Emergency surgery in the elderly patient: a quality improvement approach

Two recent documents have successfully drawn press and political attention to the problems of the elderly patient in hospital, particularly those undergoing emergency surgery 1, 2. These problems include the complex management of surgery in patients with comorbidity, disability and frailty, and the need for adequate numbers of critical care beds and physicians for the care of the elderly. Many of us practising anaesthesia have been aware of these issues for a long time, and have wondered why this important problem has, until recently, drawn relatively little attention from academics, researchers and policy makers. Much emergency surgery is in the elderly, and the commonest operations in the very elderly are hemiarthroplasty/sliding hip screw, laparotomy and amputation 3. As the population ages and the fittest patients are managed in day surgery and treatment centres, the focus for much of our major general anaesthetic practice is on the elderly patient undergoing surgery for fragility fracture or an acute abdominal condition. This group of patients can only increase; the elderly are the fastest growing group of the population, and are more likely to be admitted for emergency surgery than elective 4; unsurprisingly, emergency surgery is a significant risk factor 5, 6 and the outcomes from comparable procedures performed as an emergency are much poorer in the elderly compared with elective surgery 7. Indeed, for many elderly patients emergency surgery is a cataclysmic life event, with large numbers unable to return home or to their previous level of independence 1, 7. We owe it to these patients to do the best we can, not only because a civilised society should look after its elderly 8, but because it makes sense. The costs of less than optimal care are high 9, most importantly for the patient, but also in terms of managing complications and the increased dependency associated with poor outcomes 1, 5, 7–9. In the past 10 years, there has been a massive injection of resources into non-urgent surgery 10, resulting in improved access to hospital for elective patients. However, the problem for the elderly emergency surgical patient is not getting into hospital, but getting out again. The absolute number of patients undergoing emergency surgery is high; emergency general (bowel or vascular) surgery comprises the largest number, by a wide margin, of all surgical admissions in UK National Health Service (NHS) hospitals 10. Elderly patients undergoing emergency general and vascular surgery have the highest mortality of any common surgical procedure 11–14. Therefore, if we wish to affect surgical mortality, the patient and procedure groups to focus on are elderly patients undergoing emergency general surgery. Can care be improved for this group of patients, or do their comorbidities and the nature of the surgery mean that poor outcomes are inevitable? The most recent National Confidential Enquiry into Patient Outcome and Death (NCEPOD) report found that only 36% of patients received care that the advisors classified as good 1, and other large audits support the notion that there is room for improvement in the management of the emergency surgical patient 4, 11. It is highly unlikely that there will be a ‘magic bullet’ to improve care for these complex patients in the form of a new drug or surgical technique. Research in these cases is challenging; ‘emergency laparotomy’ is a term covering a multiplicity of potential surgical conditions-indeed, to do a database study to examine outcomes of emergency laparotomy requires searching on over 400 International Classification of Diseases-10 (ICD10) codes (personal experience). In addition, the patients are elderly, potentially septic, may have confusion or dementia and are often admitted out of routine working hours. To undertake a randomised controlled trial for this group of patients is therefore likely to be very difficult. These problems have been discussed recently in relation to critical care research 15. An alternative method is to use a quality improvement approach, with data displayed as ‘run charts’ using statistical process control techniques and small samples paired with repeated small tests of change using ‘plan-do-study-act’ (PDSA) methods 16, 17. Of course, quality improvement studies should be as rigorously evaluated as randomised controlled trials before implementation 18. What is certain, as Lord Darzi stated in the NHS plan, is that “We cannot improve what we cannot measure” 19. Although individual centre observational studies of outcome have their uses 12, e.g. to illustrate prolonged length of stay, they are open to criticism as they are influenced by local factors 20. We need to understand better the size and nature of the problem nationally and how we compare internationally, to have high quality risk-adjusted outcome data to support decision making, and to counsel patients and their families. This approach has been used with success to drive improvement in cardiac surgery 21. The need for more outcome data is the basis on which the emergency laparotomy network has been founded 22 and for the forthcoming European Surgical Outcomes Study (EuSoS). Once we are better able to understand the