Lung-protective Mechanical Ventilation for Abdominal Laparoscopic Surgeries

Study population. We performed a randomized, double-blind controlled trial at Vietnam National Cancer Hospital from January 2020 to July 2020. The trial protocol was approved by the medical ethics committee of Vietnam Military Medical University and Vietnam National Cancer Hospital, simultaneously written informed consent was obtained from all patients before inclusion.

Randomization and blinding technique. Participants were randomly assigned to receive either lung-protective ventilation (LPV group) or conventional ventilation (CV group) at a ratio of 1:1. The randomization was performed using the R program with the "runif", "as.integer", "int" and "replace" functions. As a result, a list of random numbers was created in each group. The patients were numbered according to their orders of hospital registration and then were allocated into the group of their numbers. The intervention protocol was stored in sealed, opaque numbered envelopes. An anesthesiologist who did not involve in the study set the ventilator in accordance with the protocol in the envelopes. Another anesthetist who was in charge of the patient collected data during surgery. The surgeons taking part in the procedures and patients were not informed of the ventilator setting. Physicians in post-anesthesia care unit who were not responsible for intraoperative care carried out the postoperative evaluation. The analysis of the postoperative chest X-ray was completed by a radiologist who was not involved in the study.

Standard procedure. All patients fasted for 12 hours before the procedure but still consumed clear water until 2 hours prior to surgery in order to avoid preoperative dehydration. Participants were premedicated with intravenous midazolam 1-2 mg 30 minutes before the surgery. In the operating room, a radical arterial cannula was inserted to monitor invasive blood pressure, collect blood gas sample, and measure the pulse pressure variation (PPV) index to guide intraoperative fluid therapy. The ASA standards for monitoring such as pulse oximeter, capnography, electrocardiography, thermometer were applied prior to the insertion of an epidural catheter at level T7-T12 for postoperative analgesia.

All patients received intravenous fentanyl 2 µg/kg, lidocaine 40 mg, propofol 2 mg/kg, and rocuronium 1 mg/kg for the induction. Intubation was performed 90 seconds after the administration of muscle relaxant and then 8 mg dexamethasone was injected. Anesthesia was maintained using sevoflurane of which the concentration was justified to achieve the end-tidal concentration within the range of 1,4-1,8 in oxygen and the PRST score less than 3. If the PRST score was greater than 3, then an additional bolus dose of 20-30 mg propofol and 25-50 µg fentanyl was injected along with increasing sevoflurane concentration. On the contrary, if signs of deep anesthesia were presented (PRST score=0, blood pressure decreased by more than 20% of the baseline values, bradycardia), then the sevoflurane concentration was cut down and 100 ml of crystalloid solution was rapidly infused within 2 minutes. A bolus dose of 100-200 µg phenylephrine was added if the blood pressure was still less than 20% of the baseline value in spite of these above-mentioned steps. Rocuronium was continuously infused at the rate of 10 µg/kg/min. The solution of bupivacaine 0,1% combined with fentanyl 2 µg/ml was infused via the epidural catheter at the rate of 5 ml/hour after a loading dose of 5 ml prior to skin incision. Normothermia was maintained during surgery. The pneumoperitoneum was implemented by CO2 insufflation at a pressure of 12 mmHg with room temperature in all patients. The intraoperative fluid was managed based on the goal-directed fluid therapy with a crystalloid solution. In brief, no additional fluids were provided if PPV was lower than 10%, otherwise, additional boluses of 250 ml ringer lactate solution were given over 10-15 minutes. After each bolus dose, PPV was re-assessed and further bolus was administered until reaching a PPV of lower than 10 %.

Intravenous 8 mg ondansetron was injected 30 minutes before the end of surgery to prevent postoperative nausea and vomiting. Rocuronium infusion was stopped 30 minutes before abdominal closure. The neuromuscular blockade was reversed in the postoperative care unit using intravenous neostigmine 40-60 µg/kg combined with atropine 0,5 mg. Patients were extubated when they met the extubation criteria (spontaneous tidal volume > 6 ml/kg and respiratory rate = 12-20 breath/minute, SpO2>95%, normocarbia, body temperature > 350C, positive gag reflexes and ability to follow a verbal command, hemodynamic stability without vasopressor support and ability to lift their heads and hold for 30 seconds [23]). Postoperative epidural analgesia for 48 hours was executed using bupivacaine 0,1% combined with fentanyl 2 µg/ml at a speed of infusion 5-10 ml/hour to maintain a visual analogue scale (VAS) score < 3. If the score was greater than 3, then a bolus dose 5 ml of the anesthetic solution was provided along with an increase in the speed of infusion. After extubation, patients were oxygenated via nasal cannula 3-5 ml/minute (1 ml/minute raises the FiO2 by 3%) to keep the SpO2>95%.

