Do Incentive Payments Reward The Wrong Providers? A Study Of Primary Care Reform In Ontario, Canada

Physician payment reform is under way in high-income countries globally, most commonly focusing on primary care and taking the form of capitation, while sometimes also incorporating pay-for-performance, shared savings, and bundled payment.¹ In most cases these reforms are supplanting fee-for-service in part or in whole. Fee-for-service is increasingly seen as incentivizing unnecessary care, which results in care that is volume driven rather than valuedriven and being wasteful.^2,3 A desire for value-based care that rewards higher quality and lower costs rather than volume is driving payment reforms—most prominently, the Medicare Access and CHIP Reauthorization Act of 2015—across the US in many markets and among numerous payers.⁴ Other countries such as the UK, Australia, the Netherlands, Italy, and Canada are also engaged in major payment reforms.^5–9 Central to most of these reforms is some form of risk-adjusted prospective capitation-based payment.

While its drawbacks are clear, fee-for-service payment is well understood, transparent in costs in some settings, modifiable to a degree through fee schedules, and largely predictable in terms of consequences. Capitation-based payment shifts risk to providers as they receive fixed amounts per person, and therefore providers’ behavior may be less predictable and could vary depending on the setting and how capitation is structured. Known risks include the selection of healthy patients, who are expected to use fewer services, and underservicing through lack of follow-up or availability¹⁰—both of which contributed to a backlash against capitation-based managed care in the US in the 1990s.¹¹ Many current payment reforms include case-mix adjustment, episode-related care, bundled payment, shared savings, and other features designed to mitigate these problems.

Capitation reimbursement in primary care usually features attribution of patients to physicians via formal enrollment, a defined basket of services covered by the capitated payment, and limitations on payment for in-basket services when provided outside of the patient’s enrollment group. However, in single-payer systems such as in Canada, where patients are free to seek care with any provider, payers are at risk of paying twice for capitated patients—paying a monthly fee to the enrolling physician but also paying fee-for-service outside the capitation arrangement. To avoid paying twice and to incentivize timely access to care, some payers have sought to financially reduce capitation payments by the amounts billed by other physicians in fee-for-service, a practice sometimes called negation.^12,13 The government of Ontario, Canada, introduced widespread primary care reforms in the 2000s, including practice models in which physicians were paid largely by capitation. Instead of negation, which had previously been unpopular with Ontario physicians, the government and medical association negotiated a financial incentive called the “access bonus.” Practices received a bonus of up to 18.59 percent of capitation payments, which was reduced dollar for dollar if their patients sought care outside their group for in-basket services. Emergency department (ED) visits and specialist visits were not counted as outside care.

The ostensible policy objectives for implementing the access bonus were to incentivize access to care by the enrolling physician group and to control the costs of outside use of health services by patients enrolled in the capitated payment plan. Although this payment incentivizes providing care to patients within the physician group, it is also affected by patients’ behavior, and we hypothesized that it could vary according to the services available in each community. For example, patients in rural areas typically rely on EDs as an access point for after-hours care and have few of the alternatives—such as walk-in clinics—available to patients in large urban centers.^14,15 The characteristics of physicians who received access bonus payments in Ontario are not known, nor is whether the payments were associated with the policy objectives of better access and lower costs. The impacts of negation or retention bonuses in other settings are also poorly understood. The purpose of this study was to examine the distribution of access bonus payments according to practice characteristics and to understand the relationship of these payments to primary care visits, after-hours care, ED use, and related costs.

Ontario Setting

Ontario is Canada’s most populous province, with a 2018 population of 14,411,424 people—accounting for almost 40 percent of the national population.¹⁶ Ontario has universal health coverage of medically necessary physician and hospital services for permanent residents without copayments or deductibles, and it does not limit patients’ choice of physician providers. Physician visits in family practice, walk-in clinics, and ED settings are fully insured. The single payer is the Ontario Ministry of Health and Long-Term Care. Health care is financed through income tax revenues, including federal transfers and an annual Ontario health premium, which ranges from $60 to $900 per person depending on income and is limited to people earning more than $20,000 per year.¹⁷

