Patient Safety in Surgical Residency: Root Cause Analysis and the Surgical Morbidity and Mortality Conference—Case Series from Clinical Practice
Winter 2012 - Volume 16 Number 1
Although the surgical morbidity and mortality conference (SMMC) has been a core educational venue for surgical education and quality assurance (QA), its current format focuses mainly on human errors rather than system failures, which are responsible for the vast majority of medical errors. To avoid having surgeons seemingly put on trial, root cause analysis (RCA) can be used as an effective way of analyzing system failures and of finding possible solutions for them. Preliminary data confirm the value of RCA in that respect and promise a great potential for improving patient safety away from the culture of blame. Bringing the findings of RCA to the SMMC has the advantage of having both perspectives—human errors and systems failures—thus enhancing surgical education, improving QA, and hopefully improving patient safety. However, although this seems to be a novel approach, several factors should be considered before its implementation, such as the quality of analysis, cost-effectiveness, and actual impact on patient safety. We believe that to maximize learning, sentinel events that currently require RCA should not be discussed in SMMCs until the findings of RCA are available for review. The use of some of the tools of RCA should be considered when discussing nonsentinel events during SMMCs.
Historically, the surgical morbidity and mortality conference (SMMC) has been a core educational venue, a quality-assurance (QA) tool, and a way to socialize a surgical trainee into the culture of surgery.1 Traditionally, the hospital course for a given patient is presented with the rationale for the provided care. The decisions are then critiqued by the most experienced and senior surgeon in attendance. Any perceived errors in the patient's care are highlighted. The primary premise is that we learn from our errors, and through this education, the quality of subsequent patient care is improved.2 However, along with the candid discussion of error, the surgeon's personal culpability in any error or death of a patient has been the cornerstone of the SMMC. This proclamation of accountability is the means by which a surgical apprentice is socialized into a surgeon.3 However, the hierarchy that has been created, in which a surgeon accepts full responsibility for everyone's actions, particularly in cases of gross error, has led to a culture of blame.3 Thus the SMMC is often focused on human errors and only occasionally addresses system failures that are responsible for the vast majority of medical errors. Recent evolution in health care delivery mandates a de-emphasis of personal responsibility for error. Root cause analysis (RCA) is one of the most effective ways to analyze system failures and find possible solutions to them. Preliminary data showed that RCA not only shifted analyses of adverse events toward system vulnerabilities4 but also improved patient safety.5 Using RCA results in SMMC may bring the best of the two worlds together to enhance surgical education, QA, and teaching of patient-safety skills in surgical residency.
To put matters into perspective, we consider here a case series of three sentinel events that were discussed in SMMCs and were also subjected to RCA, analyze the findings from each approach, identify what was learned, and pinpoint how it was translated into potential patient-safety practices.
During a mastectomy for breast cancer, the surgeon discovered a piece of a guidewire under the superior skin flap from a long-term central venous access catheter that was placed by a different surgeon. The retained foreign body was removed without any further complications.
A few hours after completing a laparotomy, a surgeon suspected that a laparotomy sponge might have been left behind in the patient's abdomen. The patient was taken back to the operating room for removal of the retained foreign body.
A patient underwent spinal decompression. The surgery was completed as planned on the proper side but at the wrong level, in spite of the use of intraoperative imaging modalities.
Not much was learned from the SMMC; all three surgeons accepted responsibility for their errors and indicated that they should have been more vigilant during surgery. Are surgeons the only responsible professionals for these errors, though? Do we know if other members of the surgical team could have been responsible too? To find out, we subjected all three cases to RCA as required for all sentinel events.
In the first case, RCA showed two factors to be the proximate cause for the error: 1) a process factor when a post-insertion chest x-ray was not obtained and 2) a controllable environmental factor when the surgeon failed to realize that the guidewire was much shorter than expected. At the time of the insersion of the central venous catheter, there was no policy requiring the surgeon to inspect the guidewire when it was pulled out. In view of these findings, three action plans were adopted: 1) the operating-room committee enforced a policy mandating that all such procedures be performed under fluoroscopy and be followed by the acquisition of a chest x-ray; 2) all staff involved in these procedures completed an in-service education program; and 3) random audits by the QA Department were set in motion to ensure compliance with procedures.
In the second case, RCA showed two factors to be the proximate cause for the error: 1) a process factor when the circulating registered nurse and the scrub technician did not view the laparotomy sponges as they counted them and 2) a human factor when a drift in policy was identified regarding where to place the sponges as they are being counted. An additional system issue or human resources issue was identified: The circulating nurse and the scrub technician did not perform their duties as expected. The surgeon in this case was not at fault. In view of these findings, new strategies were adopted: 1) a Devon Bag-It sponge counter bag (Covidien, Mansfield, MA) was to be used in all operating rooms. This bag looks like a shoe holder that hangs on a pole. Each slot holds one sponge that can be viewed by the scrub technician and by the circulating nurse as they are counting aloud and concurrently; 2) leadership accountability was also put in place to prevent drift from and workarounds of existing policies; 3) adequate in-service education programs were provided; and 4) random audits by the QA Department were set in motion to ensure compliance.
