Comments
Abstract
Rationale: Reducing tidal volume decreases mortality in acute respiratory distress syndrome (ARDS). However, the effect of the timing of low tidal volume ventilation is not well understood.
Objectives: To evaluate the association of intensive care unit (ICU) mortality with initial tidal volume and with tidal volume change over time.
Methods: Multivariable, time-varying Cox regression analysis of a multisite, prospective study of 482 patients with ARDS with 11,558 twice-daily tidal volume assessments (evaluated in milliliter per kilogram of predicted body weight [PBW]) and daily assessment of other mortality predictors.
Measurements and Main Results: An increase of 1 ml/kg PBW in initial tidal volume was associated with a 23% increase in ICU mortality risk (adjusted hazard ratio, 1.23; 95% confidence interval [CI], 1.06–1.44; P = 0.008). Moreover, a 1 ml/kg PBW increase in subsequent tidal volumes compared with the initial tidal volume was associated with a 15% increase in mortality risk (adjusted hazard ratio, 1.15; 95% CI, 1.02–1.29; P = 0.019). Compared with a prototypical patient receiving 8 days with a tidal volume of 6 ml/kg PBW, the absolute increase in ICU mortality (95% CI) of receiving 10 and 8 ml/kg PBW, respectively, across all 8 days was 7.2% (3.0–13.0%) and 2.7% (1.2–4.6%). In scenarios with variation in tidal volume over the 8-day period, mortality was higher when a larger volume was used earlier.
Conclusions: Higher tidal volumes shortly after ARDS onset were associated with a greater risk of ICU mortality compared with subsequent tidal volumes. Timely recognition of ARDS and adherence to low tidal volume ventilation is important for reducing mortality.
Clinical trial registered with www.clinicaltrials.gov (NCT 00300248).
Reducing tidal volume decreases mortality in mechanically ventilated patients with acute respiratory distress syndrome (ARDS); however, the effect of the timing of low tidal volume ventilation is not well understood.
In this multisite, prospective cohort study of patients with ARDS, higher tidal volumes shortly after ARDS onset were associated with an even greater risk of intensive care unit mortality compared with subsequent tidal volumes. Timely recognition of ARDS and prompt adherence to low tidal volume ventilation thereafter may be important for maximally reducing intensive care unit mortality in patients with ARDS.
Randomized trials and metaanalyses have shown that use of low tidal volumes reduces mortality in patients with acute respiratory distress syndrome (ARDS) (1–4). However, as part of routine clinical care, patients may not consistently receive this evidence-based therapy in part because of barriers in the timely recognition of ARDS and in initiating and sustaining low tidal volume ventilator settings thereafter (5–11). The potential harm of delayed initiation of low tidal volume ventilation is not fully understood. Preclinical studies and randomized trials support that higher tidal volumes, for periods as short as minutes to hours, may be harmful (12–17). Despite this evidence, one prior study, using data from ARDS Network clinical trials, could not detect an association between hospital mortality and tidal volumes received in the initial 36- to 48-hour period after ARDS onset and before study enrollment (18). This prior study’s finding may have been influenced by evaluating a select patient population who (1) were eligible and consenting for a clinical trial, (2) survived for 36–48 hours after ARDS onset before trial enrollment, and (3) were strictly managed with a low tidal volume ventilation protocol after enrollment. Hence, using a multisite prospective cohort study of patients with ARDS receiving routine medical care, our objective was to evaluate the association of initial tidal volume and change in tidal volume over time, with mortality in the intensive care unit (ICU).
On a daily basis, we prospectively screened patients for eligibility in this study, including detailed review of data in medical records and review of chest radiograph to enroll 520 patients with ARDS from 13 medical, surgical, and trauma ICUs at four teaching hospitals in Baltimore, Maryland. The ARDS inclusion criteria for enrollment were mechanical ventilation, PaO2/Fio2 ratio less than 300, and meeting the American-European Consensus Conference criteria (1) that were in effect at the time of screening for this study (2004–2007). ARDS onset was defined as the time at which a patient met all inclusion criteria. Consistent with the more recent Berlin consensus meeting (19), we use the term ARDS, rather than acute lung injury, throughout this report. Neurologic specialty ICUs and patients with ARDS with primary neurologic disease or head trauma were not eligible for this study. Because the study was designed to evaluate the association of critical illness and ICU care, in particular lung protective mechanical ventilation, on patient’s long-term mortality and functional outcomes (5, 20), the study had the following relevant exclusion criteria (Figure 1) at the onset of ARDS: (1) prior lung resection, (2) transfer from another hospital with preexisting ARDS of greater than 24 hours duration, (3) mechanically ventilated for more than 5 days before ARDS onset, (4) a physician order limiting the use of life-support therapies or preexisting comorbid illness with a life expectancy less than 6 months (e.g., metastatic cancer), (5) preexisting cognitive impairment or communication/language barriers, and (6) no fixed address.
