Annual visits to GPs by elderly patients CMAJ 1998;158:299 See response from: M.W. Rosenberg and E.G. Moore See also: Letter: Annual visits to GPs by elderly patients Letter: Correction: Annual visits to GPs by elderly patients In the article "The health of Canada's elderly population: current status and future implications" (CMAJ 1997;157(8):1025-32 [full text / résumé]), Drs. Mark W. Rosenberg and Eric G. Moore use data from the National Population Health Survey (NPHS) to explore the health of Canada's elderly population. They performed a logistic regression analysis to model the influence of chronic conditions on the likelihood of an individual visiting a GP more than once a year. I don't think that this is the best way to investigate utilization, because it is not necessary to use a dichotomous model to analyse frequency of visits. In any event, the authors must have made an error in their computations, because Table 5 shows negative values for the odds ratios. The odds for any event, p/(1 - p), where p is the probability for the event, can never be negative, because p can never be greater than 1. Consequently, the odds ratio, which is calculated by dividing the odds for an event in the index category by the odds for the event in the reference category, can range from zero to infinity, but it can never be negative. I used a generalized linear model to investigate the frequency of visits to an academic family medicine clinic in 1993.1 I found that the mean annual number of visits was greater for older patients, that women made more visits than men and that the presence of a chronic condition (such as back pain or hypertension, both of which appear in Table 5 of the article by Rosenberg and Moore) was associated with a higher frequency of visits. More complicated models than the one used by Rosenberg and Moore are needed to capture the complexity of health care utilization. Murray M. Finkelstein, PhD, MD, CM Assistant Professor Department of Family and Community Medicine Mount Sinai Hospital Toronto, Ont. murray.finkelstein@utoro nto.ca Reference Finkelstein MM. Analysis of OHIP billing data for patients attending the family medicine centre at Mt Sinai Hospital 1992–1993. DFCM Tech Rep 97MF2. Toronto: Department of Family and Community Medicine, University of Toronto; May 1997. Send a letter to the editor responding to this letter Envoyez une lettre à la rédaction au sujet de cette lettre | CMAJ February 10, 1998 (vol 158, no 3) / JAMC le 10 février 1998 (vol 158, no 3) |