PHILADELPHIA – Putting a uniform, brief cap on the number of days that psychiatric patients can remain hospitalized might have the unintended consequence of significantly boosting the rate of short-term hospital readmissions, according to results from a study of more than 12,000 patients at one U.S. hospital.
Psychiatric patients hospitalized for 4 days or fewer had a statistically significant 25% increased rate of readmission in the following 30 days in an analysis that adjusted for many possible demographic and clinical confounders, Dr. John W. Goethe said at the annual meeting of the American Psychiatric Association.
At a time when the Centers for Medicare and Medicaid Services (CMS) is poised to start penalizing hospitals with outlier short-term readmission rates, the finding highlights the complex constellation of factors that affect 30-day readmission rates. The finding also underscores the potent, rock-and-hard-place one-two punch that hospitals and clinicians will soon face when treating psychiatric patients: The pressure from payers to quickly move inpatients out the door and the penalty that CMS will levy if these patients wind up rehospitalized less than a month later.
"Readmission is a complex problem and an extremely crude proxy for the many things that can make a patient’s condition deteriorate," said Dr. Goethe, a psychiatrist and director of the center for research and education at Hartford (Conn.) Hospital.
"Readmission rates are clearly related to length of stay," and trying to assess quality of hospital care with the metric of readmission loses sight of factors contributing to readmissions that can be out of a hospital’s control, he said in an interview.
Mitchel L. Zoler/IMNG Medical Media
Dr. John W. Goethe
A length-of-stay cap fixed at something like 4 days "does not take into account a patient’s particular circumstances" and might mandate a hospitalization duration that is "inadequate to address what brought the patient into the hospital," he said. The important role that length of stay plays in readmissions should be "taken into account by policy makers who are planning financial penalties for hospitals with what [they consider] inappropriate readmission rates," Dr. Goethe said.
His study included 12,830 patients with a psychiatric diagnosis who were admitted to Hartford Hospital during January 2002-May 2008. The majority of patients, 71%, were 18-60 years old; 14% were age 17 or younger and 15% were age 61 or older. Patient records showed a length of stay of 4 days or fewer for 13% of pediatric patients, 33% of adults, and 9% of older patients. The rate of readmission within 30 days was 7% for pediatric patients, 10% for adults, and 9% among older patients.
The researchers then performed a series of proportional-hazard regression analyses that calculated the relationship between a 4-day-or-less length of stay and readmission rates. They adjusted for many baseline factors, including age, sex, race and ethnicity, type of psychiatric diagnosis, types of treatment received, and prior hospitalization history. The result showed that patients hospitalized for 5 days or more had a statistically significant 20% reduced rate of 30-day hospital readmission, compared with patients hospitalized for 4 days or fewer. Expressed conversely, patients hospitalized for 4 days or fewer had a 25% higher rate of 30-day readmission than those hospitalized for longer periods.
The analyses identified six other variables that also had a significant, independent effect on 30-day readmission rate: female sex, schizophrenia, and attention-deficit disorder linked with a reduced readmission rate, while benzodiazepine treatment, dementia, and conduct disorder all linked with an increased rate.
Results from a companion study identified other factors that significantly linked with 30-day readmission. Patients who previously had been hospitalized within 30 days, 90 days, or 2 years of the index hospitalization all faced a significantly increased risk for 30-day rehospitalization following the index admission. And patients treated with six or more psychotropic drugs had a significantly increased risk for 30-day readmission, compared with patients who received five or fewer psychotropic drugs.
"To our knowledge, this is the largest patient sample used to address this question" for psychiatric patients, Dr. Goethe said. He cautioned that the analysis did not take into account several other factors that could also affect readmission rates, such as severity of illness, socioeconomic status, social support, quality of life, and adherence to treatment.
Dr. Goethe said he had no disclosures, but his hospital’s research program has received grants from several drug companies.