Comparing primary care Interprofessional and non-interprofessional teams on access to care and health services utilization in Ontario, Canada: a retrospective cohort study

Background Many countries, including Canada, have introduced primary care reforms to improve health system functioning and value. The purpose of this study was to examine the association between receiving care from interprofessional primary care teams and after-hours access to care, patient-reported walk-in clinic visits and emergency department use. Methods We conducted a retrospective cohort study linking population-based administrative databases to Ontario’s Health Care Experience Survey (HCES) between 2012 and 2018. We adjusted for physician group characteristics as well as individual physician and patient characteristics while assessing the relationship between receiving care from interprofessional teams and the outcomes of interest. Results As of March 31st, 2015, there were 465 physician groups with HCES respondents of which 177 (38.0%) were interprofessional teams and 288 (62.0%) were non-interprofessional teams in the same blended capitation reimbursement model. In this period, there were 4518 physicians with HCES respondents, of whom 2131 (47.2%) were in interprofessional teams and 2387 (52.8%) were in non-interprofessional teams. There were 10,102 HCES respondents included in this study, of whom 42.4% were in interprofessional teams and 42.3% were in non-interprofessional teams. After adjustment, we found that being in an interprofessional team was associated with an increase in the odds of patients reporting same/next day access to care by 12.0% (OR = 1.12 CI = 1.00 to 1.24 p-value 0.0436) and a decrease in the odds of patients reporting walk-in clinic use by 16% (OR = 0.84 CI = 0.75 to 0.94 p-value 0.0019). After adjustment, there were no significant differences in patient-reported after-hours access to care and emergency department use. Conclusions Ontario has invested heavily in interprofessional primary care teams. As compared to patients in non-interprofessional teams, patients in interprofessional teams self-reported more timely access to care and less walk-in clinic use but no significant difference in self-reported access to after-hours care or in emergency department use. For jurisdictions aiming to expand physician voluntary participation in interprofessional teams, our study results inform expectations around access to care and health services utilization.

Conclusions: Ontario has invested heavily in interprofessional primary care teams. As compared to patients in noninterprofessional teams, patients in interprofessional teams self-reported more timely access to care and less walk-in clinic use but no significant difference in self-reported access to after-hours care or in emergency department use. For jurisdictions aiming to expand physician voluntary participation in interprofessional teams, our study results inform expectations around access to care and health services utilization.
Keywords: Primary care, Interprofessional teams, Patient experience, Access, Health services utilization