Figure 1. Flow of patients through study. ARDS = acute respiratory distress syndrome.
[More] [Minimize]Institutional review board approval was obtained from all participating sites with a waiver of informed consent granted for abstraction of preexisting data from the medical record. Written informed consent was obtained from survivors after they regained decision-making capacity (or from a proxy if a patient remained incapable of decision-making).
At the study site hospitals, patients receiving mechanical ventilation were rarely transferred elsewhere within the hospital (e.g., a stepdown unit) or to another healthcare facility (e.g., a long-term acute care hospital). Hence, because of its temporal proximity to the primary exposure in this analysis, the primary outcome, selected a priori, was time to death in the ICU (in days) during patients’ ICU admission for ARDS.
The primary exposure was tidal volume received while mechanically ventilated, modeled as milliliter per kilogram of predicted body weight (PBW; calculated based on patient sex and height [2]). This time-varying exposure was recorded at 12-hour increments over the entire duration of mechanical ventilation and partitioned into two parts: the first available tidal volume after ARDS onset, and the time-varying change in current tidal volume relative to this initial tidal volume.
Our analysis adjusted for 29 baseline and time-varying covariates previously identified as potential confounders (5) and obtained from patients’ medical records. Baseline variables included age, sex, body mass index, Charlson comorbidity index (21), severity of illness within 24 hours of ICU admission (Acute Physiology And Chronic Health Evaluation II score [22]), ARDS risk factor (sepsis vs. other), ICU type (medical vs. surgical), patient location before ICU admission (e.g., emergency department), year of study enrolment, and study site identifier. Time-varying covariates were obtained either daily or twice-daily. Daily covariates included organ dysfunction (Sequential Organ Failure Assessment score [23]), sedation and delirium status (Richmond Agitation and Sedation Scale [24] and Confusion Assessment Method for the ICU [25], respectively), dose of systemic corticosteroids and neuromuscular blocking agents, and net fluid balance (total fluid input minus total fluid output). Twice-daily covariates represented mechanical ventilation parameters, including positive end-expiratory pressure, PaO2, Fio2, arterial pH, actual respiratory rate, use of high-frequency oscillatory and airway pressure release ventilation modes, and static compliance of the respiratory system (3).
Descriptive statistics summarized the baseline and time-varying covariates for all subjects, with comparison between patient groups conducted using the Wilcoxon rank-sum and Fisher exact tests, as appropriate. As per the original clinical trial (1), a tidal volume of less than or equal to 6.5 ml/kg PBW was used to define adherence to the low tidal volume goal of 6.0 ml/kg PBW. Kaplan-Meier plots with log-rank tests were used for unadjusted analyses of patient survival. A multivariable Cox regression model was used to evaluate death as a function of the time-varying tidal volume (primary exposure), after accounting for the duration of mechanical ventilation and the other 28 baseline and time-varying covariates as previously described, with the time-varying covariates modeled as cumulative averages. A potential time-varying effect of tidal volume on ICU mortality over the duration of a patient’s ICU stay was evaluated by including statistical interaction of the primary exposure (as previously described) with time (measured in 12-h intervals). Statistical interaction between the two parts of the primary exposure (i.e., initial tidal volume and change in current tidal volume relative to initial tidal volume) was also evaluated. As a sensitivity analysis, a Fine and Gray proportional subhazards regression model was fit, treating ICU discharge as a competing risk. This sensitivity analysis was conducted to confirm appropriateness of the Cox regression analysis assumption of noninformative censoring of ICU discharge in evaluating the outcome of ICU mortality.
We used standard statistical diagnostic procedures to evaluate the model. To assess the linearity assumption for continuous covariates, we plotted Martingale residuals against covariate values using a nonparametric LOESS-smoother. For each covariate, we also assessed the proportional hazards assumption via graphical displays of scaled Schoenfeld residuals and via performing individual tests of proportional hazards. We assessed the influence of individual observations by comparing the relative change in the estimated regression coefficients by deleting each observation, in turn, from the model, with no observation demonstrating high influence.
To illustrate the effects of the exposure-outcome relationship, the previously described Cox regression model was used to estimate the absolute difference in the cumulative risk of mortality at 8 days after ARDS onset, for a prototypical patient having median values for all continuous covariates and mode values for all binary covariates, with various profiles of tidal volume settings assumed during the ICU stay.
There were no missing data for patient-level characteristics. For the ventilator setting data (i.e., positive end-expiratory pressure, PaO2/Fio2, and respiratory rate), there were less than 0.2% missing for the cumulative averages used in the regression model. To impute the 10% missing data for plateau pressure (used to calculate static compliance of the respiratory system [26]), we used multiple imputation (with five imputed datasets), as previously described (5). As an a priori sensitivity analysis for this imputation, we repeated all analyses using a subset of the entire dataset that excluded ventilator settings with missing plateau pressure (analysis of “complete data”). Statistical significance was defined as a two-sided P less than 0.05. All statistical analyses were completed using R statistical software (version 3.0.3) (Foundation for Statistical Computing, Vienna, Austria) and STATA 12.1 (StataCorp, College Station, TX).
