Doctors typically refer their patients to specialists that they know, creating an unofficial network. And this network may be as important to the quality and cost of care that a patient receives as any one individual doctor, hospital or medical group, a new study by Weill Cornell investigators found.
Published April 23 in Medical Care, this research suggests that accountable care organizations, hospitals, and health insurers may want to identify these physician referral networks, learn from high performing networks, and work with lower performing networks to improve the care they provide. This should lead to lower healthcare costs and better patient outcomes.
"I think that networks are becoming prominent in healthcare in several ways now, in part because of healthcare reform," said the study's lead researcher, Dr. Lawrence Casalino, the Livingston Farrand Professor of Public Health and chief of the Division of Health Policy and Economics in Weill Cornell's Department of Healthcare Policy and Research. "Those who want to form accountable care organizations might want to think about identifying how physicians grouped into specific networks perform compared to each other."
In the study, researchers reviewed Medicare claims for a random 39 percent sample of 987,000 Medicare beneficiaries in five states — Ohio, Pennsylvania Tennessee, Washington and Wisconsin — who were 65 years or older. Working with James Moody, a Duke University sociologist and specialist in social networks, the team created and applied algorithms to the claims data to identify direct and indirect connections among physicians.
Researchers identified 417 of these communities, or networks, with a mean size of 129 physicians, and found that the rate of ambulatory care sensitive admissions — potentially preventable admissions of patients with chronic diseases like congestive heart failure and asthma — varied significantly across the networks. In 2010, there were more than 3.9 million of these admissions of adults to U.S. hospitals at a cost of $31.9 billion, yet 40 percent of those cases may have been preventable with good medical care, says the Agency for Healthcare Research and Quality.
The research team found that many hospitals receive large numbers of admissions from two separate networks, and that the ambulatory care sensitive admission rate generally varies greatly between two networks at the same hospital. In other words, patients are more likely to have a potentially preventable admission to a hospital if they visit a physician in the lower performing network at the hospital than if they visit a physician in the higher performing network.
"If patients had this information, and knew which doctors are in the highest performing networks, it could influence their decision of which doctor to see," Dr. Casalino said. "Similarly, physicians might be interested in knowing which network they're in and how that network is performing."