Improving perioperative outcome

The ideal aim of perioperative medicine is to provide complete organ protection without complications, with no failures to rescue and with total avoidance of postoperative disability and loss of independence. This implicitly means no unplanned readmission to hospital or unplanned admission to the ICU. Fundamental to the success of such strategies are the highest standards in perioperative medicine. Clinical excellence in this field requires extensive expertise in various domains of medicine that include the ability to critically appraise and integrate all sources of patient information, a talent for organisation, a large knowledge base and a desire for continuous learning and the ability to control outcomes. Busy clinicians too often neglect the last two issues. This editorial offers a short update of some relevant scientific investigations in high-risk patients having noncardiac surgery, which could be incorporated into daily practice to improve outcomes. Standardisation of outcome definitions Precise definitions of perioperative events together with severity grading and outcomes are important for perioperative science, not only to avoid speculation and reliance on evidence of poor quality, but also to enable high standards in clinical practice. What cannot be measured precisely cannot be improved and what cannot be diagnosed accurately cannot be treated specifically. Meaningful comparisons of endpoints require consensus and standardisation of definitions, and should ideally encompass the full extent of surgical outcomes such as functional status and quality of recovery. Recently published definitions of adverse perioperative events such as myocardial injury after noncardiac surgery, myocardial infarction, pneumonia based on the Centers for Disease Control and Prevention criteria, delirium, including a simple screening score and acute renal failure are highly useful in this regard.1 But the proposed standards of definition have their limitations, in particular for combined endpoints such as postoperative pulmonary complications, where cardiac complications (pleural effusion) may be involved. Also, not all the definitions have been externally validated or tested. More precise definitions including biomarkers will undoubtedly be published soon. However, even without very precise definitions, high-quality care with excellent patient relevant outcomes can be achieved in everyday clinical practice, and a key part of this is organ protection. Organ protection Protecting organs during the period of perioperative stress is a priority of perioperative medicine. One classic example of this is the management of the cardiovascular system during a surgical intervention. Control of intraoperative hypotension and variation in blood pressure is important for three organ-specific outcomes in noncardiac surgery: myocardial impairment,2,3 cognitive dysfunction (delirium)4 and acute kidney injury.5 Despite this, to date, no adequately powered randomised controlled trial assessing the effect of targeted interventions against intraoperative hypotension on clinical outcome has been conducted. Standardised definitions of intraoperative events such as intraoperative hypotension are required. The best perioperative treatment of hypotension, with fluids and pharmacologic agents using a goal-oriented strategy (mini-fluid challenges, noradrenaline or phenylephrine) requires further investigation.6,7 The impact of haemodynamic trials on clinical outcome will depend on thoughtful design tailored to clinical situations such as bleeding, inflammation related to surgical trauma and patient factors and on meticulous decisions regarding control group care. Overall, the future success of trials will require not only more precise standardised descriptions of perioperative outcome, but also preoperative organ ‘wellness’. This would include preoperative cardiac biomarkers8 and attention to the preoperative cognitive state, two things that may be important co-determinants of outcome. Recent evidence for organ protection is summarised below. Cardiac protection Three findings are of note: β-blockers seem to be useful only in patients with high cardiac risk,9 and in the absence of high-quality evidence, the transfusion trigger in this group should be set at haemoglobin levels more than 80 g l−1 to avoid possible harm.10 A useful simplification of the perioperative management for patients with atrial fibrillation and anticoagulation with warfarin has been tested with success in a randomised controlled trial. The new recommendation is that no bridging with low molecular weight heparin before surgery is necessary for these patients.11 Pulmonary protection During general anaesthesia, most patients are mechanically ventilated and protective ventilation, including lower tidal volumes with low plateau pressures are effective in decreasing postoperative pulmonary complications, but the effect of positive end expiratory pressure and recruitment manoeuvres seem to be controversial in patients with noninjured lungs and without respiratory dysfunction.12 The introduction of high flow nasal oxygen in the postoperative setting is a very promising simple intervention to decrease unplanned postoperative intubation related to postoperative respiratory failure.13 Postoperative noninvasive ventilation seems to be an effective measure to avoid unplanned intubation in patients with postoperative hypoxaemic respiratory failure.14 Both measures should decrease the rate of unplanned postoperative intubation. Prevention of pulmonary complications with incentive spirometry has again been shown to be ineffective in a randomised controlled trial. It is, therefore, not a standard of care and cannot be recommended for routine use.15 The prevention of postoperative pulmonary complications after major abdominal surgery in high-risk patients using continuous positive airway pressure is the subject of a large ongoing trial (www.prismtrial.org). Major postoperative pulmonary complications prolong hospitalisation and require specific care services such as intermediate or ICUs. This makes precise preoperative risk stratification important for planning individual surgical and perioperative analgesic procedures, and the level of postoperative care. A recent risk predictor, the PERISCOPE (Prospective Evaluation of a RIsk Score for postoperative pulmonary COmPlications in Europe) model with patients from different European countries, is highly helpful for the management of high-risk patients16 and may contribute to high-quality clinical practice with adapted individualised pathways.17 Notably, the PERISCOPE model takes into account both preoperative medical history such as the presence of congestive cardiac failure or chronic liver disease and intraoperative factors, including the length of surgery and type of surgical intervention, when evaluating risk for postoperative respiratory failure. This reinforces the need to consider both pre and intraoperative factors when assigning the degree of postoperative risk. Renal protection