Surgical Safety
Andrew W. Howard, MD, MSc, FRCSC
Hospital for Sick Children, Toronto, ON
Hospitals are dangerous places, and surgical care in particular is a dangerous undertaking. People die in hospitals, under anaesthesia, during surgical procedures, and during followup care. A certain amount of morbidity and mortality is accepted, a certain amount may be inevitable because, after all, patients are sick. On the other hand, no surgeon wants preventable or avoidable deaths occurring among patients under his or her care. Measuring morbidity and mortality on a surgical service, and accounting for patient comorbidities, is difficult. The false perception that surgical adverse events equal negligence encourages some to avoid the topic. It is important for all surgeons to move to a culture of safety and to look for ways to improve patient outcomes in their hospitals and practices.
Recent developments in surgical safety are proven effective at improving the process of surgical care and preventing those deaths and complications which are avoidable. Surprisingly simple tools, such as checklists, can reduce surgical morbidity and mortality. The means by which they do so include preventing simple errors in process, but perhaps more importantly they encourage and improve team communication.
Discharging a patient safely from hospital requires not only avoiding complications, but also the prompt recognition and treatment of postoperative complications. Medical emergency teams are an example of an institutional structure designed to identify and rescue patients at risk of adverse outcomes.
A systematic review of the patient safety literature in high income countries concluded that 1 in 150 patients admitted to hospital dies as a result of an adverse event, and two thirds of such adverse occur in surgical patients (1).
A common method of estimating the incidence of adverse events (errors, complications, and deaths) is to perform a detailed review of thousands of charts randomly selected to be representative of a particular setting. Several large North American studies have looked in detail at adverse events occurring during hospitalization.
Leape’s influential study looked at 30,000 admissions in the United States and identified 3.7% of patients with disabling injuries caused by medical or surgical treatment. The most common events were drug complications, followed by wound infections, then technical complications of surgery. Forty - eight percent of the adverse outcomes were related to an operation. One of the important contributions of the Leape paper was to separate the concept of and adverse event from the concept of negligence. Some adverse events may be caused by negligence, but most are not. Falsely associating all adverse outcomes with negligence creates a culture in which it is difficult to measure, evaluate, and prevent adverse outcomes among our patients. Leape found that only 17 percent of surgical adverse outcomes were associated with negligence. The fact that the majority of surgical adverse events were NOT due to negligence makes studying them both less threatening and more compelling for surgeons.
By sampling 18000 charts in small and large hospitals from five Canadian provinces, Baker estimated an adverse event rate of 7.5 per 100 admissions in Canadian hospitals, with 37% of adverse events preventable and 20% of adverse events leading to death (2).
Gawande and colleagues looked specifically for details of surgical adverse events among a representative random sample of 15,000 admissions to hospitals in Colorado and Utah during 1992 (3). They estimated an adverse event rate of 3 percent of admissions, of which two thirds were related to surgical care. They estimated that 54% of surgical adverse events were preventable, and 5.6% of surgical adverse events resulted in death. Specific operations with the highest risk of preventable adverse events included lower extremity bypass, abdominal aortic aneurysm repair, colon resection, CABG, TURP, cholecystectomy, hysterectomy, and appendectomy. Nearly half of the surgical adverse events were from technique related complications, wound infections, or postoperative bleeding. From the surgical epidemiology, it is clear that the prevention of some of these events lies in the surgeons hands.
It is difficult to estimate a rate of surgical adverse events in Africa - the continent is too big and too diverse, and data are too sparse. Overall data are restricted to mortality rather than morbidity plus mortality. A few published estimates provide an inadequate, but still concerning picture. An early study from Zambia reported a mortality rate of 3.3 deaths per 1000 operations (4). Hansen reviewed surgical experience in Malawi and reported an avoidable surgical death once per 3000 operations, but an avoidable anaesthetic death once per 500 operations (5). Most anaesthetic deaths were related to airway issues or fluid rescusitation. Walker reported 7.7 intraoperative anaesthesia related deaths per 10,000 paediatric operations in Southwestern Uganda (6). In the majority of hospital settings in Uganda a deficiency was found (personnel, drugs, equipment or training) which prevented the delivery of safe anaesthesia (7). A systematic review estimates that in the UK an anaesthetic related mortality rate of less than 1 per 185,000 is attainable for healthy patients, and 5 to 10 times higher than this for patients with comorbidities (8) . It is very difficult to compare mortality data across countries because of differences in data collection methods, patient populations, comorbidities, definitions used, and time span examined. Jenkins summarizes all the international comparator studies which have attempted to quantify the risk of death from anaesthesia (8).
