Implementing the Dutch College of General Practitioner's guidelines for influenza vaccination: an intervention study

Influenza is a major health problem in most Western countries. In September 1993, the Dutch College of General Practitioners (NHG) issued guidelines for influenza vaccination. Although most general practitioners (GPs) are well acquainted with NHG standards, knowledge does not invariably lead to application.

To evaluate a regional intervention promoting the implementation of NHG's influenza vaccination guidelines.

In a non-equivalent control group design (pre-test 1992, post-test 1993), two general practice regions were studied. In the intervention region, Amersfoort, there were 82 practices (118 GPs, 250,000 patients) and in the control region, Arnhem, 97 practices (124 GPs, 300,000 patients). In the intervention region, all professionals involved in influenza vaccination were approached at educational meetings and by mail. Postcard material and vaccines were distributed. The main outcome measures were five organizational aspects measured by a questionnaire (registration of high-risk patients, mail prompt, vaccine in stock, special vaccination hours and vaccination by practice assistant), and the vaccination rate (number of vaccines delivered divided by the total number of regional health insurance patients).

All practices in the intervention region were involved; 78% responded to the pre-test and post-test questionnaires compared with 76% in the control region. Three of the five organizational aspects improved more in the intervention region: mail prompt by 25% (95% CI 11-38%), vaccine in stock by 29% (95% CI 16-44%), and special vaccination hours by 16% (95% CI 2-27%). Multivariate analyses failed to reveal any modifying factors. The vaccination rate increased by 21% (from 7.7% to 9.3%) in the intervention region, and by 6% (from 8.5% to 9.0%) in the control region. The mean increase in the intervention region exceeded that in the control region by 1.1 per 100 patients (95% CI 0.6-1.6). Multiple regression analysis revealed that this was an independent effect.

This complex intervention was considered to be effective. The same strategy might be appropriate for other regions and other guidelines.