Non-invasive ventilation in the recovery room for post-operative respiratory failure : a feasibility study

Background: Non-invasive ventilation (NIV) has become a standard of care in acute respiratory failure. However, little data is available on its usefulness in recovery ward patients after general surgery. The present study aimed to document the feasibility of implementing NIV in this setting, and its impact on lung function. Methods: During a 12-month period, all adult patients who underwent elective general surgical procedures under general anaesthesia during weekdays, were transferred to the recovery ward after extubation, and those who required NIV were included in this prospective observational study. NIV was applied with a bilevel device (VPAP II ST, ResMed, North Ryde, Australia). Results: 4622 patients were admitted to the recovery ward, 83 of whom needed NIV. NIV increased pH (7.38 ± .06 vs 7.30 ± .05), reduced PaCO2 (7.38 ± .06 vs 7.30 ± .05) in hypercapnic patients (44 ± 9 vs 55 ± 10 mm Hg), and increased PaO2 in non-hypercapnic patients (80 ± 10 vs 70 ± 11 mm Hg). No complications attributable to NIV occurred. Most patients improved after 1–2 NIV trials, and all were transferred to the ward the same day. Conclusions: In recovery ward patients after general surgery, NIV is seldom required. When applied, NIV seems to exert favourable effects on lung function. NIV can be safely implemented with a bilevel device in a recovery ward not accustomed to the use of ICU ventilators. The costeffectiveness of its systematic use in this setting should be assessed.

During the last decade, non-invasive ventilation (NIV) has proven to be an effective strategy to reduce intubations rate, intensive care unit (ICU) and hospital lengths of stay, morbidity and mortality in patients with either hypercapnic [1] or non-hypercapnic [2] acute respiratory failure.NIV has thus become a standard of care in the management of these patients [3,4].Beneficial effects of NIV have also been documented in post-operative patients.Indeed, surgery can induce respiratory muscle dysfunction, which in turn can lead to hypoventilation, hypoxemia, atelectasis and infections [5].Acute respiratory failure is one of the major complications of the early post-operative period [6].NIV, performed in the intensive or intermediate care unit, has been shown to improve pulmonary function and gas exchange in patients after gastroplasty [7], lung resection surgery [8], and coronary artery bypass grafting [9].However, little data have been published so far on the use of this technique outside of these specialised units after general surgery.Indeed, many such patients undergo a relatively brief stay in the recovery ward before being transferred to the general surgical ward.Recovery wards are often not equipped with intensive care ventilators, and nursing or physiotherapy staff is not trained in their use.We recently introduced NIV in our recovery ward, performed with a bilevel ventilator, and conducted a one-year prospective observational study to test the feasibility of the technique, quantify the need for its use, and evaluate its impact on pulmonary function.The present paper summarizes the results of that study.

Patients
The study was approved by the Department of Anaesthesiology's clinical investigation committee, and was in accordance with the modalities of the Helsinki Declaration.During a 12-month study period, all adult patients who underwent elective general surgical procedures under general anaesthesia during weekdays, were transferred to the recovery room after extubation, and those who required non-invasive ventilator support were included in this prospective observational study.The need for NIV was established by the physician in charge of the patient in the recovery room, based on our usual practice guidelines.A physician is constantly present in the recovery room.The guidelines recommend NIV when signs of respiratory failure are present (respiratory rate >25/min.or <10/min., and hypoxemia with arterial partial pressure of oxygen/inspired oxygen fraction ratio (PaO2/FIO2) <300 and/or arterial partial pressure of carbon dioxide (PaCO2) ≥ 45 mm Hg), after ensuring of adequate analgesia and airway secretion clearance.Briefly, the evaluation algorithm calls for pulmonary auscultation and an arterial blood gas analysis to be performed if the respiratory rate is >25/min.or <10/min., followed by a chest X-ray if PaO2/FIO2 <300, or if abnormal breath sounds are present.If signs of atelectasis are present on the chest X-ray, NIV is initiated even if the PaO2/FIO2 is >300.
On the basis of the blood gas analysis, two groups of patients were defined: "non-hypercapnic" (PaCO2 <45 mm Hg), and "hypercapnic" (PaCO2 ≥ 45 mm Hg).Exclusion criteria were: emergency surgery, transfer to the surgical ICU, tension pneumothorax, or any other classical contraindication to NIV (eg facial lesions, impaired consciousness, poor patient cooperation, haemodynamic instability) [3].

