Morales‑Quinteros et al. Crit Care (2021) 25:171 https://doi.org/10.1186/s13054-021-03570-0 RESEARCH Open Access Dead space estimates may not be independently associated with 28‑day mortality in COVID‑19 ARDS Luis Morales‑Quinteros1,2,3* , Ary Serpa Neto4,5,6,7, Antonio Artigas2,3,8,9, Lluis Blanch2,3,8,9, Michela Botta4, David A. Kaufman10, Marcus J. Schultz4,11,12, Anissa M. Tsonas4, Frederique Paulus4 and Lieuwe D. Bos4 on behalf of the PRoVENT-COVID Study Group Abstract Background: Estimates for dead space ventilation have been shown to be independently associated with an increased risk of mortality in the acute respiratory distress syndrome and small case series of COVID-19-related ARDS. Methods: Secondary analysis from the PRoVENT-COVID study. The PRoVENT-COVID is a national, multicenter, retro‑ spective observational study done at 22 intensive care units in the Netherlands. Consecutive patients aged at least 18 years were eligible for participation if they had received invasive ventilation for COVID-19 at a participating ICU dur‑ ing the first month of the national outbreak in the Netherlands. The aim was to quantify the dynamics and determine the prognostic value of surrogate markers of wasted ventilation in patients with COVID-19-related ARDS. Results: A total of 927 consecutive patients admitted with COVID-19-related ARDS were included in this study. Esti‑ mations of wasted ventilation such as the estimated dead space fraction (by Harris–Benedict and direct method) and ventilatory ratio were significantly higher in non-survivors than survivors at baseline and during the following days of mechanical ventilation (p < 0.001). The end-tidal-to-arterial ­PCO2 ratio was lower in non-survivors than in survivors (p < 0.001). As ARDS severity increased, mortality increased with successive tertiles of dead space fraction by Harris– Benedict and by direct estimation, and with an increase in the VR. The same trend was observed with decreased levels in the tertiles for the end-tidal-to-arterial ­PCO2 ratio. After adjustment for a base risk model that included chronic comorbidities and ventilation- and oxygenation-parameters, none of the dead space estimates measured at the start of ventilation or the following days were significantly associated with 28-day mortality. Conclusions: There is significant impairment of ventilation in the early course of COVID-19-related ARDS but quanti‑ fication of this impairment does not add prognostic information when added to a baseline risk model. Trial registration: ISRCTN04346342. Registered 15 April 2020. Retrospectively registered. Keywords: Acute respiratory distress syndrome, ARDS, Respiratory dead space, Dead space, Ventilatory ratio, COVID- 19, Mortality, Prognostication Background Since the outbreak of coronavirus disease 2019 (COVID- *Correspondence: luchomq2077@gmail.com 19) in the City of Wuhan, Hubei Province, China, caused 1 Intensive Care Unit, Hospital Universitari General de Catalunya, by the transmission of the novel coronavirus SARS- Grupo Quironsalud, Carrer Pedro i Pons, 1, 08195 Sant Cugat del Vallès, CoV-2, millions of individuals have been infected and Barcelona, Spain Full list of author information is available at the end of the article more than one million have died. Severe disease requiring © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 2 of 13 admission to intensive care unit (ICU) occurs in approxi- Clinical Practice–guidelines. Integrity and timely com- mately 5% of infections [1], and the most common reason pletion of data collection was ensured by the study for admission is respiratory failure requiring high-level coordinators. support. Among these patients, two-thirds meet the cri- teria for the acute respiratory distress syndrome (ARDS) Patients [2]. Consecutive patients ≥ 18 years of age were eligible for Patients with COVID-19 pneumonia meeting criteria participation in PRoVENT-COVID if they were admit- for ARDS usually present with a high respiratory drive and ted to one of the participating ICUs and had received minute ventilation, potentially due to hypercapnia and an invasive ventilation for COVID-19 ARDS. COVID- increased dead space fraction (VD/VT) [3]. In patients with 19 infection was defined by a confirmed reverse tran- ARDS, an elevated VD/VT is a predictor of death and is one scriptase-polymerase chain reaction (RT-PCR) [13]. of the few lung-specific physiological variables indepen- PRoVENT-COVID had no exclusion criteria, but for dently associated with mortality [4, 5]. Methods for estimat- the current analysis, we excluded patients transferred ing VD/VT do not require quantitative assessment of exhaled from a non-participating hospital who had been receiv- carbon dioxide, are less complicated to perform and easier ing invasive ventilation for more than 2 calendar days, to calculate at the bedside compared with calculations made patients without complete data to calculate the ­VD/VT or by volumetric capnography [6]. In recent years, the ventila- VR on the first day of ventilation, and patients with no tory ratio (VR) was proposed as an easily acquired bedside data about 28-day mortality. index of impaired ventilation that can be computed using routinely measured respiratory variables [7]. In patients Data collection with ARDS, the VR correlates well with VD/VT [7] and may Demographics and data regarding premorbid diseases function as a surrogate marker for impaired ventilation [8]. and medication were collected at baseline. Ventilator set- At least two independent groups have described series tings and parameters were collected after one hour of of patients with COVID-19-related ARDS who may have invasive ventilation and every 8 h thereafter, for the first 4 inefficient ­CO2 removal due to increased physiologic dead calendar days. In the present study, the first day of venti- space [3, 9]. However, few studies have assessed the impact lation is called ‘at start of ventilation.’ for dead space ventilation mortality in a large cohort of COVID-19 patients undergoing invasive ventilation [10]. Therefore, we aimed to assess the association between Data definition and exposure markers of impaired ventilation, such as estimated VD/VT The primary exposure of interest was the ­VD/VT calcu- and VR with 28-day mortality in patients undergoing inva- lated using the Harris–Benedict formula as described in sive ventilation because of COVID-19 ARDS. We hypoth- Eq. (1) [14]: esized that these markers of impaired ventilation are   independently associated with 28-day mortality. VD 0.863 ∗ V̇ CO2 =1− (1) VT (RR ∗ VT ∗ PaCO2 ) Methods RR is the respiratory rate in breaths per minute, ­VT the Study design and oversight tidal volume in liters, ­PaCO2 the partial pressure of car- PRoVENT-COVID is an investigator-initiated, multi- bon dioxide in mmHg, and ­VCO2 the ­CO2 production in center, observational cohort study undertaken at 22 ICUs mL/min estimated using Eq. (2): in the Netherlands. The study protocol including the statistical analysis plan is available [11]. The approved REEHB V̇ CO2 =  (2)  protocol is available in Additional File 1. A statistical 5.616 RQ + 1.584 analysis plan for the current analysis was written before assessing the database and is available online [12]. Study RQ is the respiratory quotient, assumed to be 0.8, and sites were recruited through direct contact by members ­REEHB is the rest energy expenditure calculated by the of the steering committee of PRoVENT-COVID. The unadjusted Harris–Benedict estimate using Eq. (3) [14]: institutional review boards of the participating cent-   ers approved the study protocol, and need for patient Males : REEHB = 66.473 + 13.752 ∗ weight informed consent was waived. Study coordinators con-     + 5.003 ∗ height − 6.755 ∗ age tacted the local doctors, trained data collectors to assist   Females : REEHB = 655.096 + 9.563 ∗ weight the local doctors and monitored the study according to     the International Conference on Harmonization Good + 1.850 ∗ height − 4.676 ∗ age (3) Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 3 of 13 Weight is the actual body weight in kilograms, height is time of the variables was assessed with mixed–effect lin- in centimeters and age in years. ear models with center and patients treated as random In addition, two additional estimations of ­VD/VT were effect to account for clustering and repeated measure- done considering a direct estimation [6] and the end- ments, and with 28-day vital status (alive/dead), time (as tidal-to-arterial ­PCO2 ratio [15], and the formulas are a continuous variable) and an interaction of 28-day vital described in the Additional File 1. status and time as fixed effect. Overall P values from this The secondary exposure of interest is the VR, calcu- analysis represent the overall difference among groups lated using Eq. (4) [16]: over time, and P values from interaction represent a statistical assessment of whether the trend over time V̇Emeasured ∗ PaCO2measured differed among the groups. All daily measurements of VR = (4) V̇Epredicted ∗ PaCO2predicted variables (three times a day) were aggregated as the mean per day. In addition, to compare variables at each day, the VR is the ventilatory ratio, VE measured is the measured day variable was entered as a categorical variable in the minute ventilation in mL/min, P ­ aCO2 measured is the meas- model described above, and the P value for the daily dif- ured ­PaCO2 in mmHg, VE predicted is the predicted minute ference was extracted using pairwise comparisons after ventilation in mL/min (calculated as 100 * predicted body Bonferroni correction. weight) [16], and P­ aCO2 predicted is the predicted P ­ aCO2 We examined the risk of death for each tertile of the determined as 37.5 mmHg. lung-specific physiological variables used to evaluate A post-hoc analysis was performed using Corrected whether the predictive ability of each variable varied by Minute Ventilation as additional parameter of wasted level. In addition, a simple stratification creating two ventilation. This parameter is calculated using the follow- groups according to the median of the variables was also ing formula: assessed. The comparison of the two groups was pre- sented in Kaplan–Meier curves and compared using Log- V̇E ∗ PaCO2 V̇E corr = (5) rank tests. 40 mmHg Univariable mixed-effect generalized linear mod- where 40 mm Hg is the ideal value of ­PaCO2 [17]. This is els considering a binomial distribution and with center reported in the Additional File 1. Additionally, the delta as random effect were used to estimate the unadjusted values between days were calculated and used as addi- effect of each variable on 28-day mortality. A multivari- tional parameters and reported in the Additional File 1. able mixed-effect generalized linear model considering All variables were calculated three times per day, and a binomial distribution and with center as random effect the values were aggregated as the mean in the respective were used to test the association of each of the exposures day. Primary analyses focused on the values obtained on described above with 28-day mortality. A list of candidate the day on which ventilation was started. confounders was determined a priori, and based on clini- cal relevance rather than statistical significance. The fol- Outcomes lowing baseline variables (measured at baseline or within The outcome assessed in this study was death at 28 days, 1 h after intubation or ICU admission with ventilation) defined as the mortality 28 days after the start of ven- were considered in the models: age, gender, body mass tilation. Other clinical outcomes are reported only to index, ­PaO2/FiO2 ratio, plasma creatinine, hypertension, describe the cohort but were not used to test their asso- diabetes, use of angiotensin converting enzyme inhibi- ciation with the exposures described above. tors, use of angiotensin II receptor blockers, use of a vasopressor or an inotrope drug, fluid balance, pH, mean Statistical analysis arterial pressure, heart rate, respiratory system compli- The amount of missing data was low for most of the vari- ance and PEEP. Multicollinearity was assessed through ables (Table S1 in Additional File 1). Continuous variables the analysis of the variance inflation factors, and