Effectiveness of multifaceted implementation of guidelines in primary care.

OBJECTIVE: To evaluate the effectiveness of a tailored and multifaceted approach to the implementation of nationally recommended and evidence-based guidelines in primary care within existing systems and resources.
METHODS: A non-randomised Latin square to compare guideline implementation in two neighbouring health districts covering 180 general practices. Evidence-based guidelines for the treatment of patients with asthma and angina were implemented actively in one district and passively disseminated in the other district. Outcome measures for asthma were smoking status and inhaler technique. For angina the outcome measures were: smoking status; blood pressure; aspirin prescribed, contraindicated or self-medicated; beta-blocker prescribed or contraindicated; routine hospital admission; prescribed drugs; self-reported change.
RESULTS: There were improvements in all outcome criteria between baseline and follow-up audits, regardless of whether the guideline was actively implemented or passively disseminated. The estimated increase in the proportion of records complying with guidelines was 4% [95% confidence intervals (CI): 0, 8] and was higher in intervention than in control practices. Using only records not compliant at baseline, the corresponding difference was 15% (95% CI: 7, 24). The only significant improvement associated with active implementation was smoking status in angina patients. Both prescribing and hospital admission monthly totals changed during the period of the trial, but there was no significant difference between the pattern of changes in intervention and control districts. A significantly greater proportion of health professionals saw the intervention guideline compared with the control (75% versus 25%). There was a significant correlation between self-reported change and interventions steps (P < 0.05).
CONCLUSIONS: Increases in quality markers occurred irrespective of the multifaceted implementation efforts. Some of this increase was due to the method of data collection. Nevertheless, national initiatives may have more influence than local implementation initiatives.