RESUMEN
BACKGROUND AND PURPOSE: Our earlier study showed that structured education of general practitioners (GPs) improved their practice in headache management. Here the duration of this effect was assessed. METHODS: In a follow-up observational study in southern Estonia, subjects were the same six GPs as previously, managing patients presenting with headache as the main complaint. Data reflecting their practice were collected prospectively during a 1-year period commencing 2 years after the educational intervention. The primary outcome measure was referral rate (RR) to neurological services. Comparisons were made with baseline and post-intervention data from the earlier study. RESULTS: In 366 patients consulting during the follow-up period, the RR was 19.9%, lower than at baseline (39.5%; P < 0.0001) or post-intervention (34.7%; P < 0.0001). The RR was diagnosis-dependent: the biggest decline was for migraine. Use of headache diagnostic terms showed changes generally favouring specific terminology. In particular, the proportion of patients given migraine diagnoses greatly increased whilst use of the inappropriate M79.1 (Pericranial) myalgia almost disappeared. Requests for investigations, which had fallen from 26% (of patients seen) at baseline to 4% post-intervention, resurged to 23% (mostly laboratory investigations; requests for X-rays continued to dwindle). Initiation of treatment by the GPs remained at the post-intervention level of just over 80% (up from baseline 58%). CONCLUSIONS: Improvements in GPs' practice after a structured educational programme mostly last for ≥3 years, some showing further betterment. A few measures suggest the beginnings of a decline towards baseline levels. This policy-informing evidence for continuing medical education indicates that the educational programme needs repeating every 2-3 years.