Anaesthetic management and unplanned admission to intensive care after thoracic surgery.

Abstract

Over the past three decades, the emergence of less invasive treatments and better peri-operative anaesthetic care in thoracic surgery have contributed to better clinical outcomes, despite the increasing age of patients and their higher burden of comorbidities 1, 2. Postoperative pulmonary complications, namely acute respiratory distress syndrome, pneumonia and bronchopleural fistula, occur frequently after lung cancer resection and they are largely implicated as causes of early mortality and poor long-term survival 3. The decision to admit a patient to the intensive care unit (ICU) or the high dependency unit or the recovery area are mainly based on the type of operation, the patient's comorbidities, the institutional policy as well as the availability of beds in these settings 4. Most admissions to the ICU are planned or anticipated before the scheduled case, with the aims of preventing postoperative complications and enhancing functional recovery. Unplanned or emergency ICU admissions after surgery are much less frequent and are related to unexpected peri-operative adverse events such as severe bleeding, persistent hypoxaemia, unstable hemodynamic conditions and hypothermia or delirium. The aim of unplanned ICU admission is to reverse any organ dysfunction using enhanced monitoring and nursing care, along with fluids, drugs and organ supportive therapy. For clinicians and healthcare managers, unplanned ICU admission has become a valid indicator to assess patient safety in surgical patients and can be used in cost-effective benchmarking as well as in root cause analysis of peri-operative adverse events 5. Compared with direct or planned admission to ICU, unplanned ICU admission is associated with a significantly higher risk of death beyond the expected consequences of comorbidities, age, type of surgery and emergency status. In this issue of Anaesthesia, Shelley et al. report an incidence of 2.3% unplanned ICU admission among 7431 cases of lung resection performed in 16 UK thoracic surgical centres 6. The vast majority of patients who required unplanned ICU admission in this study also required mechanical ventilation of their lungs, although the cause(s) of respiratory failure and timing of admission to the ICU were unclear or unreported. Multivariate analysis indicated that unplanned ICU admissions were less frequent in patients receiving intravenous (i.v.) anaesthesia, compared with inhalational anaesthesia, and in patients receiving thoracic epidural analgesia, compared with other analgesic techniques, including paravertebral block (PVB). In a larger prospective dataset from the American College of Surgeons National Surgical Quality Improvement Program (n = 16,696), 3.6% of patients required unplanned re-intubation of their tracheas within 30 days after lung surgery. Interestingly, the need for postoperative invasive ventilatory support was associated with patients' frailty features (advanced age, poor functional status, low haematocrit, low albumin level) and with the severity of the surgical stress (open thoracotomy, prolonged surgery) 7. Nowadays, anaesthetists are recognised as a key member of the thoracic team as in the liver transplantation team where the anaesthetists' level of experience has been shown to influence postoperative mortality and graft failure 8. Likewise, in thoracic surgery, the application of proper lung isolation techniques and of physiologically evidence-based practices may contribute to the success of the intervention and also to improve clinical outcomes 9. From several cohort studies, we know that major postoperative pulmonary complications are strongly associated with risk factors such as pre-existing low aerobic physical fitness and inspiratory muscle weakness, as well as with intra-operative positive fluid balance and elevated inspiratory driving pressure 10, 11, all factors that could be targeted by specific interventions. Indeed, implementation of pre-operative exercise training has been effective in reducing postoperative complications and shortening hospital length of stay 12. Likewise, the application of an open lung ventilatory strategy using low tidal volumes, individualised or fixed levels of PEEP and alveolar recruitment have been shown to improve lung mechanics and reduce postoperative pulmonary complications 13. So far, a strong body of evidence also supports the use of thoracic epidural anaesthesia and PVB. A lower-risk of postoperative pulmonary complications and chronic pain has been reported when epidural analgesia was provided in patients undergoing open thoracotomy 14, 15. More recently, PVBs have become more popular given their simplicity (insertion under ultrasound-guidance or visual control by the surgeon), safety and effectiveness. Compared with epidural analgesia, no difference in analgesic efficacy, similar reduction in postoperative pulmonary complications and lower-risk of hypotension and urinary retention have been reported in patients with PVBs following open thoracotomy 16. With the increasing trend for minimally invasive surgical approaches (e.g. multi-, uni-portal video- or robotic-assisted thoracic surgery), novel analgesic techniques such as intercostal and serratus anterior plane blocks have been introduced in multimodal opiate-sparing regimes with the aims to provide patient comfort while avoiding depression of sympathetic nerve activity and preserving locomotor and respiratory muscle function in order to enhance functional recovery and shorten hospital stay 17, 18. Regarding the effects of general anaesthesia, a recent meta-analysis including 23 randomised controlled trials (RCT, n = 1349) showed that inhalational anaesthesia was associated with increased intrapulmonary shunting and consequent impairment in gas exchange during one-lung ventilation, whereas the incidence of postoperative pulmonary complications was reduced – regardless of the type of analgesia – when compared with i.v. anaesthesia 19. There was a low level of heterogeneity among the trials (I2 statistic of 4%) and for every 1000 patients receiving inhalational anaesthesia, it was expected that 150 postoperative pulmonary complications could be prevented due to the administration of inhalational anaesthesia. Furthermore, trial