I thank Burt and Hales for their thoughtful comments. It is unfortunate that they assumed that our study1 was flawed by detection bias, even before seeing the data related to their assertion.
They believe that I did not search for concurrent medical diseases in a uniform fashion, without respect to CTS diagnosis. Here is the relevant information: Excluding those patients with an acute traumatic injury, 100% of the 123 patients that I diagnosed as having CTS underwent either radiography (92.7%) of symptomatic areas or laboratory studies (95.9%). Of those not diagnosed as having CTS, 98.7% underwent either x-ray (89.3%) or laboratory tests (87.3%). Only 2 patients did not undergo any studies; in both, their symptoms had nearly resolved by the time of my examination. The evaluation process was the same in all patients: x-ray films were reviewed at the time of the examination, and laboratory tests were then ordered as necessary. The same basic laboratory tests were obtained regardless of CTS diagnosis because conditions, such as arthritis, diabetes mellitus, and hypothyroidism, commonly cause rheumatic complaints in people with no symptoms suggestive of CTS. There is further evidence against significant detection bias in the 43 concurrent diseases that we found in the 170 patients that I did not think had CTS; in virtually all these cases, laboratory or x-ray testing was necessary to confirm a diagnosis.
Atcheson SG. Work-Related Carpal Tunnel Syndrome: Fix the Jobs; Don't Blame the Workers. Arch Intern Med. 1999;159(12):1371-1373. doi: