Adverse respiratory events (AREs) are still one of the most important causes of postoperative morbidity and mortality.1 The reported incidence of postoperative respiratory complications ranges from 5% to 80%, depending on the patient population and the diagnostic criteria used to define complications.2 Respiratory episodes that occur immediately after a surgical intervention are common, and ARE rates of 1.3%–6.9% have been reported after admission to a post-anesthesia care unit (PACU).3,4

Respiratory problems are common in the PACU.5 The vast majority are attributable to airway obstruction, hypoventilation or hypoxemia. Hypoxemia, irrespective of its etiology, is generally the final stage in a process leading to severe morbidity and mortality, and may be a result of hypoventilation or compromised alveolar oxygen exchange causing a rise in the alveolar-arterial gradient. Patients with significant comorbidities, particularly neuromuscular, pulmonary or cardiac dysfunction, have a higher risk of respiratory compromise, but all patients are at risk of developing postoperative hypoxemia, which occurs in 30%–50% of abdominal surgery and in up to 8%–10% of patients who require intubation and mechanical ventilation.6

In the immediate postoperative period, the most common causes of hypoventilation are airway obstruction, anesthetics, residual neuromuscular blockade (RNMB) and breathing difficulty due to poor control of incision wound pain.4,5,7–11 Postoperative oxygen exchange deficit may occur as a result of intrapulmonary shunt, lung edema and embolism.12 Pre-existing medical disorders, such as obstructive sleep apnea, may also make the patient more susceptible to airway obstruction in the PACU.13

The most common cause of airway obstruction in the PACU is pharyngeal floppiness caused by pharyngeal muscle weakness.11 Other causes include laryngeal spasm, airway edema, hematoma or presence of a foreign body.

Muscle weakness is a result of loss of pharyngeal motor tone caused by RNMB, the enduring effect of anesthetics, or narcotic drugs.10,14,15

The diaphragm recovers more quickly from the effect of muscle relaxants than the pharynx. Residual pharyngeal muscle paralysis causes the base of the tongue and the posterior oropharyngeal tissues to bunch together, producing supraglottal obstruction.16,17

The aim of this study was to determine the incidence and predictive factors of postoperative ARE in the PACU.

A total of 357 patients were admitted consecutively to the PACU during the study period and gave their consent to participate; of these, 340 participants were evaluated.

Seventeen patients were excluded: 7 were admitted to the surgical ICU; 3 could not give informed consent or had an MMSE score of <25; 3 did not undergo surgery; 1 was under 18 years of age; 1 did not speak Portuguese; 1 refused to participate; 1 underwent a neurosurgical procedure.

Median age was 57 years and median BMI was 26kg/m2. Demographic characteristics of the population, surgical details, anesthetic management and PACU data are summarized in Tables 1–3.

A total of 67 patients (19.7%) had ARE (Table 4). AREs most commonly observed were the following: inability to breathe deeply (15.9%), mild-moderate hypoxemia (5.9%) and upper airway weakness (3.2%). Other AREs recorded were upper airway obstruction (2.7%), signs of respiratory distress or impending ventilatory failure (2.1%) and severe hypoxemia (1.5%). None of the patients required reintubation and none had clinical evidence or suspicion of pulmonary aspiratory during their PACU stay. Multiple AREs were observed in 21 patients (6.2%).

The most common AREs occurred in patients who had undergone high-risk procedures (42% vs 24%, P=.003) and major surgery (37% vs 25%, P=.041).

AREs were more common in patients who were given general anesthesia (85% vs 67%, P=.004) and less common in patients given combined anesthesia (8% vs 23%, P=.004).

Five patients who received locoregional anesthesia developed inability to breathe deeply.

Neuromuscular relaxants were used in 66% of the patients: AREs occurred more commonly in these patients (79% vs 55%, P<.001). AREs were also more common in patients who received neostigmine (69% vs 48%, P=.002).

Neither duration of anesthesia nor duration of intervention affected the incidence of ARE.

AREs were more common in patients with RNMB (46% vs 11%, P<.001) and in patients with a diagnosis of hypoactive emergence (13% vs 5%, P=.015).

Multivariate analyses (Table 5) performed after adjustment for univariate predictive factors (general anesthesia, combined anesthesia, use of neuromuscular relaxant and neostigmine, hypoactive emergence, and major and high-risk surgery), revealed RNMB as an independent predictive risk factor for ARE.

A total of 20% of the patients had at least 1 ARE during their PACU stay.

The reported incidence of AREs recorded in the PACU varies widely. Observational studies have described rates of 1.3%–34%,4,25,26 depending on the definition of ARE, the duration of monitoring and the patient population. Murphy et al.26 found a rate of mild hypoxia of 22%, similar to the findings of this study, but the same author described a much lower rate in another study.