Measuring surgical morbidity and mortality remains fraught. We each know what we mean by a complication, by morbidity, and by mortality and yet we may legitimately disagree about how these terms should be applied in a particular clinical setting. Without standardized definitions it is difficult to compare complication rates across institutions and across time. A large review performed for the UK NHS in 2001 concluded that work is needed both on standardizing the definitions even of such common surgical complications as infections, deep venous thromboses, and anastomotic leaks (9). Furthermore, risk adjustment for case mix and patient comorbidity needs substantial development before complication rates can be compared across institutions (9).
The Patient Safety in Surgery Study (US) - A prospective cohort study
Even if we agree on the definitions and methodology, the process of measuring complications is by itself important. Among the more rigorous and better studied methodologies for measuring surgical complications is the methodology of the Patient Safety in Surgery study (10). This well funded study began at Veteran’s hospitals in the United States in response to concern from Congress that surgical morbidity may be too high in these institutions. The study design involves a prospective cohort of surgical patients, all of whom have baseline measures to allow risk adjustment. The first 40 consecutive patients undergoing major surgery during each 8 day cycle form the study group at each hospital. A trained nurse evaluates all complications and deaths intraoperatively or within 30 days of the index surgery. A structured data collection process ensures both completeness and standardization of the reporting of complications. Participating centres now include 128 veteran’s hospitals and 14 academic medical centres. The Patient safety in surgery study methodology has been endorsed by the American College of Surgeons and is now widely applied, under the name of the National Surgical Quality Improvement Program, or NSQIP.
Hutter compared the traditional surgical morbidity and mortality rounds (of the General Surgery Service at the Massachusetts General Hospital) with the national surgical quality improvement program methodology. Traditional morbidity and mortality rounds identified a complication rate of 6.4% and a mortality rate of 0.9%. NSQIP methodology identified a complication rate of 28.9% and a mortality rate of 1.9% at the same institution during the same time period (11). In other words, the traditional morbidity and mortality review at one of the leading hospitals in the world was missing about half the deaths and three quarters of the complications. Massachusetts General replaced it traditional morbidity and mortality process with a structured prospective review, so that they could better measure and prevent those complications which are preventable.
The process of audit might by itself reduce surgical morbidity and mortality. Khuri reported that implementation of the Patient Safety in Surgery monitoring system in private hospitals was associated with an 8.7% reduction in complications, a 9.1% reduction in infections, and a 23% reduction in renal complications . In Western Australia, surgeons take part in a voluntary system which aims to review 100% of postoperative deaths. When this review process was implemented, the state observed improvement in deep venous thrombosis prophylaxis over time, as well as an improvement in delays in care, and in the staffing level and supervision of trainees particularly for emergency surgery (12).
One of the more surprising recent results in surgery was the dramatic reduction in postoperative complications and mortality following the implementation of the WHO surgical safety checklist, as published in the New England Journal of Medicine in 2009 (13). It was surprising how well the tool worked, it was (pleasantly) surprising that the primary research included African hospitals, and it was surprising that the checklist had the same relative effect whether implemented in Toronto, Canada or Ifakara, Tanzania or one of eight other cities worldwide. The checklist is a simple tool which can be downloaded freely from the WHO. Adaptation of the checklist to suit local conditions is encouraged. http://whqlibdoc.who.int/publications/2009/9789241598590_eng_Checklist.pdf
The WHO surgical safety checklist is applied in the operating room only. It includes three different time points - sign in, time out, and sign out, corresponding to the patient entering the room, the moment before incision, and the patient about to leave the room. At each point, the surgical, anaesthetic, and nursing team verbally go through the items on the checklist to ensure that everything is in place for a safe operation. The only physical resources addressed by the checklist are a pulse oximeter (for all anaesthetics), and prophylactic antibiotics (where appropriate). As can be seen, the majority of the checklist involves simple reminders. Key points are that the checklist is done aloud, so all members of the team go through the habitual reminders, and perhaps more importantly all team members are required to communicate.