the final are presented as median (quartile 25%–quartile 75%) and model was assessed for discrimination using c-statistics, categorical variables as counts and percentages. Descrip- and for calibration using the Brier-Score. tive data are presented according to the 28-day status In addition to the odds ratio (OR) and its 95% con- (non-survivors vs. survivors), and the two groups were fidence interval, the predictive accuracy of the lung- compared using Wilcoxon rank-sum test for continuous specific physiological variables was measured by the variables, and Fisher exact tests for categorical variables. area under the receiver operating characteristics curve Trends in markers of impaired ventilation were pre- (AUC-ROC). Also, to estimate whether these vari- sented in boxplots between survivors and non-survivors ables improved predictive accuracy on top of that of over the first 4 calendar days. The direction of effect over the base model described above, the net reclassification Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 4 of 13 improvement (NRI) and the integrated discrimination Mortality by tertiles of each variable is reported in index (IDI) were assessed. Fig. 2. Tertiles were calculated separately for each vari- For the primary analysis, covariates with less than 3% of able and each day, to account for potential differences in missing values were imputed by the median value of the scaling and measurements. Mortality increased with suc- overall cohort. Since respiratory compliance was missing cessive tertiles of dead space fraction by Harris–Benedict in 8.2% of the patients (Table S1 in Additional File 1), an and by direct estimation, and of ventilatory ratio, and additional sensitivity analysis considering multiple impu- decreased with successive tertiles of end-tidal-to-arterial tation for all missing variables was conducted (described ­PCO2 ratio. in details in Additional File 1). Mortality over the first 28 days was higher in patients All continuous variables were entered after standardi- in the high group of dead space fraction by the Harris– zation to improve convergence of the models, and the Benedict estimation (16.4% vs. 12.3%; p = 0.003), but not odds ratio (OR) represents the increase in one standard different in the groups considering the dead space frac- deviation of the variable. All analyses were conducted in tion by direct estimation (15.4% vs. 13.3%; p = 0.100), and R v.4.0.2 (R Foundation, Vienna, Austria) [18], and signif- the ventilatory ratio (15.5% vs. 13.2%; p = 0.080) (Fig. 3). icance level was set at 0.05. When assessing the end-tidal-to-arterial P ­CO2 ratio, 28-day mortality was lower in the highest tertile group Results (10.7% vs. 17.1%; p < 0.001). Study population From March 1, 2020, through June 1, 2020, 31 ICUs were Predictive accuracy of markers of impaired ventilation invited for participation in PRoVENT-COVID, and 22 of The unadjusted impact of each marker of impaired ven- them met inclusion criteria. A total of 1340 individuals tilation is shown in Additional File 1: Table S4. Estimated were screened. A total of 218 were not enrolled; 62 did dead space fraction (by HB and direct method) and not have COVID-19-related ARDS, 150 never received end-tidal-to-arterial ­PCO2 ratio were associated with invasive ventilation, and 6 were excluded for other rea- 28-day mortality at the start of mechanical ventilation. sons (Additional File 1: Figure S1). Of the enrolled 927 Twenty-four hours after this, dead space fraction by Har- patients, 259 (28.7%) were non-survivors and 661 (71.3%) ris–Benedict and end-tidal-to-arterial ­PCO2 ratio were were survivors at day 28. Demographics characteris- associated with 28-day mortality. The final multivariable tics are presented in Table 1. Non-survivors were older, base risk model is shown in Table S5 and in the Addi- more often male, had a greater severity of ARDS and tional File 1. No problems were found due to multicol- higher creatinine level at admission, more frequent pre- linearity or linearity assumption (Additional File 1: Tables sented co-existing disorders like hypertension, diabetes S6 and S7). and chronic obstructive pulmonary disease, and more After adjustment for the base risk model, none of the often were using systemic steroids, metformin, beta- markers of impaired ventilation measured at the start of blockers and statins at home. Ventilatory variables in the ventilation or the following day was significantly asso- first day of ventilation and general clinical outcomes are ciated with 28-day mortality (Table 4). The inclusion of shown in Table 2, Table S2 and Table S3 (Additional File these variables did not improve the AUC-ROC compared 1). Non-survivors received higher ­FiO2, had higher levels to the base model (Fig. 4). The addition of dead space of lactate and creatinine and lower levels of pH and end- fraction by direct estimation at start of ventilation and tidal-to-arterial ­PCO2. There were no differences in the of end-tidal-to-arterial P­ CO2 ratio at start of ventilation radiographic findings between survivors and non-survi- or day 1 slightly improved the predictive accuracy of the vors (Table S15). base model in terms of IDI (Table 4). Markers of impaired ventilation Sensitivity analysis The dynamic change of markers of impaired ventilation Results after multiple imputation were similar to the pri- over the first four days of ventilation as shown in Table 3 mary analysis (Additional File 1: Tables S8 and S9, and and Fig. 1. VD/VT calculated using the Harris–Benedict Figures S2 and S3). formula was higher in non-survivors and increased more during the first four days compared to survivors. A simi- Post‑hoc analysis lar trend was found with the direct VD/VT calculation. All analyses described above were repeated for Cor- While VR was also higher in non-survivors, especially rected Minute Ventilation and yielded similar results after day 2, the end-tidal-to-arterial ­ PCO2 ratio was (Additional File 1: Tables S9, S10, S11, S12, S13, S14 and lower. Figures S4 and S5). The change in measures of wasted ventilation between day 3 and day 0 of intubation were Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 5 of 13 Table 1 Baseline characteristics of the patients according to 28-day mortality All patients Non-survivors Survivors p (n = 927) (n = 266) (n = 661) value Age, years 65.0 (57.0–72.0) 70.0 (64.0–75.0) 63.0 (55.0–70.0) < 0.001 Male gender—no (%) 668 (72.1) 206 (77.4) 462 (69.9) 0.023 Body mass index, kg/m2 27.8 (25.3–30.8) 27.7 (25.2–29.8) 27.8 (25.4–30.9) 0.227 Severity of ARDS—no (%) 0.020 Mild 186/921 (20.2) 45 (17.0) 141 (21.5) Moderate 638/921 (69.3) 180 (68.2) 458 (69.7) Severe 97/921 (10.5) 39 (14.8) 58 (8.8) Co-existing disorders—no (%) Hypertension 317 (34.2) 112 (42.1) 205 (31.0) 0.002 Heart failure 39 (4.2) 16 (6.0) 23 (3.5) 0.102 Diabetes 207 (22.3) 78 (29.3) 129 (19.5) 0.002 Chronic kidney disease 41 (4.4) 16 (6.0) 25 (3.8) 0.157 Baseline creatinine, µmol/L* 76.0 (62.0–97.0) 84.0 (65.0–112.0) 74.0 (61.0–92.0) < 0.001 Liver cirrhosis 3 (0.3) 0 (0.0) 3 (0.5) 0.562 Chronic obstructive pulmonary disease 76 (8.2) 32 (12.0) 44 (6.7) 0.011 Active hematological neoplasia 15 (1.6) 6 (2.3) 9 (1.4) 0.388 Active solid neoplasia 23 (2.5) 9 (3.4) 14 (2.1) 0.253 Neuromuscular disease 8 (0.9) 3 (1.1) 5 (0.8) 0.696 Immunosuppression 24 (2.6) 7 (2.6) 17 (2.6) 0.999 Previous medication—no (%) Systemic steroids 35 (3.8) 17 (6.4) 18 (2.7) 0.012 Inhalation steroids 108 (11.7) 40 (15.0) 68 (10.3) 0.054 Angiotensin converting enzyme inhibitor 160 (17.3) 55 (20.7) 105 (15.9) 0.084 Angiotensin II receptor blocker 101 (10.9) 35 (13.2) 66 (10.0) 0.164 Vital signs Heart rate, bpm 84.3 (73.7–97.0) 85.0 (74.5–100.3) 84.0 (73.5–96.0) 0.092 Mean arterial pressure, mmHg 80.3 (73.7–88.0) 80.0 (73.0–86.0) 80.5 (74.8–88.3) 0.061 Organ support—no (%) Continuous sedation 892 (96.4) 252 (95.1) 640 (97.0) 0.173 Inotropic or vasopressor 733 (79.2) 216 (81.5) 517 (78.3) 0.324 Vasopressor 732 (79.1) 216 (81.5) 516 (78.2) 0.283 Inotropic 41 (4.4) 17 (6.4) 24 (3.6) 0.077 Fluid balance, mL 631.5 (50.0–1428.2) 753.6 (100.0–1482.0) 593.0 (34.0–1390.0) 0.143 Urine output, mL 702.5 (370.0–1143.8) 570.0 (280.0–1075.0) 725.0 (415.0–1165.0) 0.001 Data are median (quartile 25%—quartile 75%) or No (%). Percentages may not total 100 because of rounding * Most recent measurement in 24 h before intubation, or at ICU admission under invasive ventilation Table 2 Respiratory variables at start of ventilation All Patients Non-Survivors Survivors p value (n = 927) (n = 266) (n = 661) Tidal volume, mL/kg PBW 6.4 (5.9–7.1) 6.5 (5.9–7.2) 6.4 (5.9–7.1) 0.175 PEEP ­(cmH2O) 12.7 (11.0–14.5) 13.0 (11.2–15.0) 12.7 (10.7–14.3) 0.091 Driving pressure ­(cmH2O) 14.0 (12.0–16.0) 13.7 (12.0–16.3) 14.0 (12.0–16.0) 0.746 PaO2/FiO2 130.9 (99.9–175.5) 137.2 (98.1–180.0) 128.8 (100.2–171.5) 0.262 EtCO2 36.5 (32.6–41.6) 34.7 (31.7–40.2) 37.0 (33.2–42.0) 0.001 Mechanical power (J/min) 18.4 (15.2–22.3) 18.9 (15.9–22.8) 18.3 (15.1–22.0) 0.100 Compliance (mL/cmH2O) 33.2 (26.7–40.3) 33.8 (26.7–40.7) 32.9 (26.7–40.1) 0.899 Data are median (quartile 25%—quartile 75%). ­FiO2: inspired fraction of oxygen; PEEP: positive end-expiratory pressure Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 6 of 13 Table 3 Lung-specific physiological variables in the first four days of ventilation according to 28-day mortality All Patients Non-Survivors Survivors p (n = 927) (n = 266) (n = 661) value* Dead space fraction by HB At start of ventilation 0.58 ± 0.11 0.60 ± 0.11 0.58 ± 0.11 < 0.001 Day 01 0.62 ± 0.10 0.64 ± 0.09 0.61 ± 0.11 < 0.001 Day 02 0.64 ± 0.10 0.67 ± 0.09 0.63 ± 0.10 < 0.001 Day 03 0.67 ± 0.09 0.69 ± 0.08 0.65 ± 0.10 < 0.001 p value (interaction survival × day) 0.005 Dead space fraction direct At start of ventilation 2.22 ± 0.55 2.28 ± 0.61 2.19 ± 0.52 0.022 Day 01 2.35 ± 0.54 2.40 ± 0.55 2.33 ± 0.54 0.036 Day 02 2.48 ± 0.58 2.60 ± 0.63 2.44 ± 0.54 < 0.001 Day 03 2.62 ± 0.65 2.77 ± 0.71 2.56 ± 0.61 < 0.001 p value (interaction survival × day) < 0.001 Ventilatory ratio At start of ventilation 1.72 ± 0.60 1.77 ± 0.56 1.70 ± 0.62 0.114 Day 01 1.85 ± 0.64 1.88 ± 0.53 1.84 ± 0.68 0.142 Day 02 1.99 ± 0.66 2.09 ± 0.60 1.95 ± 0.68 < 0.001 Day 03 2.12 ± 0.70 2.26 ± 0.68 2.06 ± 0.70 < 0.001 p value (interaction survival × day) < 0.001 end-tidal-to-arterial ­PCO2 ratio At start of ventilation 0.85 ± 0.18 0.81 ± 0.18 0.87 ± 0.18 < 0.001 Day 01 0.84 ± 0.14 0.81 ± 0.14 0.85 ± 0.13 < 0.001 Day 02 0.82 ± 0.14 0.78 ± 0.15 0.83 ± 0.14 < 0.001 Day 03 0.80 ± 0.15 0.77 ± 0.15 0.81 ± 0.14 < 0.001 p value (interaction survival × day) 0.455 Data are median (quartile 25%—quartile 75%) HB: Harris–Benedict *Calculated using pairwise contrasts in a mixed-effect generalized linear model considering a Gaussian distribution and with day, group and an interaction day × group as fixed effect, and with patients and center as random effect. A binomial distribution was used for binary variables and a Gaussian distribution for continuous significantly associated with outcome (Additional File possible barrier is the requirement of volumetric cap- 1: Table S10). When restricting the analyses to patients nography (or other techniques of analyzing exhaled gas) with hypercapnia at presentation ­(PaCO2 > 50 mmHg) or to measure Vd/Vt. Estimated measures for calculating prone positioning during the first day of mechanical ven- VD/VT are more frequently utilized and a wide array of tilation, we did not observe large changes in effect sizes these indices were included in this study [6, 20]. VR is (Additional File 1: Table S12). a recently validated index in patients under controlled modes of mechanical ventilation. This index was shown Discussion to be high in patients with COVID-19-related ARDS The findings of this multicenter, observational cohort [3, 9] and is known to show moderate correlation with study of COVID-19-related ARDS patients showed that VD/VT by volumetric capnography [7]. We found that estimates for dead space ventilation increased over the the VR was not significantly different between survivors first days of invasive ventilation and were significantly and non survivors at the start of ventilation and on day higher in non-survivors than survivors. However, none of 1. However, we