The determinants of this discrepancy between demand for and provision of elderly psychiatric services are not known. The major thrust of psychiatric residency programs is to train psychiatrists in the necessary knowledge and skills to practise psychiatry for all patient populations. In Canada, since 1987, the Royal College of Physicians and Surgeons of Canada has required that training in geriatric psychiatry be provided in all psychiatry residency programs (7,8). It is not clear whether this training has contributed in any way to interest in and the practice of geriatric psychiatry. Herrmann, in a 1992 study of 16 graduates of the University of Toronto who had spent 6 months on a geriatric psychiatry inpatient unit during the first year of their residency, suggested that for psychiatrists who had gone on to commit their entire practice to geriatric patients, supervisors and positive clinical experiences early in training strongly influenced their decisions (9). Despite the fact that 5 of these 16 residents had no interest in geriatric psychiatry prior to this rotation, 13 were seeing geriatric patients in their practices and 9 of them were devoting more than 50% of their time to this age group. His report did not specify whether these psychiatrists were practising in the community or in teaching centres. Lazarus reported on the graduates of a Chicago psychiatry residency program, who had had mandatory training in geriatric psychiatry (10). In 1987, these psychiatrists reported that 7% of their patients were over 65 years of age (compared with 1970 finding of 0.5% in a US survey) and attributed this to the impact of this training program. He suggests that this mandatory training experience encouraged trainees to incorporate geriatrics into their practices. He does not comment on the fact that this program may be attracting learners because they are already interested in this geriatric component. His findings, then, may reflect a selection bias rather than the influence of the program itself. This is also true for Richardson’s findings. His  report of the impact of geriatric training in a family medicine program at the University of Maryland found that most graduates devoted a significant portion of their practice time to geriatric patients (11). Watson and Jolley, in a 1989 study of British senior psychiatry registrars who chose advanced training in geriatric psychiatry, found that 45% went on practise general psychiatry (12). They did not, however, ascertain whether these 45% were seeing any geriatric patients in their practices.

There have been several studies of psychiatric trainees and their expressed interest in seeing geriatric patients. El-Guebaly, in a 1980 survey of final-year psychiatry residents, found that 50.9% were interested in geriatric psychiatry (13). Walker, in 1985, reported that 46% of senior Canadian psychiatry residents expressed an interest in a career in geriatric psychiatry (14). Draper surveyed residents in Australia and New Zealand in 1993, 70% of whom had had a psychogeriatric training experience (15). He found that 70% were interested in practising geriatric psychiatry, either as a specialty or as part of general psychiatry. None of these studies examined what aspects of training the residents found to be influential in generating this interest.

We do not know how this interest translates into practice. A high degree of interest in a resident may result in practising geriatric psychiatry. Or the high interest may have little follow-through. Or the alleged high interest may reflect social desirability bias in completing the questionnaire.

Whether prepractice interest in geriatric psychiatry is significant or necessary to ultimately practising geriatric psychiatry is not clear. Prochaska, in his transtheoretical model of behavioural change, suggests that there are 5 stages that individuals go through in modifying behaviour (as it applies to psychotherapy or addictive behaviours such as smoking) (16): precontemplation,  contemplation, preparation, action, and maintenance. All stages are necessary in the evolution of behavioural change. He would argue that interest, spanning the contemplation and preparation stages, is a necessary precondition to the development of the behaviour  (practising geriatric psychiatry), but is not sufficient to result in behaviour change.

Several studies have examined the noneducational factors that influence practice choice. Spitzer, in a 1962 study of University of Toronto graduates, found that time requirements, interest in specialized knowledge, independence, and issues related to raising a family influenced the decision to pursue a specialty (17). Ford, in a survey of 170 psychiatrists regarding their opinions of 4 clinical vignettes, found that psychiatrists in general perceived the elderly to be less ideal for their practices and to have a poorer prognosis than younger patients with similar stories, reflecting an ageist bias (18). Small, in an article on recruitment issues in geriatric psychiatry, comments on the poor financial reimbursement for geriatric psychiatric services (19).

All of the studies to date have been cross-sectional and used no, or poor, measures of influencing factors and interest in geriatric psychiatry; consequently, the results are difficult to interpret.

Given that the intent of mandatory training in geriatric psychiatry is to prepare future psychiatrists to care for this patient population and that interest in practising geriatric psychiatry is a necessary precondition for practice, we examined these issues from the perspective of senior psychiatry residents. We assumed that the link between training and practice would be the development of interest in treating this patient population. This pilot study generates hypotheses regarding factors that influence senior psychiatric trainees to treat geriatric patients in their future practices. This study was implemented in parallel with a similar study of practising geriatric psychiatrists.

Method

The Delphi technique, which was developed by Helmer and associates at the Rand Corporation in the early 1950s, was used to generate ideas and consensus on what the influential factors are for psychiatry residents. This process provides an anonymous setting in which opinions can be voiced without  the experts directly meeting each other. A group of experts having been identified, the experts provide several brief statements regarding a specific issue of interest. These viewpoints are subsequently synthesized and offered back to the group to elicit their opinions. The goal is to gain consensus. The