Preventing injuries during surgery due to technical mistakes

Blogging on Peer-Reviewed ResearchYou’ve probably heard the oft-repeated charge of “alternative” medicine advocates. If you get into a debate or conversation with one, you can almost count on seeing or hearing it before too long. Indeed, we heard a variant of this very claim yesterday coming from über-woomeister supreme Deepak Chopra. I’m referring, of course, to the rant against “conventional” medicine that medication errors claim 100,000 lives a year. Of course, as Mark pointed out, “conventional” therapies actually work, and because they work there’s risk to them. Moreover, its hospitals actually care for seriously ill patients. However, even so, medical errors are more prevalent than they should be, and this is where the difference between “conventional” and alternative medicine is most apparent. No, not because conventional medicine produces errors. What I’m referring to is how, unlike “alternative” medicine, conventional medicine is engaged in a concerted effort to minimize medical errors, be they systemic or by individual physicians.

As hard as it may be for some to believe, in many cases surgeons are spearheading such efforts, doing studies that seek to identify causes of bad outcomes and surgical errors. Not long ago, I discussed one such study, which looked at morbidity and mortality rates by the month in teaching hospitals and seemed to indicate that morbidity and mortality increased in July and August, paralleling the arrival of new interns, residents, and fellows every year. The studies had some flaws, but the effort was admirable. Now, hot off the presses in this month’s Annals of Surgery is study that seeks to characterize patterns in surgical errors. It’s coauthored by Dr. Atul A Gawande, the Harvard surgeon who’s made a name for himself with his books Complications: A Surgeon’s Notes on an Imperfect Science and Better: A Surgeon’s Notes on Performance, both of which deal with complications of surgery and how to improve medical care.

The study, Patterns of Technical Error Among Surgical Malpractice Claims: An Analysis of Strategies to Prevent Injury to Surgical Patients, took a rather clever strategy. As much as we physicians hate it, malpractice suits represent a source of data that is seldom used but can often identify glaring medical errors that can’t be ascribed, even obliquely, to differences in surgical opinion. We’re talking about wrong-side surgery errors and the like. Of course, while I find the information in this data as interesting and potentially useful as Dr. Gawande did, I would be remiss if I didn’t point out that the threat of malpractice suits can also distort medical care, leading doctors to practice “defensive” medicine, actually increasing the risk of mistakes by subjecting the patient to more tests and procedures. Be that as it may, actual malpractice suit data do provide a unique insight into surgical errors.

Such an insight is needed because, as the article points out right in the first paragraph, between one-half and two-thirds of hospital adverse events are attributable to surgery and surgical care. Also, the sorts of errors that occur in surgical care tend to be different than those that occur on medical services, making many of the studies of medication errors in hospitals not easily generalizable to surgical care. The big difference is that most surgical errors occur in the operating room and most are “technical” in nature. Technical errors are errors in which some aspect of the surgery is not done properly, and can include errors of manual skill and errors of “surgical judgment” (i.e., decision-making in the operating room) or knowledge. An example of a purely technical error includes accidentally cutting something that shouldn’t be cut, whether due to carelessness or failure to identify the anatomy properly, or a tie on a blood vessel that’s improperly tied, leading it to fall off later. These errors, if recognized promptly at the time of surgery, often cause little or no morbidity because the surgeon can fix the problem right away, such as when surgeon makes an accidental knick in the bowel, sees the hole, and sews it up right away. If unrecognized, however, such technical complications can lead to devastating complications, the hole in the bowel being a classic example, which may not be recognized if small. The patient will slowly leak stool into the abdomen and become really septic within a few days of surgery.

Another class of surgical error is the error in knowledge or judgment. Surgery is unique among medical specialties in that, while doing operations surgeons are constantly making decisions in real time and acting on them. There’s not much time to contemplate, because you can’t leave the patient under anaesthesia while you walk off to the library to look up literature on what you should do. This aspect of surgery is often what is tested on board examinations, where the candidate will be asked what he or she would do, for instance, if a certain unexpected finding were to be encountered at the time of surgery. One classic example of a “judgment” question is what to do if you find an ovarian mass in a woman undergoing a splenectomy for a hematological disease. Another is what to do if an abdominal aortic aneurysm is found at the time of surgery for a malignant polyp. (These days, if a patient has a decent-sized colon cancer, he would be likely to have had a preoperative CT scan, which would identify the aneurysm making this question a little less relevant than it was, say 20 years ago.) Errors in knowledge tend to involve doing the wrong procedure for a condition, a classic example being performing a simple cholecystectomy for an invasive gallbladder cancer.

These sorts of errors can occur at any phase of surgical care, and numerous studies have been done to try to identify causes. Such errors and adverse outcomes have been attributed to low hospital volume, breakdowns in communication, systems shortcomings, fatigue, lack of experience in trainees, and many other causes. These studies have led to many proposals of how to minimize such errors, many of which are controversial, such as referral of certain kinds of cases only to high volume centers.