Ontario Primary Care Reform

Ontario’s Ministry of Health and Long-Term Care introduced a number of voluntary alternative physician payment models starting in 2001. By 2012 about three-quarters of primary care physicians were included in these alternative models, and a similar proportion of Ontario’s population was enrolled in these models for health care.¹⁸ These models involve formal patient enrollment; a requirement that primary care physicians provide extended office hours on weekday evenings and on weekends; and blended payments to physicians that include capitation, incentives and bonuses, and fee-for-service payments. The access bonus was exclusive to two models: the Family Health Network and the Family Health Organization. Introduced in 2006, the Family Health Organization is the most popular alternative payment model, covering close to 40 percent of the Ontario population in 2012.¹⁸ In Family Health Organizations, about 70 percent of payments to providers are from capitation adjusted for age and sex, 10 percent from incentives and bonuses, and the remaining 20 percent from fee-for-service limited to codes outside a defined basket of services and shadow billing for codes within the basket, valued at 15 percent of the total fee.¹⁸ Shadow billings are partial fee-for-service payments for services included in the basket of services covered by capitation payments. Their purpose is to provide an incentive for patient visits and to track patient encounters. In 2015 only a few hundred physicians participated in Family Health Networks, and slightly more than half of Ontario primary care physicians practiced in an enrollment model based on fee-for-service or continued to practice in straight fee-for-service outside of a formal model.¹⁹

Study Data And Methods

Data Sources

The administrative data used in these analyses included data from Ontario’s health care registry, physician claims, alternate payments, patient and physician enrollment records, physician characteristics, and ED visits (see online appendix 1).²⁰ These data sets were linked using unique, encoded identifiers and analyzed at ICES, formerly known as the Institute for Clinical Evaluative Sciences.

Analyses

This study was a cross-sectional population-based analysis of Ontario’s Family Health Organization model in 2012–13, which comprised 4,052 primary care physicians and 5.9 million patients. We attributed access bonus payments made to the group to physicians within the group equally. We divided physicians into quintiles of equal size according to the percentage of the maximum potential access bonus payment they received, with quintile 1 representing the lowest and quintile 5 the highest share. We examined the characteristics of patients and physicians in each quintile. We then assessed the association between the quintiles and primary care visits, after-hours care, ED use, and related costs.

Study Variables

Household income was measured using area-level census data linked to the patient’s postal code of residence and expressed as quintiles. First-time registration for health care coverage within the past five years was used as a proxy for recent immigration, as close to 80 percent of new registrants in Ontario are international migrants.²¹ Case-mix was measured using the Johns Hopkins Adjusted Clinical Groups System,²² with at least five Aggregated Diagnosis Groups representing high comorbidity and Resource Utilization Bands 4 and 5 representing the highest quintiles of expected resource use. Rurality was measured using the Rurality Index of Ontario, with scores of 0–9 representing larger cities, 10–39 smaller cities, and 40 or more rural areas.²³ Analyses of health care use and costs were stratified by rurality to account for different patterns of care in rural areas, such as higher ED visit rates.¹⁴

Continuity of care was measured as the percentage of primary care visits over two years made to the enrolling physician and the enrolling group.²⁴ Visits after 5:00 p.m. on weekdays were assessed through a 30 percent after-hours premium (billing code Q012) and those on weekends and holidays through billing as an ED equivalent (billing code A888), which was paid in full as an out-of-basket service. The urgency of ED visits was assessed using the Canadian Triage and Acuity Scale (CTAS), with categories of 1–3 being considered more urgent and 4–5 considered less urgent.²⁵ Ambulatory visits were defined as the combined number of primary care visits and ED visits. Ambulatory costs included primary care fee-for-service, capitation, and shadow billing payments; ED hospital costs; and shadow billing costs for emergency physicians.²⁶ Ambulatory costs did not include bonus or incentive payments, the access bonus, or costs of specialist visits. Physician groups could apply to be exempt from the after-hours requirements if, for example, more than half of group physicians provided regular ED coverage, anesthesia, or obstetrical deliveries. We excluded groups with an after-hours exemption from analyses of health care use and costs.

Data Analysis

Data were presented as means with standard deviations, medians with interquartile ranges, and proportions expressed as percentages. Significance was set at $p<0.05$ and assessed using chi-square tests for proportions, analysis-of-variance tests for means, and Mann-Whitney U tests for medians. Adjusted analyses were conducted using Poisson regression, controlling for age, sex, comorbidity, expected resource use, recent registration, income quintile, and rurality. Secondary analyses were conducted stratifying by rurality to examine whether associations were independent of setting.

