Discussion: Important Details in Performing and Interpreting the Scratch Collapse Test
It is one thing not to understand the mechanism of the scratch collapse test—it may simply be that we lack the requisite knowledge of neurophysiology. Medical history is rife with such misapprehensions, a classic example being the idea that peptic ulcers should be treated with antibiotics, scorned as irrational until we saw that bacteria were causative. Even today, we do not fully understand the cause of such mundane diagnoses as ganglion cyst, trigger finger, and osteoarthritis, and yet we readily treat them, sometimes with corticosteroids, the mechanism of which we also do not fully understand.
However, in the case of the scratch collapse test, the incomprehensibility seems insurmountable. The test is based on the shoulder’s external rotators, which are somehow inhibited when the skin overlying a diseased nerve is lightly stimulated. Why the teres minor and infraspinatus would hold such revelatory power is a mystery, but it is conceivable that some yet-undiscovered feedback loop is at work. (There are indeed some pretty crazy feedback loops in nature—such as making an alligator fall asleep by rubbing its belly.) Still, that these muscles would reveal any nerve compression anywhere in the upper extremity, or even the neck, strains credulity. It is evidently these muscles alone that react this way—the authors make a point to isolate the external rotators and eliminate the action of the deltoid, even though the deltoid contributes to external rotation and shares a close brachial plexus lineage with the teres minor and infraspinatus. Are the teres minor and infraspinatus so special that they can reveal neuropathology in the lower extremity? Why not?
However, the more puzzling component of the scratch collapse test is not the “collapse,” but the “scratch.” When we lightly stroke the skin over, say, the cubital tunnel, we are stimulating the medial brachial and/or antebrachial cutaneous nerve, not the ulnar nerve. The same can be said for the carpal tunnel, when we stimulate the palmar cutaneous branch of the median nerve, the palmar cutaneous branch of the ulnar nerve, and/or the lateral antebrachial cutaneous nerve. So now we have a cutaneous nerve that is not diseased sending a feedback loop to the external rotators because a nerve underneath it—on the other side of a thick, fibrous structure—is diseased. Moreover, according to the authors, these presumably normal cutaneous nerves are so discriminating they can distinguish not just, say, cubital tunnel syndrome, but compression at the proximal versus the distal cubital tunnel.
At first glance, the masking effect of ethyl chloride might seem to support this mechanism. However, this may be the most perplexing claim of all. The authors use the scratch collapse test to diagnose pronator syndrome, but advise the use of firm pressure, as light touch is evidently not stimulating enough to trigger a response when the diseased nerve is so deep, and yet they are able to eliminate this response with a quick spray of ethyl chloride. They cannot have it both ways—either it is the overlying cutaneous nerve sending the inhibitory signal, in which case the spray might have a masking effect; or it is the underlying diseased nerve, in which case the spray should have no effect.