In order to characterize surgical errors that led to serious injury to patients, this study examined a database of closed malpractice claims. 444 such claims were examined in which 258 were identified in which an error resulted in an injury. Examining the claims, reviewers identified 135 (52%) in which it was judged that a technical error was a major contributing factor to patient injury. Error was defined according to the Institute of Medicine definition: “the failure of a planned action to be completed as intended (i.e., error of execution) or the use of a wrong plan to achieve an aim (error in planning).” Technical errors were those where reviewers believed that an error of operative technique directly contributed to the adverse outcome, either because of manual error (error of execution causing injury to viscera or vasculature, for example) or judgment or knowledge error (wrong timing, wrong selection of procedure, failure to diagnose complications, wrong site surgery).

There were 140 technical errors identified. Attending surgeons were responsible for 69% while 27% involved both the attending and trainee. Reassuringly, only 4% were due to actions by surgical residents or fellows alone. Less reassuringly, most of the technical errors caused serious injuries, with 49% resulting in permanent disability and 16% resulting in death. Not too surprisingly, given that these were severe errors, the kinds that lead to malpractice suits, 31% involved gastrointestinal surgery, 15% spine surgery, 12% gynecologic surgery, and 9% non-spine orthopedic surgery. In terms of error type, 91% of technical errors involved manual error and 35% involved judgment or knowledge errors. A summary of the types of manual versus judgment/ knowledge errors is summarized in this table:

i-6a907bbb00607e2b14d9c5d899f08ba5-table2.jpg

Perhaps the most interesting findings were the contributing factors to technical errors. Not surprisingly, 69% of technical errors involved complicating factors, with 61% having patient-related factors (difficult or unusual anatomy 25%; reoperation 20%; urgent or emergency operations 17%). There were also a number (16%) of equipment use misadventures. Not surprisingly, experienced surgeons were less likely to have equipment use misadventures than inexperienced surgeons and more likely to make errors related to reoperative surgery. The reason this is not surprising is that the anatomy is often hard to ascertain in reoperative surgery due to scarring and adhesions, and more experienced surgeons are more likely to attract referrals of more difficult patients. What one might find surprising on the surface is this comment in the discussion:

Almost three-fourths of technical errors in this study involved fully trained and experienced surgeons operating within their area of expertise and 84% occurred in routine operations, for which advanced expertise beyond a standard training program is not required or expected.

What this tells us is that it is during the common operations when errors happen most frequently. As an accompanying editorial puts it, “It is not the neophyte doing the 10-hour Whipple that leads to most malpractice claims; it is the experienced general surgeon doing a gastrectomy on a patient with three previous upper abdominal operations and a replaced left hepatic artery.” This should not be that surprising, given that highly complex operations tend to be performed in academic medical centers by highly specialized surgeons and surgical teams. What this study also suggests is that the best strategy to have a rapid impact to reduce surgical error may be to develop strategies to improve decision-making, operative planning, and team performance for common, not necessarily the highly complex, operations. Moreover, it suggests that volume- or experience-based limitations on privileging for high-complexity operations would only have an impact on a relatively small minority of surgical errors. In other words, a bigger bang for the buck in this respect would likely result from focusing on interventions to improve performance and decision-making in routine operations in emergencies and on high risk patients.

Another interesting aspect of this study was raised in the editorial, namely the question of how often experience itself invites disaster:

How often did experience itself invite disaster? I am thinking of a case of my own recently where I embarked on a difficult liver resection in an 82-year-old man who had undergone 8 cycles of chemotherapy. The lesion was in an awkward place and I knew of the increased risk. Yet, my experience made me confident that I could “get away with” a procedure that I would flunk a young surgeon for if she proposed it during a board examination. This reminds me of the comment made by the first officer of an airliner that ran off the runway in Burbank, California, after an unstabilized, too fast, nighttime approach. “I’d seen it work out before,” he said.

This is a telling observation. Pretty much every surgeon knows of a “cowboy” who loves taking on the most difficult, highest risk cases, sometimes even to the point of going too far and operating on patients who probably should not be operated on. They may be so good that they do “get away with it” most of the time, but it’s still riskier than it should be.

I think the takehome message from this study is, as for many things in surgery, that simple interventions in common problems are likely to yield the most benefit. Big cases like Whipples and liver resections are often examined at a frequency that far outstrips their actual frequency relative to common operations like hernia repairs and cholecystectomies. After all, the median number of Whipple operations done by a typical private practice general surgeon per year is zero, while the same surgeons may do a couple hundred or more laparoscopic cholecystectomies during that same time It would almost certainly pay off more dramatically to take care of fixing factors leading to errors in the common “bread and butter” operations first.