Money matters: does ‘pay-for-performance’ improve quality?

In Ontario, new ways of paying doctors have been introduced in an attempt to improve the quality of their services. 

One approach is pay-for-performance, which pays doctors for meeting certain treatment goals.

However, there is little high quality evidence that pay-for-performance improves the quality of care, and it appears to have had limited impact in Ontario so far.

Over the last ten years, Ontario has significantly changed how it pays doctors – particularly family doctors.  Prior to 2003, nearly all doctors in Ontario were paid through “fee for service”, which paid them a set amount for each medical service they provided.

Fee for service has been widely criticized, because it is thought to promote quantity over quality.  Rick Glazier, a family doctor and researcher with the Institute for Clinical Evaluative Sciences (ICES) explains that “around the world, there is virtually no high-functioning health care system that still pays doctors on a strictly fee for service basis.”

Over the past decade, the Ontario Ministry of Health and Long-Term Care has introduced a number of new payment plans, with the goal of improving quality of care and access. Brian Hutchison, professor of family medicine at McMaster University explains that “Ontario’s move away from fee for service was primarily about changing the system from being driven by the goal of providing many services, to being driven by the need to  improve peoples’ health. It was really aimed at getting Ontarians to sign up with family doctors and remove the barriers to establishing family practices where care is delivered by a group of health care professionals.”

However, this change has come at a price. Earlier this year an ICES* report found that payments to doctors increased roughly in line with inflation until 2005, when they began to increase sharply.  From 2005 to 2010, payments to doctors rose by 10% per year, compared with only a 2% increase in inflation.


One of changes that Ontario made that contributed to the increase in physician salaries was to introduce payments for family doctors who met certain care goals, such as vaccinating their elderly patients against the flu.  This approach is known as pay-for-performance. Some suggest that the costs to Ontario’s health care system have been considerable – estimated at over $50 million in 2010. However, research from other jurisdictions suggests that there is a relatively weak link between pay-for-performance and quality of care, and recent studies from Ontario appear to bear this out.

In 2002, Ontario introduced a special incentive for family doctors whose patients with diabetes received three designated monitoring tests (glycated hemoglobin levels, cholesterol levels, and eye tests).  The hope was that this incentive would encourage all physicians to ensure that their patients with diabetes were carefully monitored and appropriately treated.  However, a study on the effectiveness of this program showed the incentives had limited impact, and that the doctors who claimed the incentives tended to be those who had already been providing the three tests before the incentive was introduced.

Another Ontario-based study examined the effects of performance incentives on the delivery of preventative services (flu shots for seniors, toddler immunizations, pap smears, mammograms and colorectal screening).  It found only a modest improvement for most of these services, and no improvement for toddler immunizations.  Only the rate of colorectal cancer screening experienced a substantial increase at the same time as pay for performance was introduced.  The same study also evaluated incentives to increase some services provided by family doctors that are more complex or time consuming such as obstetrical care, hospital care, palliative care, office procedures, prenatal care and home visits. The study found that financial incentives did not lead to an increase in these services.

A major review of pay-for-performance programs in the United States found a few small positive impacts in quality associated with the incentives, but no evidence of any major quality improvements.  “Given the state of the evidence, it seems there’s been a tendency to overestimate the effects of payment mechanisms on doctors’ behavior and medical outcomes,” says Brian Hutchinson. “The general picture is that you can spend a lot of money and not achieve very much.”

The future of funding incentives & quality

While pay-for-performance as currently designed appears to have had only a small impact to date, many experts believe it can be successful in the future.  However, all the experts interviewed agreed payment reforms will not substantially improve quality of care unless several important shortcomings are addressed.

One limitation of Ontario’s approach to date has been a lack of monitoring and enforcement to ensure doctors are actually providing the services they are collecting incentives for.  For example, a recent Ontario Auditor General’s report found that many doctors groups were not providing the after-hours care stipulated under their contracts.  To be successful in the future, any incentive-based payment system must have sufficient monitoring and follow-up to ensure targets are being met.

Another limitation is raised by Rick Glazier, who explains, “In Ontario, we haven’t aligned our payment reforms with our larger goals for our health care system.” He suggests that “moving forward, incentives should be targeted at our key priorities which include reducing emergency department visits, hospital readmissions and improving timely access to primary care.” However, again the challenge remains developing the right set of incentives that will lead to the intended positive outcomes without causing any unintended negative consequences – something the literature on pay-for-performance suggests is easier said than done.

