A health risk score calculated automatically using routine data from hospital electronic medical records systems can identify patients at high risk of unplanned hospital readmission, according to a study.
The score, called the Rothman Index, may provide a useful tool for lowering the rate of avoidable repeat hospitalizations, according to the report in the September issue of the journal Medical Care.
Clinicians can use the Rothman Index to target hospital programs and supports to patients at highest risk of readmission, wrote Elizabeth Bradley, PhD, of the Yale School of Public Health, and colleagues.
The Rothman Index software uses information from the hospital EMR system to provide a continuously updated score indicating the likelihood of death or readmission within 30 days, according to a news release.
The score is calculated automatically using standard data on each patients vital signs, routine nursing assessments, skin condition, heart rhythms and laboratory tests. Lower Rothman Index scores (from a maximum of 100) indicate a higher risk of readmission.
Bradley and colleagues evaluated the ability of the Rothman Index to predict hospital readmission, based on data from more than 2,700 patients hospitalized during 2011. (During this time, physicians and nurses did not have access to the Rothman Index scores.) Of the patients in the study sample, 16% had an unplanned readmission within 30 days after hospital discharge.
The Rothman Index was strongly associated with the risk of unplanned readmission, the researchers found. For patients in the highest-risk category a score of less than 70 readmission risk was one in five. By comparison, for those in the lowest-risk category a score of 80 or higher the risk was about one in 10.
After adjustment for other factors, patients in the highest- versus lowest-risk category were more than 2.5 times as likely to be readmitted within 30 days of discharge. The Rothman Index predicted readmission across various diagnoses and medical specialties.
About 20% of Medicare patients are readmitted to the hospital within 30 days, at an estimated cost of $17 billion per year, according to background information in the study. Medicare has begun reducing payments by up to 2% for hospitals with high readmission rates.
The Rothman Index is especially valuable because it is calculated automatically from routine data, requiring no manual input from busy healthcare professionals, according to the news release. It was developed by brothers Michael and Steven Rothman in memory of their mother, who died unexpectedly four days after hospital discharge following heart surgery.
During their mothers illness, the Rothman brothers were surprised to learn that the hospitals EMR system did not generate summary patient health measures that might have alerted doctors to unrecognized complications that were present at discharge. While neither brother had medical training, both were computer scientists with experience in applying complex analytical tools to massive electronic databases.
The new study suggests that the Rothman Index could help reduce rates of unplanned readmission, identifying a group of patients two to three times more likely to be readmitted. Implemented into daily care, the score could provide a practical way for clinicians to identify patients who might be at higher risk for unplanned readmission and intervene specifically for these patients to try to avert unplanned readmission, Bradley and coauthors wrote.
We know the Rothman Index is associated with readmissions, but we do not know if it can be used to improve decision making at the bedside in terms of when patients are discharged, Bradley added in the news release. We also dont know if physicians would benefit from using it as part of determining what kinds of added supports at home and in the community might be arranged at discharge.
Answering these questions will determine if the Rothman Index can be used prospectively by clinicians to reduce readmissions and adverse events post-hospitalization.
Study abstract: http://journals.lww.com/lww-medicalcare/Abstract/2013/09000/Identifying_Patients_at_Increased_Risk_for.2.aspx.