PubMed2002
BACKGROUND: Spontaneous physician behaviour can affect guideline applicability, implementation strategies and application costs, particularly in relation to widespread pathologies chiefly managed by general practitioners (GPs). Of the array of dyspepsia management guidelines, the closest to general practice, partly owing to proposing committee composition, are the European Society for Primary Care Gastroenterology (ESPCG) guidelines. METHODS: To evaluate variability in dyspepsia management among GPs in Padua and divergence in spontaneous prescriptive behaviour from the ESPCG dyspepsia guideline, we prospectively studied the behaviour of 39 GPs over a 3-month period of outpatient activity, through questionnaires on each consultation. Test-group representativeness was preliminarily defined in terms of antisecretory drug expenditure. RESULTS: 1790 forms on dyspepsia-related consultations were studied in a population of 51,193 registered patients; 1264 patients with a history of dyspeptic pathology consulted their GP (19% duodenal ulcer (DU), 9% gastric ulcer (GU), 54% gastro-oesophageal reflux disease (GERD), 32% non-ulcer dyspepsia (NUD), 1% cholelithiasis), while 526 patients presented with symptoms of dyspepsia with no previous gastroscopy (EGDS) (42% were aged <45 years), of whom 42% had twice consulted their GP. Empirical management by prescription of symptomatic drugs was the most common procedure in DU (33%), GU (73%) and NUD (74%) relapses. Helicobacter pylori eradication therapy was prescribed in only 2% of patients with a history of organic or functional dyspepsia. 145 patients with uninvestigated dyspepsia were referred for second-level endoscopy and 43 for H. pylori testing. Forty-four percent of endoscopies prescribed for uninvestigated patients did not comply with the ESPCG guideline; full compliance would have determined a 105% rise in endoscopies. Prescriptive variability between GPs was high (based on the Goodman-Kruskal (0.41, P < 0.001) and Cramer tests (V = 0.51, P < 0.005)) and agreement between observed and expected prescriptions according to ESPCG criteria was as low as V = 0.11. On the basis of the most frequently observed behaviours, we developed three options of the ESPCG guideline and compared them to spontaneous prescriptions. CONCLUSIONS: Highest compliance emerged where the clinical approach for all patients with uninvestigated dyspepsia was symptomatic therapy at first presentation followed by a different attitude at second presentation, setting a higher cut-off age than in the guideline (which in our case proved, on mathematical calculation, to be 55 years).