Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Observational Study
. 2023 Oct;78(10):990-1003.
doi: 10.1136/thorax-2023-220262. Epub 2023 Jul 26.

Estimating the attributable fraction of mortality from acute respiratory distress syndrome to inform enrichment in future randomised clinical trials

Rohit Saha  1   2 , Tài Pham  3   4 , Pratik Sinha  5 , Manoj V Maddali  6 , Giacomo Bellani  7 , Eddy Fan  8 , Charlotte Summers  9 , Abdel Douiri  10 , Gordon D Rubenfeld  11 , Carolyn S Calfee  12 , John Gerard Laffey  13   14 , Daniel Francis McAuley  15   16 , Manu Shankar-Hari  17 ; LUNG-SAFE investigators
Collaborators, Affiliations
Observational Study

Estimating the attributable fraction of mortality from acute respiratory distress syndrome to inform enrichment in future randomised clinical trials

Rohit Saha et al. Thorax. 2023 Oct.

Abstract

Background: Efficiency of randomised clinical trials of acute respiratory distress syndrome (ARDS) depends on the fraction of deaths attributable to ARDS (AFARDS) to which interventions are targeted. Estimates of AFARDS in subpopulations of ARDS could improve design of ARDS trials.

Methods: We performed a matched case-control study using the Large observational study to UNderstand the Global impact of Severe Acute respiratory FailurE cohort. Primary outcome was intensive care unit mortality. We used nearest neighbour propensity score matching without replacement to match ARDS to non-ARDS populations. We derived two separate AFARDS estimates by matching patients with ARDS to patients with non-acute hypoxaemic respiratory failure (non-AHRF) and to patients with AHRF with unilateral infiltrates only (AHRF-UL). We also estimated AFARDS in subgroups based on severity of hypoxaemia, number of lung quadrants involved and hyperinflammatory versus hypoinflammatory phenotypes. Additionally, we derived AFAHRF estimates by matching patients with AHRF to non-AHRF controls, and AFAHRF-UL estimates by matching patients with AHRF-UL to non-AHRF controls.

Results: Estimated AFARDS was 20.9% (95% CI 10.5% to 31.4%) when compared with AHRF-UL controls and 38.0% (95% CI 34.4% to 41.6%) compared with non-AHRF controls. Within subgroups, estimates for AFARDS compared with AHRF-UL controls were highest in patients with severe hypoxaemia (41.1% (95% CI 25.2% to 57.1%)), in those with four quadrant involvement on chest radiography (28.9% (95% CI 13.4% to 44.3%)) and in the hyperinflammatory subphenotype (26.8% (95% CI 6.9% to 46.7%)). Estimated AFAHRF was 33.8% (95% CI 30.5% to 37.1%) compared with non-AHRF controls. Estimated AFAHRF-UL was 21.3% (95% CI 312.8% to 29.7%) compared with non-AHRF controls.

Conclusions: Overall AFARDS mean values were between 20.9% and 38.0%, with higher AFARDS seen with severe hypoxaemia, four quadrant involvement on chest radiography and hyperinflammatory ARDS.

Keywords: ARDS; clinical epidemiology.

PubMed Disclaimer

Conflict of interest statement

Competing interests: CS is funded by UKRI (MR/S035753/1 and MR/X005070/1) and the National Institute for Health and Care Research (NIHR133788). Work in her research group is supported by GlaxoSmithKline, the Wellcome Trust and the Cambridge NIHR Biomedical Research Centre, and she has received consultancy fees from AbbVie, Sanofi and GlaxoSmithKline. CSC is funded by National Institutes of Health R35-HL140026 and reports grant funding from Roche-Genentech and Quantum Leap Healthcare collaborative, in addition to the National Institutes of Health, and consulting fees from Vasomune, Gen1e Life Sciences, Cellenkos and Janssen. EF reports personal fees from ALung Technologies, Aerogen, Baxter, GE Healthcare, Inspira and Vasomune outside the submitted work. JGL is funded by a Future Research Leaders Award (16-FRL-3845) from Science Foundation Ireland and reports receiving consulting fees from Baxter and from GlaxoSmithKline. RS reports no conflicts. MS-H is funded by a clinician scientist fellowship from the National Institute for Health Research (CS-2016-16-011) and reports receiving grants from the NIHR, MRC, EME, HTA, Huo Foundation and highlights industry support for TRAITS research programme (a Chief Scientists Office, Scotland funded time critical precision medicine in adult critically ill patients (TRAITS Programme)).

