Health system hospitals, physicians deliver 'marginally' better care at higher cost than independent peers: study

Hospitals and physicians associated with a health system logged “substantially” higher spending and prices in 2018 with clinical quality and patient experience measures only “marginally” better than their unaffiliated peers, according to a review published this week in JAMA.

The findings outline the “enormous” proportion of care being delivered through health systems and establish them as “the dominant form of organization in U.S. healthcare,” according to the study.

The data comparing system and nonsystem performance are also a potential blow to hospital industry groups and many merger hopefuls who commonly cite cost efficiencies and reduced healthcare spending as a benefit when announcing their merger plans to the public and regulators.

“The results cannot establish causal inferences,” Harvard researchers wrote in the journal. “However, the findings are relevant for understanding where price and quality are currently the highest.”

To better characterize a U.S. healthcare delivery system that “has undergone major structural change because of horizontal and vertical consolation,” the researchers combined data from 25 different data sources—IRS forms, Medicare claims databases, American Hospital Association surveys and more—to create a national database of health systems classified by ownership and size.

They also collected data on care quality, patients’ experiences, prices and other performance measures from 2018 claims data and Medicare Consumer Assessment of Healthcare Providers and Systems survey data.

The effort yielded a total of 580 identified health systems in 2018, which included 64% of the nation’s acute care hospitals and 84% of general acute care hospital beds.

Hospitals that were part of health systems tended to be larger and more likely to be a teaching hospital, for-profit or a Medicare Accountable Care Organization participant, according to the study. They were less likely to be a critical access hospital, located in a rural area or a 340B drug pricing program participant.

Additionally, 40% of active physicians and 40% of primary care physicians were part of health systems, though there was “wide variation” in physicians' health system affiliations from state to state, the researchers wrote.

System physicians were more often located in metropolitan areas, in ZIP codes with higher social disadvantage measures, participating in Medicare ACOs or in practices billing as hospital-based outpatient departments. Most larger practices were in systems while nearly all small practices were not.

When comparing care at systems to nonsystems, the researchers found minor but still statistically significant differences in favor of systems for measures such as routine vaccinations and how closely cardiovascular and diabetes care stuck to clinical guidelines.

Patients hospitalized at systems were slightly more likely to be readmitted but less likely to die, according to the study, and less likely to receive low-value services. There were no consistent differences in clinical quality measures across different types of health systems. Patient experience measures were “very similar” for those receiving care from system and nonsystem physicians.

The wider gulf between systems and nonsystems came when reviewing commercial prices and average spending.

On average, researchers found higher commercial prices among systems for outpatient physician visits (26% higher) and visits delivered during an inpatient hospital stay (13% higher). Commercial hospital prices for admissions in 40 of the most common diagnosis groups were 31% higher on average among system hospitals.

Average spending among commercial and Medicare beneficiaries was 4.7% and 5.2% higher, respectively, when care was attributed to a health system physician, which the researchers noted was moderated by the large number of patients that start their care with a health system before flowing to another system or nonsystem provider.  

“Adjusting for practice size reduced the magnitude and significance of most health system performance differences on quality and spending measures and greatly reduced price differences for physician services among large practices,” the researchers wrote. “Still, small- and medium-sized system practices received much higher prices than non-system practices of comparable size.”

In line with prior literature, the new study suggests health systems are able to wield their “substantial” market power in local and regional areas due to their greater scale, wrote Lawrence Casalino, M.D.,Ph.D., a professor of population health sciences and of public health at Weill Cornell Medicine, in an accompanying editorial.

The data also suggest that—at least not at the moment—the increased negotiation power is leading to major quality differences and may potentially allow systems to “drive smaller competitors—and certainly the 49% of physicians still in independent practices—out of business even if the smaller competitors provide higher quality,” Casalino wrote.

With dire financial straits on the horizon and increasing competition from payers, retailers and other emerging players, the industry is on track to see more physicians and hospitals step away from independent operation, he wrote, meaning that researchers and policymakers have limited time to determine whether such a landscape is in the country’s best interest.

“In the worst case, we could wind up with a system in which quality is lower and spending higher than need be, but we would never know it because the only possible comparisons would be among large systems,” he wrote. “This dismal possibility highlights the need for continued research on—and federal and state antitrust attention to—health system and health insurer size and performance.”