Background
Moving towards value-based healthcare is a priority for healthcare systems internationally [1]. The pursuit of providing value-based health care revolves around three aims: improving the experience of care, improving the health of populations, and reducing per capita costs [2]. A strong primary care system is recognized as the cornerstone of health systems and is associated with better outcomes, improved patient experience and lower cost [3]. Many countries around the world, including Canada, have introduced primary care reform to deliver on those goals.
During the economic recession in the 1990s, there has been limited investments in primary care innovation in Canada [4]. A decade later, primary care reform initiatives started to emerge in Canada in response to various recommendations from federal and provincial committees [5,6]. In line with the Canadian healthcare reform movement, Ontario has undergone three major primary care policy initiatives: new physician payment and governance models, enrolment of patients with a primary care physician and support for the development of interprofessional teams [7]. Interprofessional teams are "groups of professionals from different disciplines who communicate and work together in a formal arrangement to care for a patient population in a primary care setting." [8] They typically include primary care physicians, nurses or nurse practitioners, and at least one other health care professional (e.g. pharmacist, social worker, dietitian or physiotherapist). Interprofessional teams are also eligible for funding an administrator or executive director.
During the last 20 years, more than 30% of primary care physicians have willingly moved from fee-forservice payment model to a blended capitation. Some of those physicians have received extra funding to set up and deliver interprofessional team-based care. Currently, the dominant blended capitation model in Ontario is called Family Health Organization (FHO). FHOs have formal patient enrollment, electronic medical records, physician-led governance and a minimum of three physicians practicing together. They offer comprehensive care, including preventive health care services, chronic disease management and health promotion, through a combination of regular physician office hours and after-hours services. FHOs were eligible to apply for additional funding for allied health professionals to join their practice and become interprofessional primary care teams called Family Health Teams.
The government's priorities in establishing interprofessional teams were to increase access to primary care and appropriate healthcare services utilization [9]. Physicians in FHO models in Ontario are required to provide afterhours access to care and receive a bonus when their patients do not seek services from physicians outside of their group, such as in walk-in clinics. The bonus is not affected if their patients visit the emergency department. Interprofessional team-based care is thought to free up some of the physicians' time by delegating tasks to other health care professionals within their scope of practice [10]. Access to quality primary care can reduce the need for unnecessary and more expensive services [11]. Treating less-urgent conditions in primary care could improve continuity of care and patient experience [12,13].
Several studies conducted in Ontario have compared capitation-based interprofessional teams to other funding and delivery models of care on specific measures of quality [14][15][16][17][18][19][20]. However, little research to date has evaluated the association between the interprofessional aspect of primary care teams and access to care and health services utilization. Our study examined the association between receiving care from interprofessional versus non-interprofessional primary care teams and patient-reported timely and after-hours access to care, patient-reported walk-in clinic visits and emergency department use. We hypothesised that interprofessional teams would be better performers on these measures given their enhanced capacity and structure. Evidence from our setting that underwent large-scale primary care reform will be relevant to other jurisdictions contemplating innovations in primary care delivery and, specifically, the adoption of interprofessional team-based primary care.

Setting
Ontario is a province in Canada and had a population on 14.7 million people in 2020 [21]. Permanent residents of Ontario are fully insured for physician primary care services through the Ontario Health Insurance Plan (OHIP) with no co-payment or deductible. Primary care organization and payment models have evolved over the course of the last 18 years. Three dominant practice models exist in Ontario-enhanced fee-for-service, nonteam blended capitation and team-based blended capitation. These models are described in detail elsewhere [7,22,23].
The focus of this study was on the dominant blended capitation model-FHO-within which physicians practice in either interprofessional or non-interprofessional teams. When patients seek primary care services outside the practice in which they are enrolled, for example in walk-in clinics, the FHO loses a bonus payment equal to the fee-for-service payments to the physician who treated the patient, to a maximum bonus of 18.59% of the practice's total capitation [24]. There is no deduction if an enrolled patient visits an emergency department for non-emergency care. FHOs are required to provide at least one three-hour block of after-hours services per week for each physician in the group, to a maximum of five three-hour blocks per week for practices with five or more physicians. Contracts define "after-hours" as Monday to Thursday after 5 p.m. or any time on the weekend-that is, any time from Friday after 5 pm through Sunday [25].

Design and population
We conducted a retrospective cohort study where we linked several population-based administrative databases to the Health Care Experience Survey (HCES) using encoded identifiers at ICES (formerly known as the Institute for Clinical Evaluative Sciences) to form data extractions and identify the population of interest (Fig. 1). The HCES is collects information to understand Ontarians' experience in obtaining primary care services and helps the Ministry of Health in planning health care programs and policies. The HCES survey is conducted continuously by the Institute for Social Research (ISR) at York University, with data being provided to the Ontario Ministry of Health every 3 months (termed a 'wave').
The study population comprised respondents to the HCES over six fiscal years (April 1 -March 31) from 2012/13 to 2017/18. The HCES targets persons 16 years and older who live in private dwellings in Ontario. People living in institutions, in households without telephones are excluded. The study included respondents from 20 quarterly waves of the HCES that were conducted between October 2012 and October 2017. The average response rate was 51% during that period. Once households were sampled in the HCES, they were removed from the sampling frame for 2 years. Respondents who responded to the survey more than once throughout the study period were excluded.
For each of the data extractions, we identified respondent to the HCES at the end of the fiscal year. To be included in the study, respondents had to be consistently in an FHO blended capitation model throughout the observation period for the fiscal year they responded to the HCES. We captured patients' characteristics at the beginning of the fiscal year they responded to the HCES. Self-reported timely access to care, after-hours access to care and walk-in clinic visits were captured during the fiscal year the patient responded to the HCES and ED visits were captured at the end of that fiscal year from health administrative data. Physician group and physicians' characteristics were captured at the midpoint of the study timeframe, March 31st, 2015 (Fig. 2).

