It's a line any hospital executive has heard before:
"My patients are sicker."
"This data isn't statistically meaningful."
"There's more to the story."
They likely came from physicians who received less-than-satisfactory performance results — whether it's surgical complication rates, cost and utilization rates, or patient outcomes. As a healthcare leader, these words seem like a knee-jerk reaction to shift blame elsewhere.
Mark Wagar, president of Northridge, Calif.-based Heritage Medical Systems, has heard a similar line or two before.
"Physicians are scientists at heart," Mr. Wagar says. "They are preselected for scientific acumen."
This scientific frame of mind, powered by a healthy dose of competition, explains the root cause of these excuses. "When a physician is presented with performance data and compared to their peers, their natural instinct is to look for what the drivers or differences are," says Igor Belokrinitsky, a healthcare strategist and partner at Strategy&.
Train physicians to see the forest, not the trees
It may be frustrating to see living, breathing patients represented by dots on a graph. This is especially true for physicians today, who now not only face internal performance scores, but external ones like Yelp, ProPublica's Surgeon Scorecard, and Medicare's Physician Compare.
Part of the frustration stems from concerns about the metrics used to track performance. Physicians know each of their data points personally, as patients. They know patients' vitals and other clinical measures, as well as their stories, setbacks and triumphs. This can make it difficult to see how a patient with significantly improved medication adherence, for example, still misses other quality benchmarks.
There is great variability in terms of patient complexity. Patients with multiple chronic illnesses, behavioral health issues or socioeconomic difficulties may have poor nutrition, lack access to resources or not comprehend their medication schedule. "It can have a dramatic impact," says Mr. Belokrinitsky. "We don't always have a good way of capturing those complexities."
However, it is highly unlikely complex patients will gravitate to the same physician rather than multiple within the same facility or organization, Mr. Belokrinitsky notes. If every physician at an organization draws from the same pool of patients, they will likely treat similar portions of complex patients, with multiple chronic diseases or other difficulties. Considering two important factors — risk adjustment and physician variability — the "My patients are sicker" excuse becomes plain unconvincing.
The first of these factors is risk adjustment. Data should be adjusted for case mix and outliers that could significantly skew data — a physician with low procedure volume at a specific facility, for example — should be thrown out.
While it's not out of the realm of possibility to have flawed data, any issue raised by physicians should be met with data clarification and actionable information.
"The 'My patients are sicker' explanation is not plausible especially when we are able to compare the practitioner with peers using the case mix index — which tells us how sick the patient is — and other risk adjustment methodologies for length of stay and mortality through vendors," says Anil Gopinath, MD, regional CMO of Presence Health Fox River Valley in Aurora, Ill.
Second, the 'My patients are sicker' explanation is improbable due to the undeniable "physician-factor" in patient care. Variability in physician skill and performance significantly impacts outcomes, as many studies have shown.
One such study, published in The New England Journal of Medicine links unnecessary testing directly to individual ophthalmologists. The likelihood patients in the study would undergo preoperative testing — which is shown to have no effect on outcomes — hinged on provider preference. In fact, the 36 percent of ophthalmologists responsible for most of the testing administered it to 74 percent of their patients.
Even beyond basic practice patterns, medicine involves a level of skill that varies among medical staff. This is aptly demonstrated by ProPublica's online Surgeon Scorecard database. While the scorecard has met some skepticism from the medical community, it clearly indicates a great deal of variation in surgical performance. More than 750 surgeons, for example, had no recorded surgical complications, while 25 percent of complications could be attributed to 11 percent of surgeons.
The idea that quality metrics are not effective is "completely outrageous," says Mr. Belokrinitsky. When case loads are adjusted for severity and outliers, it is doubtful one physician is magnet for the sickest patients, and even more so considering physician variability. Consequently, any physician who cries foul should raise a red flag for hospital leadership.
"For the most part, when there's smoke, there's fire," says Mr. Wagar about physicians who claim their patients are disproportionately complex when faced with poor performance results.
Open the conversation anonymously
Telling someone they are wrong isn't easy.
It's especially challenging if that person is trained to think and act like an expert. The key to addressing physicians' most common excuse, according to Mr. Belokrinitsky, lies in framing the conversation as collaborative, not confrontational.
Data should be anonymously presented to physicians, Mr. Belokrinitsky says, to keep meetings from becoming personal right away. This provides a safe and constructive environment for departments to talk about what behaviors drive better results. It also facilitates discussion about effective clinical pathways, so physicians can walk away with a resolve to improve standardization, Mr. Belokrinitsky says.
The next meetings should be individual. Hospital leadership should meet with low performers and show them their results, giving physicians a chance to explain their outcomes. This conversation will be much easier if it's built about cost rather than quality.
"Quality is very touchy," Mr. Belokrinitsky says. "It calls into question how good of a physician you are."
Choose metrics wisely
Some say setting the bar for performance leads physicians to avoid a complex case load. To hit their benchmarks, physicians may cherry pick the most textbook cases and leave the unique cases to the rest.
Lisa Rosenbaum, MD, details one such instance in an op-ed published by NEMJ:
"One Monday morning, rounding on a patient who needed relatively urgent coronary-artery bypass surgery, a newly appointed cardiologist in New York asked the team to call a surgical consult (some details have been changed to protect those involved). 'We can't call today,' the cardiology fellow explained patiently. 'Dr. X. is taking consults. He wouldn't touch our patient with a 10-foot pole.' The fellow scrutinized the call schedule. 'The only surgeon who might take him isn't on until Wednesday,'" Dr. Rosenbaum wrote.
Her article was a reaction to the Surgeon Scorecard, which she felt was an unfair evaluation of quality. She wrote specifically on New York since it was one of few states to publicly report cardiac surgery and percutaneous coronary intervention, which motivated the cardiologist in question to avoid complex cases.
This phenomenon — similar to what Mr. Belokrinitsky calls "turfing," or protecting one's turf by moving complicated cases to different hospital units — is preventable.
"To avoid those behaviors, before you engage with physicians, you have to think as an administrator about what the few quality metrics are that matter to you," Mr. Belokrinitsky says. Administrators must choose metrics and benchmarks most meaningful to patient care and help practitioners understand why they were chosen.
End the excuse
As hospitals begin to reward value over volume, physicians are becoming more comfortable with metrics and benchmarking, and less apt to say, "My patients are sicker."
"These days I am hearing less and less of this reasoning from physicians compared to five to seven years ago. This could be partly due to the fact that physicians are getting familiar with data reporting in their practices from third party payers," Dr. Gopinath said.
Nonetheless, it's likely not the last time a provider will attempt to explain away poor outcomes or poke holes in performance data. Hospital leadership must work with physicians to ensure data is as clean as possible, and when it is, help providers learn from it.
As Dr. Gopinath puts it, today's mantra in healthcare is "transparency." The increasing popularity of physician data from sources like Physician Compare, Yelp, the Surgeon Scorecard and even hospitals' own sites indicate performance data is here to stay. "I encourage physicians to adapt to the changing landscape in healthcare and quality reporting, rather than resist it," he says.
More articles on integration and physician issues:
Physicians emerged from retirement to treat victims of Oregon school shooting
University Hospitals physician writes break-up letter to patient
Study finds physician bargaining power counterproductive to hospital survival