Objectives Low back pain cannot be overdiagnosed, at least not in the narrow sense of the word. However, it is a common symptom, and one that is often given a diagnostic label (slipped disc, pinched nerve, instability, arthritis, degeneration, and so on), despite there being no reliable way of determining the pain source in most cases. Once the symptom is labelled, however, the consequences begin to resemble those of overdiagnosis: many will experience no clinical benefit from receiving a diagnostic label, but will feel less well because of it, and are more likely to undergo costly, invasive treatments with questionable efficacy. We will present our current understanding of overdiagnosis of low back pain, discuss how low back pain might differ from other well-known examples of overdiagnosis, and detail efforts from within our research group and beyond to develop and evaluate solutions.

Method If one were to consider the broader definitions of overdiagnosis proposed by Carter et al. (BMJ 2015;350:h869), healthcare for low back pain would have examples abound: disease mongering (‘Pain as the 5th vital sign’ campaign by US Veteran’s Affairs), overutilisation (spinal injections, opioids), overdetection (diagnostic imaging), overtreatment (spinal fusion surgery, early physiotherapy), and false positives (red flags for serious pathology). In 2013 Americans spent US$81.6 billion on care for low back pain. How did we end up here?

Results Unlike other well-known examples, overdiagnosis of low back pain appears to have little to do with altering disease definitions or thresholds, or providing screening programs for the healthy. Some people with low back pain may receive no diagnosis but are overtreated.

Conclusions We argue that many of the problems with overdiagnosis and overtreatment of low back pain arise because people enter a health system that is set up to encourage inappropriate care and discourage appropriate care.