Figures

Figure-1
Figure-1. Flowchart of patients screened and included in the models used to generate overall and subpopulation estimates of AFARDS, AFAHRF and AFAHRF-UL
AF is the proportion of individuals with the outcome of interest e.g. death that can be attributed to the exposure e.g. ARDS. For example, AFARDS= [(Deaths in ARDS – Deaths in non-ARDS)/Deaths in ARDS]. Comparisons used to generate overall estimates for AFARDS, AFAHRF, and AFAHRF-UL are shown in the black rectangles. Further details on each model, and the AF estimates generated are provided in the table. To generate subpopulation estimates, analysis was stratified by severity of hypoxaemia, maximum number of quadrants involved in the first 48 hours, and ARDS sub-phenotype. ARDS: acute respiratory distress syndrome; AF: attributable fraction; AHRF: acute hypoxaemic respiratory failure; AHRF-UL: acute hypoxaemic respiratory failure with unilateral infiltrates only
Figure-2
Figure-2. Overall and subpopulation estimates of AFARDS
Figure-2a. The fraction of deaths attributable to the ARDS exposure was ascertained using proportions. Propensity for ARDS logistic regression models were used to derive estimates for AFARDS in model-1. Bar graph shows the mortality difference between ARDS population compared with propensity matched non-AHRF controls (Model-1). Analysis was then stratified by severity of hypoxemia, maximum number of quadrants involved in the first 48 hours, and ARDS hypo/hyper-inflammatory subphenotype. Subpopulation AF ARDS estimates from model-1 are shown in the forest plot. Figure-2b. Propensity for ARDS logistic regression models were then used to derive estimates for AFARDS in model-2. Bar graph shows the mortality difference between ARDS population compared with propensity matched controls who had AHRF with unilateral infiltrates (Model-2). Analysis was also stratified by severity of hypoxaemia, maximum number of quadrants involved in the first 48 hours, and ARDS hypo/hyper-inflammatory subphenotype. Subpopulation AFARDS estimates from model-2 are shown in the forest plot. AF: attributable fraction; ARDS: acute respiratory distress syndrome; AHRF: acute hypoxemic respiratory failure; CI: confidence interval; RD: risk difference.
Figure-3
Figure-3. Overall and subpopulation estimates of AFAHRF and AFAHRF-UL
Figure-3a. Bar graphs show the mortality difference between AHRF population compared with propensity matched non-AHRF controls. AFAHRF estimates stratified by severity of hypoxaemia and maximum number of quadrants involved in the first 48 hours are shown in the forest plot. Figure-3b. Bar graphs show the mortality difference between AHRF-UL population compared with propensity matched non-AHRF controls. AFAHRF-UL estimates stratified by severity of hypoxemia and maximum number of quadrants involved in the first 48 hours are shown in the forest plot. AF: attributable fraction; AHRF: acute hypoxemic respiratory failure; AHRF-UL: acute hypoxemic respiratory failure with unilateral infiltrates only; CI: confidence interval; RD: risk difference.
Figure-4
Figure-4. Illustrative examples of sample size calculations for different AFARDS scenarios
These curves illustrate the AFARDS principle. Each curve represents the sample sizes required for different AF estimates (when control event rate is fixed at 40%). We show the estimates of AFARDS from Model-1, stratified by (a) severity of hypoxaemia, (b) maximum number of quadrants involved on chest radiography at 48 hours, and (c) sub-phenotype of ARDS. We contrast these against the common assumption that AFARDS is expected to be 100%. Dot plots represent ARDS RCTs with mortality as primary outcome identified previously in our systematic review[12]; they correspond to the actual RRR used for sample size estimation and sample size per group in these RCTs. Median (IQR) control group mortality used for sample size calculations in these RCTs was 45.0% (33.3% - 52.5%) and RRR was 29.0% (24.5% - 33.3%). Most trials aimed for 80% power and 5% alpha. The sample size per group varied between 53 to 704 patients. RCTs above a curve will have an adequate sample size to detect the predicted RRR. RRR; relative risk reduction, AF; attributable fraction, ARDS: acute respiratory distress syndrome.
See this image and copyright information in PMC

Comment in

  • Can we design better ARDS trials?
    Hammond NE, Finfer S. Hammond NE, et al. Thorax. 2023 Oct;78(10):955-956. doi: 10.1136/thorax-2023-220446. Epub 2023 Jul 26. Thorax. 2023. PMID: 37495366 No abstract available.

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Force ADT, Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, Camporota L, Slutsky AS: Acute respiratory distress syndrome: the Berlin Definition. JAMA 2012, 307(23):2526–2533. - PubMed
    1. Matthay MA, Zemans RL, Zimmerman GA, Arabi YM, Beitler JR, Mercat A, Herridge M, Randolph AG, Calfee CS: Acute respiratory distress syndrome. Nat Rev Dis Primers 2019, 5(1):18. - PMC - PubMed
    1. Saha R, Assouline B, Mason G, Douiri A, Summers C, Shankar-Hari M: Impact of differences in acute respiratory distress syndrome randomised controlled trial inclusion and exclusion criteria: systematic review and meta-analysis. Br J Anaesth 2021, 127(1):85–101. - PMC - PubMed
    1. Sackett DL: Superiority trials, non-inferiority trials, and prisoners of the 2-sided null hypothesis. Evid Based Med 2004, 9(2):38–39.
    1. Shankar-Hari M, Santhakumaran S, Prevost AT, Ward JK, Marshall T, Bradley C, Calfee CS, Delucchi KL, Sinha P, Matthay MA et al.: Defining phenotypes and treatment effect heterogeneity to inform acute respiratory distress syndrome and sepsis trials: secondary analyses of three RCTs. 2021, 8:10. - PubMed

Publication types