Family physician views about primary care reform in Ontario: a postal questionnaire

Since the inception of Canadian medicare in the late 1960s there have been recurrent cycles, albeit largely fruitless, of demands for innovation to primary care delivery models [1]. Increasingly, however, the need to review and revise the delivery of primary care has assumed new urgency and is currently seen as a core health policy issue in the developed world and beyond [2]. Primary care reform has been advocated by provincial and national commissions in Canada as fundamental to health system restructuring [3‐6]. In Ontario, two main models of reformed primary care (reformed fee-for-service and global capitation) have been initiated as pilot projects and evaluated[7]. In 2001 the Ontario government launched the Ontario Family Health Care Network, intended to enlist 80% of practitioners by the year 2004 into a redesigned model of delivering primary care. The key elements in this model include: patient rostering, capitation payment with added incentives for prevention and other targeted services, provision of out-of-hours service and tele-triage, and extensive use of electronic medical records and linkages[8].

The move towards changes in the delivery of primary care appears to be a top-down process driven by provincial Ministries of Health. What do 'grass-roots' practitioners feel about change? A study in 2001 found that only five percent of Ontario family physicians, in practice eight to ten years, believed primary care reform would have a favourable effect on their practices[9]. A subsequent study of physicians who were eligible for, but did not participate in, a primary care reform pilot project identified many concerns about the impact of a new system on: practice routines, working conditions, financial arrangements and loss of autonomy[10].

Currently in Ontario the long-standing Community Health Centre and Health Services Organization programs, and recent primary reform sites, account for only about five percent of family physician practices[1]. This small number, along with evidence that practitioners are sceptical about current reforms, suggests that the recruitment target may be unrealistic. The aims of this study are: 1) to determine whether family physicians intended to participate in the Ontario Family Health Care Network; and 2) to identify factors that may influence their decision.

A cross-sectional survey of family physicians in Ontario was carried out in July 2002. A two page questionnaire, along with a covering letter and a pre-paid return envelope, was sent to 1200 family physicians, randomly selected from a list of 5200 members of the Ontario College of Family Physicians. The college represents 65% of all practicing family physicians in the province of Ontario. Non-responders were sent a second questionnaire one month later. The questionnaire sought information about physicians (age, sex, years in practice), practice characteristics (type, size, remuneration, percent of time spent in clinical practice, use of computer, out-of-hours service) and their level of agreement with several statements related to primary care reform in Ontario (Table 1). The level of agreement ranged from: strongly agree, agree, slightly agree, slightly disagree, disagree, to strongly disagree. The wording of statements is presented in Table 1. To test for sampling bias, responder characteristics (age, sex, years in practice) were compared with the Ontario responses to the 2001 National Family Physician Workforce Survey [11].

Data analysis consisted of frequency distributions and the calculation of odds ratios. The variables were re-coded as follows: age-group [less than 44 vs. 45+ years], years in practice [less than 9 years vs. 10 years or greater], type of practice [solo vs. group, community health centre, health services organization, other], remuneration [fee-for-service vs. capitation, salary, other], size of practice [less than 1500 vs. 1500 or more patients], percent time in clinical practice and [less than 75% vs. greater than 75%]. Level of agreement was re-coded as either agreement [strongly agree, agree, slightly agree] or disagreement [slightly disagree, disagree, strongly disagree].

All analyses were conducted using procedures written in SAS[12]. Ethics approval was received from the Queen's University Health Sciences and Affiliated Teaching Hospitals Research Ethics Board.

Of the 1200 family physicians sent a questionnaire, 50 were not in practice so the final study population was 1150. 565 questionnaires were not returned and a further 8 were returned without a response so that the final response rate was 50.2% (577/1150). The mean age of responders was 44 years, with the majority of responders (39%) aged between 35–44 years (Table 2). There were slightly more men (55%) than women (45%). Responders had practiced for an average of 15 years, with 38% of family physicians reporting that they had practiced for less than 9 years. Most practices differed from the type proposed by the Network (72%). Table 3 shows a comparison of sample characteristics with those of the Ontario portion of the National Family Physician Workforce Survey. Responders to the present survey were more likely to be younger, to be women and to have practiced for less time than the Ontario responders to the national survey.

The frequency distribution of practice characteristics is presented in Table 2. The majority of family physicians (56%) worked in group practices, while the remainder worked in solo practices (30%), 'other' practices (7%), community health centres (5%), or health services organizations (3%). A substantial majority (81%) reported that they were remunerated on a fee-for-service basis – of the remainder, 10% were salaried, 5% were paid on a capitation basis and 4% were 'other'. Practice sizes were: less than 500 patients (4%); 500–1000 patients (11%), 1001–1500 patients (21%), 1501–2000 patients (26%), 2001–2500 patients (16%) and greater than 2500 patients (22%). The majority of family physicians (75%) spent between 76–100% of their time in clinical practice. All family physicians reported using a computer in their practice; 60% for billing, scheduling and registration; 30% for billing and only 10% reported that they kept full electronic records.

