Mohammad Taha1, Aroop Pal1, Jonathan D Mahnken2, Sally K Rigler1. 1. Division of General and Geriatric Medicine, University of Kansas School of Medicine, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA. 2. Department of Biostatistics, University of Kansas School of Medicine, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA.
Abstract
OBJECTIVE: To create a simple readmission risk-prediction tool that can be generated easily at the bedside by physicians, nurses, care coordinators and discharge planners. DESIGN: Retrospective cohort study. SETTING: Tertiary academic medical center. PARTICIPANTS: Inpatients aged 18 and older on general internal medicine services. MEASURES: Predictor variables included age, prior hospitalization, high-risk diagnoses, high-risk medications, polypharmacy, depression, use of palliative care and a cumulative score summing these factors (readmission risk score-RRS). The main outcome measure was 30-day readmission. Predictive values were calculated. RESULTS: Readmission increased linearly from 4.9% of those whose RRS score was 0-37.5% of those with highest risk scores (P = 0.0002). We derived a simple formula for readmission risk as 8 and 4% more for each additional readmission risk factor. The positive predictive value for RRS >0 was low, while the negative predictive value for this cutoff was 95%. CONCLUSIONS: An easily calculated 7-point score can be used to estimate readmission risk. This tool may be particularly useful for identifying lower risk patients who may not require intensive intervention, thus aiding in appropriate targeting of resources.
OBJECTIVE: To create a simple readmission risk-prediction tool that can be generated easily at the bedside by physicians, nurses, care coordinators and discharge planners. DESIGN: Retrospective cohort study. SETTING: Tertiary academic medical center. PARTICIPANTS: Inpatients aged 18 and older on general internal medicine services. MEASURES: Predictor variables included age, prior hospitalization, high-risk diagnoses, high-risk medications, polypharmacy, depression, use of palliative care and a cumulative score summing these factors (readmission risk score-RRS). The main outcome measure was 30-day readmission. Predictive values were calculated. RESULTS: Readmission increased linearly from 4.9% of those whose RRS score was 0-37.5% of those with highest risk scores (P = 0.0002). We derived a simple formula for readmission risk as 8 and 4% more for each additional readmission risk factor. The positive predictive value for RRS >0 was low, while the negative predictive value for this cutoff was 95%. CONCLUSIONS: An easily calculated 7-point score can be used to estimate readmission risk. This tool may be particularly useful for identifying lower risk patients who may not require intensive intervention, thus aiding in appropriate targeting of resources.
Authors: Stephen L Jones; Ohbet Cheon; Joanna-Grace Mayo Manzano; Anne K Park; Heather Y Lin; Josiah K Halm; Juha Baek; Edward A Graviss; Duc T Nguyen; Bita A Kash; Robert A Phillips Journal: Am J Med Qual Date: 2021-12-20 Impact factor: 1.200
Authors: David Picker; Kevin Heard; Thomas C Bailey; Nathan R Martin; Gina N LaRossa; Marin H Kollef Journal: BMC Health Serv Res Date: 2015-07-23 Impact factor: 2.655
Authors: Lian Leng Low; Nan Liu; Sijia Wang; Julian Thumboo; Marcus Eng Hock Ong; Kheng Hock Lee Journal: BMJ Open Date: 2016-10-14 Impact factor: 2.692
Authors: Adam Andruska; Scott T Micek; Yuichiro Shindo; Nicholas Hampton; Brian Colona; Sandra McCormick; Marin H Kollef Journal: Chest Date: 2015-07 Impact factor: 9.410
Authors: Lian Leng Low; Nan Liu; Sijia Wang; Julian Thumboo; Marcus Eng Hock Ong; Kheng Hock Lee Journal: PLoS One Date: 2016-12-09 Impact factor: 3.240