrated greater improvement in practice (e.g., antibiotic use in ARI) when the guideline workshops were facilitated by national opinion leaders (-28.8%) or by health facility staff opinion leaders (-30.6%), than by health system administrators at the state level (-15.0%); however, despite lower absolute impact, the state level intervention was more cost-effective because of its greater scope. In Nicaragua, Hugh and colleagues (1996) demonstrated that is was possible to train local medical leaders to facilitate a group-oriented process to improve quality of ARI treatment within their own municipalities that resulted in reductions in antibiotic prescribing of 17.4%.