The impact of length of stay and number of skilled nursing visits on hospitalization following home health discharge

Background: Medicare costs due to hospitalization exceeded $553 billion in 2009 with unplanned hospitalization expenditures estimated to be $12 billion a year. Since the implementation of the Home Health Prospective Payment System, Medicare-reimbursed home health (HH) beneficiaries have experienced dramatic reductions in the length of stay (LOS) and number of skilled nursing visits (SNV). Aims: The purpose of this study was to investigate the association LOS in HH and number of SNV with the occurrence of and time to hospitalization after discharge from HH services among Medicare reimbursed HH beneficiaries. Methods: Guided by the Quality Health Outcomes Model, the methodology included propensity score analysis, logistic regression and Cox proportional hazards regression analysis of five CMS-owned national data sets from 2009. Samples included two independent randomized samples (n=4,500) in addition to two independent randomized validation samples (n=4,500) to confirm results. Results: Between 16.3% to 17.5% were hospitalized within 90 days of discharge from HH. Validated study results show a LOS of 42 days or more in HH has significant advantages. Length of stay findings were statistically significant with more hospitalizations in the short LOS group (≤21 days) compared to the long LOS group (≥42 days) (sample 1: 18.1% vs. 14.3%; sample 2: 19.3% vs. 15.1%). A short LOS in HH and having high risk diagnoses were associated with the increased occurrence of hospitalization and reduced time to first hospitalization and death. Between 13% to 14% of beneficiaries died following HH services. More deaths occurred in the short LOS groups compared to the long LOS groups (sample 1: 17.5% vs. 9.5%; sample 2: 13.3% vs. 9.4%). Reduced time to death was also associated with the presence of high risk diagnoses, increased age, reduced ability to ambulate, and dyspnea. Being female, Hispanic, and having an inpatient stay prior to HH was associated with increased time to death. Number of SNV was not associated with the occurrence of or time to first hospitalization or the time to death. Conclusion: Additional research is needed to clarify the most effective way to structure HH services and determine readiness for discharge to maximize benefits and reduce hospitalization