ABSTRACT
Background: The COVID-19 pandemic has provided a unique opportunity to evaluate global intensive care unit (ICU) admission practices. Patients with chronic liver disease (CLD) and cirrhosis may have limited or variable access to ICU. We aimed to describe international ICU admission rates and outcomes in critically-ill patients with CLD and COVID-19. Methods: Data were combined from two international registries (SECURE-Liver and COVID-Hep) for patients with CLD and lab-confirmed COVID-19 deemed sick enough to require ICU admission by the reporting clinician. Rates of ICU admission or decline, and respective outcomes were compared by country. We performed a secondary analysis comparing ICU admissions/declines and outcomes from the United States (US) and United Kingdom (UK), the two greatest contributing countries. Results: Between 25 March 2020 and 3 February 2021, 319 patients with CLD and COVID-19 from 27 countries were deemed to require ICU care. There was considerable country-level variability in ICU decline rates (Figure 1A), although mortality following ICU admission was similar by country (Figure 1B). Rates of ICU admission differed significantly between the US (75/79, 95%) and UK (22/77, 29%) (p<0.001). However, there were no differences in the US and UK in mortality after ICU admission (42/75 [56%] vs. 10/22 [45%];p=0.468;Figure 1B) or mortality after invasive ventilation (29/59 [49%] vs. 9/17 [53%];p=1.000). Both in those requiring ICU admission and admitted to the ICU, there were no differences in age, sex, Charlson Comorbidity Index or Child Pugh Score. Only four US patients were declined ICU admission of whom 2 (50%) died compared to 55 UK patients declined ICU admission of whom 51 (93%) died. Baseline factors associated with being declined ICU admission in the UK were older age, alcohol-related liver disease, and Child B/C cirrhosis. In both US and UK cohorts, the reason for not admitting patients to ICU was due to this being deemed inappropriate (futile) by the responsible clinician, except for one case in both countries in which no ICU bed was available. Information relating to patient goals of care, longterm outcomes in survivors, and granular detail regarding organ support requirements were not available. Conclusion: Patients with CLD and critical COVID-19 were over 3-times more likely to be admitted to ICU in the US than the UK despite having similar baseline characteristics. However, the rates of mortality following ICU admission were comparable between the two countries. ICU bed availability was not a key factor in decline rates. The differing thresholds for escalation to ICU with similar post admission outcomes warrants further discussion.
ABSTRACT
Background: A number of factors can inform ICU escalation decisions, including the likelihood of survival and patient co-morbidities. This study examined prior liver transplant (LT) recipients and patients with chronic liver disease (CLD) diagnosed with SARS-CoV-2, and compared the rate of ICU admission and decline amongst those who were sick enough to require ICU care. Methods: Patient data from 12 March 2020 to 6 May 2021 was extracted using two international reporting registries (SECURE-Liver and COVID-Hep). Patients had a history of LT or CLD, laboratory-confirmed SARS-CoV-2, and were deemed ill enough to require ICU care. Patients were either admitted to the ICU, or declined admission due to inadequate capacity or because ICU escalation was deemed inappropriate. We compared patient characteristics by ICU decline, and compared ICU decline rates by LT and CLD categories with unadjusted and multivariable logistic regression. Results: 173 LT recipients were admitted to the hospital with SARS-CoV-2 (transplant year 1986-2020, median age 63, 74% male), and 66 (38.2%) were deemed unwell enough to require ICU care. Among those sick enough to require ICU care, 55 (83.3%) were admitted to the ICU and 11 (16.7%) were declined admission. Compared to those admitted to the ICU, patients declined ICU admission were significantly older (median 69 yrs vs 62 yrs, p=0.01) but otherwise similar in other characteristics. ICU decline rates in prior LT recipients (16.7%) were similar to patients with non-cirrhotic CLD (16.1%, p=0.96), but substantially lower than patients with Child A cirrhosis (31.8%, p=0.03), Child B cirrhosis (37.1%, p=0.006) and Child C cirrhosis (38.7%, p=0.004). Differences in ICU decline between LT recipients and Child B or C cirrhosis remained statistically significant after adjustment for age, sex and major co-morbidities. Among patients admitted to the ICU, mortality was higher in LT recipients compared to non-cirrhotic CLD (OR 0.31, 95% CI 0.14-0.71) but lower in LT recipients compared to Child C cirrhosis (OR 3.85, 95% CI 1.47-10.11) after adjustment for age, sex and co-morbidities (see Figure 1). Conclusion: ICU decline was less likely in LT recipients compared to patients with decompensated cirrhosis. LT recipients may be seen as gaining more benefit from ICU care, given the higher mortality amongst patients with decompensated cirrhosis. This is in line with prior data showing decompensated cirrhosis is a predictor of higher mortality in patients with SARS-CoV-2. Moreover, large investment of resources in LT recipients may make them more likely to be admitted to the ICU.