Overall, prospective screening identified 754 patients meeting inclusion criteria, of whom 234 met exclusion criteria (Figure 1). Consequently, 520 patients were enrolled in the study, of whom 38 were excluded from the analysis, 35 (7%) had no eligible ventilator settings for this analysis (e.g., exclusive use of high-frequency oscillation or airway pressure release ventilation), and 3 (<1%) had missing data on height (required for calculating PBW for the primary exposure). Thus, 482 patients, with 11,558 total ventilator settings, were available for analysis. For the sensitivity analysis of complete data (see Methods), 482 patients with 10,397 ventilator settings were available.
Tables 1 and 2 present patient characteristics and mechanical ventilation data by initial tidal volume of less than or equal to 6.5 versus greater than 6.5 ml/kg PBW (n = 154 [32%] vs. 328 [68%]) and by ICU mortality status. During their ICU stay, 169 (35%) of the 482 patients died. The entire 482 patient cohort had a median (interquartile range) age of 53 (42–63) years and Acute Physiology and Chronic Health Evaluation II score of 27 (20–33), with 56% male, 74% with sepsis as the primary ARDS risk factor, and 85% in a medical ICU (Table 1). Across all 11,558 twice-daily mechanical ventilator settings observed throughout the ICU stay for the 482 patients, the median (interquartile range) positive end-expiratory pressure was 5 (5–10) and tidal volume was 6.6 (5.9–8.0) ml/kg PBW (Table 2). More specifically, 328 (68%) of the 482 patient cohort had an initial tidal volume greater than 6.5 ml/kg PBW, with 183 (56%) of these 328 patients ever having a tidal volume of less than or equal to 6.5 ml/kg PBW thereafter and 23 (7%) of them having all subsequent tidal volumes less than or equal to 6.5 ml/kg. Additional details describing changes in tidal volume over time are provided in Figure 2.
| All Patients (n = 482) | First Tidal Volume | P Value | At ICU Discharge | P Value | |||
|---|---|---|---|---|---|---|---|
| ≤6.5 ml/kg PBW (n = 154) | >6.5 ml/kg PBW (n = 328) | Alive (n = 313) | Dead (n = 169) | ||||
| Median (IQR) age | 53 (42–63) | 51 (41–60) | 53 (43–65) | 0.015 | 51 (41–61) | 55 (45–66) | 0.033 |
| Male sex | 271 (56%) | 119 (77%) | 152 (46%) | <0.001 | 181 (58%) | 90 (53%) | 0.338 |
| Underweight (body mass index < 18.5) | 27 (6%) | 10 (6%) | 17 (5%) | 0.533 | 15 (5%) | 12 (7%) | 0.305 |
| Overweight or obese (body mass index ≥ 25) | 306 (63%) | 84 (55%) | 222 (68%) | 0.006 | 202 (65%) | 104 (62%) | 0.552 |
| Median (IQR) Charlson comorbidity index | 2 (1–4) | 3 (1–5) | 2 (1–4) | 0.224 | 2 (1–4) | 3 (2–5) | <0.001 |
| Median (IQR) APACHE II score | 27 (20–33) | 27 (21–36) | 26 (20–33) | 0.115 | 24 (19–30) | 31 (25–37) | <0.001 |
| Sepsis as acute lung injury risk factor | 358 (74%) | 121 (79%) | 237 (72%) | 0.148 | 220 (70%) | 138 (82%) | 0.006 |
| Admission to medical ICU | 410 (85%) | 144 (94%) | 266 (81%) | <0.001 | 256 (82%) | 154 (91%) | 0.007 |
| Admission to ICU from emergency department | 201 (42%) | 60 (39%) | 141 (43%) | 0.429 | 138 (44%) | 63 (37%) | 0.175 |
| Median (IQR) mean daily SOFA score | 8 (5–12) | 9 (6–12) | 8 (5–13) | 0.151 | 6 (4–9) | 13 (9–15) | <0.001 |
| Median (IQR) number of days of delirium in ICU | 1 (0–5) | 1 (0–4) | 1 (0–6) | 0.072 | 2 (0–6) | 0 (0–2) | <0.001 |
| Median (IQR) number of days of deep sedation in ICU | 3 (1–7) | 4 (1–9) | 3 (1–7) | 0.022 | 3 (1–7) | 3 (1–7) | 0.568 |