Measures and data sources Exposure
Enrolment in a FHO blended capitation model, with an interprofessional team was the exposure. The exposure variable was retrieved from a population and demographics database-the Client Agency Program Enrolment tables that identify the patient enrolment model and the physician with whom patients are enrolled. A separate file provided by the Ontario Ministry of Health (MOH) to ICES identified physicians who are part of an interprofessional team versus a non-interprofessional team.

Outcomes
The outcomes included patient-reported timely access to care, patient-reported after-hours access to care, patient-reported walk-in clinic use and emergency department use. Patient-reported timely access to care, after-hours access to care and walk-in clinic use were derived from the HCES (How many days did it take from when you first tried to see your provider to when you actually saw them or someone else in their office? (sick_3); The last time when you needed medical care in the evening, on a weekend, or on a public holiday, how easy or difficult was it to get care without going to the emergency department? (access_5); Have you been to a walk-in clinic because you were sick or for a healthrelated problem in the 12 months? (wi_1)). The HCES is a quarterly survey of a random sample of the Ontario population, 16 years and older, conducted on behalf of the MOH by the Institute for Social Research at York University. The survey focuses on Ontarians' primary care experience, including access to care, to generate regional and province-level data. The National Ambulatory Care Reporting System (NACRS) was used to derive emergency department visits.

Physicians, physicians groups and patient characteristics
All characteristics were derived from administrative databases available at ICES. Physicians' characteristics included age, sex, years since graduation, Canadian graduate status and number of years in practice. Physician group characteristics included the number of physicians per group and number of years under the capitation model. Patient characteristics included age, sex and OHIP registration (as proxy for immigration), neighborhood income quintiles, rurality, and Resource Utilization Bands [26,27].

Analysis
For the descriptive results, we generated counts and percentages for categorical variables and means and standard deviations for continuous variables to describe the characteristics of physician groups and physicians who were either in interprofessional or non-interprofessional teams in relation to the outcomes of interest. For the patient variables, we generated sample weighted descriptive statistics. The probability weights assigned to respondents in the HCES were dependent on the probability of being selected, which was determined from the sampling design.
For the outcomes, we ran sample weighted survey logistic regressions to model each of the outcomes while adjusting for the respective physician group, physician and patient characteristics. All study analyses were conducted using SAS v.9.3 and statistical significance was assessed at a p-value < 0.05.

Results
Baseline group, physician and patient characteristics comparing HCES respondents in interprofessional teams versus non-interprofessional teams As of March 31st, 2015, there were 465 FHO physician groups with HCES respondents of which 177 (38%) were interprofessional teams and 288 (62%) were noninterprofessional teams. Interprofessional teams with HCES respondents had more physicians per group as compared to non-interprofessional teams (means = 13.1 versus 8.84, respectively) and more years under the capitation model (means = 6.0 versus 4.3 respectively).

Univariate analysis
Patient-reported timely access to care and after-hours access to care comparing HCES respondents in interprofessional teams versus non-interprofessional teams HCES respondents in interprofessional teams were slightly more likely to report timely access to care (same/next day) when compared to patients in non-interprofessional teams (39.9% versus 39.1%). HCES respondents in interprofessional teams were less likely to report easy or somewhat easy access to after-hours care compared to patients in non-interprofessional teams (30.8% versus 35.2%). The results stratified by physicians charateristics are presented in Tables 3 and 4. Tables 5 and 6 present the results stratified by patient characteristics.

