Letter to the Editor: Editorial: Chance Encounters, Overdiagnosis, and Overtreatment

To the Editor,

Thank you for your Editorial, “Chance Encounters, Overdiagnosis, and Overtreatment” [5], which I read with great interest, as it resonated with my thoughts developed in a busy orthopaedic spine surgery practice over a 30-year span. I agree that inappropriate diagnostic testing results both in overdiagnosis and overtreatment, and I would particularly like to focus on one aspect of current spine practice: the increasing use of MRI, its overpowering effect on the decisions that surgeons make, and the ill-effects both on the psyches and outcomes of our patients.

The problem lies both in overuse of MRI and the timing in the course of a patient’s experience when the MRI is ordered. Because most patients with low back pain will improve spontaneously, an MRI performed very early in a patient’s experience of symptoms, particularly if a proper trial of appropriate nonsurgical treatment has not been performed, exerts a powerful nocebo effect on the patient [6]. Radiologists, who have no clinical knowledge of the patient, report purely on images; the language used in those reports affects patients’ perceptions of their conditions and the likelihood that those conditions will produce ongoing symptoms. Many incidental and age-related findings are described using alarming terminology; descriptions like disc degeneration or herniation, annular tears, neural compression, and foraminal compromise all suggest conditions that need to be fixed or problems that won’t improve in the absence of intervention, when in fact the opposite usually is true. Patients who don’t understand the (lack of) medical relevance of such terminologies often misinterpret the reports [2], they perceive doom, and often feel convinced that some intervention is required to regain some kind of “normal” status [7]. The treating surgeon, while being aware of the inconsequential nature of the findings, may still take advantage of these reports and justify the performance of a procedure that may not benefit the patient. One study found that performing MRI to assess back pain instead of plain radiographs results in a threefold increase in the use of spine surgery, with no difference in patients’ outcomes at the end of one year [4]. This can explain some of the wide variations in the use of spine surgery for back pain in various regions of the US [8]. A meta-analysis of six randomized trials with patients followed both short and long term (n = 1804) without advanced imaging did as well as those who had MRI [3]. We performed a randomized controlled trial of patients with chronic low back pain, of whom one group were provided the MRI report and a factual explanation of the report and the other group was reassured that their MRI findings were normal [6]. With the same treatment, patients who are aware of their MRI reports had a more negative perception of their spinal conditions, increased catastrophization of their medical condition, and decreased pain improvement and poorer functional status; these represent good evidence that MRIs performed too early or without good reason carry real and measurable ill effects. As the editorial suggests, the problem is not limited to low back pain; the same general findings likely apply to other orthopaedic conditions, including shoulder pain, neck pain, and others [5].

I think that the major ill-effect of inappropriately using MRIs in patients with benign, chronic conditions is the diagnostic label that it provides to a symptom. This test turns a patient who previously had occasional back pain into someone with a degenerated disc, a compressed nerve root, or an annular rupture; these frightening labels may push the patient toward unhelpful treatment. In fact, the studies I referenced earlier suggest this occurs with alarming frequency. This is only aggravated by another common modern phenomenon: so-called “cyberchondria,” which has been defined as the heightened distress caused by excessive searches of the internet for medical information [1]. I’ve seen this lead to a negative spiral of self-diagnosis, increased distress levels, progressive symptom severity, and functional impairment. Patients with this kind of distress often will continue to seek many different forms of medical and surgical therapies, and physicians will—often inappropriately—provide them, leading to paradoxically poorer health even as healthcare utilization increases.

So what is the solution? The point is not only whether an MRI is performed, but also when in the course of a patient’s symptoms that it is done. In spine surgery, we usually can arrive at a correct clinical diagnosis simply by taking a good history and performing a thorough clinical examination. If there are no red flag signs of spinal surgical emergency, then we generally ought to begin with effective, evidence-driven nonsurgical treatment, along with empathic reassurance. MRI should be reserved for those few patients in whom symptoms progress, or when there is real doubt about an important clinical diagnosis. The alternative: Routine use of MRI early in a patient’s clinical course, and before nonsurgical therapy has been tried, may result both in harm to the patient and needless costs to the medical system.