| Ever received corticosteroid | 324 (67%) | 109 (71%) | 215 (66%) | 0.298 | 190 (61%) | 134 (79%) | <0.001 |
| Median (IQR) number of days of steroid use, if any | 5 (3–9) | 5 (2–9) | 5 (3–10) | 0.502 | 6 (3–10) | 5 (2–9) | 0.045 |
| Median (IQR) cumulative dose of prednisone, if any, mg | 322 (150–726) | 300 (107–648) | 339 (153–759) | 0.202 | 355 (156–893) | 291 (126–534) | <0.001 |
| Ever received neuromuscular blockade | 117 (24%) | 45 (29%) | 72 (22%) | 0.088 | 68 (22%) | 49 (29%) | 0.095 |
| Median (IQR) number of days of neuromuscular blockade, if any | 2 (1–3) | 2 (1–4) | 1 (1–2) | 0.127 | 1 (1–2) | 2 (1–4) | 0.073 |
| Median (IQR) cumulative dose of vecuronium, if any, mg | 2 (1–23) | 3 (1–51) | 1 (1–12) | 0.089 | 1 (1–8) | 3 (1–72) | 0.036 |
| Median (IQR) cumulative ICU fluid balance, L | 9 (2–20) | 8 (1–20) | 10 (3–21) | 0.182 | 6 (0–18) | 14 (6–23) | <0.001 |
| HFOV (ever) | 57 (12%) | 20 (13%) | 37 (11%) | 0.650 | 27 (9%) | 30 (18%) | 0.005 |
| APRV (ever) | 62 (13%) | 10 (6%) | 52 (16%) | 0.003 | 51 (16%) | 11 (7%) | 0.002 |
| Median number of settings with HFOV, if any | 5 (2–11) | 4 (2–8) | 7 (3–13) | 0.382 | 8 (4–13) | 4 (2–8) | 0.082 |
| Median number of settings with APRV, if any | 10 (5–18) | 9 (5–19) | 10 (5–17) | 0.800 | 10 (6–17) | 14 (2–32) | 0.830 |
| Median days of ventilation (all settings) (IQR) | 9 (5–17) | 9 (5–17) | 9 (5–17) | 0.746 | 11 (7–19) | 7 (3–13) | <0.001 |
| ICU length of stay | 13 (8–22) | 13 (7–20) | 14 (8–23) | 0.474 | 16 (10–25) | 8 (5–16) | <0.001 |
| All Ventilator Settings† (n = 11,558) | First Tidal Volume | P Value | At ICU Discharge | P Value | |||
|---|---|---|---|---|---|---|---|
| ≤6.5 ml/kg PBW (n = 3,366) | >6.5 ml/kg PBW (n = 8,192) | Alive | Dead | ||||
| (n = 8,335) | (n = 3,223) | ||||||
| Median (IQR) positive end-expiratory pressure, per 1 cm H2O | 5 (5–10) | 5 (5–10) | 5 (5–10) | <0.001 | 5 (5–8) | 5 (5–10) | <0.001 |
| Median (IQR) PaO2/Fio2 | 199 (142–299) | 188 (118–253) | 199 (154–299) | <0.001 | 199 (160–299) | 176 (104–242) | <0.001 |
| Number of pH < 7.25 | 887 (11%) | 337 (13%) | 550 (10%) | <0.001 | 426 (7%) | 461 (20%) | <0.001 |
| Median (IQR) static compliance of respiratory system, per 10 ml/cm H2O | 31 (22–40) | 31 (24–40) | 30 (21–40) | <0.001 | 32 (23–41) | 29 (21–37) | <0.001 |
| Median (IQR) respiratory rate, per 1 breath/min | 25 (20–33) | 30 (22–35) | 24 (19–31) | <0.001 | 24 (19–32) | 29 (22–35) | <0.001 |
| Median (IQR) tidal volume (ml/kg predicted body weight) | 6.6 (5.9–8.0) | 6.0 (5.7–6.6) | 7.0 (6.2–8.2) | <0.001 | 6.7 (6.0–8.0) | 6.6 (5.9–7.8) | 0.140 |
Figure 2. Timing and direction of first change in tidal volume from initial ventilator setting after ARDS onset. The numbers of patients with their first tidal volume after ARDS onset of ≤6.5 and >6.5 ml/kg PBW, respectively, were 154 and 328. The data represented by the dots and connecting line represent the proportion of patients with tidal volume ≤6.5 ml/kg PBW, calculated based on the number of patients at that point in time that were alive and receiving mechanical ventilation with a measurable tidal volume. In patients with their first tidal volume >6.5 ml/kg, 17% had no change in tidal volume or an increase in tidal volume over all subsequent mechanical ventilator settings, whereas 39% had a decrease in the next ventilator setting with 78% ever having a decrease in tidal volume over all subsequent mechanical ventilator settings. ARDS = acute respiratory distress syndrome; PBW = predicted body weight.
