Care Partnerships Between Family Physicians and Rheumatologists

MATERIALS AND METHODS

We selected a random sample (20%, n = 478) of family physicians from the mailing list of the Quebec College of Family Physicians (N = 2393). All these family physicians were practicing within the McGill University Integrated Health Network (RUIS) at the time the mailing list was created. A survey package was sent, including a personalized cover letter, an English or French questionnaire (depending on the physician’s preference, recorded by the College), and a stamped, addressed return envelope. Two waves of followup mailings, at an interval of 2 weeks, were sent to initial nonrespondents.

The questionnaire (provided as an Appendix) asked the physician to indicate whether there is a rheumatologist (or more than one) to whom the family physician tended to refer patients. If this was the case, that family physician was considered to represent the “rheumatology care partnership” model of practice. Physicians who did not identify a rheumatologist were retained for the study to represent the “nonpartnership” model of practice. The questionnaire further asked the family physician to rate, on a 5-point scale, factors of importance related to that relationship with the rheumatologist. Data collection included elements that have been shown previously to influence shared care between family physicians and specialists: physician demographics, organizational factors, and accessibility^3,4,5.

We calculated overall descriptive statistics of the family physicians who identified rheumatology care partnerships, and compared their demographics (age, sex, year of graduation, and practice setting) with those family physicians who did not identify a partnership. Practice variables included single versus group practice, and academic versus private setting. We also performed multivariate logistic regression to explore potential factors independently associated with the existence of rheumatology care partnerships, and with the factors that family physicians and rheumatologists consider as most important in their care partnership interactions.

RESULTS

Of the 478 family physicians, 11 were excluded because they did not have an active general family practice, and 5 physicians were excluded because they held a restrictive permit of practice in Quebec and were no longer in practice at the address given. Of the remaining 462 physicians, 84 completed the questionnaire, for a response rate of 18.2%. There was a trend toward more female and more English respondents, compared to the established demographics as of 2007 (Table 1). The respondents also tended to be academic-based, as compared to the family physician demographics recorded by the College in 2007.

Of the 84 respondents, 52 (61.9%) reported having rheumatology care partnerships according to our definition. Among them, 43/52 (82.7%) indicated having patients with RA in their practice, 3 did not have such patients, 4 were unsure, and 2 did not provide an answer. Of the remaining 32 physicians who did not report care partnerships, 16 (50.0%) had patients with RA in their practice and 15 (46.9%) did not (1 physician was unsure). After adjusting for age, sex, academic status, and graduation year in a multivariate logistic regression model, those who completed the English questionnaire were more likely to report having a rheumatology care partnership (adjusted OR: 4.9; 95% CI: 1.6 to 15.1). We did not observe any other physician characteristic to be associated with partnership status.

Regarding factors of importance in interactions with rheumatologists, most respondents rated the following factors as important (score ≥ 4/5): adequate communication and information exchange (44/50, 88.0%); waiting time for new patients (40/50, 80.0%); clear and appropriate balance of responsibilities (39/49, 79.6%); and patient feedback and preferences (34/50, 68%). Around half the respondents (23/50, 46.0%) accorded high importance to their personal knowledge of the rheumatologist. Regarding their actual relationships with rheumatologists, 30/50 (60.0%) of respondents felt communication and information exchange were clear and appropriate, and 35/50 (70.0%) felt they had a clear balance of responsibilities.

Among those who rated factors of importance, stratified analyses (Table 2) indicated that male respondents, compared to females, tended to give high importance (score ≥ 4/5) to the following elements: personal knowledge of the rheumatologist and physical proximity of practice. Physicians older than 40 years were more likely to assign high importance to clear and appropriate balance of responsibilities, and patient feedback and preferences. Regarding their actual relationships with rheumatologists, male respondents were more likely to report that communication and information exchange were clear and adequate. However, in the multivariate linear regression model that adjusted for age, sex, language, academic status, and graduation year, we could not establish independent differences in the responses of the family physicians according to demographics (likely related to power issues).

