The Power of Physician Relationships

Primary care physicians’ (PCPs) relationships with specialists impacted how their patients experienced referrals and how they were treated by the specialists, a study showed.

Patients referred to specialists that their PCP had trained alongside at the same institution for at least 1 year of medical school or postgraduate education rated their care by the specialist better overall, found J. Michael McWilliams, MD, PhD, and Maximilian J. Pany, BA, both of Harvard Medical School in Boston.

In fact, the significant 9.0-percentage point higher adjusted composite rating compared with PCP-specialist referrals where there was no education overlap was on par with the specialist moving from the 50th to the 91st percentile of performance.

The specialists also were more likely to prescribe medication, order an imaging test, or both for patients of physicians with whom they went to school or fellowship.

“This study suggests potentially large gains in quality from encouraging and harnessing physician-peer relationships,” the researchers wrote in JAMA Internal Medicine.

Results were “quantitatively” similar when looking at patients who ended up being seen by a specialist their PCP hadn’t referred them to but had trained with.

While more medication or more tests wouldn’t automatically be a good thing for patients, the findings do give clues for quality improvement programs, suggested Lawrence P. Casalino, MD, PhD, of Weill Cornell Medical College in New York City.

“It may be debated whether the desire to act professionally in the eyes of colleagues should be defined as intrinsic motivation, but in any case, this motivation — whatever it is called — is, as the authors suggest, something that organizations that provide medical care could use more extensively to improve quality,” he wrote in an accompanying invited commentary.

There are implications for the “related, intertwined phenomena” of value-based purchasing, like the financial carrots and sticks of Medicare’s Merit-based Incentive Payment System (MIPS), and rapid expansion of healthcare organization size, Casalino added.

If physician intrinsic motivation is important, as the study suggested, “it may not be wise for health care organizations or for external entities like Medicare to give direct financial incentives to individual physicians, especially when they are tied — as they typically are — to narrow performance measures,” Casalino wrote, citing health economist Uwe Reinhardt’s argument that physicians’ motivation toward high-quality care need not be “bought with tips.”

But the large regional and national healthcare organizations that are becoming ever more the norm are “necessarily bureaucratic and lack human scale,” he added, which undermines the smaller practice experience in which “mutual knowledge and trust and often, on a practical level, are based on informative and professionally satisfying telephone conversations about individual patients.”

One key question that deserves more attention is, “how (if at all), can large organizations structure some degree of human scale into the settings in which physicians engage with each other and with patients?” Casalino noted.

Potential solutions in the current healthcare climate might include increasing team-based care, electronic consultations, peer coaching, and multispecialty case discussions, McWilliams and Pany pointed out.

Their study involved one from a large academic health system’s electronic health record data on referrals initiated by its PCPs from 2016 to 2019.

During that period, there were 9,920 specialist visits by 8,655 patients (62.9% female; mean age 57.4) referred to 502 specialists in 13 specialties, accounting for 50.7% of total system referrals.

Among the specialist visits in the study, 3.1% involved a PCP-specialist pair in which the two had been in medical school or postgraduate studies together for at least a year. However, the associations were stronger for physician pairs whose training had fully overlapped, whether in the same class or cohort.

The results were also consistently strong for nine of 10 patient-experience items, the researchers noted, such that the co-training pairs “not only elicited a more friendly, courteous, and concerned manner but also clearer explanations, greater engagement in shared decision-making, more time spent with patients, and objective changes in medication prescribing.”

Patients were 1.6 percentage points (95% CI 0.3-2.9) more likely to get a prescription at their specialist visit and a nonsignificant 1.2 percentage points (95% CI -0.7 to 3.0) more likely to get imaging ordered when their PCP and specialist had trained together, after adjusting for between-specialist differences observed in the absence of co-training ties. The same wasn’t true for follow-up appointment recommendations.

“Thus, the behavioral response elicited by co-training ties seems to extend well beyond a change in demeanor that might alter patient perceptions to include behavioral changes that promoted patient-centered care,” the researchers wrote.

Notably, PCPs’ concordance with their patient’s specialists in terms of physician age, sex, medical school graduation year, and having attended the same institution at a different time was not associated with better care experiences.

Casalino noted that a study limitation was its examination of only a single, elite health system, “in which the physicians are likely to have trained at elite institutions and to highly value that era of their lives.” Another was the definition of overlap between PCPs and specialists, because “co-training physicians may not even have met each other or cared for the same patients.”

“It is possible that the results would have been even stronger if the primary care physician-specialist relationship was more substantial (for example, if they had cared for multiple patients together),” he noted.