Brian Golden, chair of health sector strategy at the University of Toronto Rotman School of Management adds that “it is essential that any payment reforms address not just motivating doctors to change, but – as needed – changing the environment in which they practice so that they have the capacity to change.”

* Andreas Laupacis and Irfan Dhalla were co-authors on the February 2012 ICES Investigative Report: Payment to Ontario Physicians from the Ministry of Health and Long-Term Care, Sources 1992/93 to 2009/10.

The comments section is closed.

  • ND says:

    Something could be learned about P4P models in the UK. The models that have been studied generally don’t work in achieving the key outcomes (better access, reduced ER visits, better disease management etc.). In the UK, they generally found that when you financially incentivize specific health care services, such as cancer screening, you just increase the use of those particular services and those services that were not linked to the P4P plan are neglected. Another issue is that you just end up paying more for services that are already being rendered – greater rewards to those docs who are how have already been doing a good job/meeting P4P goals. There is also an issue of over/inappropriate use of incentivized services when the indications or practice guidelines are unclear. Certainly, reinforcement is key as well. In Quebec they’ve just started cracking down on the FMGs who are not offering after hours care. The Ministry of Health and Social Services is threatening to cut their funding for not delivering.

    Wasn’t there a plan to introduced outcomes-based funding where payment is linked to patient outcomes as opposed to services provided?

  • Wisconsin health insurance says:

    I’m sorry, but this seems like a ridiculous idea. Pay for performance? what will they give doctors bonuses for every patient that lives? If that is the case…no one would get a bonus. I’m not sure how they’d go about something like this but doctors should be paid for quality care.

  • Anonymous says:

    Doctors or business individuals?! This is the question that every Canadian ask themselves after five to ten minutes from any typical family physician visit. Five to ten minutes of her/his time but still charging the government a one full hour fee, even though everyone knows that it's never been done in the history of family physicians that one of them seen a patient for one full hour. Simply because as many patients they can see a day as more money is direct deposited into there accounts, fee for service or assembly line piece work is how physicians are paid, sometimes they bill for more hours than in 24 hours. patients could wait, most of the times for months to see there family physician, and yes I will never count them as health care providers. I will never put any physician in the same level as a nurse, nurse practitioner, or other health care providers, because they are simply not health care providers, they are business individuals with medical bachelor or master degree. Canada is the number one country in the world in human rights. It's the first desirable country for immigrants from all over the world. But when it comes to the health care system, some times a country identified as a third world could provide better health care for one simple reason, doctors have time for there patients. There is no political B.S. and there salary is not effected by the number of there daily patients visits, therefore they have all the time in the world to think, analyze and diagnose correctly to make sure that there patients are ok and they won't need to visit them again for a long time. Compare this to Canadian physicians, they always ask you to book an appointment and come to there office even after simple routine lab test just to tell you that your lab appears to be fine so they can send a one hour bill to the government. Physicians were teenagers with big dreams to save the world one patient at a time, that's initially why they choose to become who they are now. Unfortunately they did not fulfil there dreams, no one would, when the government is paying by the number of patients seen. I had the opportunity to work in a hospital and witnessed in person the system first hand. And to show the mentality deference's, I noticed one important thing the health ministry pushes to do is to call patients "clients or customers" meanwhile all over the world, patients are called patients for one reason, because they are.

    • Mark Macleod says:

      Unfortunately the facts you presented here and on the radio are not correct. The fee is not for "an hour" – it is for the visit and there is no time stipulation added to it. Doctors are not lawyers. So the doctor gets the same payment if the visit takes 10 minutes or 40 minutes unless there is a change in complexity or criteria are met that allow for an increased fee.

  • Mark Macleod says:

    sorry – I neglected to give the year for the articles, it was HBR 2009 if my memory is correct.

  • Mark MacLeod says:

    If the system of care isn’t designed for and hasn’t obtained a culture of quality, then incentives to groups or individuals within that system will either only go so far or fail. I’m leery of the current trend of scapegoating doctors for what I think is increasingly system malfunction.

    I’d refer people to the April issue of The Harvard Business Review. In this there is strong emphasis on the need for robust system development and management and that the responsibility care does not fall on an individual provider but on the organization or institution that the provider works within. To this end, the management of the diabetic becomes the responsibility of the group or institution and the performance of the group of institution is what is measured and rewarded.

    Medical knowledge is far to complex now to be held in the hands of individual providers and rests within systems and institutions. This changes the roles of physicians and all other care givers in health – to that of managers and facilitators in addition to technical acumen to really begin to “manage” the patient’s health, not to simply treat disease A and then disease B.