Ventilation protocol. Mechanical ventilation protocol was performed on the anesthesia machine GE healthcare carestation 620. The patient's ideal body weight was predefined according to formula: 45.5 + 0.91×[height(cm)-152.4] for women or 50 + 0.91×[height(cm)-152.4] for men. In both groups, the mechanical ventilation was set up at the volume-controlled mode, inspiration to expiration ratio of 1:2. After being lifted to 100% in the induction period, FiO2 was maintained at the level of 40% until extubation. Respiratory rate (starting with 18 breaths/minute) was justified to keep the end-tidal carbon dioxide (EtCO2) in the normal range of 35-40 mmHg. In the CV group, the tidal volume was set at 10 ml/kg IBW without PEEP and RM. In contrast, in the LPV group, patients were provided with a tidal volume of 7 ml/kg IBW and an intraoperative 10 cmH2O in PEEP. Simultaneously, in the LPV group, alveoli were recruited applying a stepwise increase in PEEP (from 4 to 10 cmH2O for 3 breaths, 10 to 15 cmH2O for 3 breaths, and 15 to 20 cmH2O for 10 breaths) with maximum PIP = 50 cmH2O [24]. The recruitment maneuvers were performed right after intubation, 30 minutes after CO2 insufflation, then every hour, and finally before extubation.

Data source and collection. The demographic characteristics including age, gender, height, weight, BMI, ASA physical status and history of coexisting diseases, and smoking were recorded. Vital signs (heart rate, blood pressure, SpO2, EtCO2, core temperature) were also recorded at 15 minutes intervals throughout the surgery. The volume of intravenous infusion (crystalloid, colloid solution), blood loss, urine output, total dose of anesthetics, fentanyl, and muscle relaxant were recorded as well. The dynamic compliance (Cdyn) was measured directly on the ventilator, while the static compliance (Cstat) was calculated in accordance with the pre-defined formula as Vt/(plateau pressure of the respiratory system - PEEP) with the plateau pressure measured during the normal ventilation setting using an inspiratory pause at 10% of the inspiratory time. Both types of pulmonary compliance were recorded at H0 (after intubation), H1 (30 minutes after pneumoperitoneum), H2 (1 hour after pneumoperitoneum), H3 (2 hours after pneumoperitoneum), Hkt (10 minutes after pneumoperitoneum stopped) and Hro (before extubation). Arterial blood samples were withdrawn from the radical arterial cannula for blood gas analysis before induction, 1 hour after pneumoperitoneum, and day 1 after operation. The pulmonary oxygenation index (OI) and the alveolar-arterial oxygen gradient (A-aO2) was calculated respectively as OI= PaO2/FiO2 and A-aO2= (PB-PH2O)×FiO2 -PaCO2/R - PaO2 where PB (atmospheric pressure) was 760 mmHg, PH2O (saturated vapor pressure at room temperature) was 47 mmHg, and the R (respiration quotient) was 0.8. SpO2 in the first postoperative day was measured in the room air (the nasal oxygen catheter was removed for 10 min) with the patient in bed. If SpO2 dropped to lower than 90%, the measurement was stopped and the supplemental oxygen was immediately re-provided via the nasal cannula.

Pre- and postoperative (day 1) chest radiography was obtained at the bedside and was analyzed in a blinded way by a radiologist who did not get involved in the study. Pathological chest X-ray was defined as the presence of at least one of the following: an increase in thickness of the interstitium, atelectasis, pleural effusion, localized or diffused infiltrates, and other chest radiological alterations.

Modified Clinical Pulmonary Infection Score (mCPIS). The modified original version of the Clinical Pulmonary Infection Score described by pelosi et al. (2008) was applied in the first postoperative day.

Incidence of postoperative pulmonary complications (PPCs). PPCs were defined by the European Perioperative Clinical Outcome (EPCO) definitions .

Primary and secondary endpoints. Our hypothesis was that the intraoperative lung-protective ventilation strategy could improve the pulmonary mechanics, oxygenation function, and ameliorate early postoperative pulmonary complications. The primary endpoints were the intraoperative fluctuation of Cdyn and Cstat, the intra- and postoperative changes in pulmonary oxygenation function including OI, A-aO2. The secondary endpoints were the alteration on chest x-ray, mCPIS, and the incidence of PPCs on the first postoperative day.

Statistical analysis. The sample size was calculated in accordance with the formula: n= (2×C)/δ2 +1 with δ=|µ1-µ2|/σ, where n is the sample size in each group, µ1= mean of OI in the LPV group, µ2 = mean of OI in the CV group, σ is the common standard deviation and c = 7.9 for 80% power. Based on the study of Xin Pi (2015) whose primary outcome was the OI after 2 hours of ventilation in the two groups was 382.21 ± 88.03 µg and 450.10 ± 70.29 µg respectively. Replacing µ1=382.21, µ2=450,10, σ=88,03 in the formula, n was equal to 27,5 for each group, which represented that the minimum sample size for each group was at least 28 patients.

Statistical analysis was completed using SPSS software version 20.0 (IBM, USA) on an intention-to-treat basis. Whether variables distributed normally was tested with the Kolmogorov-Smirnov and Shapiro-Wilk test. Continuous variables were compared applying either Student's t-test or the Mann-Whitney U test, depending on the characteristics of the distribution of the variable, and consequently were presented as mean ± SD or median and interquartile range (25-75%) as appropriate. As for categorical variables, the χ2 test was employed for comparison and the Fisher exact test was used for small frequencies. All the tests were two-tailed, and statistical significance was accepted at p<0,05.