An analysis of eight surgical subspecialties at Johns Hopkins hospital found that more than 80 percent of the variability in 30-day readmissions was due to individual patient factors. The overall readmission rate was 13.2 percent, including 9.6 percent of patients who underwent cardiac surgery.

Lead researcher Faiz Gani, MBBS, and colleagues collected data from all patients discharged from Johns Hopkins hospital after undergoing surgery from Jan. 1, 2009, through Dec. 31, 2013. Results were published online in JAMA Surgery on Aug. 5.

As part of the Patient Protection and Affordable Care Act, the Centers for Medicare & Medicaid Services (CMS) started the hospital readmission reductions program in 2012. CMS reduces payments to hospitals that have excessive 30-day readmissions for MI, pneumonia, heart failure and total hip and knee arthroplasty.

Surgical procedures may also be included in the coming years, according to Timothy M. Pawlik, MD, MPH, PhD, a study author and chief in Johns Hopkins’s division of surgical oncology.

“We’re fully aware that at some point CMS is going to start to penalize us for what they deem as unnecessary readmissions,” Pawlik told Cardiovascular Business. “What we wanted to look at was how do readmissions vary by provider, by specialty and by patient case mix. One of the underlying hypotheses is that one size may not fit all. If CMS comes up with some type of metric whereby a certain proportion of readmissions is going to be deemed to be excessive, that may be different for different subspecialties.”

The researchers analyzed 22,559 patients who underwent one of the following eight surgeries: cardiac; gastrointestinal; trauma; hepatopancreaticobiliary; breast, melanoma or endocrine; thoracic; transplant; and vascular.

The 30-day readmissions rates varied considerably and ranged from 2.1 percent following breast, melanoma or endocrine surgery to 24.8 percent following transplant surgery. After the researchers adjusted for patient- and surgeon-level variables, the factors associated with readmission included African American race/ethnicity, increasing comorbidity, postoperative complications and an extended length of hospital stay.

Complication rates varied depending on the surgical subspecialty and ranged from 2.1 percent for breast, melanoma or endocrine surgery to 37.0 percent for cardiac surgery. Further, the median length of stay in the hospital ranged from one day following breast, melanoma or endocrine surgery to eight days following cardiac surgery.

The researchers found 82.8 percent of the variation in readmissions was due to patient-related factors. In addition, the type of surgical subspecialty accounted for 14.5 percent of the variation, while surgeon-level factors accounted for 2.8 percent.

Pawlik said there could be issues if CMS is inflexible about its readmissions program and institutes common thresholds across different surgical subspecialties and hospitals. For instance, hospitals that serve primarily African Americans and people with preoperative comorbidities are likely to have higher readmissions rates solely based on patient-related factors that are out of the hospitals’ control. Pawlik added that certain patients might not receive treatment because doctors and hospitals would worry about getting reductions in pay.

“Physicians may be deincentivized to take on patients who are older and have more comorbidities and are at higher risk if we’re going to be penalized for that,” he said.

Pawlik said the findings were most generalizable to other academic medical centers and may not be as relevant for community hospitals that do not have as many surgical subspecialties. However, he said previous research using national, publicly available claims databases have found similar results.