Readmission and Death After Initial Hospital Discharge Among Patients With COVID-19 in a Large Multihospital System | Critical Care Medicine | JAMA | JAMA Network
[Skip to Content]
[Skip to Content Landing]
Figure.  Kaplan-Meier Curves and Hazard Ratios for 60-Day Readmission or Death Among Patients With COVID-19 and Matched Patients to Comparison Hospitalizations
Kaplan-Meier Curves and Hazard Ratios for 60-Day Readmission or Death Among Patients With COVID-19 and Matched Patients to Comparison Hospitalizations

In weighted analyses, patients who survived their hospitalizations were balanced on age, sex, week of discharge, intensive care unit admission, and length of hospitalization using coarsened exact matching. Hazard ratios for readmission or death were estimated using piecewise Cox proportional hazards regression.

A, Included 1366 patients with coronavirus disease 2019 (COVID-19) and 1799 with pneumonia. The hazard ratios for 0 to 10 days were 1.43 (95% CI, 1.09-1.87); for 11 to 20 days, 0.51 (95% CI, 0.36-0.74); for 21 to 40 days, 0.63 (95% CI, 0.45-0.88); and for 41 to 60 days, 0.57 (95% CI, 0.38-0.85).

B, Included 1430 patients with COVID-19 and 3505 with heart failure. The hazard ratios for 0 to 10 days were 1.62 (95% CI, 1.31-2.01); for 11 to 20 days, 0.55 (95% CI, 0.41-0.74); for 21 to 40 days, 0.43 (95% CI, 0.33-0.56); and for 41 to 60 days, 0.39 (95% CI, 0.28-0.55).

Table.  Characteristics of Patients With COVID-19 With vs Without 60-Day Readmission, Death, or Readmission or Deatha
Characteristics of Patients With COVID-19 With vs Without 60-Day Readmission, Death, or Readmission or Deatha
1.
Vincent  BM, Wiitala  WL, Burns  JA, Iwashyna  TJ, Prescott  HC.  Using Veterans Affairs corporate data warehouse to identify 30-day hospital readmissions.   Health Serv Outcomes Res Methodol. 2018;18:143-154. doi:10.1007/s10742-018-0178-3Google ScholarCrossref
2.
COVID-19 resources. Health Services Research & Development: US Department of Veterans Affairs. Accessed October 7, 2020. https://www.hsrd.research.va.gov/covid19.cfm
3.
Scehnet  J, DuVall  S. VA COVID-19 shared data resource update: VA informatics and computing infrastructure. US Department of Veterans Affairs. Accessed August 17, 2020. https://www.hsrd.research.va.gov/for_researchers/cyber_seminars/archives/3834-notes.pdf
4.
Clinical Classifications Software Refined (CCSR) for ICD-10-CM Diagnoses. Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality. Published 2020. Accessed August 17, 2020. https://www.hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp
5.
Wang  XQ, Vincent  BM, Wiitala  WL,  et al.  Veterans Affairs patient database (VAPD 2014-2017): building nationwide granular data for clinical discovery.   BMC Med Res Methodol. 2019;19(1):94. doi:10.1186/s12874-019-0740-x PubMedGoogle ScholarCrossref
6.
Blackwell  M, Iacus  S, King  G, Porro  G.  Cem: coarsened exact matching in Stata.   Stata J. 2009;9(4):524-546. doi:10.1177/1536867X0900900402 Google ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    EXPAND ALL
    Follow-up After Hospital Discharge From a COVID-19 Admission
    Kathleen Mullane |
    Thank you for your work in acquiring / analyzing the data for this important publication.

    So much effort is placed upon the inpatient management of COVID-19-infected individuals (seriously ill enough to be hospitalized in the first place). Limited hospital beds/HCW support forces physicians to make difficult decisions on discharging expediently to open beds for those patients presenting acutely and judged to be more seriously ill than those presently receiving inpatient care. Your manuscript importantly gives a look at what happens to those individuals discharged from the inpatient setting. As only age appeared significant
    as a predictor of mortality in your data set, it would be incredibly helpful to those physicians making these difficult decisions to better discern measurable risk factors for re-admission and death so that these decisions may be made in a more evidenced based manner and as well what interventions might be considered to reduce the risk of re-admission and/or death in these individuals.

    Post-discharge care has been disrupted due to limitations in outpatient access to health care: reduced available visit slots due to social distancing, need for isolation of the patient post-discharge, as well as physicians and other health care workers being deployed to put a Band-Aid on managing those presenting to hospitals and admitted for care thereby limiting availability of HCWs for outpatient care.

