Program reduces readmissions by nearly 10 percent among high-risk Medicare patients

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon

A program implemented at an urban academic medical center in Connecticut reduced 30-day readmissions by 9.3 percent among high-risk discharge patients who were enrolled in Medicare. The rate fell short of the goal of reducing all-cause readmissions by 20 percent.

Lead researcher Grace Y. Jenq, MD, of the Yale School of Medicine and Yale-New Haven Hospital, and colleagues published their findings online in  JAMA Internal Medicine on April 11.

“Our analysis revealed a fairly consistent and sustained but small, beneficial effect of the intervention on the target population as a whole,” the researchers wrote.

The Centers for Medicare & Medicaid Services (CMS) funded the program through its community-based care transitions program, which reimburses participants for transitional care services. Hospitals are eligible to participate if they have above average readmissions rates and partner with community-based organizations.

This program, which was known as Co-STARR (Greater New Haven Coalition for Safe Transitions and Readmission Reductions), began in May 2012 as a partnership among Yale-New Haven Hospital, the Hospital of Saint Raphael’s and the Agency on Aging of South Central Connecticut.

The researchers said that senior executives made readmission reduction a hospital-wide quality improvement priority and contributed financial support when the CMS funds were delayed.

Jenq, the medical director of inpatient services, assumed leadership of the program on a full-time basis during the intervention period.

The researchers enrolled high-risk patients using a modified risk assessment tool. They considered risk factors of 10 or more routine medications when admitted or the use of high-risk medications (insulin, anticoagulants, oral hypoglycemic agents, dual antiplatelet therapy, digoxin or narcotics); principal diagnosis of cancer, chronic obstructive pulmonary disease, stroke, heart failure or diabetes; depressive symptoms or diagnosis of depression; physical limitations; poor patient support; need for palliative care; or previous nonelective 30-day readmission in the previous six months.

Transitional care consultants followed up with patients who were discharged to their homes. The transitional care consultants were social workers from the Area Agency on Aging who were knowledgeable about local resources. They each cared for approximately 40 patients and met with the patients and/or family members during hospitalization and also performed bedside assessments of patients’ social, cognitive, functional status, and postdischarge needs.

After patients were discharged, the consultants made follow-up telephone calls at least once per week for 30 days following discharge. They also made home visits when necessary.

The researchers analyzed 10,621 high-risk patients with Medicare who were at least 65 years old and did not die during hospitalization. The mean age was 79.7 years old.

The control group included patients who were older than 54 years old who were discharged to eligible locations, lived in the target zip codes and did not have Medicare insurance.

The adjusted 30-day readmission rate decreased from 21.5 percent to 19.5 percent in the target population and from 21.1 percent to 21.0 percent in the control group, which represented a relative reduction of 9.3 percent. The number needed to treat (NNT) to avoid 1 readmission was 50.

A difference-in-differences analysis found that the odds of readmission in the target population decreased significantly more than the odds of readmission in the control population during the intervention period.

The researchers mentioned a few limitations of the study, including that they measured outcomes in the entire target population even though only 58.3 percent were enrolled in the program. They also noted that the results might not be generalizable to other hospitals and that the observational trial could not establish causality and could possibly have unmeasured confounders.