[More] [Minimize]Unadjusted survival analysis of patients with their first tidal volume after ARDS onset of greater than 6.5 ml/kg PBW demonstrated that a subsequent decrease (vs. increase) in tidal volume was associated with significantly improved survival (P = 0.008) that was not observed for patients with a first tidal volume of less than or equal to 6.5 ml/kg PBW (P = 0.446) (Figure 3). After adjusting for all covariates, an increase of 1 ml/kg PBW in initial tidal volume was associated with a 23% increase in the risk of ICU mortality (hazard ratio, 1.23; 95% confidence interval, 1.06–1.44; P = 0.008). Moreover, during the time after the initial tidal volume setting, a 1 ml/kg PBW increase in tidal volume from the initial setting was associated with a 15% increase in risk of ICU mortality (hazard ratio, 1.15; 95% confidence interval, 1.02–1.29; P = 0.019) (see Table E1 in the online supplement for full model results). There were no significant statistical interactions of the initial tidal volume with the change in current tidal volume relative to initial tidal volume (see Table E2), or of these two primary exposures variables with time (see Table E3). Sensitivity analyses evaluating ICU discharge as a competing risk (see Table E4) and evaluating missing data imputation, as previously described (see Table E1), demonstrated results consistent with the primary analysis.
Figure 3. Kaplan-Meier survival curves for increase versus decrease in tidal volume from initial ventilator setting after ARDS onset. The numbers of patients with their first tidal volume after ARDS onset ≤6.5 and >6.5 ml/kg PBW, respectively, were 154 and 328. For patients with first tidal volume >6.5 ml/kg PBW, a subsequent decrease (vs. increase) in tidal volume was associated with significant improvement in survival as illustrated by the Kaplan-Meier survival curves and the log-rank test (P = 0.008). For patients with first tidal volume ≤6.5 ml/kg PBW, a subsequent decrease (vs. increase) in tidal volume was not associated with a significant difference in survival (P = 0.446). ARDS = acute respiratory distress syndrome; PBW = predicted body weight.
[More] [Minimize]The absolute risk difference in ICU mortality comparing various example profiles of initial and subsequent tidal volume settings for a prototypical patient are summarized in Table 3. Compared with a reference case of the prototypical patient receiving 8 days of mechanical ventilation with a tidal volume of 6 ml/kg PBW, there was an estimated absolute increase in mortality (95% confidence interval) of 7.2% (3.0–13.0%) and 2.7% (1.2–4.6%) for receiving 10 and 8 ml/kg PBW tidal volume across all 8 days. In tidal volume profiles with 4 days of 6 ml/kg PBW and 4 days of 10 ml/kg PBW, the estimated absolute increase in mortality was substantially greater when the 10 ml/kg PBW tidal volume was received in the first 4 days versus in the last 4 days of the 8-day mechanical ventilation period at 4.8% (1.9–8.5%) versus 2.0% (0.6–3.9%) (Table 3).
| Case Scenario | Estimated Absolute Increase in ICU Mortality (95% CI) |
|---|---|
| Reference case with 6 ml/kg PBW tidal volume throughout mechanical ventilation | |
| Case 0: 8 d of tidal volume = 6 ml/kg PBW* | Reference |
| Cases with 10 ml/kg PBW tidal volume during mechanical ventilation | |
| Case 1: 8 d of tidal volume = 10 ml/kg PBW* | 7.2% (3.0–13.0%) |
| Case 2: 4 d of 10 ml/kg PBW, then 4 d of 6 ml/kg PBW | 4.8% (1.9–8.5%) |
| Case 3: 2 d of 10 ml/kg PBW, then 6 d of 6 ml/kg PBW | 3.1% (0.8–6.0%) |
| Case 4: 4 d of 6 ml/kg PBW, then 4 d of 10 ml/kg PBW | 2.0% (0.6–3.9%) |
| Cases with 8 ml/kg PBW tidal volume during mechanical ventilation | |
| Case 5: 8 d of tidal volume = 8 ml/kg PBW* | 2.7% (1.2–4.6%) |
| Case 6: 4 d of 8 ml/kg PBW, then 4 d of 6 ml/kg PBW | 1.8% (0.7–3.1%) |
| Case 7: 2 d of 8 ml/kg PBW, then 6 d of 6 ml/kg PBW | 1.2% (0.3–2.3%) |
| Case 8: 4 d of 6 ml/kg PBW, then 4 d of 8 ml/kg PBW | 0.8% (0.3–1.5%) |
In this multisite, prospective cohort study of patients with ARDS, higher tidal volumes shortly after ARDS onset were associated with an even greater risk of ICU mortality, compared with subsequent tidal volumes. Specifically, after adjusting for other covariates potentially associated with ICU mortality, a 1 ml/kg PBW increase in initial tidal volume or in a subsequent tidal volume setting was associated with a 23% and 15%, respectively, increase in the risk of ICU mortality. Thus, within the setting of routine clinical practice, timely adherence to the use of low tidal volumes for patients with ARDS is associated with improved survival.
