Finding Out Your Doctor Made a Mistake

Medical errors have become more common as care has become complex. Clinical pathological conferences involving review of autopsy findings, chart reviews and medical audits are older methods of identifying errors in retrospect¹. More recently, quality improvement rounds including multidisciplinary perspectives¹ and electronic clinical decision support systems have been used to identify and prevent error prospectively.

In this issue of JGIM, Bollinger et al.² describe a new way of uncovering medical error, doing so in the course of conducting research aimed at quality improvement. Pragmatic clinical trials (PCTs), a methodology that is not really new (the polio trial of the 1950s is perhaps the most famous example), are being more commonly performed once again. In the modern version of PCTs, questions can be answered with simple designs like randomizing institutions to one commonly used intervention or another, collecting lots of data about the subjects in each place and then comparing outcomes. In many cases, the interventions being compared are deemed to be so standard that research ethics boards exempt the investigators from obtaining informed consent. In the course of performing the study, medical errors are sometimes inadvertently discovered by the research team, a phenomenon known as “collateral findings”. While the research team may not be required to inform patients that they were subjects in a study (including the fact that someone other than their clinical team was looking at their medical records), they are ethically required to inform those people of the errors, especially when they can be corrected. Bollinger et al. ask an important question: what do patients think about whether and how they should be told about those errors? What are their reactions to this information? Is it welcome?

Using a diverse sample of 66 people in 3 cities across the USA, they used a hypothetical example of a PCT that compared 2 strategies for treating hypertension. In the course of doing the research, the investigators discovered that a patient had been placed on a contraindicated combination of drugs that potentially interact in a way that could lead to harm (arrhythmia). They asked the subjects in their study what they wanted to know, how they wanted to be informed, and who ought to tell them. The key finding was that patients varied considerably in their answers. Some were grateful, some were angry. Some wanted their own doctor (who likely made the mistake) to tell them. Others didn’t trust that person any more so wanted the researchers to tell them. Some wanted to sue. Some wanted lots of information about how the mistake was uncovered. Others didn’t want to know their privacy was being invaded by researchers. So aside from wanting to correct the error to prevent any harm, and being treated respectfully by someone they could trust, no one size fit all.

As I reflect on these findings, I personalized the question, and expect others will do so too. How would I (you) want to be informed? I have my own experience with medical error; here is my story. As a teenager, I was treated with radiation for acne on my face by a dermatologist. When I went to medical school, I learned that this treatment predisposed me to thyroid cancer, so in my late 20’s, I started to be monitored. Years later, a nodule was found that was aspirated several times for cytology. After one sample was read as “suspicious”, my endocrinologist suggested a hemithyroidectomy. The nodule in question was diagnosed as a follicular adenoma (benign), but there were also several microscopic foci of papillary carcinoma, a finding that would now be considered unimportant. But at that time, both the surgeon and endocrinologist recommended complete thyroidectomy; the other side was completely normal. This meant I needed to receive thyroid hormone for the rest of my life. About 5 years later, I told this story to another endocrinologist who wondered if the benign diagnosis of the initial nodule was correct. He asked if he could review the sample with a different pathologist. I agreed. The revised diagnosis of the nodule was papillary carcinoma. Based on this new information, my original endocrinologist recommended radioactive iodine treatment, which I underwent.

About 15 years later, concerns were raised about the accuracy of the second pathologist’s research, leading me to question the revised diagnosis. So I contacted the head of the Pathology Department and asked for a third reading. She didn’t reply to my email or phone calls for weeks, so I pursued her. Eventually, it was reviewed (by 2 pathologists) and the diagnosis firmly reverted to the initial benign one. So if we follow this chain of events, I received a therapy that likely didn’t help me and predisposed me to cancer, a biopsy that used the term “suspicious” that (in retrospect) led to two operations I didn’t need, and worse left me dependent on exogenous thyroid for the rest of my life. If that wasn’t enough, I received a likely false diagnosis of cancer, which then led to more therapy that now increased my risk of other cancers. All from a case of teenage acne!

When I discussed this matter with the head of the pathology department, she explained that pathological diagnoses of thyroid cancer are not exact, and that over the years there has been controversy and changing criteria. When I asked her why she evaded my request for a re-reading and discussion, she told me she felt that re-opening the matter would cause emotional distress for me. I replied that she was wrong; I found the new information and conversation helpful. I was still angry with the second pathologist (and dermatologist too), although part of me still wonders which reading is correct. But at my age, I also understand that errors and interobserver variation in pathology interpretation are common, and medical practice changes; so get over it. There is nothing I can do about this now. But in my case, I appreciated knowing the “truth”, having a full discussion of the reasons for the “error”, and being able to express my concerns. But that’s just me; I don’t expect everyone to have the same preferences. The point here is that a physician presumed to know what I wanted to know, and she was wrong. She appreciated hearing that.

Bollinger et al. teach us that, like my story, it is very hard to predict what patients want to know when retrospective error is discovered; an interesting and important line of investigation. I suspect they will continue to find there is no one correct answer. But then as clinicians, we already know this; patients handle all kinds of bad news in different ways. Nevertheless, there are likely some common themes; all want to receive the news with empathy and respect. Research is increasingly including mechanisms for informing subjects about study findings and including them in the research process. I expect questions like the ones Bollinger et al. have raised will be important to answer as we embark on this new approach to generating new knowledge.

Compliance with Ethical Standards

Conflict of Interest

Dr. Detsky has no financial conflicts of interest, though as the patient in this story, he clearly has a personal one.