Evaluating inventiveness in GP prescribing practice

Problem

At present, the Quality Outcomes Framework evaluates GP practices against a plethlora of criteria that requires substantial documentation to evidence, and can be thought of as adding to the administrative burden. The QOF criteria are also heavily dependent on the type of medical condition being treated. Is there a simpler way to characterise GP prescribing practice that uses prescription data alone, and that is independent of the type of medical condition being treated?

Approach

For this study, an open-source database of GP prescriptions in Wales (The General Practice Prescribing Data Extract) was analysed to extract two variables that were independent of the medical nature of the conditions prescribed for and that could therefore be applied to any prescription records: the number of different medications prescribed over a given period (described here as the "range"), plotted against the number of prescriptions issued per month (described here as the "rate"), for each GP surgery in the database. This approach was also applied to specific groups of medication taken over all surgeries.

Findings

Plotting the number of distinct prescriptions (the "range") against the number of prescriptions per month (the "rate") for each GP surgery in Wales gave a positive correlation, so that the higher the prescription rate, the higher the range of prescriptions offered. This can be understood in terms of the higher number of medical conditions being presented in the more busy surgeries. However, the slope of the curve decreased at the higher prescription rates: this could be thought of as "prescription fatigue", where the busier GPs lack the capacity for considering a change in medication, or suggesting several alternatives. When the data was fitted with a logarithmic curve, the points above the curve were colour-coded as "inventive", and the points below the curve were colour-coded as "cautious", and when this was plotted on a map of Wales, it could be seen that valley surgeries were more cautious than coastal surgeries. When plotted for different groups of medication, immunological medications could be seen to have a particularly low range for the rates prescribed, suggesting that this should be a priority area for research.

Consequences

This study suggests a way to measure GP prescribing practice that is independent of both the size of the practice and the nature of the medical conditions, and therefore can be applied outside the QOF framework and perhaps more generally. By plotting the range of medications prescribed against the rate of prescription per month, a key attribute of "inventiveness" (or the opposite, "caution") can be measured to help characterise the culture of a GP practice.