By Daniel L. Cohen, MD, FRCPCH, FAAP
The challenges we face in improving patient safety and decreasing healthcare-related harm and death remain urgent, compelling, and undeniable. Despite our considerable efforts over the past 15 years, and some very notable advances, mostly resulting from increased standardization of processes and the use of checklists, the overall magnitude of the problem remains enormous. A recent study has purported that as many as 250,000 people may die each year as a result of the healthcare services they receive (Makary & Daniel, 2016), and many more will be injured, possibly with long-lasting impairments. Five or six baseball stadiums filled with relatives and friends perish every year. Even if one instead uses the lower estimate of 48,000 deaths portrayed in the Institute of Medicine (IOM)’s sentinel report, To Err Is Human (Kohn, Corrigan, & Donaldson, 1999), that is still a large baseball stadium of mothers, fathers, children, grandparents, friends, and lovers. We have a public health problem of enormous magnitude.
It has been well established that harmful patient safety incidents result primarily from failures or insufficiencies in the complex structures and processes of healthcare, as well as from human errors that occur when working with these structures and processes. Attempts to identify the numerous contributing factors that collectively result in causality at the bedside have relied on investigations of incidents, commonly referred to as “root cause” analyses. Unfortunately, the quality of such investigations is highly variable, actionable recommendations are sparse and inconsistently implemented, and shared national learning between institutions has been insufficient. Often, investigations of patient safety incidents focus on identifying a few “causes” without getting at the contributing factors where opportunities for improvements really reside. All roots have rootlets—pull a carrot out of the ground and look at it—and unless the numerous contributing factors are identified, opportunities for learning will be missed. Finally, and sadly, investigations have all too often focused on identifying individuals responsible for harm—individuals to blame—and that is inappropriate and inexcusable.
Though real progress has been made in the performance of many processes and procedures, through the use of standardized checklists and training designed to improve communication and teamwork, challenges remain in the critical domains of clinical diagnosis and diagnostic accuracy. Any error in diagnosis is, ipso facto, also a delay in arriving at a correct diagnosis, resulting in delayed treatment and possible harm. The compelling issue of diagnostic error must be addressed now.
Coming to grips with diagnostic error
The National Academies of Science and the IOM, in 2015, published a monograph entitled Improving Diagnosis in Health Care (Balogh, Miller, & Ball, 2015). In that report, the committee defined diagnostic error as the “failure to (a) establish an accurate and timely explanation of the patient’s health problem(s) or (b) to communicate that explanation to the patient.” Diagnostic error was estimated to be a causal factor in 10% of patient deaths and up to 17% of adverse events. At a pragmatic level, the report focused on diagnostic error arising from failures in communication among professionals and between professionals and patients; challenges in healthcare systems that do not support diagnostic processes; insufficiencies in feedback mechanisms that advise clinicians about their diagnostic performance; and, very importantly, the lack of a professional culture that encourages transparency and disclosure, which may impede learning from harmful events and near-misses. We cannot expect to learn from diagnostic errors, and to improve our diagnostic capabilities and accuracy, if those errors are not reported and investigated to identify all contributing factors that collectively result in causality.