Patient outcomes are similar across hospital profiles

Patient outcomes don’t differ significantly across high- or low-cost hospitals, according to a study from Harvard and Yale researchers published in JAMA Network Open.

Researchers sought to establish whether the same patients admitted with the same diagnosis––either heart failure or pneumonia, in this case––at two different hospitals would have different costs based on the hospital’s Medicare payment profile. CMS payments are premised on characterizing the variation of the hospitals and the post-acute care system, but there are concerns that payment variations actually reflect differences in case mix.

Medicare beneficiaries had higher costs when they were admitted to hospitals with the highest payment profiles compared to hospitals with the lowest payment profiles for the same diagnosis. That means that hospital payments are somewhat linked to the hospital regardless of patient characteristics, researchers concluded in the study.

However outcomes, including mortality rates, were similar between the two hospital profiles.

The researchers looked at the variations of hospital payments and analyzed CMS discharge data of Medicare patients with a principle diagnosis of heart failure or pneumonia between July 1, 2013, and June 30, 2016. They further sampled 1,615 patients with heart failure and 708 with pneumonia.

Among heart failure patients, the 30-day episode payment for hospitalization at the lowest-payment hospitals was $2,118 lower than at the highest-payment hospitals. For pneumonia, the difference was $2,907.

According to the researchers, their study was unique because they were able to isolate patients admitted to two different hospitals for the same diagnosis. That meant many social determinants of health, behaviors, social context and demographics were constant.

“We were able to isolate a hospital factor in the variation in payments to hospitals,” wrote first author Harlan M. Krumholz, MD, SM, of the Yale School of Medicine, and colleagues. “This finding suggests that the variation in RSP is, in part, associated with factors that are unrelated to the patients.”

Furthermore, the care and outcomes across the two hospital profiles was similar.

“We found marked differences in payment depending on which hospital the patient received care at,” Krumholz et al. wrote. “No evidence was found that lower-payment hospitals had higher mortality rates.”

The higher-cost hospitals for heart failure did have more procedures, they noted, but the difference in the use of procedures was small between the two hospital groups. The lower-cost hospitals could also serve as a benchmark to help lower healthcare costs, the authors concluded.