The level of agreement with statements about primary care reform is presented in Table 4. A substantial majority (72%) of physicians did not expect to join the Network by 2004 or by some later date after 2004 (60%). If the intermediate category, 'slightly disagree' was included, these estimates rose to 84% and 76% respectively. The proportion of persons who reported that they never expected to join the network was 82%. Thirty six percent of responders either agreed or strongly agreed with the statement that they did not understand the Network well enough to make an informed decision about it. Forty eight percent of responders agreed or strongly agreed that the current system for the organization of primary care delivery required change and many responders (60%) agreed or strongly agreed with the statement that the financing of primary care requires change. Nevertheless, few responders agreed with the statement that capitation (10%) or rostering (23%) would improve primary care in Ontario. Many responding physicians (68%) believed that appropriate incentives would enhance prevention. Thirty percent of responders agreed or strongly agreed that patients should always have access to extended office hours, while 42% agreed or strongly agreed that patients should have access to a telephone health line. Half the responders (51.9%) agreed or strongly agreed that they would like to see computer systems replace papers systems in their practice.

Odds ratios and their associated 95% confidence intervals are presented in Table 5. Younger physicians were less likely (OR = 0.62) to agree that rostering would improve primary care and that patients should have access to extended hours (OR = 0.49). Physicians in practice 9 years or less, were 1.6 times more likely to expect to join the Network by 2004, or after 2004 (OR = 1.67). Solo practice physicians were less likely (OR = 0.65) to agree with the idea of extended hours and of computers replacing paper systems (OR = 0.41). Physicians in small practices were less likely (0.64) to understand the reforms and to favour telephone health lines (OR = 0.65). They were 1.49 times more likely to report that they expected to join the Network after 2004. Those physicians remunerated on a fee-for-service basis were less likely to agree with statements about joining by 2004 (OR = 0.25), or after 2004 (OR = 0.47), that the organization required changing (OR = 0.29), favour capitation (OR = 0.27), rostering (OR = 0.34), financial incentives for prevention (OR = 0.41), or extended access for patients (OR = 0.58). Those who derived less than 75% of their income from clinical practice, were less likely to understand the reforms (0.65), much more likely to agree that the financing of primary care required changing (OR = 2.38), agree with both capitation (OR = 1.83) and rostering (OR = 1.77), and favour extended access for patients (OR = 1.58).

Three possible explanations for a physicians intention to join the Network were assessed. Neither a lack of understanding of the reform initiative (OR = 1.5, 95% C.I.s 0.96–2.3), nor the possibility that physicians were already practising in a manner similar to that proposed by the Network, were statistically significant explanations for the stated intention to join the Network (OR = 1.3, 95% C.I.s 0.6–2.9). The main factor associated with a physicians intent to join by 2004 was related to method of payment – physicians who favour capitation were 5.5 times more likely to report that they expected to join the Network by 2004 (O.R. = 5.5, 95% C.I.s 3.5–8.7).

This study has described the views of physicians about primary care reform in Ontario. While many physicians recognize the need for change in both the organization and financing of primary care, the majority of physicians do not expect to join the Network, and half of them do not understand the Network enough to make an informed decision about participating. Physicians are divided on issues such as patient rostering and extended hours, although many support the idea of financial incentives for preventive interventions or a telephone health line. Whether or not a physician was remunerated on a fee-for-service basis strongly influenced their views on reform. Physicians on fee-for-service were less likely to: report that they would join the Network; believe that the organization of primary care required changing; support rostering or capitation; favour financial incentives for prevention; or support extended access for patients. Physicians who spent less than 75% of their work in clinical practice were more likely to agree that the financing of primary care required changing, and to favour both capitation, rostering and extended access for patients. Physicians who favour capitation were 5.5 times more likely to report that they intended to join the Network by 2004. Although many responders did not understand the Network, the only statistically significant explanation was that physicians did not support capitation.

