Multilevel analysis of the influence of patients' and general practitioners' characteristics on patented versus multiple-sourced statin prescribing in France.

BACKGROUND

The French National Health Insurance and the Ministry of Health have introduced multiple reforms in recent years to increase prescribing efficiency. These include guidelines, academic detailing, financial incentives for the prescribing and dispensing of generics drugs as well as a voluntary pay-for-performance programme. However, the quality and efficiency of prescribing could be enhanced potentially if there was better understanding of the dynamics of prescribing behaviour in France.

OBJECTIVE

To analyse the patient and general practitioner characteristics that influence patented versus multiple-sourced statin prescribing in France.

METHODOLOGY

Statistical analysis was performed on the statin prescribing habits from 341 general practitioners (GPs) that were included in the IMS-Health Permanent Survey on Medical Prescription in France, which was conducted between 2009 and 2010 and involved 14,360 patients. Patient characteristics included their age and gender as well as five medical profiles that were constructed from the diagnoses obtained during consultations. These were (1) disorders of lipoprotein metabolism, (2) heart disease, (3) diabetes, (4) complex profiles and (5) profiles based on other diagnoses. Physician characteristics included their age, gender, solo or group practice, weekly workload and payment scheme.

RESULTS

Patient age had a statistically significant impact on statin prescribing for patients in profile 1 (disorders of lipoprotein metabolism) and profile 3 (complex profiles) with a greater number of patented statins being prescribed for the youngest patients. For instance, patients older than 76 years with a complex profile were prescribed fewer patented statins than patients aged 68-76 years old with the same medical profile (coefficient: -0.225; p = 0.0008). By contrast, regardless of the patient's age, the medical profile did not affect the probability of prescribing a patented statin except in young patients with heart diseases who were prescribed a greater number of patented statins (coefficient: 0.3992; p = 0.0007). Prescribing was also statistically influenced by physician features, e.g., older male physicians were more likely to prescribe patented statins (coefficient: 0.245; p = 0.0417) and GPs practicing in groups were more likely to prescribe multiple sourced statins (coefficient: -0.178; p = 0.0338), which is an important finding of the study. GPs with a lower workload prescribed a greater number of patented statins.

CONCLUSION

There is significant variability in the prescribing of different statins among patient and physician profiles as well as between solo and group practices. Consequently, there are opportunities to target demand-side measures to enhance the prescribing of multiple-sourced statins. Further studies are warranted, in particular in other therapeutic classes, to provide a counter-balance to the considerable marketing activities of pharmaceutical companies.