did find a significant difference in the fol- these indices was independently associated with mortal- lowing days of mechanical ventilation between survivors ity when corrected for potential confounders. Therefore, and non-survivors, not only for the VR but also for the wasted ventilation, and, tentatively, increased estimated rest of dead space estimates under study when a post-hoc dead space fraction, may be a marker for disease severity analysis was performed (Table S14). This finding suggests rather than an independent predictor of outcome. the dynamic changes of these estimates over time are Despite [19] the potential clinical value, VD/VT is not much more important than a single variable at the start routinely measured in daily critical care practice. One Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 7 of 13 Fig. 1 Lung-specific physiological variables over the first four days of ventilation. Jitter boxplot of lung-specific physiological variables over the first four days of ventilation of mechanical ventilation, also because this includes the of VD/VT. It is worth noting that that the impact of car- response to optimization of ventilator settings. diac output exists only when measuring VD/VT the Eng- Recently, the end-tidal-to-arterial ­ P CO2 ratio hoff modification of Bohr’s original formula is used. In ­(PETCO2/PaCO2) has been described as another sur- the case of shunt, the increase in venous admixture will rogate for VD/VT in ARDS patients [15]. Each of these elevate the ­PaCO2 increasing dead space fraction [21]. estimations has particular limitations, and they should This contribution is of special importance when VD/VT be seen as complementary: if all point in the same is high, where physiologic dead space can be contami- direction, this likely reflects increased dead space nated by the large shunt fractions present in any type ventilation. For example, in the presence of increased of ARDS. In low cardiac output states, a decrease in intrapulmonary shunt (as in ARDS patients), rising pulmonary blood flow leads to a reduced alveolar C ­ O2 ­PaCO2 coincides with decreasing ­PETCO2. Both shunt delivery decreasing ­PECO2, thereby increasing VD/VT and low cardiac output states are known determinants [22]. In both cases, indices for increase in dead space Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 8 of 13 Fig. 2 28-Day mortality according to tertiles of lung-specific physiological variables over the first four days of ventilation. a Tertiles are < 0.54, 0.54–0.64 and > 0.64 for start of ventilation, < 0.58, 0.58–0.67, > 0.67 for day 1, < 0.62, 0.62–0.69, > 0.69 for day 2, and < 0.64, 0.64–0.71, > 0.71 for day 3; b tertiles are < 1.94, 1.94–2.32 and > 2.32 for start of ventilation, < 2.09, 2.09–2.47, > 2.47 for day 1, < 2.19, 2.19–2.65, > 2.65 for day 2, and < 2.31, 2.31–2.80, > 2.80 for day 3; c tertiles are < 1.45, 1.45–1.80 and > 1.80 for start of ventilation, < 1.57, 1.57–1.98, > 1.98 for day 1, < 1.71, 1.71–2.13, > 2.13 for day 2, and < 1.80, 1.20—2.26, > 2.26 for day 3; and d tertiles are < 0.77, 0.77—0.91 and > 0.91 for start of ventilation, < 0.79, 0.79–0.90, > 0.90 for day 1, < 0.76, 0.76–0.87, > 0.87 for day 2, and < 0.74, 0.74—0.85, > 0.85 for day 3 fraction would capture these phenomena and is hard studies in all patients with ARDS (not only those with to know each part’s relative contribution in practice. COVID19), in which dead space (VD/VT) was elevated Taken together, dead space indices reflect impaired out- during the first week after start of invasive ventilation gassing of ­CO2 because of abnormal ventilation-perfu- [4, 25]. We also described the association between these sion matching giving a good global index of a lung’s gas indices and outcome that was previously observed in exchange efficiency [23, 24]. patients with ARDS due to other causes than COVID-19 Dead space estimations were significantly increased in [4, 25]. However, in our study the investigated estimates non-survivors in the first four days of mechanical venti- did not add predictive value to a model that included lation compared to survivors. This is line with previous other known predictors for 28-day mortality, with the Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 9 of 13 Fig. 3 28-Day survival according to lung-specific physiological variables measured at the start of ventilation. Groups were created according to the median of the variables at start of ventilation; p values from Log-rank tests possible exception of ­PETCO2/PaCO2 at the start and at ventilation in non-COVID-19-related ARDS [8]. Taken day 1 of ventilation. This contrasts with several studies in together, the data suggest that markers of impaired ven- ARDS that showed increased dead space ventilation to tilation reflect disease severity but are not independent be a robust and independent predictor of mortality risk predictors of outcome, irrespective of the cause of ARDS. [4, 25–27]. Decreasing ­ PETCO2/PaCO2 was also inde- Although not the primary aim of this study, we pendently associated with mortality risk in one study observed that patients who were deceased were less [15]. Yet, our findings are in line with a previous report frequently put in prone positioning (Table S2). Prone in which we assessed the added value of markers of position facilitates shape matching, which helps mini- impaired ventilation during the first days of mechanical mizing injurious ventilation and frequently improves gas Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 10 of 13 Table 4 Predictive accuracy of lung-specific physiological variables Odds RATIO* p AUC​ Brier NRI p IDI p (95% CI) value (95% CI) Score (95% CI) value (95% CI) value Base model – – 0.751 (0.715 to 0.788) 0.167 – – – At start of ventilation + Dead space frac‑ 0.91 (0.74 to 1.11) 0.356 0.751 (0.715 to 0.788) 0.167 0.08 (− 0.07 to 0.23) 0.280 0.001 (− 0.000 to 0.165 tion by HB 0.003) + Dead space frac‑ 0.83 (0.67 to 1.02) 0.082 0.754 (0.718 to 0.790) 0.167 0.13 (− 0.02 to 0.28) 0.081 0.005 (− 0.000 to 0.025 tion direct 0.009) + Ventilatory ratio 0.90 (0.71 to 1.14) 0.394 0.752 (0.716 to 0.788) 0.167 0.07 (− 0.08 