What’s in a story? Lessons from patients who have suffered diagnostic failure

For a variety of reasons, the patient’s perception of medical care differs profoundly from the clinician’s. Aside from the obvious differential of knowledge and power, there is the unavoidable fact that patients and families are the only individuals who traverse the entire spectrum of care and the only ones who experience first-hand the effects and outcomes of assessment and treatment. As is attested by a small but growing body of first person accounts by physician-patients, the reality of the patient experience often comes as a surprise to doctors, who assume, until they become patients themselves, that the system functions much more seamlessly than it does [1–4].

Patients, too, have a body of literature on their healthcare experiences [5–8]. Many of these accounts paint a less than flattering picture. A significant portion of the dissonance they record revolves around diagnosis: misdiagnosis, delay in diagnosis, failure to diagnose, and the many factors that contribute to these failures. Like doctors, patients are often taken by surprise by the fallibilities of the healthcare system, and even more by the system’s apparent lack of responsiveness to feedback from its users. Yet it should be evident that patient accounts, medicine’s reality check, have tremendous value. They hold the keys not just to detecting system errors but to improving diagnostic performance by providing a window onto the ways that recommendations are carried out, the ways information flows, and the ways that individual clinicians’ recommendations intersect with those of others. Above all, they can shed light on the aspects of the patient-provider equation that facilitate accurate diagnosis, and on the real-life effects on patients of diagnoses both accurate and inaccurate, timely and delayed.

Over the course of what has become a second career in patient safety advocacy, I have heard many patient accounts of harm – and you only need hear a few to appreciate the repetition inherent in these stories. Among the most striking patterns are those of diagnostic error. These include systematic ways in which certain diagnoses are missed, pat conclusions that are substituted for clinical assessment when diagnoses are difficult, and delays in identification that can turn treatable conditions into crises. It is the persistence of these patterns that patients find so hard to comprehend; and it is the stories of these patterns that reveal the fault lines along which the system can give way, again and again. If ever there was a treasure to be found in complaints, this is it.

Like most patient advocates, I came to the calling of patient safety through a story of my own – the loss of my brilliant young son to complications of elective surgery. And like most stories, his illustrates more than one of the patterns listed above [9]. Lewis was a healthy 15-year-old when we took him to the hospital for a procedure to correct a structural defect of the chest wall, pectus excavatum. While the surgery was hardly minor, we did not appreciate that it had a risk-benefit calculus that might not be appropriate for a child with a relatively mild defect (as Lewis’s was). It was not the surgery per se from which our son died, however, but the pain control regimen that accompanied it. Residents and nurses failed for over 30 hours to recognize signs of a perforated ulcer brought on by the NSAID painkiller ketorolac, which Lewis was given as an opioid-sparing adjunct to his epidural narcotics. Ironically, the well-known NSAID side effect of gastric perforation was never considered, because residents and nurses focused immediately and unswervingly on the possibility of narcotic-induced ileus as an explanation for Lewis’s sudden, severe abdominal pain.

Autopsy revealed the error. Our vibrant boy had not only suffered a giant duodenal ulcer and severe peritonitis; he had also lost more than half his blood as the ulcer eroded into the underlying artery. The signs of these conditions were blindingly obvious in retrospect. It was clear that the diagnosis of narcotic-related constipation was a poor fit with the 5-out-of-5 pain, weakness, and wildly deranged vital signs that Lewis had exhibited. Yet residents and nurses had been so sure of themselves that they had treated our requests for help as little more than a nuisance and casually disregarded our request for an attending physician. What could account for this?

I think there are many answers. In Lewis’s case, residents and nurses exercised little to no critical thinking and fell prey to almost every cognitive bias in the book – anchoring, belief perseverance, and groupthink, among others [10]. An inflexible hierarchy that provided inexperienced caregivers with little support and little incentive to change the plan may have given a quick and easy diagnosis even more appeal. Lack of familiarity with the side effects of common drugs and with the signs of shock and sepsis – knowledge deficits all too common across the healthcare continuum – were also important factors. Above all, lack of respect for the voice of the patient meant that input from me and from Lewis was disregarded, even though we were polite, highly educated, very worried patients who presented no barrier to communication. All these factors can – and in our case did – lead to complacency, failure to look at trends in the patient’s condition, and failure even to include the real problem on the differential list.

Failure to rescue – cases like Lewis’s – is a significant cause of death in US hospitals [11]. Although, it is often thought of as a nursing issue, it is also a major diagnostic failing. Hospitals have come up with a number of tools to try to prevent these catastrophes, including computer decision aids, algorithms, electronic alerts, rapid response and other specialized teams. One of the most important interventions, however, is less often used. This is direct patient and family involvement, including such areas as patient education – especially information about navigating the system and danger signs to watch for in the patient’s condition – bedside rounding, access to the medical record, and patient-activated rapid response. These are all important strategies for helping the patient and family ensure that they can get access to the diagnostic expertise they need, and that diagnosis and care are in accord with their understanding of the patient’s symptoms and treatment plan.

Diagnosis is born in a relationship, and it forms the basis for everything else that happens in healthcare. Perhaps more than any other aspect of medicine, it is knowledge-based, culture-based, and personal. The vast majority of diagnostic errors and delays reported by patients concern not issues of process, but problems in the patient-provider relationship. More specifically, diagnosticians’ attitudes and preconceptions toward individual patients can and do result in failure to listen, really listen, to what the patient is saying. As George Bernard Shaw is often (perhaps apocryphally) quoted, “The single biggest problem in communication is the illusion that it has taken place.” So it is with diagnosis. If there is a single thing that diagnosticians need to learn, it is to listen in a meaningful way and to communicate what they have heard. In a profession whose success depends upon the success of relationships, this ought to be taught, and taught well, as a basic tenet. Disregarding the knowledge of patients and families is a dangerous act. This is the fundamental lesson from Lewis’s story, and many others.