Background. Diagnostic blocks are used in different ways for the diagnosis of spinal pain, but their validity has not been fully evaluated.
Methods. Four clinical protocols were analyzed mathematically to determine the probability of correct responses arising by chance. The complement of this probability was adopted as a measure of the credibility of correct responses.
Results. The credibility of responses varied from 50% to 95%, and was determined less by the agents used but more by what information was given to patients and if the agents were fully randomized for each block.
Conclusions. Randomized, comparative local anesthetic blocks offer a credibility of 75%, but randomized, placebo-controlled blocks provide a credibility of 95%, and are thereby suitable as a criterion standard for diagnostic blocks.