Discharge Criteria Key Clinical Questions
01. Which clinical criteria do hospitals use to determine when it is safe to discharge COVID-19 patients to home?
a. HOMERuN institutions
Sites varied widely in their guidance, generally falling into one of several categories: 1) little or no guidance (e.g., use clinical judgment); 2) general guidance concerning disease stability and improvement (e.g., hemodynamically stable, and fever or other symptoms are improving, and oxygen requirement is declining or resolved); 3) guidance resembling the Centers for Disease Control and Prevention (CDC) guidelines for the removal of home quarantine using the non–test-based strategy; and 4) detailed clinical guidance or algorithms based on age, comorbidities, immunocompromise, lab values, need for supplementary oxygen, and stability of vital signs (temperature, oxygen saturation, and respiratory rate). The most commonly mentioned criteria were symptom improvement, temperature, and oxygen requirement, although specific definitions varied. For example, some sites specified afebrile for a specific time (range: 24-72 hours), while others simply required afebrile at discharge. Similarly, some sites required specific oxygen saturation levels (range >90%-94%) or supplementation levels (range: 2-4 liters), while others simply required these to be stable or at baseline at the time of discharge. Relatively few sites (36%) addressed laboratory criteria, age (36%), high-risk comorbidities (32%), or infectious disease consultation (18%) as criteria for discharge.
b. What is Known
Some sites (e.g., ChristianaCare, Mount Sinai) have general guidance that incorporates clinical improvement, temperature, and oxygen requirement, while other sites (e.g., Johns Hopkins and University of Michigan) provide more specific guidance algorithms to standardize care (acknowledging the absence of evidence to substantiate this approach). None of these guidelines are evidence-based, only incorporating first principles regarding general and disease-specific risk factors for severe disease and markers of clinical stability.
The literature on risk factors for poor post-discharge outcomes is limited. We identified three studies (two in pre-publication, one published), using different study designs, definitions of readmission, and data collected as potential predictors. The data are summarized below:
| STUDY | SETTING, DESIGN | N | OUTCOME, FREQUENCY | RISK FACTORS |
| Somani, et al.1 | 5 NYC hospitals (Mount Sinai system)
Retrospective cohort |
2864 | Return to hospital within 14 days (ER visit or readmission)
3.6% |
Lower BMI COPD HTN ~less anticoagulation Less likely admitted to ICU Shorter LOS |
| Parra, et al.2 | 1 tertiary hospital in Spain
Nested Case-Control (matched by age, sex, time period) |
61 cases and 61 matched controls | Readmitted within 3 weeks
4.4% |
Immunocompromise ~HTN Shorter LOS Fever in 48 hours prior to discharge ~neutrophil/lymphocyte ratio |
| Richardson, et al.3 | 12 NYC hospitals (Northwell system)
Retrospective cohort |
2081 patients discharged alive | Readmitted by end of study period
2.2% |
Older age ~DM ~HTN |
| Kingery, et al.4 | Quaternary referral hospital and community hospital in NYC | 1344 | Readmission, ED visit, or mortality within 30 days
16.5%, 9.8%, 2.4% |
Readmission: Older age Diabetes Inpatient dialysisMortality: Older age Asian race |
| Saab, et al.5 | 1 Tertiary hospital in CA (UCLA) | 99 | Readmission or ED visit within 30 days
5%, 5% |
*Fever within 24h of discharge, O2 requirement, lab abnormalities were not associated, but small numbers |
| Lavery, et al.6 | Premier database from 865 hospitals | 126,137 | Same hospital readmission within 2 months
9% |
Discharge to SNF DC home with services Age per decade >65 Previous admission within 3 months prior COPD Heart Failure Diabetes CKD Shorter LOS Less likely if ICU Less likely if non-invasive ventilation |
| Verna, et al.7 | Chargemaster data from 297 hospitals across 40 states | 29,659 | Readmission within 30 days
3.6% |
Discharge to SNF Age > 60 Shorter LOS Never on O2 Northeast region Medicaid or Medicare CKD CVD HTN less risk Diabetes Smoking Index presentation with AKI, CHF, DIC, VTE Less likely if ICU |
| Donnelly, et al.8 | VA Corporate Data Warehouse, 132 VA hospitals | 2179 | Readmission or death within 60 days
19.9%, 9.1% |
Readmission: LOS < 7 daysMortality: Older age Mechanical ventilation Pressor use |
*NYC: New York City; ER: emergency room; BMI: Body Mass Index; COPD: chronic obstructive pulmonary disease; HTN: hypertension; ICU: intensive care unit; LOS: length of stay; DM: diabetes mellitus
~ indicates borderline statistical significance
Notable findings include the following:
- The overall rate of readmission is low in these cohorts, with two exceptions (Kingery and Donnelly). This may be due to several factors:
a. These patients have less biopsychosocial complexity than typically admitted general medicine patients.