Non-invasive ventilation protocol
The choice between continuous positive airway pressure (CPAP) or pressure support was made according to current recommendations of opting for the former in nonhypercapnic and the latter in hypercapnic respiratory failure [3,4].NIV was applied by a trained respiratory therapist.Our recovery room is not equipped with high-end, multi-ventilator mode ICU ventilators based on proportional solenoid valve technology [10].Therefore, NIV was applied with a bilevel turbine-type device, the VPAP II ST (ResMed, North Ryde, Australia), which can be used either in CPAP or pressure support mode.CPAP was initiated with a positive end-expiratory pressure of 5 cmH2O.Pressure support was initially set at 10 cmH2O inspirator pressure and 5 cmH2O positive end-expiratory pressure (PEEP).Both modes were subsequently titrated according to patient tolerance, pulse oxymetry oxygen saturation (SpO2) and respiratory rate.A backup rate ("ST" mode on the VPAP II ST) was applied in those patients in whom initial respiratory rate was <10/min.The target duration of the initial NIV trial was 45 min.Mean arterial blood pressure, respiratory and heart rates, dyspnea assessed by a visual analogue scale, and arterial blood gases were collected before, at the end (while still on NIV), and 30 minutes after this initial trial.Subsequently, NIV was applied if the initial criteria for its application were again met.Despite NIV, endotracheal intubation was performed if the following commonly accepted criteria [11] used routinely in our clinical guidelines [12], were met.Major criteria: respiratory arrest, loss of consciousness, severe agitation, haemodynamic instability (systolic blood pressure <70 mm Hg or >180 mm Hg, heart rate <50/min.).Minor criteria: respiratory rate >35/min.and higher than admission value, arterial pH <7.30 and lower than admission value, PaO2 <45 mm Hg despite oxygen supplementation, neurological deterioration, weak cough reflex with secretion accumulation.Intubation was performed if one major or two minor criteria were present [11,12].If intubation was required, or if the patient's condition was judged too unstable for same-day ward transfer, the patient was transferred to the surgical ICU.

Results
During the study period, 4622 patients were transferred to the recovery room following elective surgical procedures, 83 of whom required NIV.Of those 83 most underwent thoracic, cardiovascular or abdominal surgery (table 1).The main clinical characteristics of the patients having received NIV are summarised in table 2. The male/female ratio was 47/36.More than 50 % of patients had a history of a chronic respiratory condition, obstructive disease being the most preva- lent.The last documented 1 s forced expiratory volume / forced vital capacity ratio (FEV 1.0/FVC) in this group was 58 ± 9% (mean ± SD).Following the criteria defined above, of the 83 NIV patients, 60 (72%) were non-hypercapnic and 23 (28%) hypercapnic.The predominant indications for NIV were altered blood gases and signs of atelectasis on the chest X-ray (table 2).Non-hypercapnic patients received mostly CPAP, hypercapnic patients being treated with pressure support with or without the backup rate.