Perioperative renal dysfunction is quite often neglected or reduced to surrogate endpoints such as oliguria and polyuria, which are poor markers of organ dysfunction.18 Even minor perioperative increases in serum creatinine may be associated with increased mortality,19 therefore, early surveillance of renal function and early intervention, including adaptation or avoidance of drugs (NSAIDS, diuretics, antihypertensive drugs) and fluids that may adversely affect renal function are essential in the perioperative setting. However, the best approach for diagnosis and surveillance of acute kidney injury still depends on classical measurements of biomarkers such as creatinine, urea and creatine kinase in patients with rhabdomylysis.20 Newer approaches such as surveillance of early but nonspecific altered blood gas analyses (hyperchloraemic, nonlactic acidosis) are controversial. Similarly, early altered urinary biomarkers such as tissue inhibitor of metalloproteinase 2 and insulin-like growth factor binding protein 7 (TIMP2 * IGFBP7),21 and ultrasonographic Doppler investigations (renal resistive index22) have not been investigated for accurate discrimination, external validation and long-term benefit in large cohorts undergoing noncardiac surgery. There is uncertainty regarding the feasibility of perioperative renal protection; remote ischaemic preconditioning seems to be effective in cardiac surgery23 but needs to be tested in noncardiac surgery. The renoprotective or harmful effects of intravenous fluids (NaCl 0.9% and colloids) are still controversial.24,25 Brain protection Postoperative delirium is a major burden for older patients, relatives and healthcare providers. Patients who suffer have increased lengths of hospital stay, increased readmission rates and often a discharge to longer term institutional care or a nursing home. It is also a risk factor for posttraumatic stress disorder.26 Postoperative delirium is a silent epidemic in elderly patients, with some effective prevention strategies but lacking high-quality prediction models and specific drug treatments that work. Potentially, reduced exposure to (deep) general anaesthesia may decrease the incidence of postoperative delirium.27 Delirium is a well known condition at the end of life: preoperative delirium, often combined with advanced dementia, should be an alarming sign that a patient may not have a sufficient physical and cognitive reserve for a successful surgical intervention. Another perioperative neuropsychological problem is postoperative opioid dependency, which calls for new acute pain strategies and prolonged follow-up.28 Minimally invasive surgical techniques, such as video-assisted thoracoscopic surgery, substantially decrease postoperative pain29 and regional anaesthesia combined with non-opioid analgesia may contribute to decreased opioid dependency. Control The control of outcomes is essential for high-quality care during the perioperative period. There is a consensus that good management in surgery and anaesthesia reduces mortality, disability, loss of independence and loss of health-related quality of life, even in high-risk patients with considerable postoperative complication rates. The timing and type of outcome to be measured remain controversial, although it is clear that mortality alone is not a sufficient indicator of standards of postoperative care. For many hospitals in Europe, it is quite easy to estimate in-hospital or 30-day mortality and an adjustment with available prediction models is feasible.30 Increased transparency of mortality outcome data could be a first step in a pan-European quality improvement process. However, early mortality is only a surrogate (but important) outcome and is only the tip of the iceberg of long-term outcome. Postoperative mortality is much higher after 1 year and depends on age, urgency of surgery and comorbidities.31 Scientific reports are seldom available for frail patients at the end of life. These patients with impaired sensorium but with surgical indications,32 are highly prevalent in high-income countries and may decrease the overall success of perioperative management.33 These situations should be identified promptly during the preoperative evaluation and the preoperative interventional plan should also consider palliative interventions to avoid stressful and futile surgical and nonsurgical procedures. Hopefully in the years to come, we will become wiser about the factors involved and be more accurate in our approach. Management Perioperative medicine is multidisciplinary, needing a high degree of coordination. But this aspect and its influence on outcomes have rarely been investigated except in studies using the ‘enhanced recovery after surgery’ concept. The impact of a number of logistical risk factors on outcome has not been adequately examined. Featuring in this list are availability of operating theatres and postoperative intermediate or critical care, and availability of the right personnel, including the proportion of those with clinical expertise in specific domains. These may include dedicated perioperative teams for high-risk surgical patients that provide a chain of care along the whole perioperative pathway. Strategies to avoid drug errors and infections acquired as an inpatient are associated with improved outcomes; however, organisational and economic aspects may limit widespread use of simple safety measures. These measures may include prefilled syringes,34 hand hygiene35 and checklists.36 More studies on the barriers to implementation of these simple measures in the perioperative period would be of high interest. One gleam of light in this logistic setting was the finding that a toilet in the recovery room decreases the time to mobilisation37 and this finding could be a further measure in the concept of enhanced recovery! Conclusion In the last years, three avenues of progress can be identified which are suitable for implementation in clinical practice: The availability of more precise and validated models of risk prediction allowing identification of specific at-risk patient groups. Electronic health record systems may support the integration of such predictive models. Higher standards of clinical diagnosis are now available. These should be used in not only in research but also in clinical practice to reduce diagnostic errors. Outcome measures should be available, at least during the in-hospital stay for quality and safety purposes. A caveat is that these outcomes may underestimate postoperative complications. Considerable progress has been made in the pulmonary setting of perioperative medicine. Hopefully, these scientific advances will be adopted to provide benefit to all. Acknowledgements related to this article Assistance with the editorial: none. Financial support and sponsorship: none Conflicts of interest: none. Comment from the editor: this editorial was checked and accepted by the editors but was not sent for external peer review. MC is an Associate Editor and BW is a Deputy Editor-in-Chief of the European Journal of Anaesthesiology.