Below is an article that I wrote several years ago, that is now more true today than when I wrote it. At the time, we had a flamingly incompetent Chief of Staff (called Dr. Bigshot, since he remains very prominent in politics at our hospital) and the staff of our hospital was all given an article by Dr. Guwande from the New Yorker regarding the virtues of checklists in saving lives. My apologies for not being able to give you the exact reference for this article, as I threw it in the wastebin. I have no problems with checklists. I have a serious problem with assuming that checklists are what saved the airline industry and that people would be saved if only we used checklists. So, I re-post my article. The next post carries on with the same theme, now written contemporarily. BYW, Dennis, I found most of my grammatical errors, but feel free to inform me of others.
Several years ago, tort reform became the cry of the medical profession. We felt that our profession was being destroyed by a litigious culture that was strongly supported by a government that seemed to thrive off of a healthy legal industry. We lost that battle. In return, the law industry laid claim that the health care industry was careless and did not attend properly to quality control or error reduction. In turn, we responded with multiple programs. There were state and national programs that were initiated, such as the 100,000 lives campaign (I await eagerly the 250 million lives campaign). Even in Pierce County, our medical society invited various quality control pundits to speak to us. The rallying cry was to become like the airline industry. After all, did not the airline industry take an intensely complex system, and produce methodological algorithms (such as checklists) to eliminate human error? As I learned in flight surgery school, the number one cause for airline fatalities was a loss of situational awareness on the part of the pilot. Checklists helped to reduce routine operational error, thus, decreasing the one aspect of a fatal error.
The article by Atal Guwande in the New Yorker further fosters this idea that if only the health care industry model itself after the airline industry, then error reduction would significantly fall, and lives would be saved. I certainly agree with Dr. Guwande that checklists can serve some useful purposes in our profession. Yet, I also see certain problems with what he proposes. The first problem discusses differences between the airline industry and medicine, that disallow the airline model. The second details the evidence that Dr. Guwande himself provides claiming that checklists can solve many of our woes.
First, what are the differences between medicine and the airline industry? There are a number of issues that I can list.
1. We can’t control the circumstances. In the airline industry, if bad weather hits, the airlines shut down. We can’t do that. We “fly” in any circumstances. If a patient arrives in immediate need of surgery when the operating rooms are already filled and the patient already has multiple system organ failure, we aren’t allowed to “stop all flights (surgeries)” and wait, in order to get control of the situation.
2. We don’t aim for 100% survival. Ultimately, all of our patients will die, which is 0% survival. Unlike airplanes, we have a poor means of predicting personal survivability. We can quote population statistics, which do not apply to a given individual. Checklists or not, eventually everybody will die on us. In fact, we have very poor means of measuring when we are actually successful in medicine, as it is not necessarily survivability at low cost without complications.
3. We cannot set the circumstances for surgeons or health care personnel like we can with pilots and flight attendants, airline mechanics, etc. I would love to have the same working circumstances as a surgeon as a pilot usually lives. There are strict controls of working hours, and time that a pilot is allowed in the cockpit. We have no such controls. Yet we know that human error is our biggest source of health care error, just like situational awareness is the biggest problem in the airline industry. Establishing mandatory retirement ages, mandatory work-hours, mandatory spontaneous drug testing would kill the industry. I have operated countless times high on antihistamines in the symptomatic treatment of seasonal URIs, yet such drugs would have grounded me in the airline industry. Are we willing to have our health care personnel subjected to such demanding regulation as the airline industry has done? Why not? The object is to eliminate human error, and such airline regulations would accomplish that.
4. Human systems back-up cannot compare. A pilot has not only a second backup (the copilot) always at his side, but also the capabilities of autopilot. Generally, we virtually never have a second physician (with the same expertise) simultaneously participating in a case. Auto-doctors remain to be invented.
5. This leads to a brutally serious question…why have auto-doctors not been invented yet? Autopilots work because one can “figure out” most of the system’s issues and expected problems in the operation of an aircraft. The “machinery” (the human body) that we work with is infinitely more complex than the machinery (the airplane) that the airline industry works with, and the expected problems vastly greater. While Dr. Guwande tends to disparage the “art” of medicine, heralding the virtues of scientific medicine, it remains without question that the complexities of medicine demand both intuitive as well as methodological decisions, and the intuitive decisions cannot be check-listed. An equivalent comparison would be to devise an airplane that is so complex, the ground support personnel never really understand how the airplane works, or exactly what the proper procedures are to repair. The pilot could never be sure whether pushing the joystick to the right would move the appropriate wings or flaps in the proper direction, and would be told that any control panel action would have only an 80% or less response rate, as well as a highly unpredictable nature of whether all the monitors or gauges on the control panel were ever monitoring the correct information. Yet, we live with this all the time in medicine.