Limitations

This study had a number of limitations that are important for interpretation. First is the cross-sectional nature of the study, which limits causal inference. Longitudinal approaches could be highly recommended in the study of primary care reform and would be an appropriate next step for the current work. Lack of randomization means that important unobserved factors, such as provider and patient attitudes, could have affected the results. Residual confounding is also a possibility—for example, when controlling for rurality among groups with very different distributions. For these reasons, the study did not assess whether the access bonus scheme was achieving its objectives.

Second, administrative data contain only superficial information on patients’ demographic characteristics, socioeconomic status, or health care needs and are missing important information about the duration or complexity of visits and health attitudes and behaviors. These data also exclude nonbillable forms of access (such as telephone calls and email) and exclude people without provincial health care coverage (such as refugees, foreign students, and undocumented migrants). Active members of the military and the Royal Canadian Mounted Police, as well as indigenous populations living on reserve, may receive health services that are not included in these data.

Finally, it is possible that capitation practices that were paid 15 percent of fee-for-service claims might not have billed as completely as those in fee-for-service that were paid 100 percent. However, that limitation would be expected to apply equally to all of our study groups.

Study Results

Each of the access bonus quintiles included approximately 800 primary care physicians (exhibit 1). The proportion of the maximum possible access bonus payment ranged from 0.0 percent in the lowest quintile to 82.9 percent in the highest, with corresponding mean payments of $0 and $36,570, respectively. Physicians in the highest quintile were of similar age and had practiced a similar number of years as those in the lowest quintile but were more likely to be male and graduates of Canadian medical schools. They were far more likely to practice in smaller cities and rural areas (exhibit 2). Roster sizes (mean numbers of patients) were similar across quintiles (exhibit 1).

SOURCE Authors’ analysis of the following data sets for 2013 at ICES (formerly known as the Institute for Clinical Evaluative Sciences): Generalized Alternate Payments, Architected Payments, Ontario Health Insurance Plan, Corporate Provider Database, ICES Provider Database, and the Rurality Index of Ontario. NOTES Physicians in quintile 1 received the lowest percentage of the maximum potential access bonus payment, and physicians in quintile 5 received the highest percentage. The rurality measurement is explained in the text. There were no physicians in rural areas in quintile 1 or quintile 2. The percentages for the quintiles do not sum to 100 because of missing data.

There were just over one million patients in each quintile, and patterns of age and sex were similar across quintiles (exhibit 3). The proportion of recent immigrants in the highest quintile was less than one-quarter that in the lowest quintile. The highest quintile also had the lowest proportion of people with high comorbidity, as measured by number of Aggregated Diagnosis Groups, and a low proportion of people with high expected resource use, as measured by Resource Utilization Bands.

Visits and costs were analyzed after excluding 450 physicians (11.1 percent) who were exempted from providing after-hours care (data not shown). Groups could request an exemption from after-hours care if at least half of their members provided inpatient, intrapartum, or ED care. The proportion of patients exempted was similar across quintiles except in the highest—in which a lower proportion was exempted (exhibit 4). Overall, patients had an annual average of 2.93 primary care visits, 2.09 of which were to their enrolling physician and 2.29 to that physician or someone else in their enrolling group. Patients of physicians in the highest quintile had the lowest mean number of visits to the enrolling physician, enrolling group, and overall, while patients of physicians in the lowest quintile had the highest mean number of visits to the enrolling physician, enrolling group, and overall.

Continuity of care overall was moderately high, with 71.3 percent of primary care visits overall to the enrolling physician and 78.2 percent to the enrolling group. Patients of physicians in the highest quintile had the highest levels of continuity of care, and patients of physicians in the lowest quintile had the lowest levels. Overall, 6.5 percent of primary care visits had a billing code indicating service after 5:00 p.m. on a weekday, and 3.3 percent had a code for service on a weekend or holiday. Physicians in the highest quintile had the lowest proportion of all visits after hours, while physicians in the lowest quintile had the highest proportion. Overall, almost a quarter of the patients (23.0 percent) visited the ED each year, with the highest proportion in the highest quintile and the lowest proportion in the lowest quintile. When ED visits were categorized by urgency, the same pattern was seen, with higher rates of both high- and low-urgency visits found in the highest quintile.

