Bacterial Pneumonia in Brain-Dead Patients: Clinical Features and Impact on Lung Suitability for Donation*. Issue 12 (December 2020)

• Record Type: Journal Article
• Authors: Poignant, Simon; Boutrot, Maxime; Ehrmann, Stephan; Remerand, Francis; Venhard, Jean-Christophe; Wolf, Manuel; Boulain, Thierry; Barbier, François
• Is Part Of: Critical care medicine. Volume 48:Issue 12(2020)
• Publication Date: 2020-12
• Subjects: antimicrobial therapy -- brain death -- intensive care unit -- lung donor -- pneumonia -- prevention; Critical care medicine -- Periodicals; Soins intensifs -- Périodiques; 616.028
• DOI: 10.1097/CCM.0000000000004631
• Languages: English
• ISSNs: 0090-3493

Abstract

Abstract : Objectives: To appraise the epidemiological features of bacterial pneumonia and its impact on lung suitability for donation in brain-dead patients managed with protective ventilatory settings. Design: Retrospective observational study. Setting: Six ICUs from two university-affiliated hospitals. Patients: Brain-dead adult patients managed in the participating ICUs over a 4-year period. Interventions: None. Measurements and Main Results: Among the 231 included patients, 145 (62.8%) were classified as ideal or extended-criteria potential lung donors at ICU admission and the remaining 86 patients having baseline contraindication for donation. Culture-proven aspiration pneumonia and early-onset ventilator-associated pneumonia occurred in 54 patients (23.4%) and 15 patients (6.5%), respectively (overall pneumonia incidence, 29.9%). Staphylococcus aureus and Enterobacterales were the most common pathogens. Using mixed-effects Cox proportional hazard models, age (adjusted hazard ratio, 0.98; 95% CI [0.96–0.99]), anoxic brain injury (3.55 [1.2–10.5]), aspiration (2.29 [1.22–4.29]), and not receiving antimicrobial agents at day 1 (3.56 [1.94–6.53]) were identified as independent predictors of pneumonia occurrence in the whole study population. Analyses restricted to potential lung donors yielded similar results. Pneumonia was associated with a postadmission decrease in the PaO2 /FIO2 ratio and lower values at brain death, in the whole study population (estimated marginal mean, 294 [264–323] vs 365 [346–385] mm Hg in uninfected patients; p = 0.0005) as in potential lung donors (299 [248–350] vs 379 [350–408] mm Hg; p = 0.04; linear mixed models). Lungs were eventually retrieved in 31 patients (34.4%) among the 90 potential lung donors with at least one other organ harvested (pneumonia prevalence in lung donors (9.7%) vs nondonors (49.2%); p = 0.0002). Conclusions: Pneumonia occurs in one-third of brain-dead patients and appears as the main reason for lung nonharvesting in those presenting as potential lung donors. The initiation of antimicrobial prophylaxis upon the first day of the ICU stay in comatose patients with severe brain injury could enlarge the pool of actual lung donors. Abstract : Supplemental Digital Content is available in the text.