Objective. To develop a protocol for determining when an individual is
adequately trained to locate the ten der points relative to fibromyal
gia in an exam. Methods. The error distance for each tender point was
established by polling individuals with experience in conducting tende
r point exams. Bayesian statistical methods were employed to form a pr
otocol for defer mining an individual's proficiency in locating the te
nder points. A predictive distribution was utilized to find the probab
ility of remaining trained at locating tender points. Also, the probab
ility of classifying at least 11 tender points as tender (mild) under
different ''locating'' criteria and different number of points that ar
e truly tender was computed. Results. Critical values indicating the n
umber of tender points needed in the qualification process for various
standards of reliability-80%, 85%, and 90%-are presented. To be certi
fied after 3 subjects have been examined in the 80%, 85%, and 90% crit
eria, one has to correctly identify 48, 50, and 52, respectively, out
of the 54 possible tender points. Conclusion. We believe that at least
3 subjects should be examined before certification is granted using a
ny of the 3 criteria-80%, 85%, and 90%. In our example, when using the
85% criterion, the qualification process required 7 subjects to certi
fy an individual.