Patient-reported walk-in clinic visits and emergency department use comparing HCES respondents in interprofessional teams versus non-interprofessional teams
HCES respondents in interprofessional teams reported a lower percent of walk-in clinic visits compared to patients in non-interprofessional teams (19.7% versus 28.2%, respectively) ( Table 7). A higher percent of HCES respondents in interprofessional teams had emergency department visits as compared to patients in noninterprofessional teams (26.7% versus 23.5%, respectively) ( Table 8). The results stratified by physician charateristics are presented in Tables 9 and 10.

Multivariate analysis Association between enrollment in an interprofessional team and the outcomes
When we examined timely access to care while adjusting for physician group, physician and patient characteristics, we found that being in an interprofessional team was associated with an increased odd of patient-reported timely (same/next day) access to care of 12% (OR = 1.12 CI = 1.00 to 1.24 p-value 0.0436) and decreased odds of self-reporting walk-in clinic use of 16% (OR = 0.84 CI = 0.75 to 0.94 p-value 0.0019). We did not find significant differences after adjustment between interprofessional and non-interprofessional teams in patient-reported after-hours access to care or in emergency department use (Table 11).
When we stratified the analyses by sex and by rurality, we did not find a consistent pattern across the outcomes when comparing interprofessional teams with noninterprofessional teams (results not included but can be made available upon request).