Discussion
To summarise, the results of this study show that NIV with a bilevel device can be safely applied to patients in the recovery room after elective surgery under general anaesthesia.Moreover, it seems to improve gas exchange without the need for repeated trials and/or its prolonged application.
Before discussing these results further, some of the limitations of the study should be emphasised.First, this was not a randomised prospective trial, nor a controlled study.Therefore no firm conclusion can be drawn regarding the role of NIV in avoiding re-intubation in the post-operative setting.Nonetheless, none of the 83 patients had to be re-intubated, and one can hypothesise that NIV might have contributed to this outcome in at least some patients.Second, the study excluded patients undergoing emergency or major cardiac surgery, as well as those who were transferred directly from the operating room to the ICU.Therefore, our study population was skewed towards more stable patients, as reflected by the low incidence of need for NIV.This was partly due to institutional organisation constraints, since our recovery room only manages patients who have already been extubated, and are expected to stay for <24 hours before ward transfer.However, respiratory problems do still occur in this population, as shown by the present study.Third, NIV was applied by a trained respiratory therapist.In our recovery room, respiratory therapists are on call to apply NIV.However, this is clearly not the case in many institutions.Therefore our findings that NIV can be applied safely in this setting must be extrapolated with caution; proper training and experience are prerequisites for the safe application of NIV, wherever it is used.
Several studies have documented the favourable effects of NIV in acute respiratory failure, most having observed that numerous trials of NIV were often required, at times uninterrupted for several hours, before weaning was possible [1,2,13].The post-operative setting is different; respiratory failure after successful extubation often stems from causes such as shallow breathinginduced atelectatic lesions, which result from diaphragmatic function impairment or pain, anaesthetic and/or analgesic drugs-induced hypoventilation, and/or peri-operative fluid overload [5,[7][8][9].Contrary to the mechanisms underlying acute respiratory failure in the classical studies on NIV, most of these factors are usually rapidly reversible in this context.This probably explains why in two-thirds of the patients only one application of the technique was sufficient, and why the remaining one-third responded favourably to one or two additional trials.Supporting this assumption is the fact that the rapidly favourable outcome was also present in those patients with a history of prior respiratory disease.
Another notable difference is the very low dyspnea score of the patients, given that dyspnea is usually a prominent finding in acute or acute on chronic respiratory failure [14].This most likely resulted from the analgesic medication that these patients had received, in particular opiates, which are known to effectively decrease dyspnea [15].
NIV can be performed with either an ICU ventilator or a bilevel pressure support device.Compared to the latter, the former is more powerful, exhibits more adjustable features (eg trigger type and sensitivity, pressurisation slope, cycling criteria) and provides extensive monitoring and alarm capabilities [16].However, it is usually not available outside the ICU, and has a steep learning curve, especially for non-ICU staff.An alternative is to use bilevel-type devices.Indeed, such ventilators, whose performance has improved in recent years [10,17,18] are easier to use, take up less space, and are considerably less costly.These characteristics have made them an interesting option in the treatment of patients with moderately severe respiratory failure [19].

Conclusion
In conclusion, the present feasibility study shows that a brief application of NIV with a bilevel device can be safely applied in the recovery room to patients following general surgery, and presenting signs of respiratory failure.It also demonstrates that the incidence of acute respiratory fail-ure is very low in this patient population.Therefore, trials should now be performed to explore whether the beneficial pathophysiological effects of NIV can have a favourable impact on the outcome of post-operative patients with signs of respiratory insufficiency.

Swiss Medical Weekly
The many reasons why you should choose SMW to publish your research Med Wochenschr (1871-2000) Swiss Med Wkly (continues Schweiz Med Wochenschr from 2001) Editores Medicorum Helveticorum

Table 3
tinued.Very little change in these variables was noted, except for respiratory rate, which was lower in the hypercapnic group and increased with NIV.NIV also lowered PaCO2 and increased pH in the hypercapnic patients, and improved arterial oxygenation in the non-hypercapnic ones.In 55 patients (66%), only one application of NIV was necessary.Twenty-eight patients (34%) required a

Table 3
Results of the initial application of NIV.second NIV trial within two hours after the first.No complications were noted during NIV.Leaks around the mask occurred in 35 (42%) patients, but were not severe enough to interfere with the proper administration of the technique.No patient required either intubation or transfer to the ICU.All patients were transferred to the ward the same day.