Comparison of the demographic characteristics and comorbidities of patients with COVID-19 who died in Spanish hospitals based on whether they were or were not admitted to an intensive care unit

Abstract

Objective: To describe and compare the demographic characteristics and comorbidities of patients with COVID-19 who died in Spanish hospitals during the 2020 pandemic based on whether they were or were not admitted to an intensive care unit (ICU) prior to death.

Methods: We performed a secondary analysis of COVID-19 patients who died during hospitalization included by 62 Spanish emergency departments in the SIESTA cohort. We collected the demographic characteristics and comorbidities, determined both individually and estimated globally by the Charlson index (ChI). Independent factors related to ICU admission were identified and different analyses of sensitivity were performed to contrast the consistency of the findings of the principal analysis.

Results: We included the 338 patients from the SIESTA cohort that died during hospitalization. Of these, 77 (22.8%) were admitted to an ICU before dying. After multivariate adjustment, 3 out of the 20 basal characteristics analyzed in the present study were independently associated with ICU admission: dementia (no patients with dementia who died were admitted to the ICU: OR = 0, 95%CI = not calculable), active cancer (OR = 0.07; 95%CI = 0.02-0.21) and age (< 70 years: OR = 1, reference; 70-74 years: OR = 0.21; 95%CI = 0.08-0.54; 75-79 years: OR = 0.21; 95%CI = 0.08-0.54; ≥ 80 years: OR = 0.02; 95%CI = 0.01-0.05). The probability of ICU admission significantly increased in parallel to the ChI, even after adjustment for age (ChI 0 points: OR = 0, reference; ChI 1 point: OR = 0.36; 95%CI = 0.16-0.83; ChI 2 points: OR = 0.36; 95%CI = 0.16-0.83; ChI >2 points: OR = 0.09; 95%CI = 0.04-0.23). The sensitivity analyses showed no gross differences compared to the principal analysis.

Conclusions: The profile of COVID-19 patients who died without ICU admission is similar to that observed in the usual medical practice before the pandemic. The basal characteristics limiting their admission were age and global burden due to comorbidity, especially dementia and active cancer.

Keywords: COVID-19; Características clínicas; Clinical characteristics; Comorbidities; Comorbilidades; Emergency departments; SARS-CoV-2; Servicios de urgencias.

Figure 1

Distribution of patient age, Charlson comorbidity index and the time elapsed from admission to death of the patients according to whether they were previously admitted to the ICU or not. ICU: Intensive Care Unit.

Figure 2

Estimation of the effect of the different basal characteristics of the deceased patients with COVID-19 included in the multivariate model upon the probability of admission to the ICU before death. *No patients with dementia were admitted to the ICU; the confidence interval therefore cannot be calculated. Ref.: reference; CI: confidence interval; ICU: Intensive Care Unit. The variables in boldface showed statistically significant differences (p < 0.05).

Figure 3

Estimation of the age-adjusted effect of comorbidity burden assessed by the Charlson index upon the probability of admission of the patients with COVID-19 to the ICU before death. ICU: Intensive Care Unit.

Figure 4

Segmented analysis according to care burden in the emergency department during the pandemic of the age-adjusted effect of comorbidity burden (Charlson index) upon the probability of admission of the patients with COVID-19 to the ICU before death (interaction p-value: 0.33). ICU: Intensive Care Unit; HED: Hospital Emergency Department.