The Association of Nursing Home Characteristics and Quality with Adverse Events After a Hospitalization

Terry S Field; Hassan Fouayzi; Sybil Crawford; Alok Kapoor; Cassandra Saphirak; Steven M Handler; Kimberly Fisher; Florence Johnson; Ann Spenard; Ning Zhang; Jerry H Gurwitz

doi:10.1016/j.jamda.2021.02.027

The Association of Nursing Home Characteristics and Quality with Adverse Events After a Hospitalization

J Am Med Dir Assoc. 2021 Oct;22(10):2196-2200. doi: 10.1016/j.jamda.2021.02.027. Epub 2021 Mar 27.

Authors

Affiliations

¹ University of Massachusetts Medical School, Worcester, MA, USA; Meyers Primary Care Institute, Worcester, MA, USA. Electronic address: terry.field@umassmed.edu.
² University of Massachusetts Medical School, Worcester, MA, USA; Meyers Primary Care Institute, Worcester, MA, USA.
³ University of Massachusetts Medical School, Worcester, MA, USA.
⁴ Meyers Primary Care Institute, Worcester, MA, USA.
⁵ University of Pittsburgh, Pittsburgh, PA, USA.
⁶ Qualidigm, Wethersfield, CT, USA.
⁷ University of Massachusetts, Amherst, MA, USA.

PMID: 33785310
DOI: 10.1016/j.jamda.2021.02.027

Abstract

Background/objectives: We previously found high rates of adverse events (AEs) for long-stay nursing home residents who return to the facility after a hospitalization. Further evidence about the association of AEs with aspects of the facilities and their quality may support quality improvement efforts directed at reducing risk.

Design: Prospective cohort analysis.

Setting and participants: 32 nursing homes in the New England states. A total of 555 long-stay residents contributed 762 returns from hospitalizations.

Methods: We measured the association between AEs developing in the 45 days following discharge back to long-term care and characteristics of the nursing homes including bed size, ownership, 5-star quality ratings, registered nurse and nursing assistant hours, and the individual Centers for Medicare & Medicaid Services (CMS) quality indicators. We constructed Cox proportional hazards models controlling for individual resident characteristics that were previously found associated with AEs.

Results: We found no association of AEs with most nursing home characteristics, including 5-star quality ratings and the composite quality score. Associations with individual quality indicators were inconsistent and frequently not monotonic. Several individual quality indicators were associated with AEs; the highest tertile of percentage of residents with depression (4%-25%) had a hazard ratio (HR) of 1.65 [95% confidence interval (CI) 1.16, 2.35] and the highest tertile of the percentage taking antipsychotic medications (18%-35%) had an HR of 1.58 (CI 1.13, 2.21). The percentage of residents needing increased assistance with activities of daily living was statistically significant but not monotonic; the middle tertile (13% to <20%) had an HR of 1.69 (CI 1.16, 2.47).

Conclusions and implications: AEs occurring during transitions between nursing homes and hospitals are not explained by the characteristics of the facilities or summary quality scores. Development of risk reduction approaches requires assessment of processes and quality beyond the current quality measures.

Keywords: Nursing homes; adverse events; hospitalization; quality indicators; transition of care.

Publication types

Research Support, U.S. Gov't, P.H.S.

MeSH terms

Activities of Daily Living*
Aged
Hospitalization
Humans
Medicare*
Nursing Homes
Prospective Studies
United States

Grants and funding

R01 HS024422/HS/AHRQ HHS/United States