6. The economics are different. If the airline industry is asked to institute an industry-wide change, they would raise rates to passengers to pay for that. We cannot do that anymore in the health care industry. In fact, our pay would either remain stagnant or cut, in spite of the elimination of error.
7. Training and retraining. We call retraining CME, yet CME only remotely pertains to our practice of medicine. A flight simulator has never been invented for the health care industry, probably for reasons explained in #5. Our expertise comes solely from experience, coupled with the maintenance of an innovative mindset. When we increase physician educational demands and demonstration of competence through increased testing, the net result is not increased competence among physicians, but a decreased number of physicians, who drop out rather than re-test. This doesn’t mean that we can’t learn from the airline industry. It only means that we need to be very cautious in selecting what methodological algorithms we acquire from the airline industry, and then be highly selective in exactly which circumstances or activities would be well served by these algorithms. It is possible that some systems in medicine would actually be harmed by blindly applying the airline industry methodology of error prevention.
What about Dr. Guwande’s claims that checklists can significantly reduce errors in medical care? Dr. Guwande discusses his thesis with unbridled enthusiasm. In a most unscientific manner, he fails to discuss multiple variables that should have been examined, especially since his thesis of the virtues of checklists are now being mandated throughout hospital systems in the USA. Which variables did Dr. Guwande follow? Survival? Costs? Turnover rates of health care personnel? Patient and family satisfaction? Days of hospitalization? His studies of checklists were limited to highly specific and controlled circumstances, such as the management of central lines. This is a relatively non-complex system compared to many systems seen in medicine. Does he propose that all operating systems will be helped by check-listing? Does he have evidence for that? Newly enacted checklists tend to eventually breed familiarity, which in turn lead to loss of effectiveness. Dr. Guwande has only short-term follow-up of his check-list system, so it is not surprising to see short-term improvements. What do you suppose we will see after ten years of check listing and familiarity itself leads to error? I suspect it will lead to even more detailed check-lists, probably orchestrated by a computer program, rather than a human, such as the nurses that Dr. Guwande used in his catheter study. This in turn will not only drive up the costs of medical care, but also the depersonalization of medical care.
Outside of checklists, the failure to communicate has been identified as the other great source of medical error. There is a great amount of truth to this, and check-lists certainly serve the function of forcing a brief episode of communication among the team, many of whom often don’t even know each other’s name, let alone the most rudimentary facts about the other people on the team in the room. But, we don’t dare tread on that. We must remain scientifically impersonal. Yet, when I work with a team that has known me for years, typically, minimal communication ever occurs about the patient or medical care we are rendering, save for occasional teaching points for the team (we do talk about other things!). We know how each other does things, and we expect things to be done that way. This is true for nurses and techs in the OR or recovery room, as well as experienced nurses on the wards. Sadly, regimented communication cannot fix the problem of operational harmony, something that only time and experience with each other as a team can fix. This is why “teams” are probably more important than checklists. Another communication issue, handwriting, was fixed thirty-some years ago with computer-order entry, quite the norm in Chicago, IL where I trained, but still unknown in these parts.
Dr. Bigshot comments that resistance to checklists is an “ego” issue. I doubt it. True, there are ego issues when one has a nurse policing the doctor. Not even the airline industry has stooped that low, having a stewardess tell the pilot to push the rudder right rather than left when the airplane is going down. But that is exactly what is happening in medicine. You can escape hierarchical disorientation by being independent, which is exactly what Dr. Bigshot has done. Hospital-bound doctors like surgeons and intensivists don’t have that luxury. Is it egoistic to ask questions pertaining to the efficacy of checklists? I don’t think so. Many of us could have easily gone into research rather than clinical medicine. Our training teaches us to ask questions, look for alternative solutions, explore the unthinkable, to agonize over a solution that doesn’t exist in a textbook, journal article, or on a checklist. Yesteryear, that made you a good physician. Nowadays, it makes you a non-team-player, radical, disruptive, or perhaps, worst of all, egoistic.
We will turn to checklists. We will love them with religious devotion. The Joint demands it. We will comply. Yet, it feels like we are driving just another stake into our coffin. R.I.P.