    To manage acutely ill individuals, only to have nearly 30% need readmission or perish on discharge is more than depressing for healthcare providers working so very hard to provide exemplary care.

    Again, thank you for this important publication!
    CONFLICT OF INTEREST: Investigator for Gilead (Remdesivir), Roche-Genentech (Tocilizumab) and Moderna (OWS vacicne trials)
    READ MORE
    Views 38,239
    Citations 0
    Research Letter
    December 14, 2020

    Readmission and Death After Initial Hospital Discharge Among Patients With COVID-19 in a Large Multihospital System

    Author Affiliations
    • 1Department of Learning Health Sciences, University of Michigan, Ann Arbor
    • 2Division of Pulmonary and Critical Care Medicine, University of Michigan, Ann Arbor
    • 3VA Center for Clinical Management Research, HSR&D Center of Innovation, Ann Arbor, Michigan
    JAMA. Published online December 14, 2020. doi:10.1001/jama.2020.21465

    Although more patients are surviving severe coronavirus disease 2019 (COVID-19), there are limited data on outcomes after initial hospitalization. We therefore measured the rate of readmission, reasons for readmission, and rate of death after hospital discharge among patients with COVID-19 in the nationwide Veterans Affairs (VA) health care system.

    Methods

    We identified index hospitalizations for COVID-19 among veterans at 132 VA hospitals (admitted March 1–June 1, 2020; discharged March 1–July 1, 2020) in the VA’s Corporate Data Warehouse.1 Definitions included definite hospitalizations for COVID-19, in which patients were diagnosed during hospitalization, and probable hospitalizations, in which patients were diagnosed during the 14 days preceding or 7 days following hospitalization.2,3

    We also identified comparison cohorts of hospitalizations for non-COVID pneumonia and heart failure during the same time frame, using the Agency for Healthcare Research and Quality’s Clinical Classification Software Refined diagnosis groupings.4 We extracted demographics, intensive care unit (ICU) use, length of hospitalization, receipt of invasive mechanical ventilation, and receipt of vasopressors.5 We applied weights from coarsened exact matching to balance survivors of COVID-19 and control hospitalizations on age, sex, week of discharge, length of hospitalization, and ICU use.6

    We measured readmission and death to 60 days after discharge among survivors of COVID-19 hospitalizations, determined the most common readmission diagnoses, and measured use of ICU, mechanical ventilation, and vasopressors during readmission.1 We compared characteristics of COVID-19 survivors who experienced vs did not experience 60-day readmission or death using Pearson χ2 and Wilcoxon rank sum tests. We compared rates of outcomes between matched survivors of COVID-19 and control hospitalizations using Rao-Scott–corrected χ2 tests. Statistical tests were 2-tailed, with P < .05 considered significant. We generated Kaplan-Meier curves for readmission or death to 60 days after the initial hospital discharge. We also estimated hazard ratios and 95% CIs for 0 to 10, 10 to 20, 20 to 40, and 40 to 60 days after discharge using piecewise Cox proportional hazards regression. Analyses were performed using SAS version 9.4 (SAS Institute Inc) and Stata MP version 15.1 (StataCorp). The study was deemed exempt by the Ann Arbor VA institutional review board.

    Results

    There were 2179 index hospitalizations for COVID-19, of which 678 patients (31.1%) were treated in an ICU, 279 (12.8%) were mechanically ventilated, 307 (14.1%) received vasopressors, and 1775 (81.5%) survived to discharge.

    Within 60 days of discharge, 354 patients (19.9%) who survived COVID-19 hospitalization were readmitted, 162 (9.1%) died, and 479 (27.0%) were readmitted or died. Survivors with 60-day readmission or death were older but otherwise similar to survivors without readmission or death (Table). Of those readmitted, the most common readmission diagnoses were COVID-19 (30.2%), sepsis (8.5%), pneumonia (3.1%), and heart failure (3.1%). During readmission, 22.6% were treated in intensive care, 7.1% were mechanically ventilated, and 7.9% received vasopressors. Index admissions totaled 27 496 hospital days, whereas readmissions after COVID-19 resulted in 3728 additional hospitalization days.

    Of the index hospitalizations, 2156 patients had pneumonia and 4269 had heart failure, of whom 97.8% and 98.3% survived to discharge, respectively. After excluding patients who died during hospitalization or could not be matched, 1799 with pneumonia and 3505 with heart failure who survived were included in the weighted comparisons with patients with COVID-19. COVID-19 survivors had lower rates of 60-day readmission or death than matched survivors of pneumonia (26.1% vs 31.7%; P = .006) and heart failure (27.0% vs 37.0%; P < .001). However, COVID-19 survivors had higher rates of readmission or death within the first 10 days after discharge than matched survivors of pneumonia (13.4% vs 9.7%; P = .01) and heart failure (13.9% vs 8.8%; P < .001) (Figure).