In other aspects of care for critically ill patients, such as receipt of antibiotics in septic shock, timely and appropriate initial therapy reduces hospital mortality (27, 28). With respect to ARDS, randomized trials and metaanalyses have shown that use of low tidal volumes reduces mortality (1–4). Moreover, in a randomized trial of abdominal surgery patients mechanically ventilated in the operating room for an average of 5.5 hours, reduced tidal volumes (along with positive end-expiratory pressure and recruitment maneuvers) significantly decreased major postoperative complications (including acute respiratory failure requiring mechanical ventilation) and hospital length of stay (29). Similarly, our analysis demonstrated that earlier use of low tidal volumes was associated with improved ICU survival. Such findings may be explained by higher tidal volumes, even when used for only minutes to hours, overstretching alveoli, releasing inflammatory mediators systematically, and resulting in pulmonary and extrapulmonary organ dysfunction (12–17).
A single prior study, using data from ARDS Network clinical trials, did not find an association of hospital mortality with higher tidal volumes in the 36 to 48 hours preceding strict protocolized implementation of low tidal volume ventilation (18). Differences in study design may explain the discrepancy between our findings and the prior study. For instance, by evaluating patients enrolled in the ARDS Network trials, any patients dying after ARDS onset, before enrollment, were excluded. Also, in the ARDS Network trials, all patients after enrollment were strictly managed with a low tidal volume protocol; however, in the usual care setting of our study, among those who did not have an initial tidal volume less than or equal to 6.5 ml/kg PBW, 44% never received tidal volumes less than or equal to 6.5 ml/kg PBW thereafter and only 7% always received tidal volumes less than or equal to 6.5 ml/kg PBW thereafter. Finally, there were differences in patient populations, with more stringent enrollment criteria excluding sicker patients in the ARDS Network trials versus our prospective cohort study.
The findings of this research draw attention to the need for early use of low tidal volume ventilation. In our study, approximately two-thirds of patients with ARDS had their initial tidal volume above 6.5 ml/kg PBW. To facilitate early use of low tidal volumes, timely recognition of ARDS is required along with communication of patient tidal volumes in terms of milliliter per kilogram PBW (7, 30). Moreover, given a high frequency of obesity, the calculation of tidal volume in milliliter per kilogram should be based on PBW, calculated using accurate height measurements rather an actual body weight. For mechanically ventilated patients, daily reevaluation for the onset of ARDS and the appropriateness of tidal volumes is important. Given the mortality benefit of low tidal volume ventilation and challenges in timely recognition of ARDS, there may be benefit for all mechanically ventilated patients of ICU-wide protocols that default to 6 ml/kg PBW, with a specific physician order required for use of higher tidal volumes (6, 12, 31, 32). Such an approach may play a role in preventing the development of ARDS in mechanically ventilated patients in addition to reducing mortality in those with ARDS (12). Moreover, having better integration between electronic medical records and mechanical ventilators may allow for improved setting of tidal volumes or creating alerts to notify clinicians of potentially harmful ventilator settings (33).
This study has potential limitations. First, this study was observational in design; hence, we cannot prove causation between the magnitude and timing of tidal volumes and ICU survival because there are both measured and unmeasured differences in patient groups with higher versus lower tidal volumes. However, a randomized trial evaluating delayed delivery of low tidal volume ventilation to patients with ARDS would not be ethical to conduct. Moreover, causality is plausible given the dose-response effect observed in this study, along with the consistency of our findings with both preclinical studies and randomized trials (12–17). Second, missing data on plateau pressure (used to calculate static compliance of the respiratory system [26]) have potential to bias study results; however, the primary analysis with multiple imputation of missing data and a secondary analysis of complete data showed similar results, which is reassuring. Third, only teaching hospitals from a single city were included in this research and there were exclusion criteria for patient enrollment which may limit the generalizability of these findings. However, four hospitals with 13 ICUs were included in the study, with substantial variability in routine medical care delivered, and the eligibility criteria were relatively limited compared with prior randomization controlled trials of low tidal volume ventilation, which aid in generalizability of our findings. Fourth, the data for this study are from 2004–2007. Since this study period, clinical practices with respect to adherence to low tidal volume ventilation and other aspects of ICU care may have changed and potentially modified these findings. However, the consistency of these results to both older and more recent preclinical and clinical data (12–17, 29) may support their continued importance to clinical care for patients with ARDS. Fifth, this study only collected ventilation data twice per day and did not capture instances in which there were more frequent adjustments to tidal volume settings, which may have understated our characterization of adherence to low tidal volume ventilation. Lastly, although a priori research has demonstrated the long-term survival benefit of lung protective ventilation (5), our current analysis evaluating the timing of low tidal volume ventilation only focused on ICU mortality. Hence, future research should evaluate long-term effects.