When we considered primary care and ED visits together, we found that patients of physicians in the highest quintile of access bonus payments had the lowest mean number of ambulatory visits, while those in the lowest quintile had the highest mean number. This pattern remained unchanged after adjustment. Mean and median ambulatory costs were similar across quintiles. Compared with the lowest quintile, in the highest quintile unadjusted ambulatory costs were slightly lower, but adjusted costs were higher.

For the secondary analyses stratified by rurality, the results for larger and smaller cities were largely consistent with the main findings (data not shown but available on request). Only 7.1 percent of physicians practiced in rural areas, and there were no rural physicians in quintiles 1 or 5. The patterns in rural quintiles 2–4 were inconsistent across measures, likely because of small numbers. For larger and smaller cities, the pattern of adjusted mean ambulatory visits was the same as in the overall findings, with the highest rates in quintile 1 and the lowest in quintile 5. The pattern of adjusted ambulatory costs was similar in smaller cities to the overall findings, with the highest costs in quintiles 4 and 5. For larger cities the adjusted ambulatory costs for quintile 5 were slightly lower than those in the other quintiles.

Discussion

Primary care payment reform can be fraught with trade-offs, risks, and unintended consequences.

Primary care payment reform can be fraught with trade-offs, risks, and unintended consequences. Ontario’s access bonus was ostensibly designed to incentivize access with the enrolling physician and contain ambulatory costs. It was substantial in amount, averaging over $17,000 per physician and exceeding $36,000 for physicians in the highest quintile. These funds flowed disproportionately to physicians outside large cities. Physicians receiving the access bonus served patients with less than average comorbidity and expected resource use. Although physicians receiving the greatest proportion of the access bonus had higher continuity of primary care visits, they also had the fewest primary care visits, provided the least after-hours care, and had the highest rates of ED visits. The adjusted cost of ambulatory visits, including primary care and ED visits, was highest among those receiving the greatest proportion of the access bonus. Had the costs of the access bonus itself been included, costs would have been even higher in the highest quintile.

Several considerations are taken into account when designing payment reform, including incentivizing desired provider and patient behavior. In the case of the access bonus, the government payer wished to avoid paying twice for the same service—once in capitation and again in fee-for-service for use of services outside the enrollment group. In theory, a bonus payment for access would also reward physicians who were available to their patients, thereby reducing the need for outside care. In practice, however, incentives often flow to those already exhibiting the desired behaviors, sometimes with little evidence of change in outcomes.^27–30 The literature on the impact of financial incentives is mixed, with findings from the Quality and Outcomes Framework in the UK indicating improvement and better equity for some measures, but slowing of the changes over time as well as trade-offs in performance between measures such as timely access and continuity of care.³⁰ Other assessments of pay-for-performance have shown only very small improvements or no improvement overall.^27,28,31,32

Physicians whose patients had the highest outside use of services also had the highest visit rates within their enrollment practices and groups. This pattern likely reflects greater complexity of care needs and the greater availability of primary care services such as walk-in clinics in larger cities, which demonstrates that health care services of all kinds are often driven by supply.^33–35 Conversely, smaller cities and rural areas have fewer alternative sources of care and therefore have less outside use, apart from that of EDs—which often serve as key points of primary care contact in many smaller communities. Our findings may also reflect different patterns of ED use¹⁴ and cultural differences in health-seeking behaviors between rural and urban patients.³⁶

Payment reforms should align with health system goals to the extent possible. Contrary to this principle, physicians in this Ontario capitation-based model would be more likely to receive the access bonus if they encouraged their patients to visit the ED instead of a walk-in clinic. This is particularly problematic given that Canada has some of the highest rates of ED use among high-income countries.³⁷ A previous study in Ontario found that among reformed practices that adopted the new models of care, 60 percent of after-hours telephone messages directed patients to the ED, while only 32 percent informed patients of their own after-hours clinic.³⁸ Specialist care is also not penalized in the access bonus, which discourages collaboration among family physicians with specialized skills while incentivizing referral to more costly specialists. Although continuity of care has known benefits to both health outcomes and health system costs,^39–41 patients sometimes prioritize convenience over other considerations, and penalizing physicians for those choices may be ineffective and unfair.