expected morbidity and mortality outcomes for specific patient groups, we must aim to reduce variation in outcome. We do know that the anaesthetic management of these emergency abdominal cases is highly variable 23. We need to have data to be able to identify high-performing centres and learn from what they do well, and to identify poorly performing centres with the aim of helping them to improve. Continuous measurement and feedback for improvement has been used by the National Surgical Quality Improvement Program (NSQIP) in the US, with impressive success 24; following implementation in the Veterans Administration hospitals, there was a 46% relative improvement in risk-adjusted surgical death rate over an 11-year period. The delivery of quality care can be broken down into three areas: structure; process; and outcomes 25. Quality improvement programmes frequently use driver diagrams to define the outcome aims and to identify the primary drivers that will lead to that outcome 26; behind the primary drivers are secondary drivers which are the areas, often focused on process, that can be worked on to improve the desired outcome. A driver diagram for improving outcomes in the elderly undergoing emergency surgery is shown in Fig. 1. Of course, quality care cannot be delivered if the structures to support it are not in place. This includes adequate numbers and availability of operating theatres and critical care beds staffed with senior personnel, competent to deliver the complex care needed by these patients. Drivers for improving outcomes for elderly emergency surgical patients. The outcome is on the left of the diagram, then shown are the primary drivers followed by the secondary drivers. This is not an exhaustive list but serves to provide a structure to consider areas to target for quality improvement. *SCIP, Surgical Care Improvement Project 27. If we take as an example one primary driver from Fig. 1, postoperative care, we can easily see a target area for improvement. Using the NSQIP database, in an important study Khuri et al. 28 found that the occurrence of a complication within 30 days postoperatively was the most important determinant of decreased postoperative survival after surgery, more so than pre-operative risk and intra-operative factors. This study strongly suggested that quality and improvement processes in surgery should be directed towards the prevention of postoperative complications. Postoperative complications are associated with a particularly high mortality in patients aged 80 years and over 6, 7, 12, 13. We have UK reports to acknowledge that postoperative care is an area to target 29, 30, especially as many high risk patients are managed outwith a high dependency or critical care area 14, 30, 31. In a more recent study 32, Ghaferi et al. found that the difference between hospitals with high and low surgical mortality was not in the incidence or degree of postoperative complications sustained, but the way in which the patients were managed or ‘rescued’ when these complications occurred. ‘Segmentation’, a term used in quality improvement, is a useful way of thinking about managing the postoperative patient 33. If we divide postoperative care into two segments, ward care and critical care, we can work firstly on ensuring that patients go into the correct area of care. This can be done by using some form of scoring process that takes into account both patient and surgical factors 34, 35. We can then look at process improvement in each segment; this could mean concentrating on better earlier detection and response to complications, particularly targeted at higher-risk patients on the ward, for example with surgical outreach teams 36. In the UK, there may of course be a third segment: care of patients who would benefit from a critical care bed if one was available. This patient group may particularly benefit from formal risk scoring and heightened postoperative monitoring. One area of concern identified by the most recent NCEPOD study was the paucity of medical input from physicians care of the elderly 1. Only 42.5% of patients who died following surgery for fractured neck of femur received input from a care of the elderly consultant, falling to a staggering 6.7% for patients over 80 years of age undergoing acute abdominal surgery. Identification of high-risk elderly patients and early referral to a care of the elderly team as part of a structured programme has shown benefit for elective surgery 37; these benefits may be transferrable to emergency patients 38. It would also be useful to define quality parameters against which optimal delivery can be measured; these should include areas pertinent to the elderly emergency patient but often neglected, such as delirium scoring 39. Evidence based care should be the basis of any high quality service, and the reliable delivery of that care should be paramount for these very high-risk patients. A recent consensus statement on patient safety by the Association of Surgeons 40 defined key areas of evidence based practice that should be applied to all surgical patients. These include venous thromboembolism assessment and prophylaxis, management of sepsis using care bundles, good peri-operative fluid management, dynamic monitoring of cardiac output, appropriate and