Mortality Rate in ARDS Is Higher in Those Meeting NIH ARDSNet Exclusion Criteria

Abstract

Background:

ARDS mortality is lower among subjects participating in randomized controlled trials (RCT) versus the general ARDS population (23-29% vs. 35-46%). 1 The need to exclude subjects with inordinately high mortality risk is necessary in RCTs to prevent masking the impact of potentially effective treatments. We inquired whether these differences would be apparent in subjects managed clinically using the ARDSNet ventilation protocol based upon ARDSNet eligibility or exclusion criteria.

Methods:

The SFGH ARDS quality assurance database used for tracking lung protective ventilation (LPV) practices was queried to identify 1, 986 subjects meeting Berlin Taskforce criteria for ARDS from July 2002 through December 2017. Subjects had been differentiated by enrollment criteria into an ARDSNet trial (eg, exclusions based on comorbidities, moribund conditions). Mortality was assessed at 90 days after ARDS onset by two-sided chi-square test with Yates correction. Cox proportional hazard model was used to compare the risk of death between subjects eligible for or excluded from enrollment into a RCT. Alpha was set at 0.05.

Results:

Day 90 mortality was significantly lower in those meeting eligibility criteria vs. exclusion criteria (29% vs 52%, P< 0.001). After adjusting for potential confounding variables, the mortality risk for those meeting RCT eligibility criteria was approximately half that of those excluded from enrollment: HR (95% CI) : 0.55 (0.47-0.64), P < 0.001 (Fig).

Conclusions:

Among subjects not enrolled into an RCT, but managed using the ARDSNet ventilator protocol, mortality differences were apparent based upon standard ARDSNet RCT eligibility and exclusion criteria. This likely reflects differences in the presence of comorbidities. Moreover, the relatively precipitous drop in survival over the first week of ARDS underscores the importance of excluding this cohort from RCTs as it suggests insufficient time for potentially beneficial treatments to demonstrate efficacy.

1. Pais F, et al. Influence of clinical factors and exclusion criteria on mortality in acute respiratory distress syndrome between observational studies and randomized controlled trials. Respir Care 2018;63(8):1060-1069.

RCT 0 = Exclusion Cohort, RCT 1 = Eligibility Cohort