WASHINGTON – More than one in six patients who undergo a lower extremity arterial endovascular or surgical procedure are readmitted within 30 days, according to a large national study.
The annual total cost of these early readmissions is high, in excess of $360 million. But because there turned out to be surprisingly little difference in readmission rates between hospitals, 30-day readmissions may not be a rational quality measure on which to base institutional reimbursement or withholding of payment for peripheral arterial interventions, Eric A. Secemsky, MD, said at the annual meeting of the American College of Cardiology.
Forty-seven percent of patients had an endovascular procedure, 42% had surgery, and the remainder had hybrid procedures in which both endovascular and surgical interventions took place during the same admission. Patients with hybrid procedures contributed data to both treatment groups.
In-hospital mortality occurred in 2.5% of patients.
Of the patients who survived to discharge, 21,589, or 17.4%, were readmitted within 30 days. The early readmission rate was higher following endovascular procedures, at 18.7%, than the 16.1% rate in the surgical group. The average cost of a readmission was $15,876. Death during readmission occurred in 4.2% of patients.
The median rate ratio – a measure of the amount of variance in readmission rates between hospitals – was 1.12. That’s a low figure.
“If the median rate ratio is lower, like here, it says there’s not a lot of interhospital variability across the country. So overall this burden seems to be pretty uniform across the institutions included in our analysis,” Dr. Secemsky explained.
This observation drew the attention of session comoderator Naomi M. Hamburg, MD.
“It’s interesting that you didn’t see a lot of heterogeneity across hospitals, because we often think of readmissions as a potentially modifiable quality metric. Do you think it’s modifiable, or is this just the nature of the disease?” asked Dr. Hamburg of Boston Medical Center.
It’s the disease process, Dr. Secemsky replied.
“We were surprised by the lack of hospital variation,” he added. None of the institutional characteristics examined, including teaching hospital status, bed size, and procedural volume, had a significant impact on readmission rates.
But that doesn’t mean there aren’t opportunities to whittle down those readmissions, according to Dr. Secemsky.
He noted that the high readmission rates were driven by procedural complications such as graft or stent failure. Indeed, procedural complications accounted for 29% of all early readmissions. The procedural complication rate was about 20% following endovascular procedures and 39% after surgery. It’s likely that identification and implementation of best practices could trim those high rates. Unfortunately, however, the nationwide database relies upon ICD-9 codes, which don’t provide the granular level of detail required to home in on specific best practices. That will require further studies, according to Dr. Secemsky.
A distant second on the list of causes of early readmission was peripheral atherosclerosis, meaning persistent claudication or rest pain. This accounted for 8.8% of readmissions. Rounding out the top five causes of readmission were sepsis, which was the reason for 6.7% of readmissions; diabetes with complications, at 4.7%; and heart failure, at 4.6%.
The strongest predictors of readmission included having renal disease at baseline, Medicare rather than private insurance, and discharge to a subacute nursing facility or home with home care.
Dr. Hamburg commented that a focus on reducing readmissions for sepsis as well as for skin and soft tissue infections, which accounted for 2.1% of 30-day hospitalizations, could be fruitful.
Dr. Secemsky reported having no financial conflicts regarding his study.