Related Experiment Videos
Pre-randomization decisions and group stratification in a randomized controlled trial to improve prescribing
Martine E C van Eijk1, Arsenio H P Paes, Arijan Porsius
1Department of Pharmacoepidemiology and Pharmacotherapy, Utrecht Institute of Pharmaceutical Sciences, Utrecht, the Netherlands. M.Eijk@dgvinfo.nl
Pharmacy World & Science : PWS
|September 28, 2004
Summary
Stratifying primary care practice groups by purpose, formulary use, and feedback data improved intervention effectiveness. These characteristics are crucial for optimizing prescribing behavior changes in physician groups.
Area of Science:
- Healthcare Management
- Clinical Pharmacy
- General Practice Research
Background:
- Primary care practice groups are key settings for interventions aimed at improving prescribing.
- Stratification prior to randomization in randomized controlled trials (RCTs) can enhance the precision of intervention effect estimates.
- Understanding practice group characteristics is essential for tailoring interventions to maximize impact.
Purpose of the Study:
- To identify and evaluate characteristics of primary care practice groups that are relevant for stratification before randomization in an RCT.
- To determine which practice group characteristics influence the effectiveness of an educational intervention designed to rationalize prescribing.
Main Methods:
- Structured telephone interviews were conducted with representatives of primary care practice groups.
- Pre- and post-intervention prescription data were collected and analyzed.
- Poisson regression was used to estimate incidence rate ratios for evaluating the relevance of identified characteristics.
Main Results:
- Representatives from 40 out of 61 (66%) practice groups participated.
- Three key characteristics for stratification were identified: practice group purpose (information exchange vs. binding consensus), formulary use, and feedback data use.
- These characteristics significantly modified the intervention's effect on prescribing behavior, indicating that the overall reduction in targeted drug prescribing was not uniform across all strata.
Conclusions:
- Assessing practice group characteristics enables the definition of factors for stratified randomization.
- Stratified randomization based on these factors confirmed their influence on the intervention's effectiveness in changing prescribing patterns.
- Pre-randomization assessments are valuable tools for designing effective interventions to improve physician practice patterns.