The results of implementation of the checklist were investigated by comparing 3733 patients at baseline with 3955 patients operated on after implementation of the checklist. Overall surgical complications fell from 11% to 7%, and mortality fell from 1.5% to 0.7%/ Implementation of the checklist in an institution took from one week to one month and was not costly. Complications were prospectively identified using the structured methodology of the NSQIP previously mentioned (14). Only major complications were considered (eg. renal failure, transfusion more than 4 units, cardiac arrest, reintubation, pneumonia, surgical site infection, sepsis). There were some methodological issues with the study. There was no control group (hospitals without a checklist), and the outcomes could not be evaluated in a blinded fashion. Not all of the hospitals used the checklist for all cases, and not all hospitals used every item on the checklist. Despite these shortcomings, this study strongly suggests that a simple checklist can make surgery safer.
A second major study regarding the implementation of surgical safety checklists was published in the New England Journal of Medicine in 2010 (15). The DeVries study was performed in Holland at 2 large academic medical centres and 4 teaching hospitals. The patient safety system implemented was checklist based, but not restricted to the operating room setting. Because half of surgical adverse events occur outside the OR, the study included checklists for preop, in the operating theatre, recovery room or ICU, and postoperative ward care until discharge. The checklist was much more comprehensive than the WHO one and included imaging, equipment, handoff, and prescription components. Strengths of the DeVries study included the use of a control group (a group of hospitals which did not implement the checklist), as well as an evaluation of ‘dose-response’.
Implementation of the comprehensive checklist in Holland decreased complications from 27.3 to 16.7 per 100 patients, and decreased mortality from 1.5 to 0.8 per 100 patients. No decreases were seen in control hospitals. The complication rate varied depending on how many items on the checklist had been filled. For patients with better than median checklist completion the complication rate was 7.1 per 100 operations, for patients with worse than median checklist completion the complication rate was 18.8 per 100 operations. Because the checklist was more complicated and needed to be used throughout the hospital, the implementation involved some process changes and was an expensive and difficult process. An accompanying editorial suggested that the methodology of the Dutch study was sufficient that surgical safety checklists “...have crossed the threshold from a good idea to standard of care” (16).
Weiser et al reported on the use of checklists in emergency surgery. They used a subset of the patients in the Haynes study, eliminating elective surgery and looking only at urgent cases. They found that checklists were equally effective in the emergent setting - reducing complications from 18.4% to 11.7% and reducing death from 3.7% to 1.4% (17). Stahl showed that ‘lost’ items during handover in trauma and surgical intensive care was reduced from 20% to 3.6% by using a checklist for postoperative care and a structured handover (18). DeVries performed a comprehensive chart review of settled malpractice claims in Holland and reported that 30% of the claims were about items which were on the Dutch surgical patient safety system checklist (19). Semel used economic modeling to estimate that the adoption of surgical checklists would be significantly cost saving for hospitals (20), and Verdaasdonk showed that an equipment checklist for endoscopy reduced technical equipment problems (positioning, connections, defects or malfunctions) from 49 in 30 cases to 23 in 30 cases (21). One checklist study which did not show improvement in patient outcomes was that of Sewell, in which the WHO safe surgery checklist was implemented in an orthopaedic room. Complications did not change (8.5% pre and 7.6% post) and nor did mortality (1.9% pre and 1.6% post) but the staff felt that team communication had improved. In Sewell’s study 95% of the surgery was scheduled work with teams who worked together all the time, it was suggested that this might be a reason and that the orthopaedic setting might be a reason for seeing less result from the checklist (22). Finally, Conley had performed qualitative work establishing that a persuasive, engaged, and adaptive leader is necessary in order to successfully implement a surgical safety checklist (23).
