Evaluation of risk-adjusted home time after hospitalization for heart failure as a potential hospital performance metric

Ambarish Pandey, Neil Keshvani, Mary S. Vaughan-Sarrazin, Yubo Gao, Gregg C. Fonarow, Clyde Yancy, Saket Girotra

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

IMPORTANCE Thirty-day home time, defined as time spent alive and out of a hospital or facility, is a novel, patient-centered performance metric that incorporates readmission and mortality. OBJECTIVES To characterize risk-adjusted 30-day home time in patients discharged with heart failure (HF) as a hospital-level quality metric and evaluate its association with the 30-day risk-standardized readmission rate (RSRR), 30-day risk-standardized mortality rate (RSMR), and 1-year RSMR. DESIGN, SETTING, AND PARTICIPANTS This hospital-level cohort study retrospectively analyzed 100% of Medicare claims data from 2 968 341 patients from 3134 facilities from January 1, 2012, to November 30, 2017. EXPOSURES Home time, defined as time spent alive and out of a short-term hospital, skilled nursing facility, or intermediate/long-term facility 30 days after discharge. MAIN OUTCOMES AND MEASURES For each hospital, a risk-adjusted 30-day home time for HF was calculated similar to the Centers for Medicare & Medicaid Services risk-adjustment models for 30-day RSRR and RSMR. Hospitals were categorized into quartiles (lowest to highest risk-adjusted home time). The correlations between hospital rates of risk-adjusted 30-day home time and 30-day RSRR, 30-day RSMR, and 1-year RSMR were estimated using the Pearson correlation coefficient. Distribution of days lost from a perfect 30-day home time were calculated. Reclassification of hospital performance using 30-day home time vs 30-day RSRR was also evaluated. RESULTS Overall, 2 968 341 patients (mean [SD] age, 81.0 [8.3] years; 53.6% female) from 3134 hospitals were included in this study. The median hospital risk-adjusted 30-day home time for patients with HF was 21.77 days (range, 8.22-28.41 days). Hospitals in the highest quartile of risk-adjusted 30-day home time (best-performing hospitals) were larger (mean [SD] number of beds, 285 [275]), with a higher volume of patients with HF (median, 797 patients; interquartile range, 395-1484) and were more likely academic hospitals (59.9%) with availability of cardiac surgery (51.1%) and cardiac rehabilitation (68.8%). A total of 72% of home time lost was attributable to stays in an intermediate- or long-term care facility (mean [SD], 2.65 [6.44] days) or skilled nursing facility (mean [SD], 3.96 [9.04] days), 13% was attributable to short-term readmissions (mean [SD], 1.25 [3.25] days), and 15% was attributable to death (mean [SD], 1.37 [6.04] days). Among 30-day outcomes, the 30-day RSRR and 30-day RSMR decreased in a graded fashion across increasing 30-day home time categories (correlation coefficients: 30-day RSRR and 30-day home time, −0.23, P < .001; 30-day RSMR and 30-day home time, −0.31, P < .001). Similar patterns of association were also noted for 1-year RSMR and 30-day home time (correlation coefficient, −0.35, P < .001). Thirty-day home time meaningfully reclassified hospital performance in 30% of the hospitals compared with 30-day RSRR and in 25% of hospitals compared with 30-day RSMR. CONCLUSIONS AND RELEVANCE In this study, 30-day home time among patients discharged after a hospitalization for HF was objectively assessed as a hospital-level quality metric using Medicare claims data and was associated with readmission and mortality outcomes and with reclassification of hospital performance compared with 30-day RSRR and 30-day RSMR.

Original languageEnglish (US)
Pages (from-to)169-176
Number of pages8
JournalJAMA Cardiology
Volume6
Issue number2
DOIs
StatePublished - Feb 2021

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine

Fingerprint

Dive into the research topics of 'Evaluation of risk-adjusted home time after hospitalization for heart failure as a potential hospital performance metric'. Together they form a unique fingerprint.

Cite this