In this issue of Journal of Neuroscience Nursing, Peacock and colleagues1 report the findings of a retrospective study that evaluated the ability of a readmission risk calculator to predict 30-day readmissions in a neurocritical care population. This article is an excellent example of researchers exploring ways to predict and reduce hospital readmissions.
Background
Reducing preventable readmission is a quality of care measure for the Centers for Medicare and Medicaid Services as well as third-party payers. Currently, there are financial penalties that target 30-day readmissions. The range of 30-day readmission rates for neurological patients have been reported to be between 4.2% and 7.4% among spine surgery patients, 11.5% for mixed neurosurgical patients, 14.4% in stroke patients, and 11.4% for subarachnoid hemorrhage patients.
There are several models that predict general patient population readmission risks. As an example, LACE (Length of stay, Acuity of admission, Charlson comorbidity index, and number of Emergency visits in preceding 6 months) can quantify the risk of death or unplanned readmissions and is used throughout the world to predict 30-day readmissions. However, there are no statistical models to predict readmissions among neurology patients.
Purpose
The purpose of the study was to evaluate a readmission risk calculator to predict 30-day readmission by retrospectively examining discharges from neurocritical care units. There were 2 aims: 1) compare differences in median readmission score between neurocritical care unit patients who were readmitted within 30 days and those who were not readmitted, and 2) evaluate the hospital-wide readmission risk tool and its ability to predict 30-day readmissions in neurology patient populations and describe the characteristics of patients who were admitted compared with those who were not.
Methods
A retrospective review of 340 adult patients admitted both emergently and electively to the Mayo Clinic Florida, an academic tertiary care hospital, was conducted between February 2014 and February 2015.
Data Analysis
Median values of continuous variables were reported with their respective interquartile ranges (IQRs) and counts and percentages of categorical variables. Univariate analysis was performed to determine the associations of independent variables with readmission after initial discharge, and Fisher and Wilcoxon rank sum tests were used as nonparametric tests.
Results
Of 340 patients, most were women (n = 180, 53%) and white (n = 254, 75%), with a median age of 65 years (IQR, 51-77 years). 63% were admitted because of an emergency, and the median length of stay was 5 days (IQR, 2-10 days). The overall median initial readmission score was 8 (IQR, 4-10), and most patients were given a low risk for readmission (n = 224, 66%). The most prevalent diagnosis at admission was acute ischemic stroke (26%), followed by neoplasm (14%) and hemorrhagic stroke (12%). The analysis excluded patients who died (n = 57) and patients who did not have a readmission score (n = 4).
The median time between readmission was 9 days (IQR, 2-18 days), and the median length of stay after readmission was 4 days (IQR, 3-6 days). 97% of readmissions were emergent; 1 patient was readmitted for elective surgical procedure. The median new readmission score was 12 (IQR, 9-15), with 27% having a low risk score compared with 65.8% on their first admission. The findings suggest that the current risk prediction score was inaccurate in predicting readmissions in neurocritical intensive care units.
Implications
Preventing hospital readmissions is a Medicare quality benchmark and a challenge for hospitals across the United States. Nurses, as educators, are in the position to perform interventions to prevent readmission at multiple points during a hospitalization. Developing a readmission risk calculator specific to neuroscience patients could alert the nurses and other healthcare providers about potential high-risk patients for readmission.
Reference