Physician diagnostic and reporting practices for gastrointestinal illnesses in three health regions of British Columbia.

OBJECTIVES: To estimate seasonal proportions of patient visits due to acute gastrointestinal illness (GI), assess factors influencing physicians' stool sample requests, their understanding of laboratory testing protocols and adherence to provincial stool request guidelines in three British Columbia (BC) health regions.
METHODS: During a one-year period, eligible physicians were mailed four self-administered questionnaires used to estimate proportions of patients diagnosed with GI, related stool sample requests in the preceding month, and to assess factors prompting stool sample requests.
RESULTS: The response rate overall for the initial comprehensive questionnaire was 18.6%; 7.4% responded to all four questionnaires. An estimated 2.5% of patient visits had a GI diagnosis; of these, 24.8% were asked to submit stool samples. Significant (p < 0.05) regional and seasonal variations were found in rates of GI and stool sample requests. Top-ranked factors prompting stool sample requests were: bloody diarrhoea, recent overseas travel, immunocompromised status, and duration of illness > 7 days; "non-patient" factors included: laboratory availability, time to receive laboratory results, and cost. Physicians' perceptions of which organisms were tested for in a 'routine' stool culture varied.
INTERPRETATION: BC physicians appear to adhere to existing standardized guidelines for sample requests. This may result in systematic under-representation of certain diseases in reportable communicable disease statistics.