The limitations of this study ought to be considered before any conclusions be drawn. These results may be limited by the representativeness of the sample and by the reliability and validity of the questions used to determine views about reform. The representativeness of this sample is influenced by sampling bias and response bias. Our survey was based on a 20% random sample of the Ontario College of Family Physicians, that represents 65% of practicing family physicians in Ontario. This sampling frame excluded general practitioners who are not certified in Family Medicine, but who are eligible to participate in the reforms. These physicians are likely to be older than certified family physicians and may hold different views from those sampled in this study. The extent that this exclusion may bias these results is unknown, although that it may be minimal is suggested by the finding that non-certified general practitioners were equally likely to join a pilot primary care reform site as those who were members of the Ontario College of Family Physicians [10]. Responders in our survey were more likely to be younger, to be women and to have practised for less time, than physicians who responded (53%) to the Ontario portion of the National Family Physician Workforce Survey. While, we acknowledge the possibility of sampling bias, we feel that the views of younger physicians may be more relevant because they are more likely to be the ones who are targets of the reform initiative. The low response rate (50%) was expected as it is well known that busy clinicians frequently do not complete questionnaires. Nevertheless, our response rate is comparable to that of the National Family Physician Workforce survey (53%) and a recently published survey of Ontario family physicians (47%) [11, 13].

Second, the way that the questions were phrased could bias the results. The questions used to determine the views of physicians were developed by the authors for this study and were not tested for their reliability and validity. Nevertheless, the questions were developed by the authors, whose expertise ranged from questionnaire design (DH), to primary care research (MG) and family medicine (SS, MG) and reflect the stated objectives of the Ontario initiative. We believe they had face validity. Interpretation of these results was based on the extent that responders agreed or strongly agreed with the statements. Depending on whether the "slightly agreed" category was included, interpretation of the results may differ – for example the majority agree with all but three statements (those about whether physicians expected to join the network and about capitation). A third limitation is that the survey was conducted a few months after the initiative was launched. The diffusion of innovations tends to have an initial latent phase before rapidly accelerating adoption. While it is possible that our survey only reflects this latency period, the fact that more than a year after the Network launch, only 5% of eligible physicians have joined, suggests that we have captured long-term intentions.

What are the implications of these findings for reform initiatives in Ontario? The majority of family physicians reported that they were unlikely to participate in the Network because they were opposed to capitation and patient rosters. In a study of the pilot phase of Ontario primary care reform physicians in Ontario's capitated Health Service Organizations were no more likely to join than fee-for-service physicians[10]. In contrast, we found that a physicians support for capitation was associated with their intent to join the Network. Capitation payment is a key element in primary care restructuring not only in Canada but also in many foreign jurisdictions[14]. It was the funding method officially endorsed by the Health Services Restructuring Commission in 1999 and is the dominant element in the Network remuneration scheme[15]. Nevertheless, in Ontario almost 95% of physicians are currently paid by fee-for-service[1]. For them, capitation may be associated with the unpopular principle in Ontario's Health Service Organizations where payment for patients attending a source of primary care outside the practice is withheld for that month. It has also been suggested that capitation may lead to loss of autonomy [16]. Finally, it may be that Ontario family physicians disapprove of capitated practice because they just don't like change, a view supported by their reaction to many elements in contemporary health reform [9]. Opposition to capitation is not confined to Ontario practitioners. In the USA, physicians who were introduced to partial capitation funding had strong negative views of the method and, while these views tended to moderate with time, capitation continued to be rated far below fee-for-service [17, 18].

Currently the vast majority of family physicians operate on a fee-for-service basis, with a significant number in solo practice. The decision not to be part of the network does not mean the physician will not be providing service. This may inhibit the incorporation of innovations, such as nurse practitioners or the electronic medical record, into existing practices with attendant implications for health human resources and quality of care respectively. Moreover, existing models of practice lack the capacity to address distributional issues. If all citizens were rostered to specific practices, there would be competition for patients. Some physicians now located in the urban south might be forced under such a capitated payment scheme to migrate to less serviced areas in order to acquire an adequate list of patients.

Many physicians reported that they did not fully understand the Network and may be unaware that it allows for a blended payment scheme including both capitation and fee-for-service. Since physicians may equate primary care reform largely with a switch to capitated payment, the successful recruitment to the Network will require the education of physicians about payment. Borrowing from strategy used to introduce Health Services Organizations, government suggested a process to compare the fee-for-service billings of potential recruits with their possible earnings under Network rules [1, 16]. Additionally, they need to emphasize that a selection of retained fee-for-service billing codes and the system of target achievement bonuses brings the Network's payment formula closer to a blended scheme favoured by many physicians [1, 19]. The key message, however, is that experience from many countries confirms that primary care reform does not succeed without the active support of the physicians involved [14]. The alternative is to see the Network fail to attract widespread participation, as did its Health Services Organization and Community Health Centre predecessors [16, 20]. Since the launch of the Ontario Family Health Network, government has introduced a variant of fee-for-services with a number of incentives, known as family health groups that are proving more popular than the family health networks.

The results of this survey, conducted five months after the initial offering of primary care reform agreements to all Ontario physicians, suggest that a 80% enrolment target is unrealistic.