Compliance of general practitioners with a guideline-based decision support system for ordering blood tests.

BACKGROUND Guidelines are viewed as a mechanism for disseminating a rapidly increasing body of knowledge. We determined the compliance of Dutch general practitioners with the recommendations for blood test ordering as defined in the guidelines of the Dutch College of General Practitioners. METHODS We performed an audit of guideline compliance over a 12-month period (March 1996 through February 1997). In an observational study, a guideline-based decision support system for blood test ordering, BloodLink, was integrated with the electronic patient records of 31 general practitioners practicing in 23 practices (16 solo). BloodLink followed the guidelines of the Dutch College of General Practitioners. We determined compliance by comparing the recommendations for test ordering with the test(s) actually ordered. Compliance was expressed as the percentage of order forms that followed the recommendations for test ordering. RESULTS Of 12 668 orders generated, 9091 (71%) used the decision-support software rather than the paper order forms. Twelve indications accounted for >80% of the 7346 order forms that selected a testing indication in BloodLink. The most frequently used indication for test ordering was "vague complaints" (2209 order forms; 30.1%). Of the 7346 order forms, 39% were compliant. The most frequent type of noncompliance was the addition of tests. Six of the 12 tests most frequently added to the order forms were supported by revisions of guidelines that occurred within 3 years after the intervention period. CONCLUSIONS In general practice, noncompliance with guidelines is predominantly caused by adding tests. We conclude that noncompliance with a guideline seems to be partly caused by practitioners applying new medical insight before it is incorporated in a revision of that guideline.