to 0.22) 0.350 0.001 (− 0.000 to 0.067 0.004) + end-tidal-to-arte‑ 0.93 (0.76 to 1.15) 0.521 0.739 (0.701 to 0.778) 0.168 − 0.13 (− 0.28 to 0.106 − 0.009 (− 0.015 to 0.001 rial ­PCO2 ratio 0.03) − 0.003) Day 01 + Dead space frac‑ 1.05 (0.86 to 1.29) 0.619 0.749 (0.712 to 0.786) 0.170 − 0.07 (− 0.22 to 0.362 − 0.000 (− 0.003 to 0.864 tion by HB 0.08) 0.004) + Dead space frac‑ 0.90 (0.74 to 1.11) 0.326 0.750 (0.713 to 0.787) 0.170 0.05 (− 0.10 to 0.20) 0.495 0.001 (− 0.003 to 0.606 tion direct 0.006) + Ventilatory ratio 0.99 (0.78 to 1.25) 0.938 0.749 (0.712 to 0.786) 0.170 − 0.00 (− 0.15 to 0.996 0.000 (− 0.003 to 0.994 0.15) 0.003) + end-tidal-to-arte‑ 0.87 (0.71 to 1.08) 0.206 0.743 (0.703 to 0.782) 0.170 − 0.09 (− 0.25 to 0.257 − 0.006 (− 0.012 to 0.038 rial ­PCO2 ratio 0.07) − 0.000) HB: Harris–Benedict; CI: confidence interval; AUC: area under the curve; NRI: net reclassification index; IDI: integrated discrimination index *Represents the odds ratio for the lung-specific physiological variables in the multivariable model All models are mixed-effect models with centers as random effect and considering a binomial distribution All continuous variables were entered after standardization to improve convergence of the model, and odds ratio represent the increase in one standard deviation of the variable Fig. 4 Receiver operating characteristics curve of the base model and with the inclusion of lung-specific physiological variables Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 11 of 13 exchange through better V/Q matching resulting in less The results of this study indicate that estimations for shunt and improved C ­ O2 clearance. Therefore, it could increased dead space may not be independently asso- be postulated that prone positioning confounds the rela- ciated with mortality. The observed effect sizes were tion between surrogates of dead space ventilation and remarkably similar to those observed in non-COVID- outcome. A post-hoc analysis, however, did not show a 19-related ARDS. This contrasts several reports that stronger association between these surrogates and out- have hypothesized that profound endothelial injury and comes in patients who did not receive prone position, coagulopathy may be central mediators of lung injury yielding this explanation less likely. in COVID-19 [34]. We acknowledge that we did not In the current study in patients with COVID-19-related measure these processes in this study, but we do pro- ARDS, impaired ventilation was already present in the vide evidence that COVID-19-related ARDS appears to first days of invasive ventilation. The studied estimations be similar to non-COVID ARDS with respect to Vd/Vt. for VD/VT further increased during the first days of inva- This implies that dead space and its estimates should be sive ventilation, especially in patients who did not sur- understood as a readily available marker of ARDS sever- vive. Altered hemostasis and thrombosis are postulated ity. Whether a high dead space identifies an enriched to be a key element of ARDS, with the endothelium play- patient population with a higher prevalence of vascular ing a key role by promoting microthrombogenesis [28– injury, and who might benefit from treatments aimed at 30]. Endothelial infection and activation and disorders of restoring normal pulmonary perfusion is unknown. Pre- the microvasculature have been described in the patho- vious data suggest that some drugs with anticoagulant genesis of COVID-19 [19], and perfusion defects in the properties may decrease VD/VT in patients with ARDS pulmonary arterial circulation are frequently observed [35], making this an attractive hypothesis to consider. [31]. Autopsy findings include pulmonary vascular At the moment, no data exists on the measurement of microthrombi [32] in addition to diffuse alveolar damage. physiologic dead space in COVID-19-related ARDS by These findings could lead to high dead space fraction. integrating volumetric capnography plots of eliminated The strengths of this study include the size of the ­CO2 concentration versus the respective expired tidal multi-center cohort, careful data collection and with volume of a single breath. Since volumetric capnogra- few missing data, the pre-specified analysis plan, and phy offers a more in-depth representation of the kinet- the evaluation of multiple estimations for impaired ven- ics of ­CO2 elimination per breath, the application of tilation. The central limitation of this study is that we longitudinal or time-series models to analyze the effect did not quantify dead space ventilation directly by volu- of ­CO2 elimination impairment on outcome warrants metric capnography or another technique. This was not further research. possible in the setting of a pandemic, where the critical care systems were overwhelmed with patients. A sec- ond limitation is the observational nature of the study. Conclusions Therefore, this study does not provide insight into poten- Estimates for wasted ventilation are abnormal at the start tial mechanisms that may contribute to the association of invasive ventilation in patients with COVID-19-related between high dead space estimations and mortality in ARDS and worsen during consequent days. Ventilation COVID-19-related ARDS patients. Another important impairment seems to be more extensive in non-survi- aspect to take into account is the aspect of the instru- vors than in survivors, but they may not yield prognostic mental dead space. Use of heated humidifiers of HMEs information when added to a baseline risk model. In the (heat and moisture exchangers) is heterogeneous in the absence of bedside capnography, dead space estimates clinical practice, and different HMEs have different dead may serve as an important tool to assess the severity of space volumes. Instrumental dead space may significantly COVID-19-related ARDS along with other variables such affect the total dead space, mainly when using low tidal as oxygenation abnormalities and respiratory mechanics. volume ventilation, and we have commented on this pre- viously [33]. Abbreviations For the estimated VD/VT computed by the