To the Editor:
I read with interest the detailed review on intraoperative protective mechanical ventilation.1 On the basis of their analysis of existing scientific data, the authors recommend to ventilate healthy lungs with a combination of low tidal volume (VT, 6 to 8 ml/predicted body weight [PBW], no or minimal positive end-expiratory pressure (PEEP more than or equal to 2 cm H2O), and no recruitment maneuver (RM). In case of a peripheral oxygen saturation (Spo2) less than 92%, they further recommend to increase the inspired oxygen fraction (Fio2) up to 0.6 without a simultaneous increase in PEEP, to increase PEEP to maximally 6 cm H2O should a Fio2 of 0.8 be required, and to consider a single RM. They designate this ventilatory strategy as “protective.” I am concerned with these recommendations because they are not necessarily supported by scientific data and may well result in “nonprotective” ventilation in individual patients.
My first concern relates to the undifferentiated use of the term “healthy” lungs. Baseline function and morphology and compensatory capacity differ substantially between the “healthy” lung of a young and that of an advanced-age individual. In addition to such intrinsic differences between “healthy” lungs, differences in patient physical fitness and body habitus, intraoperative positioning, and type of surgery (to name just a few factors) will be associated with entirely different intraoperative impacts on “healthy” lungs. It is unrealistic to assume that an identical ventilation strategy will be equally “protective” in the case of a 75-yr-old obese patient undergoing major abdominal surgery intermittently requiring Trendelenburg position and in that of a 20-yr-old normal weight patient undergoing peripheral surgery in the recumbent position. As the impact on the lungs differs considerably between conditions, even for “healthy” lungs, nonindividualized recommendations are unwarranted.
When excluding nonpulmonary causes, the most likely cause of intraoperative desaturation is progressive lung collapse related to the consistent 15 to 25% decrease in functional residual capacity (FRC) associated with any induction of general anesthesia and formation of atelectasis in the majority of patients.2,3 It is accepted knowledge that increasing of the Fio2 is associated with an increased risk of absorption atelectasis.3,4 It makes little physiologic sense to recommend as first-line treatment of desaturation, an intervention that facilitates the formation of atelectasis. The increase in Fio2 will acutely normalize the Spo2, but it may eventually worsen oxygenation even further.4 The increase in Fio2 in response to a decrease in Spo2 constitutes a symptomatic rather than causal treatment. Causal treatment would be an attempt at reopening collapsed lung areas.5–7
The authors base their conclusions and recommendations heavily on the findings of the PROtective Ventilation using High versus Low PEEP (PROVHILO) trial.8 This publication was coauthored by four of the authors of the present review.1 In the PROVHILO trial,8 both study groups were ventilated with a VT of 8 ml/kg PBW. The “higher PEEP” group received PEEP 12 cm H2O with RMs after intubation and any disconnection from the ventilator, and just before extubation, the “lower PEEP” group received PEEP less than or equal to 2 cm H2O and no RM (i.e., the ventilation strategy recommended in the present review). Several of the study limitations that might have obscured any treatment effect (e.g., protocol deviations, insufficient frequency of RM, routine use of high PEEP in the “higher PEEP” group, abrupt withdrawal of 12 cm H2O PEEP at the time of extubation and restoration of spontaneous respiration) had been addressed in the accompanying editorial9 and in related correspondences.10,11 The incidences of postoperative pulmonary complications (PPCs) were unusually high and almost identical in both groups (40 and 39% in the higher and lower PEEP groups, respectively). Thus, the authors’ expressed preference for the ventilatory strategy of the lower PEEP group (i.e., an “intraoperative ventilation strategy should include a low tidal volume and low PEEP, without recruitment manoeuvres”) has no scientific basis. As a matter of fact, considering the high PPC rate in both groups, it is justified to conclude that both ventilation strategies were not lung protective.
The authors of the present review1 and of the PROVHILO trial8 emphasize the increased risk of hypotension associated with an RM. First, even in the “lower PEEP” group of the PROVHILO trial, hypotension (defined as systolic arterial pressure less than 90 mmHg for more than 3 min) was observed in 36% of patients. Second, approximately 40% of patients of both groups had thoracic epidural analgesia. It should not come as a surprise that indiscriminate use of RMs in the presence of thoracic analgesia and PEEP 12 cm H2O will be associated with an increased risk of hemodynamic instability.
Relevant in the context of lung-protective ventilation, but not addressed anywhere, is the four times greater need for intraoperative rescue intervention for desaturation (Spo2 less than 90%) in the “lower PEEP” group compared with the “higher PEEP” group (2 vs. 8%; relative risk [95% CI], 0.34 [0.18 to 0.67]; P = 0.0008) in the PROVHILO trial.8 As the rescue intervention was performed in the absence of evidence of airway problems, severe hemodynamic impairment. or ventilator malfunction, it is likely that the desaturation requiring intervention developed on the basis of intrinsic worsening of pulmonary oxygenation capacity. Distal airway collapse caused by the combination of low VT, very low to no PEEP, and no RM might have contributed to the desaturations. Noteworthy, the rescue interventions included progressive increases in PEEP and RMs.
