The Residents Did Not Miss Many? Are You Kidding?

It is noteworthy that the AJNR has published a study assessing the consequences of misinterpretations of neuroradiology CTs by on-call radiology residents. If the results had shown significant misses, such an article could become fodder for plaintiff attorneys seeking settlements from teaching hospitals in medical malpractice cases alleging misdiagnosis of neuroradiologic studies such as CTs and MR images of the brain and spine. I can hear the cross-examination now: “Isn't it true, doctor, that the authoritative journal, the AJNR, published an article stating that residents routinely misinterpret neurodiagnostic imaging studies on call? Isn't it malpractice to have these studies read by residents without supervision, doctor?”

This would indeed be a frightening situation; fortunately, the accusation is not true. In this issue, Lal et al (page 124) testify that on-call residents do very well in the specific and focused task of interpreting emergency neuroimaging studies. The images evaluated in this study were mostly CT scans of the brain obtained emergently because of acute neurologic problems, and these are arguably among the most important studies regarding subsequent clinical management. The on-call radiology residents in this study had an enviably low miss rate (0.9%) of significant findings on CT, and it was even rarer for patient outcome to have been negatively affected (0.08%). These figures are heartening.

One interesting fact peculiar to this study, and probably not amplifiable to generalizations, was the lower rate of misses by junior as opposed to senior residents, with three being the greatest number of errors made by a resident. Thirty-three of the misses did not influence patient treatment or outcome; only two misses had a potentially serious effect. In one case, an atypical pattern of basal ganglia calcification was misinterpreted as possible hemorrhage. As a result, the patient did not receive thrombolysis for an early stroke, the clinical results of which tend to be uncertain. The other error involved missing a cerebellar stroke, which can be difficult to diagnose because the posterior fossa beam-hardening artifacts on CT are notorious for interfering with diagnostic accuracy. Overall, this is a good track record.

Of course, those charged with teaching residents suspected this would be the case. Residents, by the time they take emergency call, have been exposed to enough didactic and other teaching material to identify abnormalities. Residents tend to be overly cautious and probably notice minutiae that an attending might not, even though the minutiae may be of no clinical significance. I never have had a problem with a house officer noticing too much on an examination; it is a problem when findings are missed. We want the residents to see and document everything, even at the expense of a long dictation.

There are a few residents I have known over the years who, fortunately for the authors, were not involved in this study. Residents learn at different rates, require different handling, and are capable of accepting different degrees of responsibility after the same number of years of training. There are one or two who resist training of any sort. Some believe they already know enough, some are not receptive to training, and some are not motivated to learn.

The number of problem trainees is low because, in general, we are able to select motivated house officers. Even an excellent resident cannot be expected to make the correct radiographic interpretation 100% of the time—a fact the law recognizes. Even in malpractice suits, the law only requires that a physician be reasonably prudent and apply the “standard of care” to every case in which they are involved. The law does not mandate that the physician be 100% correct 100% of the time, but that the physician must do his or her best. This is what we strive to do and what we train the residents to emulate, knowing full well that 100% accuracy is not attainable.

Nevertheless, it behooves us to be aware of individuals in the population who feel that residents-in-training might represent a risk to patients. One particularly vocal and extremely well-organized group headquartered in the Tampa Bay area is the Association for Responsible Medicine. I recommend visiting this organization's web pages at http://www.a-r-m.org/tortquest.htm and http://www.a-r-m.org/armact.htm to see how one cadre of patients views the quality of care rendered by residents in teaching institutions. At the second web site, be sure to read the first case history listed at the bottom of the page, titled “Teaching Malpractice.”

Mistakes do happen in medical care, and misinterpretations, though infrequent, are at some level unavoidable. Are residents more likely to make mistakes than attending physicians, or for that matter, private practice physicians? I do not think it is inevitably so. This article supports that view and allows us to continue to train residents and fellows appropriately.