Do hypooncotic fluids for shock increase the risk of late-onset acute respiratory distress syndrome?

Objective In patients with shock, late-onset acute respiratory distress syndrome (ARDS) carries poor prognosis. Hypooncotic fluids may improve kidney function preservation, whereas hyperoncotic fluids may in theory decrease the risk of late-onset ARDS. Our objective was to determine whether predominant or exclusive use of crystalloids and/or hypooncotic colloids for shock resuscitation influenced the risk of late-onset ARDS. Participant and settings International prospective cohort of consecutive adults who were free of ARDS on admission and who received fluid resuscitation for shock in 115 intensive care units (ICUs) during a 4-week period. Measurements and results Severity scores, hemodynamic status, indication for fluids, risk factors for ARDS, plasma expander use, transfusions, and late-onset ARDS were recorded prospectively. Logistic regression models were tested to determine whether predominant or exclusive use of hypooncotic fluids was associated with higher incidence of late-onset ARDS. Of 905 patients, 81 [8.9%; 95% confidence interval (CI) 7.2–11.0] developed ARDS, with no difference between patients given only hypooncotic fluids (10.4%; 95% CI 7.6–13.7) and the other patients (7.7%; 95% CI 5.5–10.5; p = 0.16). Late-onset ARDS was significantly associated with sepsis [odds ratio (OR) 1.90; 95% CI 1.06–3.40], worse chest X-ray score at fluid initiation (1.55; 95% CI 1.27–1.91), positive fluid balance (1.06 per l; 95% CI 1.02–1.09), and greater transfusion volume (1.14 per l; 95% CI 1.01–1.29). The proportion of hypooncotic fluids in the plasma expander regimen was not associated with late-onset ARDS (1.01 per %; 95% CI 0.99–1.01). Conclusions Based on this observational study, there is no evidence that in patients with shock the use of hypooncotic fluids increases the risk of late-onset ARDS. This finding needs to be confirmed.