Discussion
We linked the HCES to administrative databases to examine the association between receiving care from interprofessional primary care teams and patient-reported timely access and after-hours access to care, patientreported use of walk-in clinics and emergency department use. We found that HCES respondents receiving care from interprofessional teams self-reported more timely access to care and less walk-in clinic use. We did not find a significant difference in patient-reported afterhours access to care or in emergency department visits.
The professional management and clinical structure available through interprofessional teams, such as having  an Executive Director and allied health professionals can theoretically support access to care. Although more timely access to care among patients in interprofessional teams is not an expectation in the contractual agreement between teams and the Ministry of Health, previous evidence indicates that enhanced interprofessional team structure can support the availability of the primary care provider by shifting some of their duties to other team members [28][29][30][31][32][33]. The evaluations of Patient-Centered Medical Homes in the United States related to timely access to care suggest that greater availability of providers can free more of their time for patient encounters [34]. Our findings of generally low timely access to care are comparable to other reports that found only 43% of Canadians report that they were able to have same-or next-day appointment at their regular place of care and identified that Canada continues to perform below the average on timely access to care when compared to other counties included in the Commonwealth Fund International Health Surveys [24].
Our findings showed a non-significant difference in patient-reported after-hours access to care between interprofessional and non-interprofessional teams. The provision of after-hours care is an expectation that all FHOs need to meet as part of their contractual agreement with the Ministry of Health [32]. Although some interprofessional teams operate out of multiple locations, the after-hours services only need to be offered at one location, which may not be convenient for many of the enrolled patients. Also, only one physician is required to be available during each after-hours block which might not be sufficient evening and weekend availability to meet patients' needs. Previous evidence that compared a slightly different after-hours access to care measure (asking if respondents providers have an after-hours clinic as opposed how easy or difficult was it to get care without going to the emergency department) found that respondents in interprofessional teams self-reported more after-hours access to care [18].
Although both interprofessional and noninterprofessional teams get penalised equally if their patients visit a walk-in clinic, our finding of significantly lower patient-reported walk-in clinic visits by HCES respondent among interprofessional teams may be explained by the higher patient-reported timely access to care in interprofessional teams, which can contribute to the lower walk-in clinic use. Patients may be less likely to seek care elsewhere if their provider is accessible to them in a timely manner. Additionally, the enhanced administrative structure of interprofessional teams can support reinforcing to patients the need to refrain from walk-in visits as part of being on the group roster. Our findings of a non-significant difference in emergency department use between interprofessional and noninterprofessional teams is consistent with evidence from Canada that looked at utilization in relation to interprofessional team-based care and found differences in quality but not in healthcare utilization [19,20,35,36].
Some of our findings are not fully consistent with an Ontario provincial analysis where throughout the investigated years (2014 to 2017) timely access to care ranged between 44.3 and 39.9% (compared to 39.5% in our study population), easy or somewhat easy after-hours access to care ranged between 48.0 and 46.0% (vs. 33% in our sample) and walk-in clinic use ranged between 29.6 and 30.5% (vs. 24% in our study) [37]. Those differences can be explained by the slightly different timeframe, inclusion of respondents from all primary care models and slightly larger sample that includes people who declined to have their data linked (6%) for the provincial analysis. Additionally, for the timely access to care question, the provincial analysis included respondents with and without a family doctor whereas our study includes only respondents with a family doctor. Through a personal communication with the Ministry of Health representative who is responsible for the survey, we have confirmed that our study results can be mainly explained by those differences.
Interprofessional teams in Ontario had access to several quality improvement initiatives that hypothetically can contribute to improved outcomes over noninterprofessional teams. The Association of Family   [38]. Timely access to care and emergency department use were among the measurement areas monitored through this initiative [39]. The Quality Improvement and Innovation Partnership (QIIP) was another province wide qualityimprovement program implemented between 2008 and 2010 to support interprofessional teams to improve the care they provide [40]. The learning collaboratives used the Institute for Healthcare Improvement's Breakthrough Series learning model and interprofessional teams were provided with a quality improvement coach who supported and mentored participants throughout the program [41]. Improved access to care was one of  the supported quality improvement areas through QIIP [9]. Those investments should theoretically be reflected in better outcomes among interprofessional teams. The government's first priority in establishing interprofessional teams was to increase access to primary care and health services utilization [32]. Our results show that interprofessional teams perform better than noninterprofessional teams in some but not all aspects related to access to care and health services utilization. Our study has limitations. First, this is an observational study that cannot address causation. It is also cross-sectional so it is not possible to distinguish whether the outcomes examined were pre-existing or were the result of joining or not joining an interprofessional team. Self-reported timely and after-hours access to care are subject to limitations as measures of performance, respondent recall bias being one of them. People living in institutions, people with non-residential phone numbers, and people with invalid/missing household addresses in the Registered Persons Database (RPDB) are not captured in the HCES. Respondents who were unable to speak English or French or were not healthy enough (physically or mentally) to complete the interview were not surveyed. Second, there are other unmeasured factors that might contribute to the decision of having a walk-in clinic visit or using the emergency department that this study cannot capture. These could include personal preference or judgment during the time the service was needed. Third, access to care can be measured in many different ways. The access questions we investigated in this study provide a specific perspective restricted to timely and after-hours access to care. Previous evidence suggests that different measures of timely access are needed to understand health care system performance. 50 Fourth, joining interprofessional team-based care was voluntary and our findings could be influenced by some unmeasured factors for physicians who chose to join this model of primary care delivery. Fifth, team composition in terms of allied healthcare professional was not available through administrative databased. Nonetheless, we aimed to capture all measured factors that can be traced through administrative databases. Finally, administrative databases have not been originally collected for research purposes, which presents a limitation in generating and interpreting the information. However, all the databases used for deriving the emergency department measure used in this study have been validated in the Ontario context.

Conclusion
Ontario has made a major investment in interprofessional team-based care. As compared to patients in non-interprofessional teams, patients in interprofessional teams self-reported more timely access to care and less walk-in clinic use but there was no significant difference in self-reported access to after hours to care and in emergency department use. Our findings can inform other jurisdictions aiming to expand voluntary participation in interprofessional team-based primary care regarding expectations about the relationship between primary care policy, organization and delivery and patient experience and health services utilization. Careful consideration should be given to contractual and policy levers that can incentivise interprofessional team-based care in delivering on intended outcomes such as improving health services utilization.