Applicability of diagnostic recommendations on dementia in family practice

Objective. To evaluate the applicability of the Dutch dementia guideline's recommendations, including the diagnostic criteria used by family practitio ners, and to explore characteristics in both patients and family practition ers which are associated with the use of these recommendations. Design. An observational study was set up with a sample of 64 family practi tioners who were instructed to use the Dutch national dementia guideline on incident-suspected dementia patients. The applicability was expressed as t he percentage of recommendations applied. The use of diagnostic criteria was checked by comparing the family practiti oners diagnoses with the diagnoses received by integrating the registered s ymptoms according to the DSM-III-R criteria. Associations between the numbe r of recommendations applied, and demographic and clinical features were ex plored. Main outcome measures. (i) guideline applicability (ii) integrated use of D SM-III-R criteria. Results. 207 patients were included. The average application rate of the gu ideline's 31 diagnostic key recommendations was 86% or 24.8 (SD 3.6). The f amily practitioners diagnoses were consistent with the expected DSM-III-R d iagnoses in 26% of the cases (kappa =0.1). A greater number of patients in a practice was positively associated with the use of recommendations. A nee d for referral by the family practitioners, and patients' denial of dementi a were negatively associated with the use of recommendations. The presence of dementia and a patient's age were negatively associated with the use of the DSM-III-R criteria. Conclusions. The applicability of the diagnostic recommendations of the nat ional Dutch dementia guideline in a representative sample of family practit ioners was promising. Nevertheless, the diagnostic criteria of the DSM-III- R, which were pan of the dementia guideline, provided little or no guidance to the family practitioners in their diagnostic decision-making. Clinical and demographic variables explained some of the variation in the use of rec ommendations.