timely delivery of antibiotics, and good communication and handover. It is unlikely that any of us would argue with these measures. However, we may know what to do but we just do not do it reliably enough. Most healthcare processes run at 60–80% reliability, if prompts and checks are not built into the system 41. This is the case with these important therapies. The 2009 NCEPOD report 42, which reviewed the care of all patients (both medical and surgical) who died within 4 days of hospital admission, found neglect of venous thromboembolism and antibiotic prophylaxis, poor fluid and electrolyte management, and poor communication between and within clinical teams, in 13.5% of cases. The use of bundles and checklists are specifically designed to improve the reliable delivery of evidenced based care and there is now a growing body of evidence to demonstrate their efficacy in healthcare 43, 44. Good communication and teamwork are particularly important in the emergency setting. Teams are less likely to work together regularly, and crisis situations are more likely to develop, in the emergency theatre. Teams should be trained in standardised communication methods and the use of briefings and the communication aspects of the checklist become more important. There is evidence supporting the importance of communication and teamwork on patient outcome 45. Quality improvement in healthcare is not ‘rocket science’. A key part of it is doing the right things in a reliable and timely fashion. De Vries et al. 46 comment that as more than half of surgical errors occur outside the operating theatre, it is likely that a more substantial improvement in surgical outcomes can be achieved by targeting the entire surgical pathway. Their study showed that introduction of a checklist system throughout the entire surgical pathway reduced mortality and complications significantly in an already high performing healthcare system. The checklist provided a comprehensive framework for the surgical pathway, minimised information loss during transfers and promoted interdisciplinary communication. As is known to happen with checklists and bundles, improving delivery in one area of the pathway triggers wider improvement and that was noted in this study. A good example of doing the basics well and producing excellent results for the elderly emergency patient is illustrated by the results of the Peterborough fractured neck of femur pathway. Data from Dr Foster (http://www.drfosterintelligence.co.uk/; accessed July 2010) show that Peterborough achieves an exceedingly low relative risk of death for observed to expected deaths, compared with other English and Welsh hospitals. When the unit’s process of care is examined, it appears to do nothing very different to most of the rest of us, just clearly doing it better. Another example of concentrating on process with constant feedback on outcomes is that of Intermountain Healthcare in Salt Lake City, recognised as one of the world’s best healthcare providers 47. This organisation has pioneered applied outcomes research; it has dynamic guidelines constantly revised by doctors directly involved in patient care, and with regular feedback of outcomes to teams who then design their own next improvement targets. Finally, any attempt to improve quality should be focused around the patient 48 and the whole pathway of care. There is evidence in elective surgery that patient centred design improves outcome 49 and the success of enhanced recovery programmes owes much to the protocolisation of the whole pathway. Although care is more complex for the emergency patient, there are aspects of enhanced recovery that would transfer well to emergency surgery: the design of an evidenced based pathway of care; the use of goal-directed therapy; multidisciplinary involvement in pathway design; and frequent audit of results. However, without good long-term risk-adjusted data and patient reported outcomes, we cannot fully understand the impact and implications of major emergency surgery on elderly individuals and their families. Many of these patients will not survive surgery; indeed, NCEPOD 3 showed that for the over 90s, 25% of patients suffered complications early and/or died on or before the second postoperative day. There may be some cases where quality of care is about allowing the patient to die with dignity. The elderly patient undergoing emergency surgery, particularly intra-abdominal, has a high mortality. Work is required on all areas of the patient pathway and process to achieve improvement including nutritional assessment, medication review, management of dementia and delirium and rehabilitation facilitated by good pain management 1, 2, 8. The numbers of patients are large, so small changes in morbidity and mortality could have a big impact, both on outcome and on the economic consequences of prolonged length of stay, increased dependency and complications. As anaesthetists, our role is to be part of a multidisciplinary team considering all ways to improve the quality of peri-operative care for these highest-risk patients. This Editorial is based on work undertaken as a Quality Improvement Fellow at the Institute of Healthcare Improvement, Boston, USA, funded by the Health Foundation, UK. No competing interests declared.