Mortality risk for hospitalized COVID-19 patients tracked with intensive care unit (ICU) demand, a study affirmed.
Across the 88 U.S. Veterans Affairs (VA) hospitals with at least 10 COVID-19 ICU patients, the proportion of COVID-19 patients who died rose from 22.9% in March 2020 to a peak of 25.0% in April, then progressively fell to under 13% by July and August (P<0.001).
Mortality in the ICU during surges when COVID-19 took up at least 75% of available beds carried an adjusted hazard ratio of 1.94 compared with periods when it accounted for no more than a quarter of beds (P<0.001), reported Dawn Bravata, MD, of the VA Medical Center in Indianapolis, and colleagues in JAMA Network Open.
ICUs appeared to be able to absorb lesser degrees of demand with numerically but not statistically greater impact on coronavirus deaths. ICU strain also didn’t correlate with COVID-19 deaths outside the ICU.
“If these results represent true causality, they provide additional support for public health strategies to ‘flatten the curve,'” noted Lewis Rubinson, MD, PhD, of Morristown Medical Center in New Jersey, in an accompanying editorial.
However, other concurrent trends in care were happening, with more unproven therapies being utilized early in the pandemic period and more effective ones being used later, he noted. Also, ICUs might have admitted less severe cases when demand was low, Rubinson added.
Further studies might examine things like staffing ratios and availability of mechanical ventilation experts to get at causality, he pointed out.
The VA study covered 8,516 veterans admitted from March through August 2020, with follow-up through November 2020.
Other health systems have suggested similar improvements after the initial surge in the U.S., such as the decline of overall COVID-19 mortality at one group of New York City hospitals from 26% in March to 8% in August.
However, the uncontrolled growth in cases over the past months since then overwhelmed hospital systems in Southern California in particular, and specter of more transmissible variants raise concerns that the progress in mortality rate could be jeopardized.
“Public health officials and hospital administrators may seek to prevent high COVID-19 ICU demand to optimize outcomes for patients with COVID-19,” Bravata’s group recommended.
Using ICU load thresholds to trigger local interventions to slow community transmission “should be broadly considered,” Rubinson noted. Transferring ICU patients out to less crowded hospitals might be another solution, albeit with significant risks, he added. “Therefore, the redistribution system, if undertaken, must be well designed and implemented to ensure that the solution is not associated with new adverse events.”
Last Updated January 20, 2021
The study was supported by the U.S. Department of Veterans Affairs (VA) Health Services Research and Development Service Precision Monitoring to Transform Care Quality Enhancement Research Initiative.
Bravata disclosed support from the VA.
Rubinson disclosed relevant relationships with Ventec Life Systems.