The seven factors that could predict readmissions

Model could be used to target and prevent patients from being readmitted

Researchers from Brigham and Women's Hospital have created a risk prediction model to assess which hospitalized patients are at the greatest risk of being readmitted for preventable reasons, according to a study published in JAMA Internal Medicine.

To create the model, researchers used data collected between 2009 and 2010 on about 9,200 Brigham and Women's patients. Researchers then studied the characteristics of the patients who were readmitted to the hospital or to one of its three partners for avoidable reasons within 30 days.

Of those patient characteristics—such as age and whether the patient required a caretaker—the researchers identified the seven that best predicted which patients would be readmitted, including:

  • Number of admissions within the prior year;
  • Length of hospital stay;
  • Sodium levels at discharge;
  • Number of procedures during first admission;
  • Whether the patient was discharged from the oncology department;
  • Hemoglobin levels at discharge; and
  • Whether the admission was elective or non-elective.

Each predictor is assigned a point value, with the more heavily linked predictors getting two points or more, and the rest getting one. For example, a patient who stays in the hospital for five days or more after being admitted is assigned two points. A patient with seven points or more when discharged is at an 18% risk of being readmitted within 30 days for a preventable condition, according to Medpage Today.

"The strength of this model is its simplicity," lead author Jacques Donzé says, adding that a physician can "easily run through [the model] at a patient's bedside prior to discharge."

The study adds to a growing body of research seeking to identify how hospitals can reduce readmissions and therefore, limit the Medicare reimbursement cuts that now accompany high readmissions.

"If a patient is determined to be at high-risk for readmission, a return trip to the hospital could be prevented by providing additional interventions such as a home visit by a nurse or pharmacist consultation," Donzé says. Researchers are now testing the model's accuracy at several facilities worldwide (Donzé et al., JAMA Internal Medicine, 3/25; Phend, Medpage Today, 3/25; Evans, Modern Healthcare, 3/25 [subscription required]; Brigham and Women's Hospital release, 3/25).

How do you compare on readmissions?

The interactive table below, produced by the Advisory Board's Data and Analytics Group, displays the previously published and corrected adjustment factors for CMS’s Readmissions Reduction Program. The estimated revenue impact applies the corrected Adjustment Factor to an estimate of total FY13 revenue.

Next in the Daily Briefing

Truce in Minnesota: Hospitals, nurses agree on staffing study

Read now

You May Also Like

ABC Homepage

Advisory Board

Analytics and Implementation Support

Health Care Facility Quickview

Main section