DISCUSSION

Rheumatic diseases are commonly encountered by family physicians. A cross-sectional study in Norway⁹ found that 45% of patients with RA sought advice from their family physician. However, family physicians may have limited experience in rheumatology care, as they feel knowledge in this field is not their priority¹⁰. In a study involving presentation of a fictional scenario¹¹, when asked about what sources of information the family physician would use, 19% said journals or textbooks and 8% said a discussion with a colleague, versus 73% who would refer to a rheumatologist. Another study examined the main reasons for rheumatology referral by family physicians. The most common medical reasons cited were uncertainty about the diagnosis, a need for advice about treatment, and a need for a diagnostic or therapeutic procedure. Other reasons included the need to comply with a standard of care, a patient request, the physician’s desire to learn, to obtain patient education, or to reassure or motivate the patient¹².

Most family physicians, in our study, had a specific rheumatologist to whom they tend to refer patients. Regarding factors of importance related to interaction with the rheumatologist, the reasons for selecting a particular specialist, from our study, in rank order according to importance rating, were similar to those found by Forrest, et al⁵ in the United States. In previous studies^5,13, personal knowledge of the physician was considered to be the prime reason for selecting a particular specialist. However, many of our respondents, especially women (who represent a growing number of family practitioners), did not cite this reason as important. In Quebec there is an increasing number of female medical students graduating each year¹⁴. Our study presents an interesting comparison, in that two-thirds of male physicians rated personal knowledge of a rheumatologist as high in importance (score > 4/5), while almost two-thirds of female physicians assigned this factor low importance (score < 2/5). Female physicians also tended to rate the “proximity of the specialist to the family physician’s practice” component very low.

The previous literature does suggest that differences may exist between male and female physicians regarding factors of most importance when referring to a specialist. One study of family physicians in the United States showed a slight trend for male versus female family physicians to place more importance on patient convenience and location of the specialists’ office¹⁵. One potential explanation is that the demographics of female physicians’ practices may be different (e.g., they may treat more socioeconomically disadvantaged and vulnerable patients, or a more elderly clientele, for whom financial and transportation barriers are important). It is well-documented that female physicians may spend more time communicating with patients¹⁶, and possibly this allows them to better understand the financial and transportation considerations that patients have when they are being referred to a specialist.

Kinchen, et al, in the study of family physicians in the United States¹⁵, observed a slight trend for male physicians, when choosing a specialist, to place greater importance (as compared to female physicians) on previous experience with a specialist. That finding may be in part driven by the striking preference that male family physicians reported in ensuring that specialists returned patients to the care of the family physician.

Communication and information exchange regarding patient issues were rated as the most important dimensions of care partnerships by our survey respondents. Though around two-thirds of respondents expressed satisfaction with rheumatologists regarding this aspect, a third did not. A survey assessing communication between generalists and specialists in the Netherlands¹⁷ found similar dissatisfaction. Indeed, half of generalists who responded to that survey felt their questions were not addressed adequately by the specialist, and that the specialist’s report contained insufficient details on treatment and followup.

Sharing of responsibilities was also considered to be very important by our respondents. Our study also showed that the majority of family physicians were satisfied on this front, and that responsibilities were clear and appropriately balanced. However, room for improvement remains.

We acknowledge the limitations of our study. The response rate for our questionnaire was not high enough to safely consider our sample as being representative of all family physicians in Quebec. Although the profile of respondents shared many similarities with the family physician population of Quebec (Table 1), there may have been a bias toward women (although since our survey was administered in 2010, the difference may simply be changing demographics) and toward academic-based practitioners. Our respondents might have been biased toward physicians interested in new models of care management and/or musculoskeletal diseases. Despite these potential biases, we believe our data provide some new and useful insight into the interactions between family physicians and rheumatologists.

Many of our family physician respondents have a rheumatology care partnership, according to our definition. Factors of importance to family physicians in this context include adequate information exchange on patient issues, short waiting time for referrals, and a clear and appropriate balance in responsibilities.