    Pay for performance relies on 2 assumptions. First that what you are measuring and rewarding is a true proxy for the health improvement that you desire to achieve. Secondly that performance payment is a means to convert an uncommon behaviour to a common behaviour that becomes cultural normative which then no longer requires remuneration to support it. The literature would suggest that both are difficult to do well. The experience of the British Health Trusts and pay for performance suggests that it takes a long time to change behaviour and removing incentive gradually results in a reversion to previous behaviours.

  • Ron says:

    The push for quality and cost control via pay-for-performance is trapped between two catchphrases. The first has been the driving credo in the health policy field over the last decade or more: “You can’t manage what you don’t measure”. Born in the world of business schools, this management idea has a compelling intuitive quantitative attraction. On the other hand, there is the notion that “Not everything that can be counted counts and not everything that counts can be counted”. This idea is rooted in sociology, attributed to William Bruce Cameron (1957), reacting to quantitative research and suggesting context and qualitative factors are no less, if not more, important to conclusions and outcomes. Let’s take glycated hemoglobin levels, cholesterol levels, and eye tests as the example. Empirical research suggests that these activities can be linked to quality care and positive outcomes. And yet, these are indicators, mere quantitative proxies for the care process. I think most can agree the critical factor in diabetes management is the physician and patient relationship – collaborative discussion and self-management education in response to test results. So it is little surprise that current incentive programs tend to capture those physicians who by the identified indicators are practicing quality care, and it might be fair to assume these same physicians are managing quality in terms of the patient relationship. But for other physicians taking up the financial incentives, one might simply be reimbursing for measured indicators while not actually affecting the physician-patient process of collaboration and self-management education. This might help further explain mixed research results; that we pay for measurable indicators since we cannot yet effectively measure the qualitative process. We are reaching the end-point in current quality initiatives, and the next step is more careful examination of how system and organizational context, the selection of future health professionals and their education produce a quality “culture”. Such an approach is resistant to reductionist quick fixes and ‘bolt-on’ incentives.

  • David says:

    These are examples of pay for performance ? Seems like pay for performing more procedures not results.

    How about having the OHIP fee tied directly to patient satisfaction ? If your patient satisfaction scores are 90% you get a 10% reduction in OHIP fees ?

    How about physicians in Hospital having OHIP tied directly to readmissions rates, HSMR, lengths of stay and infection rates ?

    That is pay for performance.

    • andreas says:

      Thanks David.

      Your last sentence illustrates how difficult it is to get the incentives right. Many, if not all, of those outcomes are influenced as much (if not more) by the overall quality of the hospital, as they are by the quality of care of an individual physician (although physicians obviously play a major role in hospitals). Also, readmission isn’t necessarily a bad thing. For example, some people with severe heart failure WILL need to be readmitted no matter how good their care is, and incentives to decrease appropriate readmissions could do more harm than good.


    • Marie says:

      I agree with the incentives part.
      What if we went to a semi-private health care similar to that in the U.S.
      This would give us a little more freedom on which doctors we actually want to see, as opposed to getting stuck with the only doctor taking new patients in your area.
      The doctors that have a good reputation will have many more patients that want to visit them than would doctors with poor reputation.
      The incentive then would be for the doctors to give the best possible patient care they can, this would keep their patients happy and get them to come back for follow ups.
      If the patient is disatisfied with the doctor and they actually can go elsewhere for a second opinion, these doctors would see a significant drop in their business.
      If the patients don’t come to see them they won’t be able to bill OHIP or any third party insurance.
      I have a feeling this would force doctors to actually care about their patients cause they actually have something to lose, where now they know if a patient leaves they can easily be replaced.

  • David says:

    It would seem intuitive that change in reimbursement drives change in behaviour. A variety of factors require elucidation: was the quantum sufficient to be an incentive? is the overall income high enough that incentives lose theirr appeal? was the information on compliance available and visible? It would seem a reasonable argument that if the income at risk were sufficiently significant, incentives would have an effect. Seems to have worked with fee for service.

    By the way, your reCAPTHCHA heiroglyphics are really tough to divine.


Jeremy Petch


Jeremy is an Assistant Professor at the University of Toronto’s Institute of Health Policy, Management and Evaluation, and has a PhD in Philosophy (Health Policy Ethics) from York University. He is the former managing editor of Healthy Debate and co-founded Faces of Healthcare

Andreas Laupacis

Editor-in-chief Emeritus

Andreas founded Healthy Debate in 2011. He is currently the editor-in-chief of the Canadian Medical Association Journal (CMAJ)

Republish this article

Republish this article on your website under the creative commons licence.

Learn more