Specialist Payment Models: Triple the Cardiac Testing, No Difference in Outcomes

How specialists are paid changes how much testing their patients receive, but not whether those patients live longer or avoid hospital. That was the message running through an Institute of Health Economics (IHE) policy webinar on specialist payment models, held on 20 August 2026, which paired new Alberta research with a broader look at what payment reform can actually achieve.

Triple the testing, no measurable difference in outcomes

Sandeep Chand, an MSc student in Community Health Sciences at the University of Calgary, opened with a single question: does the way we pay heart specialists change the care that patients with heart failure actually receive?

His team followed a defined group of Alberta adults newly identified with heart failure and referred to a specialist, tracking each for five years. The condition affects roughly 2% of Alberta adults and leads to about 6,000 hospital admissions a year. More than 100,000 Canadians are diagnosed with it annually, and heart disease overall costs Canada about $7.8 billion a year, with roughly 12% of that going to physician payments.

Ad

The backdrop is a province where most doctors work on fee-for-service. About 83% of Alberta physicians are paid entirely this way, billing for each service, against a national average of about 72%. The main alternative is salary. Because a physician’s decisions can steer up to three-quarters of what a health system spends, the payment model carries real weight.

A previous Alberta study found fee-for-service cardiologists submit about twice as many outpatient imaging claims as salaried ones. Chand wanted to know whether that extra testing changed costs or changed what happened to patients. His team ran a systematic review, published this year and searching six databases through August 2025, alongside a retrospective analysis of Alberta records comparing patients whose main specialist was paid fee-for-service with those whose specialist was salaried over five years.

After propensity-score matching on age, sex, location, socioeconomic factors, medication and other conditions, the study settled on 627 well-matched pairs drawn from 237 specialists and 7,385 patients. The headline result was stark:

“Patients whose specialist was paid fee-for-service received cardiac tests at around three times the rate of patients whose specialist was salaried.”

The testing was mostly echocardiography, nuclear imaging and stress tests, with some angiography. Yet there was no significant difference in heart failure hospitalization, emergency department visits or all-cause mortality. Chand summarized the finding directly:

“Patients seen by fee-for-service specialists consistently receive substantially more cardiac testing and cost the system more, with no difference we could detect in hospitalization, emergency visits and survival.”

A robust finding, a murkier cost picture

The testing result held across every analysis the team ran. Across four versions, the matched group, a cardiologist-only cohort, a broader cohort, and a four-to-one matching, each showed significantly more cardiac testing in the fee-for-service group, between roughly 1.75 and 3.5 times. The direction never flipped, and none of the clinical outcomes was significant in any column.

Costs were less uniform. Testing costs ran significantly higher in the fee-for-service group in all four analyses, at $51 per patient over five years against $16 in the salary group. Hospitalization costs, about $6,400 per patient versus $3,250, and overall costs, about $6,900 versus $3,700, close to double, were significantly higher only in the primary and four-to-one analyses. Chand suggested fee-for-service patients may simply have had more resource-intensive admissions.

He was careful about what the work does and does not show. The findings, he said, should not be read as proving fee-for-service causes higher use or poorer care, and the study could not judge whether any individual test was clinically appropriate. As far as the team knows, it is the first study to examine both cost and clinical outcomes in cardiology by payment model.

What specialist payment models can realistically do

Amity Quinn, a health economist and assistant professor at the University of Calgary, stepped back from the single study to ask what the wider evidence shows. Her 2020 systematic review found the strongest and most consistent effects of specialist payment reform were on utilization, while evidence on quality, patient outcomes and costs was far less consistent. No universally superior model has emerged since.

Quinn’s framing question was not which model Alberta should adopt but what payment reform can realistically accomplish. Specialists shape referrals, diagnostic testing, procedures and treatment pathways, and hold substantial professional autonomy outside traditional management structures. Payment policy, she argued, is one of the few direct levers a government has.

The pattern repeats across her own work. In an Alberta study of patients with chronic kidney disease and diabetes, patients seeing salaried physicians were sicker at the outset; once matched, there was no significant difference in quality indicators, costs or visits, and variation between physicians mattered more than the payment model. In cardiac imaging, the payment model was strongly tied to testing, with imaging rates roughly twice as high among fee-for-service physicians and nearly three times as high among those who could also bill for interpreting the images. Substantial physician and site level variation remained even after adjustment.

Quinn offered the growth of cardiac diagnostic testing in Alberta from 1995 to 2020 as the clearest illustration. Testing rose sharply after the province let private community based facilities open in the late 1990s, paid physicians in those settings on fee-for-service, and let cardiologists bill for both seeing patients and interpreting imaging. The result was a system where more testing could be rewarded while accountability stayed weak. The lesson, she said, is that payment does not operate in isolation:

“Fee-for-service can exist without this being the result. What appears to matter more is how the incentives are interacting with organizational structures.”

Who is accountable, and who holds the risk

The panel discussion, moderated by John Sproule, IHE’s senior policy director, turned to accountability. Cardiologist Glen Sumner noted the Alberta setting is unusual in how widely cardiac testing is available. He pointed to CIHI data showing fee-for-service cardiologist billings up 350%, or 21.9% a year, since 2007-08, growth he said was not fully explained by the fee-for-service model but partly driven by it and by the availability of testing. In a fee-for-service system, he observed, all the volume risk sits with the government payer.

Jason Sutherland, director of the Centre for Health Services and Policy Research (CHSPR) at the University of British Columbia, asked what information should flow to providers, provider organizations and the ministry at the micro, meso and macro levels to know whether cardiac service delivery is meeting population needs. Sumner answered that accountability today is largely peer-to-peer rather than between payer and provider, and that Alberta Health does not behave like an insurer with robust checks on each test.

On reform itself, Sutherland was blunt about the politics. Substantive change, in his experience in the Nordic countries and Ontario, has always been led by government, yet the structure for paying for care has not changed in decades in several provinces. He argued Canada’s health system is “bedevilled by access problems” and that governments should pay for access rather than trying to do everything with complex models. Asked about the US integrated systems often held up as examples, he was sceptical:

“We are a million miles away from looking like a Kaiser Permanente integrated delivery system.”

What problem are we trying to solve

The thread running through the hour was that payment reform works best when aimed at a specific problem. Quinn closed with what she called the more productive question:

“Instead of asking should physicians be paid fee-for-service or salary or through a blended model, I think our first question should be: what problem are we actually trying to solve?”

If the goal is to reduce low-value imaging, she said, payment will need to be combined with tools that have more teeth than Choosing Wisely guidelines. Her takeaway: payment models matter, but their most consistent effects are on utilization, not necessarily on health outcomes or spending. Reform is most likely to help when it is tied to a clearly defined objective, supported by credible data and accountability structures suited to the specialty, and combined with non-financial tools.

Sumner made the same point from the payer’s side, noting that physicians control 70 to 75% of healthcare spending while their own payments amount to only about 12 to 15% of it. He pointed to Alberta’s data assets, a linked administrative record system covering a province of 5.2 million people, as the basis for the kind of predictive analytics used by integrated systems such as Kaiser, Geisinger and Intermountain to hold providers to expected outcomes. Chand added that a simple patient rating system, of the kind used by delivery apps, could bring patient experience into the accountability picture.

Source: Institute of Health Economics, “Specialist Payment Models: Design and Impact” policy webinar, 20 August 2026. The systematic review is available on ScienceDirect.