All Canadian physicians (family physicians/general practitioners and specialists) (N=61,751) were surveyed using a combination of mailed and electronic questionnaires, distributed and completed between February and June 2004. In an effort to achieve the most accurate representation of Canadian Physicians, an additional one-page questionnaire was sent to all physicians not responding to the original document in July-August 2004.
All second year medical residents and all medical students in Canada were surveyed electronically during April - August 2004. Students and second year medical residents were contacted by email, forwarded through their university medical programs.
The 2004 National Physician Survey (NPS) was carried out as a self-reported survey, sent to all licensed physicians in Canada. The NPS mail and email list was generated from the Canadian Medical Association (CMA) Masterfile. The CMA Masterfile includes all physicians in Canada holding a medical license and is compiled and updated on a daily basis with information received from provincial licensing bodies, associations, College of Family Physicians of Canada (CFPC) and Royal College of Physicains and Surgeons of Canada (RCPSC) membership listings, and individual physicians. Email addresses were verified for 34.0% of all physicians. This group received survey communications by email, including invitations to complete the questionnaire online. The remaining 66.0% of physicians received all NPS communications by mail. In total, 61,751 physicians in all provinces and territories were asked to reply to the 2004 NPS, 31,965 family physicians/general practitioners (FPs) and 29,786 specialists.
Responses from all questionnaires returned by June 30th, 2004 were captured into an electronic database file. The file created by physicians completing the online questionnaire and the file created by the returned paper responses were merged to create a single NPS database.
Unique identification numbers were assigned for the NPS project to ensure that physician responses would remain confidential and anonymous, to enable subsequent mailings of the questionnaire to physicians who had not yet replied, and to apply the same numbers to future NPS surveys for longitudinal analysis.