In the last case, RCA showed process factors to be the sole proximate cause for the error: The translation of the pre-incision marking for the proper level on the spine turned out to be a difficult task because of anatomic considerations. In view of these findings, policies were created to deal with potentially difficult situations in which preoperative marking with clamps and spinal needles at the intervertebral levels may be used before an incision is made: 1) intraoperative radiology consultation was to be considered when necessary; 2) a final confirmatory intraoperative radiograph was to be obtained and the proper site was to be confirmed before the wound was closed; 3) all surgeons were provided with in-service education and were proctored for proper execution of the policies; and 4) random audits by the QA Department were set in motion to ensure compliance with policies.
The traditional format of SMMCs, particularly for sentinel events, has cultivated a culture of blame in which the surgeon is expected to accept full responsibility for everyone's actions. Such an approach may not be in the best interest of patients, given that the vast majority of errors result from system failures rather than human errors. To complicate matters, professionals from other disciplines that play a pivotal role in the safe conduct of operations almost never attend SMMCs. Therefore, surgeons may be aware of only one part of the story instead of knowing all of the events that led to errors. RCA brings together professionals from all disciplines for a round-table analysis in a friendly, blame-free, and transparent environment. Findings from such critical analysis can then be used to develop mechanisms to avert errors, which, it is hoped, will translate into increased patient safety.
Although RCA is currently required for all sentinel events, the results of the analyses are not shared among professionals because they are considered privileged and confidential material, which means that no one other than the parties involved will ever learn from them. It is no wonder, then, that surgeons often resent having to accept new policies; they are not often given the rationale behind the policies. Sharing RCA results with all surgeons in yet another protected SMMC may narrow the gap between policy makers and practicing surgeons by making possible, in a safe environment, a mutual discovery of the forces that are driving policy changes in efforts to prevent litigation. Such a process is likely to provide a single platform for addressing medical errors and their causes, whether human error, system error, or combination of both. Although discussion of error analysis as an aspect of personal failure is a potent stimulus for education, there is a strong belief that correcting adverse events should not be done through any assessment of blame or personal culpability. Such opportunity enables learners to find the best measures in a friendly environment to avoid future errors and could play a major role in enhancing patient safety. Once the culture of blame is eliminated or minimized, such discussions should probably be considered for near-miss events, particularly in the setting of SMMCs, where members of all disciplines will be able to learn from error analysis and move forward with further education, QA, and potential enhancement of patient safety.6
In spite of its potential benefits, RCA is a complex and labor-intense process with many problems of its own. Research has shown that there is a wide variability in the quality of RCAs between institutions. Many RCAs are performed incorrectly or incompletely and do not produce usable results.7 Furthermore, RCAs are time consuming and very costly, and there have not yet been any studies of their effectiveness in reducing risk or improving patient safety.7
Data from the US Department of Veterans Affairs (VA), the pioneer of RCA, show that more than 7000 RCAs had been completed as of 2009. Experts estimate that each RCA requires 20 to 90 person-hours to complete.7 At an average of 55 person-hours each, RCAs conducted by the VA have required a total of 385,000 person-hours. Even with a very conservative average estimated cost of $25 per person-hour, those RCAs cost the VA $9,625,000. To date, there have been no evaluations of the cost or cost-effectiveness of RCAs compared with other tools for mitigating hazards.7
Because RCA is currently required by QA rules for all sentinel events, it would be appropriate for the time being to delay the discussion of any sentinel event in an SMMC until the RCA results are available. Such an approach will provide more insight into the events that led to the error, particularly with input from other disciplines involved in the care of the patient. For near-miss events when RCA is not required, some principles of RCA can still be used during SMMCs to further surgical education, improve QA, and increase the potential for enhancement of patient safety. Once the idea is applied, enough data can be generated to measure its efficacy, cost-effectiveness, and its impact on patient safety.
Although combining RCA with SMMCs appears to be a novel tool for minimizing errors, enhancing surgical education, and improving patient safety in a friendly, blame-free environment, several factors should be considered, such as quality of the analysis, cost-effectiveness, and actual impact on patient safety. QA professionals, Surgery Department leaders, and surgery educators should work hand in hand to experiment with and fine-tune modalities that can achieve their goals in an efficient and cost-effective manner.
The author(s) have no conflicts of interest to disclose.
Katharine O'Moore-Klopf, ELS, of KOK Edit provided editorial assistance.
1. Gore DC. National survey of surgical morbidity and mortality conferences. Am J Surg 2006 May;191(5):708–14.