The Clinical Course of Coronavirus Disease 2019 in a US Hospital System: A Multistate Analysis

Abstract

There are limited data on longitudinal outcomes for coronavirus disease 2019 (COVID-19) hospitalizations that account for transitions between clinical states over time. Using electronic health record data from a hospital network in the St. Louis, Missouri, region, we performed multistate analyses to examine longitudinal transitions and outcomes among hospitalized adults with laboratory-confirmed COVID-19 with respect to 15 mutually exclusive clinical states. Between March 15 and July 25, 2020, a total of 1,577 patients in the network were hospitalized with COVID-19 (49.9% male; median age, 63 years (interquartile range, 50-75); 58.8% Black). Overall, 34.1% (95% confidence interval (CI): 26.4, 41.8) had an intensive care unit admission and 12.3% (95% CI: 8.5, 16.1) received invasive mechanical ventilation (IMV). The risk of decompensation peaked immediately after admission; discharges peaked around days 3-5, and deaths plateaued between days 7 and 16. At 28 days, 12.6% (95% CI: 9.6, 15.6) of patients had died (4.2% (95% CI: 3.2, 5.2) had received IMV) and 80.8% (95% CI: 75.4, 86.1) had been discharged. Among those receiving IMV, 35.1% (95% CI: 28.2, 42.0) remained intubated after 14 days; after 28 days, 37.6% (95% CI: 30.4, 44.7) had died and only 37.7% (95% CI: 30.6, 44.7) had been discharged. Multistate methods offer granular characterizations of the clinical course of COVID-19 and provide essential information for guiding both clinical decision-making and public health planning.

Keywords: COVID-19 hospitalizations; age-stratified mortality; clinical course; coronavirus disease 2019; intensive care unit; longitudinal trajectory; mechanical ventilation; multistate analysis.

Figure 1

Framework for a multistate analysis of transitions between clinical states among hospitalized patients with coronavirus disease 2019. At each time point, patients were categorized into one of 15 mutually exclusive and exhaustive states: 1) emergency department, 2) inpatient floor, 3) intensive care unit (ICU) admission without invasive mechanical ventilation (IMV), 4) noninvasive ventilation (NIV), 5) IMV in the ICU, 6) NIV after IMV, 7) ICU after IMV, 8) inpatient floor after ICU admission but no IMV, 9) inpatient floor after IMV, 10) discharge without ICU admission, 11) discharge with a history of ICU admission but no IMV, 12) discharge with a history of IMV, 13) death, 14) death with a history of ICU admission but no IMV, and 15) death with a history of IMV. The figure depicts all of the possible transitions patients could make from each state. Patients were not restricted to starting from state 1; those who were directly admitted to the hospital or transferred from another hospital started from the state in which they were first observed.

Figure 2

Longitudinal outcomes among hospitalized patients with coronavirus disease 2019 entering 3 specific clinical care states (multistate analyses), BJC HealthCare Hospital system, St. Louis, Missouri, 2020. The figure shows the proportion of patients estimated to be in each care state at any given time point, accounting for the transitions patients made between different clinical states over time. A) Outcomes following initial admission to a hospital (n = 1,577); B) outcomes following admission to the intensive care unit (ICU) (n = 571); C) outcomes following noninvasive ventilation (NIV) (n = 343); D) outcomes following intubation (n = 214). ED, emergency department; IMV, invasive mechanical ventilation.

Figure 3

Clinical trajectories of hospitalized patients with coronavirus disease 2019 over the course of their hospital stay (n = 1,417), BJC HealthCare Hospital system, St. Louis, Missouri, 2020. Alluvia are color-coded by patient outcome at 28 days, and their width represents the number of patients. Only patients with 28 days of observation time were included (inclusive of time after discharge or death). ICU, intensive care unit; IMV, invasive mechanical ventilation; NIV, noninvasive ventilation.

Figure 4

Instantaneous hazards of intensive care unit (ICU) admission, noninvasive ventilation (NIV), intubation, discharge, and death at different time points since admission among hospitalized patients with coronavirus disease 2019 (n = 1,577), BJC HealthCare Hospital system, St. Louis, Missouri, 2020.

Figure 5

Estimated durations of overall patient stay (n = 1,577) (A), intensive care unit (ICU) stay (n = 571) (B), noninvasive ventilation (NIV) (n = 343) (C), and invasive mechanical ventilation (IMV) (n = 214) (D) among hospitalized patients with coronavirus disease 2019 (multistate analyses), BJC HealthCare Hospital system, St. Louis, Missouri, 2020. Dots represent the median values; the surrounding boxes span the 25th and 75th percentiles; and the violin plots show kernel density plots spanning the full range of values. Notably, kernel density plots extend below 1 because of estimation algorithms, but no patients had a length of stay less than 0 in any state.

Figure 6

Transition intensities for transitions from the inpatient floor, intensive care unit (ICU), noninvasive ventilation (NIV), and invasive mechanical ventilation (IMV) clinical states among hospitalized patients with coronavirus disease 2019, BJC HealthCare Hospital system, St. Louis, Missouri, 2020. The figure depicts the instantaneous hazard of potential transitions from an initial starting clinical state to the next subsequent clinical state. Values shown on the x-axes represent the amount of time since the patient initially entered a particular clinical state. A) Transitions after entering the inpatient floor state (i.e., from state 2 to either state 3, 10, or 13) (n = 1,577); B) transitions after entering the ICU state (i.e., from state 3 to either state 4, 5, 8, or 14) (n = 571); C) transitions after entering the NIV state (i.e., from state 4 to either state 5, 7, or 14) (n = 343); D) transitions after entering the IMV state (i.e., from state 5 to either state 6, 7, or 15) (n = 214).

Figure 7

Cumulative incidence of intensive care unit admission (A), noninvasive ventilation (B), intubation (C), and death (D) by 28 days among hospitalized patients with coronavirus disease 2019, according to patient subgroup (n = 1,577), BJC HealthCare Hospital system, St. Louis, Missouri, 2020. Results were obtained in stratified competing-risk analyses using the Aalen-Johansen method. The reference line (vertical dashed line) aligns with the estimate for the overall population. Bars, 95% confidence intervals (CIs).

Figure 8

Age-stratified adjusted estimates of the cumulative incidence of intensive care unit admission (A), noninvasive ventilation (B), intubation (C), and death (D) among hospitalized patients with coronavirus disease 2019, by time period (n = 1,577), BJC HealthCare Hospital system, St. Louis, Missouri, 2020. Marginal estimates were obtained from Poisson models adjusting for sex, race/ethnicity, comorbidity, and whether the patient had come from a long-term care facility, with a time offset. Bars, 95% confidence intervals (CIs).