    Discussion

    In this national cohort of VA patients, 27% of survivors of COVID-19 hospitalization were readmitted or died by 60 days after discharge, and this rate was lower than matched survivors of pneumonia or heart failure. However, rates of readmission or death were higher than pneumonia or heart failure during the first 10 days after discharge following COVID-19 hospitalization, suggesting a period of heightened risk of clinical deterioration. Study limitations include the inability to measure readmissions to non-VA hospitals and an older, male-predominant study population, who may be at higher risk of severe manifestations of COVID-19. Public health surveillance or clinical trials focused exclusively on inpatient mortality may substantially underestimate burdens of COVID-19.

    Section Editor: Jody W. Zylke, MD, Deputy Editor.
    Back to top
    Article Information

    Corresponding Author: John P. Donnelly, PhD, Department of Learning Health Sciences, University of Michigan, 2800 Plymouth Rd, NCRC Bldg 14, #G100, G014-130, Ann Arbor, MI 48109 (jpdonn@med.umich.edu).

    Accepted for Publication: October 13, 2020.

    Published Online: December 14, 2020. doi:10.1001/jama.2020.21465

    Correction: This article was corrected online on December 30, 2020, to adjust the numbers and percentages of women as reported in the Table.

    Author Contributions: Drs Donnelly and Prescott had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Donnelly, Prescott.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Donnelly.

    Critical revision of the manuscript for important intellectual content: Wang, Iwashyna, Prescott.

    Statistical analysis: Donnelly, Wang.

    Obtained funding: Prescott.

    Administrative, technical, or material support: Prescott.

    Supervision: Prescott.

    Conflict of Interest Disclosures: Dr Donnelly reported receiving grants from the National Heart, Lung, and Blood Institute (NHLBI) and personal fees from the Annals of Emergency Medicine. Dr Iwashyna reported receiving grants from VA Health Services Research and Development. Dr Prescott reported receiving grants from the Agency for Healthcare Research and Quality and the Department of Veterans Affairs. No other disclosures were reported.

    Funding/Support: Drs Donnelly and Iwashyna are supported by grant K12-HL138039 from the NHLBI. Dr Prescott is supported by grant R01-HS026725 from Agency for Healthcare Research and Quality. This work was supported by grant IIR 17-045 from the VA Health Services Research and Development (Dr Iwashyna). This material is the result of work supported with resources and use of facilities at the Ann Arbor VA Medical Center.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Disclaimer: This article does not represent the views of the Department of Veterans Affairs or the US government.

    References
    1.
    Vincent  BM, Wiitala  WL, Burns  JA, Iwashyna  TJ, Prescott  HC.  Using Veterans Affairs corporate data warehouse to identify 30-day hospital readmissions.   Health Serv Outcomes Res Methodol. 2018;18:143-154. doi:10.1007/s10742-018-0178-3Google ScholarCrossref
    2.
    COVID-19 resources. Health Services Research & Development: US Department of Veterans Affairs. Accessed October 7, 2020. https://www.hsrd.research.va.gov/covid19.cfm
    3.
    Scehnet  J, DuVall  S. VA COVID-19 shared data resource update: VA informatics and computing infrastructure. US Department of Veterans Affairs. Accessed August 17, 2020. https://www.hsrd.research.va.gov/for_researchers/cyber_seminars/archives/3834-notes.pdf
    4.
    Clinical Classifications Software Refined (CCSR) for ICD-10-CM Diagnoses. Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality. Published 2020. Accessed August 17, 2020. https://www.hcup-us.ahrq.gov/toolssoftware/ccsr/ccs_refined.jsp
    5.
    Wang  XQ, Vincent  BM, Wiitala  WL,  et al.  Veterans Affairs patient database (VAPD 2014-2017): building nationwide granular data for clinical discovery.   BMC Med Res Methodol. 2019;19(1):94. doi:10.1186/s12874-019-0740-x PubMedGoogle ScholarCrossref
    6.
    Blackwell  M, Iacus  S, King  G, Porro  G.  Cem: coarsened exact matching in Stata.   Stata J. 2009;9(4):524-546. doi:10.1177/1536867X0900900402 Google ScholarCrossref
    ×