Patients presenting ARE in the PACU are more likely to have undergone high-risk surgical interventions, defined according to the RCRI.20 This index was initially designed to predict major cardiac episodes, but it is also applicable to other post-operative complications.27

AREs were also more common after major surgery; these patients may have received different anesthetic management, closer monitoring, and probably a longer surgical intervention. They were also more likely to have a poor preoperative status, which may also have affected the incidence of ARE.

Abdominal, vascular and thoracic surgeries were also defined as the major procedures by Warner,28 who found that postoperative hypoxemia and acute respiratory failure mainly occurred after abdominal or thoracic surgery.

The overall incidence of RNMB in the study population was 18%. Patients with ARE had a higher rate of RNMB than patients without ARE.

It is interesting to note that more patients with ARE had received neostigmine. Most anesthesiologists systematically administered a muscle relaxant reversal agent, but some may have used subjective criteria.

Many centers in Europe have stated that they do not use reversal agents.29–31 However, in view of the current evidence, this practice is not recommended in the standard management of these patients.32 In the hospital where this study was performed, reversal agents are generally administered to all patients as a standard procedure. This may explain the higher incidence of ARE in patients who received neostigmine.

Delayed PACU discharge due to ARE is difficult to evaluate, and depends on individual factors in each center, such as staffing resources, PACU size and availability of beds. A recent study suggested that AREs caused a delay in discharge from the recovery room,33 but we found no differences between the 2 study groups, both of which had a short mean PACU stay; the similarity in PACU stay in patients with or without ARE may be explained by the availability of intermediate surgical care units.

AREs were more common after general anesthesia: this was not unexpected and has been documented in previous studies.4 However, the protective effect of general anesthesia compared to combined anesthesia in our study was surprising. It may be that administration of NMBA is rarer in combined anesthesia and analgesia is more effective if not combined with narcotics.

No relation was found between duration of the surgical procedure and AREs; however, Canet34 suggested that duration of both anesthesia and procedure may be related with postoperative pulmonary complications after PACU discharge. These episodes may have been missed, since the period of observation in our study was the duration of PACU stay.

Hypoactive emergence, defined as a RASS score of ≤–2, was more common in patients with ARE, suggesting that these patients were more heavily sedated on PACU admission.35

Compared to the findings of Radtke et al.,35 our population had a higher rate of emergence problems (12%), mainly due to hypoactive emergence in 6.8% of patients.

Abdominal surgery is a risk factor that has been associated with hypoactive emergence,36 and may have contributed to the association observed between high-risk surgery and the incidence of RNMB. Abdominal surgery can also be related with the higher rate of hypoactive delirium in RNMB patients. Another possibility is that the relationship between ARE and hypoactive emergence may reflect the level of sedation of the patient, which increases the chance of hypoxemia and atelectasis in a non-ventilated patient.

The most common ARE during PACU stay was inability to breathe deeply, recorded in 16% (54 of 340) of all patients and 81% (54 of 67) of the patients with ARE. This high rate may be due to the association of RNMB and a hypoactive status. Mild-moderate hypoxemia was the second most common ARE, appearing in 6% (20 of 340) of all patients and in 29.9% (20 of 67) of those with ARE. An association between RNMB and postoperative hypoxemia has been described in other studies, but this study was not designed to identify the underlying mechanism of neuromuscular block (NMB)-induced hypoxemia. Murphy et al.21 affirmed that deterioration of the ventilatory response to hypoxia and respiratory muscle weakness are important causes of postoperative hypoxemia, but these causes cannot be evaluated in an ordinary clinical examination. Upper airway obstruction, another common cause of ARE, occurred at a rate of 3% (9 of 340) in the study population, accounting for 13% of all AREs. Inhaled anesthetics and narcotics can be considered as relevant factors, but RNMB may also play an important role.

Our study has several limitations that must be mentioned. First, the sample is small and the study was conducted in a single center, making it difficult to generalize our results beyond our hospital. Second, it was a prospective, observational study. No interventions were made in anesthetic practice before, during, or after surgery, and all treatments were administered according to the criteria of the clinician who was in charge of the PACU that day.

Furthermore, the definition of ARE contains several subjective criteria that may have affected diagnosis. Finally, interoperative NMB status was not monitored, and this may have affected postoperative respiratory function. This would also limit the interpretation of the influence of neostigmine administration and the duration of NMB. Although intraoperative acceleromyography has been shown to reduce the risk of ARE in the PACU,26 this benefit is less pronounced when reversal agents are systematically administered.37

To conclude, AREs are a common occurrence in the PACU, and are more common after high-risk surgical interventions and major surgery. Postoperative RNMB, hypoactive emergence and the administration of neostigmine and interoperative muscle relaxants were predictive factors for developing ARE. RNMB was an independent risk factor for postoperative AREs in the PACU.

Fernando Abelha and Alice Santos contributed to conception and design. Fernando Abelha contributed to analysis and interpretation. Alice Santos and Daniela Xará contributed to preparation of the significant intellectual content of the paper.

The authors have no conflict of interests.