Critical study of prescriptions allows irrational prescribing to be identified. A pioneering study of health centres in Ghana (19) found that if the prescriptions given for the four most frequent grounds for consultation (75% of consultations) had followed the recommendations of the national health authorities, the cost of the drugs prescribed would have been reduced by 70%. The drugs prescribed were too expensive (paracetamol instead of aspirin, parenteral rather than oral administration); or they were prescribed in excessive quantities (injections and tablets of the same drug). It is nevertheless very difficult to measure the proportion of treatments that follow standard schedules of treatment. It is therefore preferable to use indirect indicators (Table 14).

The International Network of Rational Use of Drugs (INRUD) has also put forward a list of basic indicators of rational prescribing, some of which apply to prescribing itself and others to patient care and the health services (20). These indicators have been tested in 12 countries (21).

The optimal values indicated are approximate. They permit recognition of prescribing habits that differ widely from them, but they do not permit assessment of prescribing habits that are similar to them. There is no optimal value for the average duration of consultations or dispensing. Very short durations indicate a problem, while long durations do not give precise information.

Table 14. Proposed indicators of rational prescribing

Variance from these indicators may be used to set national priorities for rational prescribing, depending on the type of problem that appears most important and the health services concerned. Indicators also permit later measurement of the effect of remedial actions.

Some complementary indicators permit evaluation of the cost of irrational prescriptions (20): mean cost of drugs per prescription, percentage of the cost of drugs spent on antibiotics, and percentage of the cost of drugs spent on injections.

By use of these indicators, one can calculate the total excess cost of prescriptions that order too many drugs, too many antibiotics, or too many injections in relation to a reference situation. Measurement of excess costs can help to identify the most pressing problems. And, knowing that many prescriptions are neither medically nor economically justified provides a strong incentive to act.

More detailed studies might measure the consequences of bad prescribing by assessing the care or the health conditions that more rational prescribing would have prevented. Indicators must be adapted to the type of prescription studied. For example, the consequences of prescribing antibiotics and antiparasitic drugs may be measured by the prevalence of resistant strains and the need for recourse to drugs of second choice.

The consequences of irrational prescribing have costs, but these costs do not necessarily have any very precise meaning. They depend on the behaviour of patients and the evolution of diseases, and some events may be very widely spaced in time. Indirect costs need not be measured to identify the scale of problems.