Improvement of perioperative care for better outcomes after surgery

In this issue of the European Journal of Anaesthesiology, Rama-Maceiras et al.1 and Clarke et al.2 present interesting studies on complications of peri-operative care. The first study evaluated unplanned surgical re-operation in a tertiary hospital and the second mortality after emergency laparotomy. Quality of peri-operative care is an essential part of patient safety3 but is difficult to measure correctly. Many experts agree that post-operative 30-day mortality,4,5 post-operative unplanned return to the operating theatre6,7 and unplanned ICU admission8,9 are important indicators of the quality of peri-operative care, at least for major surgery. Advantages of these indicators are their ease of availability. However, interpretation of results using these indicators is tricky when used without reporting on, and adjustments for, case-mix. Furthermore, institutional factors such as structure, patient throughput, quality measures and adherence to these measures, and local skills, need to be known. Case-mix recording is particularly important because, for instance, outcomes of elective and non-elective surgery are quite different;10 these may explain partly the high mortality rate of 19% reported by Clarke et al.2 after emergency laparotomy. These three indicators of peri-operative care are not independent outcomes after surgery. Patients with unplanned return to the operating theatre have an increased post-operative mortality (16 and 23%) compared with patients without unplanned return (2.3 and 2.9%).1,6 Patients with unplanned admission to ICU after surgery also have an increased mortality (40%) compared with patients who do not require ICU admission (3.6%).8 Post-operative mortality is probably determined by the pre-operative health and fitness of patients, the extent of surgical intervention, modification of the stress response by the anaesthetic technique and the quality of peri-operative care. Experts have suggested that the most important risk factor for mortality is the pre-operative physical status of the patient. Clarke et al.2 add further evidence for the validity of this relationship; high pre-operative lactate concentration and predicted mortality using the P-POSSUM scoring system are, at least, in univariate analysis, associated with increased mortality. Unplanned return to the operating room is a major burden for the patient, personnel and organisation. Rama-Maceiras et al.1 reported an unplanned re-operation rate of 3.3% in a large tertiary hospital including transplantation surgery, with a considerable rate of non-elective patients (22.4%). The rate is higher than reported in a global population (1.8%),11 but slightly lower than in patients undergoing general surgery (3.5%),6 confirming the importance of precise case-mix reporting. The particular value of the study of Rama-Maceiras et al.1 is the identification of the causes of re-operation, mainly bleeding (27.3%) and infection (31.2%). With this knowledge, multi-faceted and multi-disciplinary bleeding12 and infection control quality improvement programmes13 during the peri-operative period can be started or reinforced to reduce unplanned re-operations. Unplanned ICU admission is also a major burden for all involved parties. Clarke et al.2 reported that 2.4% of patients after emergency laparotomy with the pathway through the post-anaesthesia care unit and wards had an unplanned return to ICU with a high mortality rate, probably often related to sepsis.14 After major thoracic oncology surgery, 8.6% had an unplanned return to ICU;9 after different types of surgery and an elective post-operative stay in ICU, 11.6% were readmitted.8 Most readmissions were related to respiratory, neurological or cardiovascular diseases and were more than 2 days after ICU discharge in 75% of cases. Again, multi-faceted and multi-disciplinary improvement programmes during the peri-operative period may contribute to decreasing unplanned ICU admissions. The impact of post-operative complications on post-operative mortality is considerable and was estimated to be more important than pre-operative patient risk factors and intra-operative factors.15 This important impact of post-operative complications on mortality is supported by the study of Ghaferi et al.,4 who investigated variation in post-operative mortality among American hospitals. With similar total and major complications, post-operative mortality varied between 3.5 and 6.9%. Rama-Maceiras et al.1 and Clarke et al.2 suggest that post-operative care may have been suboptimal. In both studies, a three pathways' system for post-operative care was described: a) recovery room and ward pathway, b) an intermediate pathway and c) an intensive care pathway. Pathways (a) and (c) are well established and effective for specific defined populations (e.g. day surgery and critically ill patients).16,17 Clarke et al.2 suppose that pathway (a) may not be the best for patients after emergency laparotomy.2 With an increasing number of older patients, most European countries are more frequently confronted with more fragile patients with serious pre-existing co-morbidities and emergency interventions, who need prolonged post-operative monitored care. However, some surgical interventions have become less invasive (e.g. laparoscopy, minimal invasive and endovascular procedures), leading to a need for less intensive care beds but more normal ward beds.18 The intermediate pathway is a grey zone with different local solutions concerning structure and organisation; a harmonisation with clear criteria about the definition of an intermediate care unit is required. Some preliminary basic considerations are available,19 but need to be adapted for post-operative intermediate care. Patients' outcome may be different if they stay in intermediate care units under the care of anaesthesiologists and ICU physicians rather than under the care of surgical teams. In the future, scientific work is required to define the best post-operative pathway for an individual patient undergoing surgery, using simple checklists. There is some evidence that, after elective brain tumour surgery, most patients can be treated safely without admission to ICU20–22 and similar observations have been reported for elective cardiac23,24 and thoracic25 surgery. Cost-effectiveness of the different post-operative pathways needs to be tested in large samples after establishment of best post-operative pathways. Independent of the post-operative pathway and highly relevant is the acceleration of the return of normal function to avoid complications and, if there is any post-operative complication, it must be recognised early and treated correctly. Such a strategy could decrease ‘failure to rescue’ rates. Post-operative ‘failure to rescue’ has been identified as a major factor between hospitals with high and low post-operative mortality.26 Both studies published in this journal have major limitations, probably the most important of which is that results of both investigations cannot be generalised. Therefore, we need, as a first step, a large but simple and standardised dataset across Europe to investigate peri-operative care. A prospective European Surgical Outcomes Study (EuSOS) should start soon to include adults undergoing inpatient non-cardiac surgery in 150 European hospitals [http://www.euroanaesthesia.org/sitecore/Content/Research/Research_Network-CTN/CTN_Call_for_Centres.aspx]. This study, which is supported by the European Society of Anaesthesiology and the European Society of Intensive Care Medicine, should include about 10 000 patients. Results of this study could be generalised and should open avenues for optimal post-operative structures, pathways and skills. In recent years, excellent care levels during surgery by anaesthesiologists and during critical illness by ICU physicians have been reached with improved outcomes; in future years, new scientific evidence will increase the level of care in the peri-operative setting with, hopefully, improved outcomes. Acknowledgements The author received support from the Swiss National Foundation (SNF; K-23K1-122264/1), Swiss Accident Company and the Bangerter-Rhyner Foundation. The funding agencies had no role in the preparation, review or approval of the manuscript. The author has no conflict of interest.