Topics
More on Quality and Safety

Spotting preventable readmissions

Scorecard detects one-in-four high-risk patients before discharge

With hospitals nationwide feeling the burden of Medicare penalties for avoidable 30-day readmissions, C-suite managers have been frantically searching for ways to pinpoint these high-risk patients early on. A recent analysis suggests a low-cost and simple scorecard that detects one out of four of these patients.

With the help of an algorithm that was based on administrative data (SQLape), investigators from Harvard Medical School, looked at potentially avoidable 30-day readmissions at three hospitals in Boston’s Partners Healthcare network. They used the data to formulate a scoring system that included 10 risk factors that hospitals may want to use to spot high-risk (and expensive) patients.

[See also: Hospitals' readmission penalties revised by Medicare]

Jacques Donze, MD, from Harvard Medical School and Partners Healthcare, and his associates initially looked at more than 10,000 eligible discharged patients, nearly 2,400 of which were readmitted within 30 days. Among those 2,400 patients, they identified 879 readmissions that were potentially avoidable. They found seven markers that helped predict readmission:

  • Low hemoglobin at discharge
  • Discharge from a cancer unit
  • Low sodium at discharge
  • The occurrence of any ICD-9 coded procedure during the hospital stay
  • Non-elective admission
  • One or more hospital stays during the previous year
  • A length of stay of five or more days

“This easy-to-use model enables physicians to prospectively identify approximately 27 percent of the patients as high risk of having a potentially avoidable readmission and would allow targeting intensive transitional care interventions …,” according the JAMA Internal Medicine report on the analysis.

Donze pointed out during a recent interview with Healthcare Finance News that his team wanted to create a scoring system that was not only simple but that could be used before patients were discharged. Existing scorecards have used “billing data and ICD-9 that are only available a few weeks after the patient is discharged …,” Donze said. To make a score useful for readmission assessment, it has to be available before the discharge of the patient, Donze emphasized. 

There are a number of methods that may be used to reduce the risk of readmission in these high-risk patients, noted Jeffrey L. Schnipper, MD, senior author of the JAMA paper, but they’re expensive and resource intensive.

“You probably only want to do them when you know you’re going to get the biggest bang for your buck,” he said. And that’s the value of a scoring system, he said – it gives hospitals the information needed so that they are resorting to use these expensive methods only on the highest-risk patients.