Policy Implications

There may be a need for different payment reform models in different settings such as small rural towns and large urban centers. Providers in rural settings, where almost all use of services and after-hours care outside of the enrollment group is provided in EDs, should be neither rewarded nor penalized for that pattern of care when few viable alternatives for care are available. In major urban centers, where walk-in clinics are plentiful and many people commute to work, a different model may be needed. A model for large cities could allow dual enrollment of patients with a practice close to home and another practice close to work or provide access bonus payments to practices that could demonstrate timely access to care for urgent problems and after-hours care. A more radical approach would be to eliminate incentives or penalties for outside enrollment group use and instead make timely access and after-hours access requirements of the payment reform model, together with transparent reporting of patient experience in accessing care. Policy makers in other jurisdictions may also want to carefully consider the impact of primary care payment reform on other sectors. For example, the Ontario access bonus was seen as desirable in primary care, but when examined across the whole health system, it inadvertently provided a financial incentive for physicians to advise their patients to use EDs in preference to lower-cost walk-in clinics. Finally, policy makers should consider prospective monitoring and evaluation of payment reforms to ensure that they are achieving their goals and to implement midcourse corrections as needed.

Conclusion

Ontario’s primary care access bonus was paid to the physicians who provided patients with the least after-hours care and whose patients had the highest ED visit rates and highest adjusted ambulatory costs. Payment reform may need to be designed and implemented differently for diverse settings such as small rural communities and densely populated downtown cores. Financial incentives should be prospectively evaluated and frequently revisited to ensure relevance, alignment with system goals, efficiency, and equity.

ACKNOWLEDGMENTS

This study was supported by the Innovations Strengthening Primary Health Care through Research (INSPIRE-PHC) program, which was funded through the Health Systems Research Fund of the Ontario Ministry of Health and Long-Term Care (MOHLTC). This study was also supported by ICES (formerly known as the Institute for Clinical Evaluative Sciences), which is funded by an annual grant from the MOHLTC. The opinions, results, and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by ICES or the MOHLTC is intended or should be inferred. Richard Glazier and Tara Kiran were funded as clinician scientists in the Department of Family and Community Medicine at the University of Toronto and at St. Michael’s Hospital in Toronto. Kiran is also supported by the Canadian Institutes of Health Research and Health Quality Ontario as an embedded clinician researcher. At the time the research for this article was conducted, Michael Green was supported by the Clinical Teachers’ Association of Queen’s Chair in Applied Health Economics/Health Policy.