b. These patients tended to have longer lengths of stay than typical inpatients.
c. Once patients have fully recovered from COVID-19, they tend not have recrudescent disease. This may explain why ICU admission predicted lower readmission. - The presence of hypertension was a risk factor in most (but not all) of these studies. Other possible risk factors include chronic obstructive pulmonary disease, diabetes mellitus, COPD, CKD, CVD, heart failure, smoking, and immunocompromise.
- Being febrile at discharge might be a risk factor for readmission but most studies were too small to evaluate this.
- Lower length of stay was associated with a higher readmission rate in 5 of these studies. This suggests that at least some patients were discharged too soon, when they were still at risk of worsening disease.
- It is notable that the Somani study found a trend toward less anticoagulation at discharge as a risk factor. This was not found in the Parra Ramírez study, but in that study, 16% of readmissions were due to a thrombotic event.
- Inflammatory markers and vital signs other than temperature (e.g., O2 requirements at discharge) were not significant risk factors in the few studies that looked at them, but most of these studies were small. One study found that never being on supplemental O2 was a risk factor for readmission, likely because these patients were discharged early with mild disease which then worsened in week 2.
- Discharge to SNF or home with services was sometimes associated with readmission, not because they provide poor post-discharge care but rather because these are markers of clinical fragility.
Notably, none of the larger studies compared the readmission rates in those with and without specific discharge criteria, and none of them compared length of stay and readmission rates in patients who met particular combinations of discharge criteria.
c. What is Not Known
Our findings reveal an urgent need for empiric research on clinical factors (including improvement in temperature and oxygen saturation, severity of disease, comorbidities, and laboratory markers), alone and in combination, that might predict poor post-discharge outcomes, including readmission and death. It may be that simple scores are effective. Alternatively, more complicated scores or criteria may be better to minimize length of stay and post-discharge complications. In the absence of such data, discharge practices with high consensus in this brief synthesis may provide useful guidance for hospitals to consider as they develop and refine protocols for a prolonged COVID-19 pandemic.
d. Spring 2021 Update
- Few sites have specific written guidelines for discharge criteria
- Decision-making has become more holistic (clinical gestalt) and trend-based (e.g., improvement in oxygen requirement), rather than meeting specific criteria at discharge
- General guidelines include decreasing oxygen requirement, signs and symptoms improving, clinically stable
- Clinicians have become more comfortable discharging patients earlier, e.g., still on oxygen, as long as they are otherwise clinically improving
- Clinicians take symptom onset, comorbidities, and lab markers into account, but they are not the main drivers of when to discharge, unlike earlier in the pandemic
- A study is currently underway by our group to better determine the factors that determine clinical stability at discharge
Key Clinical Questions
- Somani S, Richter F, Fuster V, et al. Characterization of patients who return to hospital following discharge from hospitalization for COVID-19. medRxiv. 2020: p. 2020.05.17.20104604.
- Parra, LM, Cantero M, Morras I, et al. Hospital readmissions of discharged patients with COVID-19. medRxiv. 2020: p. 2020.05.31.20118455.
- Richardson S, Hirsch JS, Narasimhan M. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA. 2020;323(20):2052-2059.
- Kingery JR, Bf Martin P, Baer BR, et al. Thirty-day post-discharge outcomes following COVID-19 infection. J Gen Intern Med. 2021 Jun 7 [Epub ahead of print].
- Saab FG, Chiang JN, Brook R, et al. Discharge clinical characteristics and post-discharge events in patients with severe COVID-19: a descriptive case series. J Gen Intern Med. 2021;36(4):1017-1022.
- Lavery AM, Preston LE, Ko JY, Chevinsky JR, et al. Characteristics of hospitalized COVID-19 patients discharged and experiencing same–hospital readmission — United States, March–August 2020. MMWR Morb Mortal Wkly Rep. 2020;69(45):1695-1699.
- Verna EC, Landis C, Brown RS Jr, et al. Factors associated with readmission in the US following hospitalization with COVID-19. Clin Infect Dis. 2021 May 20:ciab464.
- Donnelly JP, Wang XQ, Iwashyna TJ, Prescott HC. Readmission and death after initial hospital discharge among patients with COVID-19 in a large multihospital system. JAMA. 2021;325(3):304-306.