Harris–Ben- ARDS: Acute respiratory distress syndrome; AUROCC: Area under the receiver edict formula, we assumed an RQ of 0.8 for V ­ CO2 cal- operating characteristic curve; COVID-19: Coronavirus disease 2019; FiO2: culation based on a previous study [6]. Although the RQ Fraction of inspired oxygen; HB: Harris–Benedict; ICU: Intensive care unit; IBW: Ideal body weight; IDI: Integrated discrimination improvement; MERS: Middle may vary among ARDS patients, a recently previous work East respiratory syndrome; PaCO2: Partial pressure of carbon dioxide in arterial showed a good correlation between the VR (which also blood; PECO2: Expired partial pressure of carbon dioxide; PEEP: Positive end depends on the V ­ CO2) and dead space measured with expiratory pressure; PaO2: Partial pressure of oxygen in arterial blood; PaO2/ FiO2: Arterial oxygen tension to fraction of inspired oxygen; PRoVENT-COVID: volumetric capnography [7]. PRactice of VENTilation in Patients with Coronavirus Disease 2019; SARS: Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 12 of 13 Severe acute respiratory syndrome; V̇CO2: Carbon dioxide consumption; V̇Ecorr: fellowship. The remaining authors have disclosed that they do not have any Corrected minute ventilation; VR: Ventilatory ratio; VD/VT: Dead space fraction. potential conflicts of interest. Author details Supplementary Information 1 Intensive Care Unit, Hospital Universitari General de Catalunya, Grupo The online version contains supplementary material available at https://​doi.​ Quironsalud, Carrer Pedro i Pons, 1, 08195 Sant Cugat del Vallès, Barcelona, org/​10.​1186/​s13054-​021-​03570-0. Spain. 2 Universidad Autonoma de Barcelona, Barcelona, Spain. 3 Institut D’Investigació, Innovació Parc Taulí I3PT, Sabadell, Spain. 4 Department of Intensive Care & Laboratory of Experimental Intensive Care and Anaesthesi‑ Additional file 1. Impaired ventilation is not associated with 28-day ology (L·E·I·C·A), Amsterdam UMC Location AMC, Amsterdam, The Nether‑ mortality in COVID-19 ARDS lands. 5 Department of Critical Care Medicine, Hospital Israelita Albert Einstein, São Paulo, Brazil. 6 Australian and New Zealand Intensive Care Research Centre (ANZIC‑RC), Monash University, Melbourne, Australia. 7 Data Analytics Research Acknowledgements and Evaluation (DARE) Centre, Austin Hospital and University of Melbourne, Not applicable. Melbourne, Australia. 8 Critical Care Center, Corporacion Sanitaria Universitaria Author’s information: A complete list of members of the PRoVENT-COVID Parc Taulí, Sabadell, Spain. 9 CIBER Enfermedades Respiratorias (ISCiii), Madrid, collaborative group is provided: Spain. 10 Division of Pulmonary, Critical Care, and Sleep Medicine, NYU School STEERING COMMITTEE: (in alphabetic order) A.G. Algera, L.S. Boers; L.D.J. of Medicine, New York, NY, USA. 11 Nuffield Department of Medicine, Oxford Bos; M. Botta; J. Pillay, D.A. Dongelmans, M.W. Hollmann; J. Horn; F. Paulus; A. University, Oxford, UK. 12 Mahidol‑Oxford Tropical Medicine Research Unit Serpa Neto; M.J. Schultz; A.M. Tsonas, A.P. Vlaar. (MORU), Mahidol University, Bangkok, Thailand. STUDY COORDINATORS: (in alphabetic order) M. Botta; A.M. Tsonas. Investigators: (in alphabetic order): J.P. van Akkeren; A.G. Algera; C.K. Received: 21 January 2021 Accepted: 8 April 2021 Algoe; R.B. van Amstel; Artigas A; O.L. Baur; P. van de Berg; A.E. van den Berg; D.C.J.J. Bergmans; D.I. van den Bersselaar; F.A. Bertens; A.J.G.H. Bindels; M.M. de Boer; S. den Boer; L.S. Boers; M. Bogerd; L.D.J. Bos; M. Botta; J.S. Breel; H. de Bruin; S. de Bruin; C.L. Bruna; L.A. Buiteman–Kruizinga; O. Cremer; R.M. Deter‑ mann; W. Dieperink; D.A. Dongelmans; H.S. Franke; M.S. Galek–Aldridge; M.J. References de Graaff; L.A. Hagens; J.J. Haringman; S.T. van der Heide; P.L.J. van der Heiden; 1. Wu Z, McGoonan JM. Characteristics of and important lessons from the N.F.L. Heijnen; S.J.P. Hiel; L.L. Hoeijmakers; L. Hol; M. W. Hollmann; M.E. Hoogen‑ Coronavirus Disease 2019 (COVID-19) outbreak in China: summary of a doorn; J. Horn; R. van der Horst; E.L.K. Ie; D. Ivanov; N.P. Juffermans; E. Kho; E.S. report of 72,314 cases from the Chinese Center for Disease Control and de Klerk; A.W.M.M Koopman-van Gemert; M. Koopmans; S. Kucukcelebi; M.A. Prevention. JAMA. 2020;323:1239–42. Kuiper; D.W. de Lange; N. van Mourik; Morales-Quinteros L; S.G. Nijbroek; M. 2. Wang D, Hu B, Hu C, et al. Clinical characteristics of 138 hospitalized Onrust; E.A.N. Oostdijk; F. Paulus; C.J. Pennartz; J. Pillay; L. Pisani; I.M. Purmer; patients with 2019 novel coronavirus-infected pneumonia inWuhan, T.C.D. Rettig; J.P Roozeman; M.T.U. Schuijt; M.J. Schultz; A. Serpa Neto; M.E. China. JAMA. 2020;323:1061–9. Sleeswijk; M.R. Smit; P.E. Spronk; W. Stilma; A.C. Strang; A. M. Tsonas; P.R Tuin‑ 3. Liu X, Liu X, Xu Y, et al. Ventilatory ratio in hypercapnic mechanically man; C.M.A. Valk; F.L. Veen-Schra; L.I. Veldhuis; P. van Velzen; W.H. van der Ven; ventilated patients with COVID-19 associated ARDS. Am J Respir Crit Care A.P.J. Vlaar; P. van Vliet; P.H.J. van der Voort; L. van Welie; H.J.F.T. Wesselink; H.H. Med. 2020;201:1297–9. van der Wier-Lubbers; B. van Wijk; T. Winters; W.Y. Wong; A.R.H. van Zanten. 4. Nuckton TJ, Alonso JA, Kallet RH, et al. Pulmonary dead-space fraction as a risk factor for death in the acute respiratory distress syndrome. N Engl J Authors’ contributions Med. 2002;346:1281–6. LMQ, LDB, AA, ASN, MJS and FP carried out the concept and design. LMQ, LDB 5. Kallet RH, Zhuo H, Liu KD, et al. The association between physiologic and ASN carried out the statistical analysis. FP and MJS obtained funding for dead-space fraction and mortality in subjects with ARDS enrolled in a the present study. LDB and ASN carried out the supervision of the manuscript. prospective multi-center clinical trial. Respir Care. 2014;59:1611–8. MB, AT participated in the coordination of the study. All authors contributed in 6. Beitler JR, Thompson BT, Matthay MA, et al. Estimating dead-space frac‑ the acquisition, analysis, interpretation of data and drafting of the manuscript. tion for secondary analyses of acute respiratory distress syndrome clinical All authors contributed to the critical revision of the manuscript for important trials. Crit Care Med. 2015;43:1026–35. intellectual content. All authors read and approved the final manuscript. 