NOTES

• 1 Ritchie A, Marbury D, Verdon DR, Mazzolini C, Boyles S. Shifting reimbursement models: the risks and rewards for primary care. Medical Economics [serial on the Internet]. 2014 Apr 8 [cited 2019 Feb 11]. Available from: http://medicaleconomics.modernmedicine.com/medical-economics/content/tags/aca/shifting-reimbursement-models-risks-and-rewards-primary-care?page=full Google Scholar
• 2 Miller HD. From volume to value: better ways to pay for health care. Health Aff (Millwood). 2009;28(5):1418–28. Go to the article, Google Scholar
• 3 James BC, Poulsen GP. The case for capitation. Harvard Business Review [serial on the Internet]. 2016 [cited 2019 Feb 11]. Available from: https://hbr.org/2016/07/the-case-for-capitation Google Scholar
• 4 CMS.gov. MACRA [Internet]. Baltimore (MD): Centers for Medicare and Medicaid Services; [last modified 2018 Sep 21; cited 2019 Feb 11]. Available from: https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/Value-Based-Programs/MACRA-MIPS-and-APMs/MACRA-MIPS-and-APMs.html Google Scholar
• 5 NHS Employers. Quality and outcomes framework [Internet]. Leeds (UK): NHS Employers; [cited 2019 Feb 11]. Available from: http://www.nhsemployers.org/your-workforce/primary-care-contacts/general-medical-services/quality-and-outcomes-framework Google Scholar
• 6 Australian Institute of Health and Welfare. Primary health care in Australia [Internet]. Canberra (Australia): AIHW; [last updated 2016 May 24; cited 2019 Feb 11]. Available from: https://www.aihw.gov.au/reports/primary-health-care/primary-health-care-in-australia/contents/about-primary-health-care Google Scholar
• 7 Van Weel C, Schers H, Timmermans A. Health care in the Netherlands. J Am Board Fam Med. 2012;25(Suppl 1):S12–7. Crossref, Medline, Google Scholar
• 8 Salvadori P. Primary care in Italy [Internet]. Utrecht (The Netherlands): European Forum for Primary Care; c 2019 [cited 2019 Feb 11]. Available from: http://www.euprimarycare.org/column/primary-care-italy Google Scholar
• 9 Hutchison B, Levesque JF, Strumpf E, Coyle N. Primary health care in Canada: systems in motion. Milbank Q. 2011;89(2):256–88. Crossref, Medline, Google Scholar
• 10 Frakt AB, Mayes R. Beyond capitation: how new payment experiments seek to find the “sweet spot” in amount of risk providers and payers bear. Health Aff (Millwood). 2012;31(9):1951–8. Go to the article, Google Scholar
• 11 Bodenheimer TS. Capitation’s uncertain future. J Gen Intern Med. 2001;16(4):281–2. Crossref, Medline, Google Scholar
• 12 Hutchison BG, Birch S, Hurley J, Lomas J, Stratford-Devai F. Effect of a financial incentive to reduce hospital utilization in capitated primary care practice. Hamilton (ON): McMaster University Centre for Health Economics and Policy Analysis; 1994. (Working Paper Series No. 94-2). Google Scholar
• 13 Alternative Relationship Plan Program Management Office. Funding models: blended capitation model [Internet]. Calgary (AB): ARP PMO; c 2019 [cited 2019 Feb 11]. Available from: https://www.arppmo.org/clinical-arp/funding-models/ Google Scholar
• 14 Hines A, Fraze T, Stocks C. Emergency department visits in rural and non-rural community hospitals, 2008 [Internet]. Rockville (MD): Agency for Healthcare Research and Quality; 2011 Jun [cited 2019 Feb 11]. (HCUP Statistical Brief No. 116). Available from: https://www.hcup-us.ahrq.gov/reports/statbriefs/sb116.jsp Google Scholar
• 15 Hodgins MJ, Wuest J. Uncovering factors affecting use of the emergency department for less urgent health problems in urban and rural areas. Can J Nurs Res. 2007;39(3):78–102. Medline, Google Scholar
• 16 Ontario Ministry of Finance. Ontario fact sheet February 2019 [Internet]. Oshawa (ON): Ministry of Finance; [last modified 2019 Feb 8; cited 2019 Feb 11]. Available from: http://www.fin.gov.on.ca/en/economy/ecupdates/factsheet.html Google Scholar
• 17 All dollar amounts in this article are Canadian dollars.
• 18 Hutchison B, Glazier R. Ontario’s primary care reforms have transformed the local care landscape, but a plan is needed for ongoing improvement. Health Aff (Millwood). 2013;32(4):695–703. Go to the article, Google Scholar
• 19 Marchildon GP, Hutchison B. Primary care in Ontario, Canada: new proposals after 15 years of reform. Health Policy. 2016;120(7):732–8. Crossref, Medline, Google Scholar
• 20 To access the appendix, click on the Details tab of the article online.
• 21 Ray JG, Vermeulen MJ, Schull MJ, Singh G, Shah R, Redelmeier DA. Results of the Recent Immigrant Pregnancy and Perinatal Long-term Evaluation Study (RIPPLES). CMAJ. 2007;176(10):1419–26. Crossref, Medline, Google Scholar
