Secular Trends in ARDS: A 20-Year Analysis of the National Inpatient Sample

Mortality from ARDS remains high, but large-scale longitudinal analyses in the pre-COVID-19 era are limited.

How has in-hospital mortality changed over time among patients with mechanically ventilated ARDS in the US between 2000 and 2019?

We performed a retrospective cohort study using the National Inpatient Sample from 2000 to 2019. Adult, nonelective admissions with a diagnosis of ARDS and concurrent invasive mechanical ventilation were included. The cohort was stratified by International Classification of Diseases coding era (International Classification of Diseases Ninth Revision, Clinical Modification [ICD-9-CM]: 2000 to quarter 3, 2015; International Classification of Diseases 10th Revision, Clinical Modification [ICD-10-CM]: quarter 4, 2015 to 2019). Outcomes included in-hospital mortality, length of stay, and hospitalization charges. Hospitalization charges were adjusted for Consumer Price Index for hospital services to 2019 dollars. Adjusted trend analyses were performed separately for each coding era using logistic and linear regression models.

Among 205,393 mechanically ventilated ARDS admissions (ICD-9-CM: n = 146,888; ICD-10-CM: n = 58,505), mortality declined during the ICD-9-CM period (OR, 0.96 per year; 95% CI, 0.95-0.97; P < .001) but increased during the ICD-10-CM period (OR, 1.05 per year; 95% CI, 1.01-1.08; P = .004). The LOS and hospitalization charges decreased during the ICD-9-CM era but showed no significant improvement in the ICD-10-CM period, likely reflecting the rising cost of managing patients with more complex ARDS during the ICD-10-CM period. The ICD-10-CM cohort had a greater comorbidity burden and were more likely to have a pulmonary etiology for ARDS.

Our results show that this 20-year analysis reveals diverging trends in ARDS outcomes across coding eras. Improvements in mortality during the ICD-9-CM era were not sustained in the ICD-10-CM period. The apparent changes in mortality trend during the ICD-10-CM period are likely artifacts of coding and more severe case-mix. These findings underscore the ongoing burden of ARDS and highlight the need for consistent case definitions and identification to standardize care delivery.