One is left wondering why checklists work so well. The checklist seems like such a simple tool - it is likely that many of the effects are indirect. For example, DeVries et al showed that anastomotic leaks were reduced during their checklist implementation (16) - although such overtly technical outcomes should have much more to do with the surgeon than with the team or the anaesthetic. A surgeon may operate in an environment with fewer stresses and distractions, the team communication may be better, and it may be the indirect effects of checklist use which are important in many of the benefits.
A team going through a checklist verbally is prompted to communicate in a structured manner. It is possible that encouraging team communication is an important means by which checklists improve outcomes.
Nagpal performed a systematic review regarding information transfer and communication in the surgical setting. A comprehensive search identified 3131 citations of which 38 studies met inclusion criteria, 29 were quantitative and 9 were qualitative studies. Information transfer failures were found to be very common in surgical care, and were distributed across the continuum of care from the outpatient setting through admission, preoperative, operative, recovery room or intensive care, and postoperative ward care and persisting into transfer back to primary care. Information transfer was found to lead to errors in care, and these errors in care were found to lead to harm. Checklists were found to improve information transfer (24).
Davenport evaluated 44 veterans hospitals and 8 academic hospitals to look for predictors of surgical morbidity. Higher measures of communication and collaboration were found to be predictive of lower surgical morbidity (25). Perceptions of safety climate and working conditions were not predictive.
An idea brought in from the aviation industry (like checklists themselves!) is the idea of crew resource management. This idea sees the entire team as a resource - any team member may possess a piece of information which could prevent an adverse event. Communication must occur effectively across a hierarchy, if the critical information known to only one member is to be used effectively by the team. It is possible that a checklist, or a time out, can flatten the hierarchy and encourage more effective team communication (26).
African surgical settings are unique in team composition. Anaesthetic care, and surgical care, are provided sometimes by non specialist physicians, and sometimes by non-physician providers in many African countries (27). Ensuring a safe surgical system in Africa includes the challenge of meeting human resource requirements, including by appropriate use of non physician providers, while at the same time ensuring that the highest possible standard of care is met. It is likely that a specific and African focus on surgical teamwork will need to be developed to optimize communication, supervision, and safety where surgical and medical task shifting is necessary.
Complications happen in surgery. Although it is important to aim to reduce the rate of complications, it is also important to identify complications, and to respond to them promptly and appropriately.
Ghaferi studied over 84000 US patients who underwent general or vascular surgery in 2005 to 2007 in hospitals participating in the National Surgical Quality Improvement Program. They ranked hospitals according to mortality quintile, and compared the complication rates in high mortality versus low mortality hospitals. The death rates varied from 3.5% in the low mortality hospitals, to 6.9% in high mortality hospitals. The complication rates, however, did not differ very much. Major complications occurred in 16.2% of the surgical admissions in low mortality hospitals, and 18.2% of the surgical admissions in high mortality hospitals. What did differ substantially was the death rate per complication. In low mortality hospitals, only 12.5 % of patients with a major complication of surgery died - whereas in high mortality hospitals 21.4% of patients with major complications died (28). In general and vascular surgery, it seems that mortality is driven by failure to rescue after a complication, rather than simply by the rate of complications itself. This has implications both for how surgeons and trainees practice, and for how hospital care is structured.