• 22 Johns Hopkins ACG System. Who we are and what we do [Internet]. Baltimore (MD): Johns Hopkins University; c 2019 [cited 2019 Feb 11]. Available from: https://www.hopkinsacg.org/ Google Scholar
• 23 Kralj B. Measuring rurality—RIO2008_BASIC: methodology and results [Internet]. Toronto: Ontario Medical Association; 2009 [cited 2019 Feb 8]. Available from: https://www.oma.org/wp-content/uploads/2008rio-fulltechnicalpaper.pdf Google Scholar
• 24 Breslau N, Reeb KG. Continuity of care in a university-based practice. J Med Educ. 1975;50(10):965–9. Medline, Google Scholar
• 25 Bullard MJ, Unger B, Spence J, Grafstein E. Revisions to the Canadian Emergency Department Triage and Acuity Scale (CTAS) adult guidelines. CJEM. 2008;10(2):136–51. Crossref, Medline, Google Scholar
• 26 Wodchis WP, Bushmeneva K, Nikitovic M, McKillop I. Guidelines on person-level costing using administrative databases in Ontario [Internet]. Toronto: University of Toronto; 2013 May [cited 2019 Feb 11]. (Health System Performance Research Network Working Paper). Available from: https://tspace.library.utoronto.ca/bitstream/1807/87373/1/Wodchis%20et%20al_2013_Guidelines%20on%20Person-Level%20Costing.pdf Google Scholar
• 27 Kiran T, Wilton AS, Moineddin R, Paszat L, Glazier RH. Effect of payment incentives on cancer screening in Ontario primary care. Ann Fam Med. 2014;12(4):317–23. Crossref, Medline, Google Scholar
• 28 Lavergne MR, Law MR, Peterson S, Garrison S, Hurley J, Cheng Let al. A population-based analysis of incentive payments to primary care physicians for the care of patients with complex disease. CMAJ. 2016;188(15):E375–83. Crossref, Medline, Google Scholar
• 29 Rosenthal MB, Frank RG, Li Z, Epstein AM. Early experience with pay-for-performance: from concept to practice. JAMA. 2005;294(14):1788–93. Crossref, Medline, Google Scholar
• 30 Campbell SM, Reeves D, Kontopantelis E, Sibbald B, Roland M. Effects of pay for performance on the quality of primary care in England. N Engl J Med. 2009;361(4):368–78. Crossref, Medline, Google Scholar
• 31 Scott A, Sivey P, Ait Ouakrim D, Willenberg L, Naccarella L, Furler Jet al. The effect of financial incentives on the quality of health care provided by primary care physicians. Cochrane Database Syst Rev. 2011;(9):CD008451. Medline, Google Scholar
• 32 Kiran T, Victor JC, Kopp A, Shah BR, Glazier RH. The relationship between financial incentives and quality of diabetes care in Ontario, Canada. Diabetes Care. 2012;35(5):1038–46. Crossref, Medline, Google Scholar
• 33 Cassidy A. Health Policy Brief: Geographic variation in Medicare spending. Health Affairs [serial on the Internet]. 2014 Mar 6 [cited 2019 Feb 11]. Available from: https://www.healthaffairs.org/do/10.1377/hpb20140306.633790/full/ Google Scholar
• 34 Dartmouth Medical School, Center for the Evaluative Clinical Sciences. Dartmouth atlas of health care in Michigan [Internet]. Hanover (NH): Dartmouth Medical School; 2000 [cited 2019 Feb 11]. Available from: http://archive.dartmouthatlas.org/downloads/atlases/michigan_atlas.pdf Google Scholar
• 35 Dartmouth Medical School, Center for the Evaluative Clinical Sciences. Dartmouth atlas of health care in Pennsylvania 1998 [Internet]. Hanover (NH): Dartmouth Medical School; 1998 [cited 2019 Feb 11]. Available from: http://archive.dartmouthatlas.org/downloads/atlases/pennsylvania_atlas.pdf Google Scholar
• 36 Kellehear A. Country health: another side of the rural crisis. Regional Journal of Social Issues. 1988;22:1–8. Google Scholar
• 37 Berchet C. Emergency care services: trends, drivers, and interventions to manage the demand [Internet]. Paris: OECD Publishing; 2015 [cited 2019 Feb 20]. (OECD Health Working Papers No. 83). Available for download from: https://www.oecd-ilibrary.org/social-issues-migration-health/emergency-care-services_5jrts344crns-en Google Scholar
• 38 Howard M, Randall GE. After-hours information given by telephone by family physicians in Ontario. Healthc Policy. 2009;5(2):106–15. Medline, Google Scholar
• 39 Starfield B, Shi L, Macinko J. Contribution of primary care to health systems and health. Milbank Q. 2005;83(3):457–502. Crossref, Medline, Google Scholar
• 40 Cabana MD, Jee SH. Does continuity of care improve patient outcomes? J Fam Pract. 2004;53(12):974–80. Medline, Google Scholar
• 41 Gill JM, Mainous AG 3rd, Nsereko M. The effect of continuity of care on emergency department use. Arch Fam Med. 2000;9(4):333–8. Crossref, Medline, Google Scholar