Evaluating the impact of educational interventions on use of antithrombotics in Australia

• Published in: Pharmacoepidemiology and drug safety 2008-02, Vol.17 (2), p.160-171
• Main Authors: Mandryk, John A., Wai, Angela, Mackson, Judith M., Patterson, Craig, Bhasale, Alice, Weekes, Lynn M.
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Aged; Anticoagulants - therapeutic use; antithrombotics; Australia; Databases, Factual; Dipyridamole - therapeutic use; Drug Utilization - statistics & numerical data; Drug Utilization - trends; Education, Medical, Continuing; educational intervention; evaluation; Fibrinolytic Agents - therapeutic use; general practitioners; Humans; Linear Models; Middle Aged; Platelet Aggregation Inhibitors - therapeutic use; Practice Patterns, Physicians' - statistics & numerical data; Practice Patterns, Physicians' - trends; prescribing data; Retrospective Studies; Ticlopidine - analogs & derivatives; Ticlopidine - therapeutic use; time series; Warfarin - therapeutic use
• ISSN: 1053-8569; 1099-1557
• DOI: 10.1002/pds.1536

Purpose The National Prescribing Service (NPS) aims to improve prescribing and use of medicines consistent with evidence‐based best practice. In particular, specific educational programmes were commenced in 2003 to improve general practitioner (GP) prescribing of antithrombotics. This report assesses the impact of these multiple educational interventions in terms of changes in prescribing rates. Methods Monthly prescribing data (July 1996–December 2005) were obtained from a national claims database, and yearly data (2001–2006) from a GP–patient encounter database. The target group was all GPs in Australia and interventions were active (voluntary) and passive (mail‐outs). Responses to the interventions were measured by changes in the mean number of antithrombotic prescriptions (for ticlopidine, clopidogrel, warfarin and dipyridamole) per 1000 consultations for each GP each month. These data were analysed using seasonally adjusted piecewise linear dynamic regression. The data from the GP–patient encounter database were reported as mean prescribing rates per 100 GP encounters. Results NPS interventions either had an effect in the expected direction or had no discernable impact. Prescribing appeared to have decreased for dipyridamole, clopidogrel and ticlopidine, although the decline was only statistically significant for dipyridamole. Prescribing of warfarin continued to rise steadily despite NPS efforts. Conclusions The NPS antithrombotics programme appears to have had modest success, but such evaluations raise questions about whether a focus on outcomes at a national level is appropriate, given likely concealment of effects at local levels. Lessons learned should be applied in the evaluation of other programmes aimed at influencing prescribing. Copyright © 2007 John Wiley & Sons, Ltd.