Facing up to medical error

Abstract

Medicine can be very dangerous for your health. This is a bitter realisation that every physician, practitioner, and researcher in the health sciences should make sooner or later. However, few dare to admit it openly, and only a handful of devotees decide to make out of this realisation a field of careful scientific study. In many other disciplines, studying and understanding errors is considered a standard process. No need to blame individuals. It is the system, stupid. You find where the system fails, you correct it, and things get better. This is how manufacturers of aeroplanes, space shuttles, washing machines, and robots usually improve performance and reduce the chances of future malfunction. Conversely, medical error and medical harm have often been accompanied by an aura of secrecy, a sense of personal failure, poor communication, defensiveness, and denial. The accompanying expectations of the public for some utopian perfection in the art and science of medicine, alongside litigation threats for documented imperfections, have not helped make things easier for doctors and their patients. In Medical Error and Harm: Understanding, Prevention and Control, Milos Jenicek tries to bring some order to the messy, but fascinating, science that deals with what can go wrong in medicine. He even comes up with a Greek term for this scientific field, lathology—ie, the study of lathos (Greek for error). Jenicek is an experienced and engaging writer, one of the most prolific authors of the past 35 years to address the interface between clinical epidemiology and evidence-based medicine. In his overview of the challenges in lathology, he draws heavily from his expertise in these two domains that are crucial for making quantitative progress on these important issues. However, much of the relevant supporting literature comes from other domains, as diverse as operational research, manufacturing, technology development, psychology, management, sociology, computer science, and ergonomics. In these disciplines, errors have usually been approached with less awe than in medicine. This wide-ranging expertise may have useful lessons to teach us when trying to understand, prevent, and control, medical errors and the harm that they may cause. Contrary to what one would expect from a clinical epidemiologist, this is not a book with lots of numbers. This might be good news for the general reader, but I actually felt a bit embarrassed at how few numbers appear in Jenicek's book. There are two reasons for this, besides the style of the author. First, there are relatively few studies on medical error and harm, compared with their potentially huge impact. Second, most of the seminal investigations have covered specific settings in specific countries. Jenicek mostly discusses the evidence qualitatively, but the references that he provides are adequate for the interested reader who might want to dig further into them. There are even fewer quantitative studies on what to do about reducing medical errors—a regrettable dearth when one wishes to make decisions based on evidence. The multifaceted nature of medical error makes evidence difficult to obtain, apply, and generalise. While making a plea for more quantitative studies using the time-honoured methods of epidemiology and experimental research, Jenicek realises that error is a field where quantitative studies based on large numbers may provide only part of the necessary information for the study of medical error. Therefore, he also endorses a variety of qualitative research approaches and the in-depth analysis of single case studies, as well as an array of theoretical logical constructs that may be applicable to modelling the error processes. Medical error is not the same as medical harm. As Jenicek points out, error does not necessarily translate to harm. I would add that perfectly correct medicine (the perfect application of what we deem to be correct based on currently accepted evidence) may also result in major harm, because of the imperfection of our knowledge and because of the probabilistic rules that permeate health outcomes. One can then always blame imperfect knowledge, and that luck, life, nature, whatever, was not on our side. However, error has the extra layer that something recognisably went wrong. This is a great opportunity that should not be missed, since we can learn from our errors and this may help us to avoid repeating them in the future. Jenicek also places a lot of emphasis on the basic distinction between human (individual) and system error and shows how different schools have placed privileged preference on either of these two aspects. Human and system error are interconnected, but also distinct; improvements in detection of error, analysis of its structure and causes, and eventually prevention and control may need to tackle both. Jenicek makes a valiant effort to impose some order on the diverse nomenclature of lathology and delves into the difficult task of taxonomy of this subject. This is a formidable task, given the range of disciplines that have approached the study of error, often with quite esoteric nomenclature. The proposed disentanglement is often not very parsimonious, and the taxonomies and long lists and schemas can become tiring at times. However, if we want to make some genuine progress in the field, we should probably be able to use a common language, and Jenicek's effort in this regard is worthy of consideration, even if some of his choices inevitably reflect his own personal preferences. Medical errors cannot be eliminated or avoided, so we have to live with them and hopefully try to reduce their frequency and, more importantly, their impact. Although we can learn about error from a wide variety of approaches, I still think that it is time to obtain more hard-core randomised evidence on how to reduce not necessarily the frequency of errors, but their impact on major clinical endpoints, such as death and major disability. Error is a surrogate endpoint. While usually more errors will translate to more harm, it is possible that in some cases more errors may result in less harm and better outcomes. If that were the case, my preference would be for health systems that make more errors. This may sound weird, but if you are not convinced, let's randomise the first patient and see what happens. Regardless of whether such paradoxical situations exist or not, in the meantime and while we await more evidence, I would definitely wish to be taken care of by doctors and health systems that recognise, admit, and occupy themselves more with the errors that we all make.