In conclusion, in this multisite, prospective cohort study of patients with ARDS, higher tidal volumes shortly after ARDS onset were associated with an even greater risk of ICU mortality compared with subsequent tidal volumes. Hence, timely recognition of ARDS and timely adherence to low tidal volume ventilation are important considerations for maximally improving survival for patients with ARDS.
The authors thank all patients who participated in the study and the dedicated research assistants and coordinators who assisted with data collection and management for the study, including Dr. Nardos Belayneh, Ms. Kimberly Boucher, Dr. Abdulla Damluji, Mrs. Rachel Evans, Ms. Thelma Harrington, Dr. Praveen Kondreddi, Mr. Robert Lord, Ms. Stacey Murray, Dr. Susanne Prassl, Ms. Arabela Sampaio, Ms. Kristin Sepulveda, Dr. Shabana Shahid, Dr. Faisal Siddiqi, Ms. Michelle MacDonna, and Ms. Jennifer Titus. They also thank anonymous peer reviewers for providing thoughtful critiques that strengthened the paper.
| 1. | Brower RG, Matthay MA, Morris A, Schoenfeld D, Thompson BT, Wheeler A; The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301–1308. |
| 2. | Fan E, Needham DM, Stewart TE. Ventilatory management of acute lung injury and acute respiratory distress syndrome. JAMA 2005;294:2889–2896. |
| 3. | Putensen C, Theuerkauf N, Zinserling J, Wrigge H, Pelosi P. Meta-analysis: ventilation strategies and outcomes of the acute respiratory distress syndrome and acute lung injury. Ann Intern Med 2009;151:566–576. |
| 4. | Petrucci N, Iacovelli W. Lung protective ventilation strategy for the acute respiratory distress syndrome. Cochrane Database Syst Rev 2007;(3):CD003844. |
| 5. | Needham DM, Colantuoni E, Mendez-Tellez PA, Dinglas VD, Sevransky JE, Dennison Himmelfarb CR, Desai SV, Shanholtz C, Brower RG, Pronovost PJ. Lung protective mechanical ventilation and two year survival in patients with acute lung injury: prospective cohort study. BMJ 2012;344:e2124. |
| 6. | Umoh NJ, Fan E, Mendez-Tellez PA, Sevransky JE, Dennison CR, Shanholtz C, Pronovost PJ, Needham DM. Patient and intensive care unit organizational factors associated with low tidal volume ventilation in acute lung injury. Crit Care Med 2008;36:1463–1468. |
| 7. | Dennison CR, Mendez-Tellez PA, Wang W, Pronovost PJ, Needham DM. Barriers to low tidal volume ventilation in acute respiratory distress syndrome: survey development, validation, and results. Crit Care Med 2007;35:2747–2754. |
| 8. | Weinert CR, Gross CR, Marinelli WA. Impact of randomized trial results on acute lung injury ventilator therapy in teaching hospitals. Am J Respir Crit Care Med 2003;167:1304–1309. |
| 9. | Young MP, Manning HL, Wilson DL, Mette SA, Riker RR, Leiter JC, Liu SK, Bates JT, Parsons PE. Ventilation of patients with acute lung injury and acute respiratory distress syndrome: has new evidence changed clinical practice? Crit Care Med 2004;32:1260–1265. |
| 10. | Kalhan R, Mikkelsen M, Dedhiya P, Christie J, Gaughan C, Lanken PN, Finkel B, Gallop R, Fuchs BD. Underuse of lung protective ventilation: analysis of potential factors to explain physician behavior. Crit Care Med 2006;34:300–306. |
| 11. | Rubenfeld GD, Cooper C, Carter G, Thompson BT, Hudson LD. Barriers to providing lung-protective ventilation to patients with acute lung injury. Crit Care Med 2004;32:1289–1293. |
| 12. | Serpa Neto A, Cardoso SO, Manetta JA, Pereira VG, Espósito DC, Pasqualucci MdeO, Damasceno MC, Schultz MJ. Association between use of lung-protective ventilation with lower tidal volumes and clinical outcomes among patients without acute respiratory distress syndrome: a meta-analysis. JAMA 2012;308:1651–1659. |
| 13. | Lellouche F, Dionne S, Simard S, Bussières J, Dagenais F. High tidal volumes in mechanically ventilated patients increase organ dysfunction after cardiac surgery. Anesthesiology 2012;116:1072–1082. |
| 14. | Imai Y, Parodo J, Kajikawa O, de Perrot M, Fischer S, Edwards V, Cutz E, Liu M, Keshavjee S, Martin TR, et al. Injurious mechanical ventilation and end-organ epithelial cell apoptosis and organ dysfunction in an experimental model of acute respiratory distress syndrome. JAMA 2003;289:2104–2112. |
| 15. | Parker JC, Hernandez LA, Longenecker GL, Peevy K, Johnson W. Lung edema caused by high peak inspiratory pressures in dogs. Role of increased microvascular filtration pressure and permeability. Am Rev Respir Dis 1990;142:321–328. |
| 16. | Dreyfuss D, Basset G, Soler P, Saumon G. Intermittent positive-pressure hyperventilation with high inflation pressures produces pulmonary microvascular injury in rats. Am Rev Respir Dis 1985;132:880–884. |
| 17. | Stüber F, Wrigge H, Schroeder S, Wetegrove S, Zinserling J, Hoeft A, Putensen C. Kinetic and reversibility of mechanical ventilation-associated pulmonary and systemic inflammatory response in patients with acute lung injury. Intensive Care Med 2002;28:834–841. |
| 18. | Checkley W, Brower R, Korpak A, Thompson BT; Acute Respiratory Distress Syndrome Network Investigators. Effects of a clinical trial on mechanical ventilation practices in patients with acute lung injury. Am J Respir Crit Care Med 2008;177:1215–1222. |
| 19. | Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, Camporota L, Slutsky AS; ARDS Definition Task Force. Acute respiratory distress syndrome: the Berlin Definition. JAMA 2012;307:2526–2533. |
| 20. | Bienvenu OJ, Colantuoni E, Mendez-Tellez PA, Dinglas VD, Shanholtz C, Husain N, Dennison CR, Herridge MS, Pronovost PJ, Needham DM. Depressive symptoms and impaired physical function after acute lung injury: a 2-year longitudinal study. Am J Respir Crit Care Med 2012;185:517–524. |
| 21. | Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–383. |
| 22. | Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818–829. |
| 23. | Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, Reinhart CK, Suter PM, Thijs LG. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996;22:707–710. |
| 24. | Ely EW, Truman B, Shintani A, Thomason JW, Wheeler AP, Gordon S, Francis J, Speroff T, Gautam S, Margolin R, et al. Monitoring sedation status over time in ICU patients: reliability and validity of the Richmond Agitation-Sedation Scale (RASS). JAMA 2003;289:2983–2991. |
| 25. | Ely EW, Inouye SK, Bernard GR, Gordon S, Francis J, May L, Truman B, Speroff T, Gautam S, Margolin R, et al. Delirium in mechanically ventilated patients: validity and reliability of the confusion assessment method for the intensive care unit (CAM-ICU). JAMA 2001;286:2703–2710. |
| 26. | Seeley EJ, McAuley DF, Eisner M, Miletin M, Zhuo H, Matthay MA, Kallet RH. Decreased respiratory system compliance on the sixth day of mechanical ventilation is a predictor of death in patients with established acute lung injury. Respir Res 2011;12:52. |
| 27. | Kumar A, Ellis P, Arabi Y, Roberts D, Light B, Parrillo JE, Dodek P, Wood G, Kumar A, Simon D, et al.; Cooperative Antimicrobial Therapy of Septic Shock Database Research Group. Initiation of inappropriate antimicrobial therapy results in a fivefold reduction of survival in human septic shock. Chest 2009;136:1237–1248. |
| 28. | Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, Suppes R, Feinstein D, Zanotti S, Taiberg L, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006;34:1589–1596. |
| 29. | Futier E, Constantin JM, Paugam-Burtz C, Pascal J, Eurin M, Neuschwander A, Marret E, Beaussier M, Gutton C, Lefrant JY, et al.; IMPROVE Study Group. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med 2013;369:428–437. |
| 30. | Needham DM. Patient safety, quality of care, and knowledge translation in the intensive care unit. Respir Care 2010;55:922–928. |
| 31. | Halpern SD, Ubel PA, Asch DA. Harnessing the power of default options to improve health care. N Engl J Med 2007;357:1340–1344. |
| 32. | Gajic O, Dara SI, Mendez JL, Adesanya AO, Festic E, Caples SM, Rana R, St Sauver JL, Lymp JF, Afessa B, et al. Ventilator-associated lung injury in patients without acute lung injury at the onset of mechanical ventilation. Crit Care Med 2004;32:1817–1824. |
| 33. | Pronovost PJ, Bo-Linn GW. Preventing patient harms through systems of care. JAMA 2012;308:769–770. |
Supported by the National Institutes of Health (P050HL73994, R01HL088045, and K24HL088551). D.M.N. received support from a Clinician-Scientist Award from the Canadian Institutes of Health.
Author Contributions: All authors contributed to the conception and/or design of this study. D.M.N., V.D.D., C.S., and J.E.S. contributed to the acquisition of data. All authors contributed to the analysis and interpretation of data. D.M.N. drafted the manuscript, and all authors critically revised it for important intellectual content and approved the final version to be submitted.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201409-1598OC on December 5, 2014
Author disclosures are available with the text of this article at www.atsjournals.org.
