Measurement of changes in antihypertensive drug utilization following primary care educational interventions

Fiona E Horn, John A Mandryk, Judith M Mackson, Sonia E Wutzke, Lynn M Weekes, Rob J Hyndman
(2007) Pharmacoepidemiology & Drug Safety 16(3), 297-308

DOI

Abstract: Purpose To measure changes in drug utilization following a national general practice education program aimed at improving prescribing for hypertension.

Methods A series of nationally implemented, multifaceted educational interventions using social marketing principles focusing on prescribing for hypertension, was commenced in October 1999, and repeated in September 2001 and August 2003. The target group was all primary care prescribers in Australia and interventions were both active (voluntary) and passive. Newsletter and prescribing feedback was mailed in October 1999, September 2001 (newsletter only) and August 2003. Approximately a third of general practitioners (GPs) in Australia undertook at least one active educational activity (clinical audit, educational visit or case study) during the period October 1999-April 2004. National dispensing data from 1996-2004 were analysed using time series methodology with a decay term for intervention effect, to assess trends in prescribing of various classes of antihypertensives. In particular the program aimed to increase the prescribing of thiazide diuretics and beta blockers.

Results Consistent with key intervention messages, the program achieved an increase in low-dose thiazide and beta blocker prescribing. The rate of prescribing of low-dose thiazides doubled from 1.1 per 1,000 consultations in October 1999 to 2.4 per 1,000 in October. 2003. Beta blocker utilisation showed a more modest but significant increase over the time of the study, with the change in observed versus expected rate of prescribing increasing by 8% by April 2004. Therapeutic options for treating hypertension changed markedly in the time of the study with the advent of ACE inhibitor / Angiotensin II antagonists and thiazide combination products. It is important, therefore, to interpret the results in light of these changes.

Conclusions A national education program aimed at GPs was successful in improving prescribing for hypertension. Lessons learned will be applied in evaluation of future NPS programs and are also applicable to analysis of other interventions aimed at influencing prescribing behaviour.

Keywords: antihypertensives; primary care; educational interventions; time series analysis; decay; regression methods.