Figure 3 of the review1 schematically outlines the expected effects of various combinations of VT and PEEP on the extent of normally aerated and overinflated and atelectatic lung regions. According to this schematic, the combination of VT and low PEEP (less than or equal to 2 cm H2O) can be expected to result in atelectasis of approximately 35 to 40% at the basal lung areas. A young healthy patient having undergone peripheral surgery is likely to postoperatively compensate for this extent of atelectasis. This cannot be expected in the elderly having undergone open abdominal surgery who starts off with an unfavorable closing capacity/FRC ratio and age-related reduced cardiorespiratory function and compensatory mechanisms. In this context, the authors introduce the concept of so-called intraoperative permissive atelectasis as an alternative approach to the application of PEEP. As this concept is not referenced, it is impossible to judge whether the theoretical benefit of such approach will outweigh the documented adverse effects of perioperative atelectasis (i.e., impaired oxygenation, focus of infection).
The authors of the review1 state that a recent individual patient data meta-analysis12 supports the interpretation that a low VT rather than PEEP combined with lung RM is responsible for lung protection in the intraoperative period. This statement is not entirely correct. The meta-analysis showed that in patients receiving a VT of less than or equal to 7 ml/kg PBW, PPCs were significantly fewer in those ventilated with a PEEP of at least 5 cm H2O compared with patients receiving a VT of more than 10 ml/kg PBW. This finding suggests a potential benefit of PEEP more than 5 cm H2O when applied during ventilation with lower VT.13
It is noteworthy that a ventilation strategy discouraged by the authors (i.e., routine application of PEEP 6 to 8 cm H2O, and RMs performed every 30 min after tracheal intubation, with a baseline VT of 6 to 8 ml/kg PBW) had been designated “protective ventilation” in the Intraoperative Protective Ventilation (IMPROVE) study.14 This ventilation strategy was associated with a combined 7-day postoperative incidence of pulmonary complications, atelectasis, and pneumonia of 25.5% (compared with 61% after nonprotective ventilation with a VT of 10 to 12 ml/kg PBW, no PEEP, and no RM). The outcome of the protective ventilation group compares favorably with the 5-day PPC rate of 39% reported by the authors of the PROVHILO trial8 for a comparable patient population, using a ventilation strategy of comparably low VT, but very low PEEP and no RM. On the basis of these findings, three of the coauthors of the IMRPOVE publication14 recommended just a year ago in Anesthesiology to ventilate healthy lungs with a VT of 6 to 8 ml/kg PBW and PEEP of 6 to 8 cm H2O and perform RMs every 30 to 45 min and following events resulting in lung derecruitment (i.e., any event associated with disconnection of the endotracheal tube).15 In essence, the authors of both the PROHIVLO8 and the IMPROVE trials14 recommend a ventilation strategy that is primarily based on the findings of their own study. It is confusing and unsatisfactory for the practitioner to be confronted with recommendations published in the same journal within the same year contradicting each other in central aspects of ventilatory care (i.e., application of PEEP, performance of RMs). “The impression given by all these studies is that we are still left with uncertainty (…) Does one size of PEEP or recruitment fit all? Thus, there are more steps to test in finding an optimal ventilation technique during and after anesthesia. They should preferably be considered before any large multicentre study tries another short-cut.16 ”
This comment addresses a basic problem of randomized trials in this area. Individual patients will not be receiving the level of PEEP or the number of RMs they objectively require, but a preset level of PEEP and number of RMs exclusively determined by random allocation. Often, ventilation strategy does not take into account patient’s age, body weight, type of surgery, intraoperative positioning, and Fio2. There will be patients who do require (and tolerate) PEEP up to 12 cm H2O and frequent RMs; and there will be those who neither require (nor tolerate) such amounts of PEEP and frequency of RMs. Those patients will be harmed if they happen to be randomly allocated to the study protocol that applies a ventilatory strategy contrary to their individual ventilatory needs.
Only a strictly individualized approach to perioperative ventilatory strategy can be expected to reduce such drawbacks.4,13 The clinically relevant questions are not of the black or white type, like PEEP or not, “low” or “high” PEEP, RM or not, “infrequent” or “frequent” RMs. The challenge is to select the amount of PEEP and the frequency of RMs optimal for individual patient care. Individual ventilatory strategy should be based on the accepted knowledge of (i) a consistent 15 to 25% decrease in FRC and an approximately 90% incidence of atelectasis after preoxygenation with 100% O2 and induction of anesthesia,2,3 (ii) the tendency for progressive intraoperative lung collapse,3,17 (iii) the usefulness of PEEP in support of keeping the lungs open and retarding redevelopment of atelectasis,15,18,19 and (iv) the usefulness of RM in support of reopening collapsed lung areas.5–7,18 In general, thus, there is no physiologic rationale for principally not applying PEEP and not performing RMs. Regular observation of lung compliance, plateau airway pressure, required Fio2, and hemodynamic behavior should form the basis for selecting the “best PEEP” and the “best” frequency of RMs. Patient characteristics and surgical factors as listed earlier will influence the choice of intraoperative ventilatory strategy. One size, be it PEEP or RM, and possibly even VT, will predictably neither fit nor benefit all.
The author declares no competing interests.