7. Sinha P, Calfee CS, Beitler JR, et al. Physiologic analysis and clinical perfor‑ mance of the ventilatory ratio in acute respiratory distress syndrome. Am Funding J Respir Crit Care Med. 2019;199:333–41. This article did not receive sponsorship for publication. 8. Morales-Quinteros L, Schultz MJ, Bringué J, et al. Estimated dead space Declarations fraction and the ventilatory ratio are associated with mortality in early ARDS. Ann Intensive Care. 2019;9:128. Availability of data and supporting materials 9. Schenck EJ, Hoffman K, Goyal P, et al. Respiratory mechanics and gas Morales-Quinteros, Bos and Serpa Neto had full access to all data in the study exchange in COVID-19 Associated Respiratory Failure. Ann Am Thorac and take responsibility for the integrity of the data and the accuracy of the Soc. 2020;17:1158–61. data analysis; the members of the Steering Committee for PRoVENT–COVID 10. Fusina F, Albani F, Bertelli M, et al. Corrected Minute Ventilation Is Collaborative Group vouch for the accuracy and completeness of the data and Associated With Mortality in ARDS Caused by COVID-19. Respir Care. for the fidelity of the study to the protocol. 2020;66:619. 11. Boers NS, Botta M, Tsonas AM, et al. Practice of Ventilation in Patients with Declarations Novel Coronavirus Disease (PRoVENT-COVID): rationale and protocol for a national multicenter observational study in The Netherlands. Ann Transl Ethics approval and consent to participate Med. 2020;2020(8):1251. Not applicable. 12. Paulus F. https://​sites.​google.​com/​view/​prove​nt-​covid/​prove​nt-​covid?​ authu​ser=0. Accessed 12 Jan 2021. Consent for publication 13. Prokop M, van Everdingen W, et al. CO-RADS - A categorical CT assess‑ Not applicable. ment scheme for patients with suspected COVID-19: definition and evaluation. Radiology. 2020;296:E97–104. Competing interests 14. Harris JA, Benedict FG. A biometric study of human basal metabolism. Dr Bos receives funding from the Dutch lung foundation (longfonds), from Proc Natl Acad Sci USA. 1918;4:370–3. the Innovative Medicine Initiative and from Amsterdam UMC via the AUMC Morales‑Quinteros et al. Crit Care (2021) 25:171 Page 13 of 13 15. Kallet RH, Lipnick MS. End-tidal-to-arterial pco2 ratio as signifier for physi‑ 27. Raurich JM, Vilar M, Colomar A, et al. Prognostic value of the pulmonary ologic deadspace ratio and oxygenation dysfunction in acute respiratory dead-space fraction during the early and intermediate phases of acute distress syndrome. Respir Care. 2021;66:263–8. respiratory distress syndrome. Respir Care. 2010;55:282–7. 16. Sinha P, Fauvel NJ, Singh S, et al. Ventilatory ratio: a simple bedside meas‑ 28. Chang JC. Hemostasis based on a novel “two-path unifying theory” ure of ventilation. Br J Anaesth. 2009;102:692–7. and classification of hemostatic disorders. Blood Coagul Fibrinol. 17. Wexler HR, Lok P. A simple formula for adjusting arterial carbon dioxide 2018;29:573–84. tension. Can Anaesth Soc J. 1981;28:370–2. 29. Chang JC. Thrombogenesis and thrombotic disorders based on “two- 18. R Core Team (2019). R: A language and environment for statistical com‑ path unifying theory of hemostasis”: philosophical, physiological and puting. R Foundation for Statistical Computing, Vienna, Austria. https://​ phenotypical interpretation. Blood Coagul Fibrinol. 2018;29:585–95. www.R-​proje​ct.​org/. 30. Chang JC. Acute respiratory distress syndrome as an organ phenotype 19. Varga Z, Flammer AJ, Steiger P, et al. Endothelial cell infection and of vascular microthrombotic disease: based on hemostatic theory and endotheliitis in COVID-19. Lancet. 2020;395:1417–8. endothelial molecular pathogenesis. Clin Appl Thromb. 2019;25:1–20. 20. Siddiki H, Kojicic M, Li G, et al. Bedside quantification of dead-space 31. Beenen LFM, Bos LB, Scheerder MJ, et al. Extensive pulmonary perfusion fraction using routine clinical data in patients with acute lung injury: defects compatible with microthrombosis and thromboembolic disease secondary analysis of two prospective trials. Crit Care. 2010;14:R141. in severe Covid-19 pneumonia. Thromb Res. 2020;135–137:135–7. 21. Sinha P, Flower O, Soni N. Deadspace ventilation: a waste of breath! Inten‑ 32. Wichmann D, Sperhake JP, Lütgehetmann M, et al. Autopsy findings and sive Care Med. 2011;37:735–46. venous thromboembolism in patients with COVID-19: a prospective 22. Mosing M, Kutter AP, Iff S, et al. The effects of cardiac output and pulmo‑ cohort study. Ann Intern Med. 2020;173:268–77. nary arterial hypertension on volumetric capnography derived-variables 33. Schultz MJ, Bos LDJ, Paulus F. Instrumental dead space in ventilator man‑ during normoxia and hypoxia. J Clin Monit Comput. 2015;29:187–96. agement - authors’ reply. Lancet Respir Med. 2021;9:e23. 23. Suarez-Sipmann F, Bohm S, Tusman G. Volumetric capnography: the time 34. Patel BV, Arachchillage DJ, Ridge CA, et al. Pulmonary angiopathy in has come. Curr Opin Crit Care. 2014;20:333–9. severe COVID-19: physiologic, imaging, and hematologic observations. 24. Lucangelo U, Bernabè F, Vatua S, et al. Prognostic value of different dead Am J Respir Crit Care Med. 2020;202:690–9. space indices in mechanically ventilated patients with acute lung injury 35. Liu KD, Levitt J, Zhuo H, et al. Randomized clinical trial of activated pro‑ and ARDS. Chest. 2008;133:62–71. tein C for the treatment of acute lung injury. Am J Respir Crit Care Med. 25. Kallet RH, Alonso JA, Pittet JF, et al. Prognostic value of the pulmonary 2008;178:618–23. dead-space fraction during the first 6 days of acute respiratory distress syndrome. Respir Care. 2004;49:1008–14. 26. Kallet RH, Zhuo H, Ho K, et al. Lung injury Etiology and other factors Publisher’s note influencing the relationship between dead-space fraction and mortality Springer Nature remains neutral with regard to jurisdictional claims in pub‑ in ARDS. Respir Care. 2017;62:1241–8. lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions