Paper
Wednesday, 19 July 2006
This presentation is part of : Acute Care Initiatives
Predicting Early Hospital Readmission for a Cohort of Adult Inpatients Using the Probability of Repeated Admission (PRA) Instrument
Nancy L. Novotny, MS, RN, Department of Medicine, University of Illinois, Peoria, IL, USA
Learning Objective #1: a) list potential benefits of accurate prediction of medical inpatients who will experience early readmission, and b) describe earlier attempts and limitations of such prediction.
Learning Objective #2: Verbally summarize: a) limitations and possible application of the study's findings, and b) useful "next steps" for conduct of follow-up research.
Background:  Up to 25% of adults experience early readmission.  Pre-discharge identification of patients likely to be readmitted could help target interventions that reduce readmissions, costs, and exposures to hospital-associated risks.  The Pra score, a valid predictor of elderly hospital readmissions within 4 years, was used to estimate prediction of adults’ readmission within 1 month.
 
Method:  A cohort of internal medicine service adults in a Midwestern tertiary medical center with life expectancy > 6 months was interviewed and provided releases of information June-July 2005.  Before discharge trained assistants abstracted records and derived Pra scores using age, gender, admissions and doctor/clinic visits in the last year, subjective health status, caretaker availability, and diabetes or coronary artery disease (CAD). The 75th percentile was used to classify patients likely (≥ .48) versus unlikely (<.48) to be readmitted.  Local hospital records were queried for subjects’ early readmissions. 
 
Results: Of 156 consecutive patients, 35 were discharged before screened, while 60% (72/121) were eligible and consented. The sample was 56% (40/72) female; with median age of 53 (18 - 97), admissions = 1 (1 - 14), and doctor/clinic visits = 10 (0 - 100).  Caretaker available for 90% (67/72), 33% (24/72) had diabetes, and 18% (13/72) had CAD.  Health status was poor/fair 60%, good/very good 37%, and excellent 3%. Readmissions occurred with 11%.  Predictive values included: sensitivity 75% (6/8), 95% Confidence Interval (CI): 0.45 - 1.05; specificity 69% (44/64), CI: 0.57 - 0.80; positive predictive value 23% (6/26), CI: 0.07 - 0.39; negative predictive value 96% (44/46), CI: 0.88 - 1.04; and odds ratio 6.143, CI: 1.28 - 28.75.
 
Conclusions:  Multi-institutional replication with larger samples is needed to confirm Pra’s ability.  Within this heterogeneous sample, readmission of high (vs low) Pra patients was 6 times more likely.  Pra’s promising predictive ability may add valuable discharge planning information.

See more of Acute Care Initiatives
See more of The 17th International Nursing Research Congress Focusing on Evidence-Based Practice (19-22 July 2006)