sequential analysis indicated that the total number of patients enrolled (n = 554) was larger than the required sample size and, therefore strongly supports the positive clinical impact of inhalational anaesthesia. In agreement with these findings, an analysis of a multicentric cohort of patients undergoing non-cardiac surgery (n = 124,497), including 4880 thoracic interventions, demonstrated a dose-dependent protective effect of inhalational anaesthesia against postoperative pulmonary complications, along with a reduced 30-day mortality rate and a decrease in hospital costs amongst those patients exposed to higher doses of inhalational anaesthetics 20. There are multiple factors that may influence what agents anaesthetists use for maintenance of anaesthesia. Most anaesthetists prefer to use volatile anaesthetic agents instead of i.v. agents to deliver general anaesthesia in major surgery. The inhalational anaesthetic technique offers a simple and reliable way to control the depth of anaesthesia; to protect the organs against ischaemia-reperfusion injury; to potentiate the effects of neuromuscular blocking agents and to decrease airway resistance 21, 22. In experimental models of sepsis, exposure to halogen-containing hydrocarbons has been shown to improve survival, whereas propofol anaesthesia is associated with increased mortality along with larger systemic release of endotoxins 23, 24. Likewise, in thoracic surgery with one-lung ventilation, the early pulmonary protective effect of inhalational anaesthesia is accompanied by lower levels of alveolar inflammatory mediators, when compared with i.v. anaesthesia 25. In contrast, regarding the long-term effects, the impairment in cell-based immunity induced by inhalational anaesthesia may promote ‘seeding' of tumour cells at the time of cancer resection, and could explain a higher rate of cancer recurrence and a reduction in long-term survival, when compared with total i.v. anaesthesia 26. Since hypoxic pulmonary vasoconstriction is inhibited by volatile anaesthetics, some anaesthetists preferentially select total i.v. anaesthesia to avoid (or to treat) hypoxaemia due to intrapulmonary blood shunting. Another reason to select an i.v. agent is the potential need to continue sedation to allow mechanical ventilation in the ICU, a place where inhaled anaesthetic agents are usually not used. Is it possible to explain the conflicting results regarding anaesthesia-related lung protection? Shelley et al.'s findings challenge common clinical practice in thoracic anaesthesia 6, whereas there is a growing body of scientific evidence supporting the organ protective effects of volatile anaesthetic agents 19, 20. Although the results from RCTs and observational studies are most frequently in agreement, despite some variations in their effect size 27, there are numerous examples of discrepant results in the medical literature (e.g. hormone replacement therapy and cardiovascular events) 28. Facing these contradictions in the literature, thoracic anaesthetists are left to wonder which paper's findings to implement and which studies offer the most reliable, but also the most relevant, results. The reader should not automatically assume the superiority of RCTs if rigorous methods are applied in observational studies. Analysis of large cohorts offers the unique opportunity to test a number of hypotheses at low cost, within a short period of time, in a large population sample and representative of the ‘real clinical world'. On the other hand, RCTs test the effects of a treatment in a relatively homogenous population with few confounding factors but less generalisability. In RCTs, patients' conditions and surgical characteristics are usually well balanced and study outcomes are clearly defined with less risk of misclassification. The Achilles heel of observational studies is the presence of confounding factors related to population characteristics, as well as to treatment and outcomes. In the article by Shelley et al., although unplanned ICU admissions were mainly related to the need for mechanical ventilation, the reasons why patients were transferred directly from the operating theatre or later from the surgical ward (or recovery area) to the ICU, were not stated 6. Adjustment of the multivariate analysis with age, sex and surgical features supported an association between i.v. anaesthesia and the occurrence of unplanned ICU admission. However, other potential risk factors such as the severity of pre-existing cardiopulmonary diseases and muscular weakness, ventilatory strategies (e.g. driving pressure, amount of PEEP), the amount of i.v. fluids given and the duration of surgery were not reported, and could therefore not be included in the imputation models of multivariate analysis. In the operating theatre, what leads a clinician to select either i.v. or inhalational anaesthesia? Presumably, patients with a higher pre-operative risk profile due to advanced respiratory, cardiac or renal diseases would be the best candidates to receive volatile anaesthetic agents, given their preconditioning and protective organ effects. Conversely, switching from inhalational to i.v. anaesthesia could be indicated if patients become hypoxaemic due to intrapulmonary shunting, and/or if they experience intra-operative complications that justify the continuation of mechanical ventilation after surgery. In most RCTs, (very) high-risk patients are often excluded, although they might be the group that potentially benefit the most from organ protective strategies. Shelley and his colleagues should be congratulated for analysing a large multi-centre database and choosing a study outcome that is easy to measure and of major importance to patients, peri-operative physicians and healthcare managers. The association between unplanned ICU admission and the type of anaesthesia could possibly be explained by differences in case-mix owing to the severity of patient's illnesses and by the complexity and duration of surgery. We should therefore consider these results as hypothesis-generating, rather than definitive. In the end, we need to remain open to the continual stream of clinical evidence that practitioners place at our disposal, but at the same time rely on rigorous clinical trials to guide our daily practice. No external funding or competing interests declared.