COVID-19 patients admitted to intensive care units (ICUs) at US Department of Veterans Affairs (VA) hospitals during peak coronavirus patient surges were twice as likely to die than those treated during low-demand periods, an observational study published today in JAMA Network Open suggests.
VA researchers studied 8,516 COVID-19 patients, 94.1% of them men, admitted to ICUs at 88 veterans hospitals from Mar 1 to Aug 31, 2020, with 30 days of follow-up.
Compared with COVID-19 patients treated in the ICU during times of low ICU demand (≤25%), those treated during times of 25% to 50% demand had an adjusted hazard ratio (aHR) of 0.99. But the aHR jumped to 1.19 when ICU demand was 50% to 75% and to 1.94 when demand was greater than 75%.
The death rate did not increase with increasing ICU demand in patients treated in the noncritical setting. The overall all-cause death rate varied over time, ranging from a high of 25.0% in April to a low of 12.5% in July.
Patient load was defined as the number of COVID-19 patients in the ICU compared with the typical ICU bed counts at each hospital, while ICU demand was defined as the mean number of coronavirus patients in the ICU during the patient's stay divided by the upper limit of COVID-19 patients in that unit. COVID-19 ICU loads ranged from 1% to more than 100%, at which point hospitals increased critical care bed capacity by repurposing other facility areas as ICUs.
COVID-19 ICU patient loads varied over time, with prevalence rates at some hospitals peaking in March and others in July. The percentage of COVID-19 patients receiving care during times of low coronavirus ICU load (≤25%) increased over time, from a low of 51.0% in March to a peak of 91.8% in August.
The percentage of coronavirus patients treated during times of peak COVID-19 ICU load (100% or higher) fell steadily, from 6.3% in March, to 1.1% in April, to 0% in May through August. The percentage of COVID-19 patients receiving care during times of high coronavirus ICU demand (higher than 75%) fell from 24.4% in March, to 20.2% in April, to 4.8% in May. It then increased to a peak of 17.4% in July before falling to 5.8% in August.
The proportion of patients receiving care on general wards climbed after the beginning of the pandemic (55.2% in March) to a peak of 67.6% in July. Mean patient age was 67.9 years.
Triage, tracking, coordination may ease strain
The authors said that because higher COVID-19 patient ICU demand was associated with increased death rates both early (March to May) as well as later on in the pandemic (June to August), it suggests that overwhelmed hospitals were linked to increased coronavirus ICU death rates.
"This cohort study found that although facilities augmented ICU capacity during the pandemic, strains on critical care capacity were associated with increased COVID-19 ICU mortality," they wrote.
The researchers called for future research to elucidate the magnitude to which patient factors such as illness severity or hospital problems like understaffing contributed to the tie between COVID-19 ICU strain and poor outcomes in critically ill coronavirus patients and to determine if patients who would have normally been admitted to the ICU under nonpandemic conditions were instead admitted to a general ward when ICUs were overwhelmed by coronavirus patients.
The authors also suggested that hospitals within a healthcare system or a geographic region work together to triage severely ill COVID-19 patients to facilities with higher ICU capacity to ease strain on other facilities. "Tracking COVID-19 ICU demand may be useful to hospital administrators and health officials as they coordinate COVID-19 admissions across hospitals to optimize outcomes for patients with this illness," they wrote.