One model for improving survival in hospitals is the concept of a medical emergency team who can intervene before a patient is critically ill. Bellomo reported on a before/after controlled study of a medical emergency team in a teaching hospital in Melbourne, Australia (29). The medical team was ICU based but attended any patient in the hospital in difficulty or at risk. The team travelled from the ICU to evaluate a patient any time a member of the nursing or medical staff reported a patient they were concerned about, or a patient whose heart rate, blood pressure, respiratory rate, oxygen saturation, level of consciousness, or urine output fell outside objective acceptable criteria. The team consisted of the duty intensive care unit fellow and an ICU nurse, with backup from the ICU attending. Major in hospital adverse outcomes (eg. stroke, MI, renal failure, sepsis) were evaluated before and after implementation of the medical emergency team. Before implementation of the team there were 301 adverse outcomes per 1000 admissions, after implementation of the team there were 127 adverse outcomes per 1000 admissions. Deaths dropped by 36%. Substantial decreases were seen in the incidence of acute renal failure (88%), respiratory failure (79%), stroke (78%) and sepsis (74%), while intensive care unit admissions dropped by 44%. Medical emergency teams are now commonplace in many hospitals, and may provide an important mechanism for the rescue of some patients suffering from operative complications.
Wrong site surgery is rare - but seems to be among the most preventable of complications. In a systematic review the frequency of wrong site surgery was estimated as varying between 0.09 per 10,000 to 4.5 per 10,000 cases in different practice settings. The causes of wrong site surgery were almost universally failures in communication. The WHO checklist entry on marking the site of surgery should address communication issues around wrong site surgery. The systematic review cautioned that the prevention of wrong site surgery is not yet proven(30). An eight year experience with a neurosurgical checklist at the Mayo clinic has recorded a zero incidence of wrong site surgery in over 6000 operations (31).
(1) de Vries EN, Ramrattan MA, Smorenburg SM, Gouma DJ, Boermeester MA. The incidence and nature of in-hospital adverse events: a systematic review. Qual Saf Health Care 2008 Jun;17(3):216-223.
(2) Baker GR, Norton PG, Flintoft V, Blais R, Brown A, Cox J, et al. The Canadian Adverse Events Study: the incidence of adverse events among hospital patients in Canada. CMAJ 2004 May 25;170(11):1678-1686.
(3) Gawande AA, Thomas EJ, Zinner MJ, Brennan TA. The incidence and nature of surgical adverse events in Colorado and Utah in 1992. Surgery 1999 Jul;126(1):66-75.
(4) Heywood AJ, Wilson IH, Sinclair JR. Perioperative mortality in Zambia. Ann R Coll Surg Engl 1989 Nov;71(6):354-358.
(5) Hansen D, Gausi SC, Merikebu M. Anaesthesia in Malawi: complications and deaths. Trop Doct 2000 Jul;30(3):146-149.
(6) Walker IA, Obua AD, Mouton F, Ttendo S, Wilson IH. Paediatric surgery and anaesthesia in south-western Uganda: a cross-sectional survey. Bull World Health Organ 2010 Dec 1;88(12):897-906.
(7) Hodges SC, Mijumbi C, Okello M, McCormick BA, Walker IA, Wilson IH. Anaesthesia services in developing countries: defining the problems. Anaesthesia 2007 Jan;62(1):4-11.
(8) Jenkins K, Baker AB. Consent and anaesthetic risk. Anaesthesia 2003 Oct;58(10):962-984.
(9) Bruce J, Russell EM, Mollison J, Krukowski ZH. The measurement and monitoring of surgical adverse events. Health Technol Assess 2001;5(22):1-194.
(10) Khuri SF, Henderson WG, Daley J, Jonasson O, Jones RS, Campbell DA,Jr, et al. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007 Jun;204(6):1089-1102.
(11) Hutter MM, Rowell KS, Devaney LA, Sokal SM, Warshaw AL, Abbott WM, et al. Identification of surgical complications and deaths: an assessment of the traditional surgical morbidity and mortality conference compared with the American College of Surgeons-National Surgical Quality Improvement Program. J Am Coll Surg 2006 Nov;203(5):618-624.
(12) Semmens JB, Aitken RJ, Sanfilippo FM, Mukhtar SA, Haynes NS, Mountain JA. The Western Australian Audit of Surgical Mortality: advancing surgical accountability. Med J Aust 2005 Nov 21;183(10):504-508.
(13) Haynes AB, Weiser TG, Berry WR, Lipsitz SR, Breizat AH, Dellinger EP, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med 2009 Jan 29;360(5):491-499.
(14) Khuri SF, Henderson WG, Daley J, Jonasson O, Jones RS, Campbell DA,Jr, et al. Successful implementation of the Department of Veterans Affairs' National Surgical Quality Improvement Program in the private sector: the Patient Safety in Surgery study. Ann Surg 2008 Aug;248(2):329-336.
(15) de Vries EN, Prins HA, Crolla RM, den Outer AJ, van Andel G, van Helden SH, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med 2010 Nov 11;363(20):1928-1937.
(16) Birkmeyer JD. Strategies for improving surgical quality--checklists and beyond. N Engl J Med 2010 Nov 11;363(20):1963-1965.
(17) Weiser TG, Haynes AB, Dziekan G, Berry WR, Lipsitz SR, Gawande AA, et al. Effect of a 19-item surgical safety checklist during urgent operations in a global patient population. Ann Surg 2010 May;251(5):976-980.
(18) Stahl K, Palileo A, Schulman CI, Wilson K, Augenstein J, Kiffin C, et al. Enhancing patient safety in the trauma/surgical intensive care unit. J Trauma 2009 Sep;67(3):430-3; discussion 433-5.
(19) de Vries EN, Eikens-Jansen MP, Hamersma AM, Smorenburg SM, Gouma DJ, Boermeester MA. Prevention of surgical malpractice claims by use of a surgical safety checklist. Ann Surg 2011 Mar;253(3):624-628.
(20) Semel ME, Resch S, Haynes AB, Funk LM, Bader A, Berry WR, et al. Adopting a surgical safety checklist could save money and improve the quality of care in U.S. hospitals. Health Aff (Millwood) 2010 Sep;29(9):1593-1599.
(21) Verdaasdonk EG, Stassen LP, Hoffmann WF, van der Elst M, Dankelman J. Can a structured checklist prevent problems with laparoscopic equipment? Surg Endosc 2008 Oct;22(10):2238-2243.
(22) Sewell M, Adebibe M, Jayakumar P, Jowett C, Kong K, Vemulapalli K, et al. Use of the WHO surgical safety checklist in trauma and orthopaedic patients. Int Orthop 2011 Jun;35(6):897-901.
(23) Conley DM, Singer SJ, Edmondson L, Berry WR, Gawande AA. Effective surgical safety checklist implementation. J Am Coll Surg 2011 May;212(5):873-879.
(24) Nagpal K, Vats A, Lamb B, Ashrafian H, Sevdalis N, Vincent C, et al. Information transfer and communication in surgery: a systematic review. Ann Surg 2010 Aug;252(2):225-239.
(25) Davenport DL, Henderson WG, Mosca CL, Khuri SF, Mentzer RM,Jr. Risk-adjusted morbidity in teaching hospitals correlates with reported levels of communication and collaboration on surgical teams but not with scale measures of teamwork climate, safety climate, or working conditions. J Am Coll Surg 2007 Dec;205(6):778-784.
(26) Erickson DR, Boulanger B. Patient safety: above and beyond the checklist. J Urol 2011 Apr;185(4):1177-1178.
(27) Chu K, Rosseel P, Gielis P, Ford N. Surgical task shifting in Sub-Saharan Africa. PLoS Med 2009 May 19;6(5):e1000078.
(28) Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J Med 2009 Oct 1;361(14):1368-1375.
(29) Bellomo R, Goldsmith D, Uchino S, Buckmaster J, Hart G, Opdam H, et al. Prospective controlled trial of effect of medical emergency team on postoperative morbidity and mortality rates. Crit Care Med 2004 Apr;32(4):916-921.
(30) Devine J, Chutkan N, Norvell DC, Dettori JR. Avoiding wrong site surgery: a systematic review. Spine (Phila Pa 1976) 2010 Apr 20;35(9 Suppl):S28-36.
(31) Lyons MK. Eight-year experience with a neurosurgical checklist. Am J Med Qual 2010 Jul-Aug;25(4):285-288.
