ABSTRACT
Background and aims: We studied use of do not resuscitate (DNR) orders in the Brain Attack Surveillance in Corpus Christi (BASIC) study before and during the COVID-19 pandemic. Methods: All hospitalized stroke cases were ascertained in Nueces County, Texas, USA during an equal time period before the pandemic (January, 2019-Feb, 2020) and during the pandemic (March, 2020-April, 2021). We compared use of DNR orders before and during the pandemic using logistic regression adjusted for demographic and clinical variables including initial stroke severity (NIHSS score). Nueces County is geographically isolated making complete case capture likely. Cases were validated by stroke physicians using source documentation. Results: There were more cases during the pandemic (N=716) than pre-pandemic (N=681). Median NIHSS score was 5 (IQR 9) during the pandemic and 4 (IQR 9) pre-pandemic (p=0.03). During the pandemic 18.0% of stroke patients had DNR orders compared with 13.3% prepandemic (p=0.016). Other demographic and risk factors were similar in the two time periods. In models adjusted for age, sex, race-ethnicity, NIHSS score, diabetes, hypertension, current smoking and stroke history, DNR orders were not more common in the pandemic compared with pre-pandemic (p=0.2), but stroke severity (NIHSS score) remained significantly higher during the pandemic (p<0.01). Conclusions: In this population-based study, greater use of DNR orders were seen during the pandemic than before the pandemic. The greater use of DNR orders may be due, in part, to the worse stroke severity presenting to hospitals during the pandemic.