Based on the limitations that the partial pressure arterial oxygen (PaO2)-fraction of inspired oxygen (FiO2) ratio shows in the classification of severity in patients with acute respiratory distress syndrome (ARDS),1 recently, Martos-Benítez et al.2 assessed the severity of ARDS using the FiO2-adjusted PaO2 and positive end-expiratory pressure (PEEP): PaO2/(FiO2 x PEEP) or P/FPE ratio.3

Due to the differences between “traditional” ARDS and ARDS due to COVID-19 (C-ARDS) like the underlying cause, mechanism of pulmonary lesion, clinical presentation, and therapeutic management,4 we believe that the hypothesis that the utility of the P/FPE ratio to predict mortality in C-ARDS is not the same for patients with “traditional” ARDS.

A cohort of 507 patients with C-ARDS treated at Hospital Universitario Marqués de Valdecilla, Santander, Spain, and another cohort of 217 patients with “traditional” ARDS previously published by Martos-Benítez et al.2 were studied using the inclusion and exclusion criteria described by them. In both cases, the registries used had been accepted by the local Research Ethics Committee of each center. In the Spanish series, informed consent from the patients/legal representatives (written or over the phone) was required. In the Cuban series consent was not required due to the retroactive nature of the data collected. The severity of ARDS was evaluated within the first 24 h after starting invasive mechanical ventilation with the patient while in the supine position. All individuals were categorized as having mild, moderate or severe ARDS according to the Berlin definition.5 Based on the P/FPE ratio, patients were categorized into mild (40 < P/FPE ratio ≤ 60), moderate (20 < P/FPE ratio ≤ 40) or severe ARDS (P/FPE ratio ≤ 20).

Descriptive statistics and a multivariate logistics regression analysis to explore the impact of ventilatory configurations and respiratory indices on in-hospital mortality were used. The area under the operator receiving characteristics curve (AUC) was used to assess the performance of respiratory indices. A multivariate logistics regression analysis was used to explore the impact of respiratory indices with the following primary endpoint: 28-day mortality rate. To interpret the results the following covariables were included: the characteristics that showing significant differences in the bivariate analysis would not show any muticollinearity issues (assessed with a Variance Inflation Factor (VIF) < 3). The models analyzed are expressed as odds ratio (OR) with its 95% confidence interval (95%CI).

The characteristics of the 506 patients from the cohort of patients with C-ARDS are shown on Table 1. The median of the PaO2/FiO2 ratio in the C-ARDS cohort was 110 mmHg (p25–75: 90–134 mmHg) vs 187 mmHg (p25–75: 117–221 mmHg) in the “traditional” ARDS cohort. The median of the P/FPE ratio was 9.5 (p25–75: 7.2–12.16) in the C-ARDS group vs 21.6 (p25–75: 10.2–33.2) in the traditional” ARDS cohort. Same as it happened in the cohort of patients with “traditional” ARDS described by Martos-Benítez et al.2 in patients with C-ARDS, the P/FPE ratio was reclassified in most patients into a different category of severity compared to the Berlin classification (Table 2). In the logistics regression multivariate analysis, the P/FPE ratio was not associated with in-hospital mortality in the C-ARDS cohort (odds ratio [OR], 0.97; 95%CI, 0.93–1.01; P = .187), but it was indeed associated with a lower mortality rate in the cohort of patients with “traditional” ARDS (OR, 0.93; 95%CI, 0.89−0.97; P = .001). Same as it happened with “traditional” ARDS, conduction pressure (OR, 1.09; IC95%:1.00–1.20; P = .046) and the SOFA value (OR, 1.75; 95%CI, 1.26–2.43; P < .01) kept a significant correlation with mortality. In the C-ARDS cohort, the P/FPE ratio in term of AROC was 0.54 (95%CI, 0.49−0.58) vs 0.52 (95%CI, 0.47−0.56) of the PaO2/FiO2 ratio. In the “traditional” ARDS cohort, the P/FPE ratio in term of AROC was 0.72 (95%CI, 0.65−0.78) vs 0.63 (95%CI, 0.54−0.69) of the PaO2/FiO2 ratio.

These results prove that the P/FPE ratio behaves different when it comes to predicting mortality between patients with “traditional” ARDS and those with C-ARDS. Also, the clinical-epidemiological characteristics of both groups, and the reasons for this difference can be explained by the severity of hypoxemia (wrongfully called “happy hypoxemia”), and the heterogeneity of pulmonary damage in patients with C-ARDS compared to patients with “traditional” ARDS due to the microthrombotic component of COVID-19, and the predominant change of the hypoxic pulmonary vasoconstriction reflex.6 This heterogeneity could lead to lower values of the P/FPE ratio in patients with C-ARDS.7

The addition of the PEEP value to the PaO2/FiO2 ratio seems appealing because it takes into account the compliance of the respiratory system, as well as pulmonary recruitment.8 Therefore, it can be useful to identify patients with ARDS who may benefit from individual therapies.9,10 However, in patients with C-ARDS, both the PaO2/FiO2 and the P/FPE ratios do not perform well when it comes to predicting the in-hospital mortality rate. We should double down to identify better predictors in this group of patients.