EXPERIMENTAL EVALUATION OF THE EFFECTS OF DRUG INFORMATION ON ANTIBIOTIC PRESCRIBING - A STUDY IN OUTPATIENT CARE IN AN AREA OF SRI-LANKA

The intervention level of epidemiology is useful for studying effects in health systems research. Due to practical and ethical reasons, it is often difficult to apply experimental methods such as classical randomized clinical trials in the field. However with alternative approaches such as 'randomization by group' some of these problems can be overcome. Drug information has since long been considered as an instrument to influence physicians, however evaluation of its effects is a new field of research. In the present study the impact of drug information on prescribing behaviour of its effects is a new field of research. In the present study the impact of drug information on prescribing behaviour was evaluated in an outpatient setting in Sri Lanka. The study include 15 state health institutions (45 prescribers) with a common drug formulary. Groups of prescribers were randomized into two interventions; newsletters and newsletters reinforced by a group seminar, and one control group. The target topic was' rational prescribing of antibotics'. Some 18 766 randomly selected outpatient drug prescriptions were studied. Antibiotics (and sulphonamides) were prescribed to 33.2% of the patients. An overall trend towards a decrease in proportion of patients prescribed antibiotics in the two intervention groups was seen, although the difference was not significant (p > 0.05) compared to the control group. This is similar to the effect of written information on prescribing in other studies. A mean difference of -7.4% in written, -7.3% in written + seminar and -0.4% in the control group was shown. The general antibiotic prescribing pattern did not change in any of the three groups. Penicillin was the most commonly prescribed antibiotic and tetracycline was only rarely prescribed to children. This experiment indicates the feasibility of drug information intervention studies in developing countries. Problems related to the integrity of individual physicians, lack of centralized prescribing statistics and old methods of prescription forms could be overcome. Suggestions are made regarding future research in this area. The weaker than expected effect of information may be due to lower overall antibiotic prescribing than in other Asian countries, and that information was not targeted at identified misuse.