Methods: Data from three previous studies of people with dentine hypersensitivity were used (n=353): people who self-reported the condition and participants in two randomised controlled trials of desensitising dentifrices (one of active treatment vs. negative control and the other comparing two active treatments). The data were pooled then randomly divided into two halves. Short forms were derived using regression and item impact methods using the first half and tested using the second half of the data. Ten and 15 item short forms were derived by each method by selecting the best two and three items in each subscale.
Results: The 10 and 15 item versions of the regression short form detected impacts in 37% and 61% of participants respectively, compared to 68% and 93% using the item impact method. All short forms had an internal consistency (a) > 0.81 and test retest reliability (ICC) > 0.77. All correlated with the long form (all r > 0.93, P < 0.001) and with the effect of the mouth on everyday life (all r ≥ 0.64, P < 0.001). None detected a treatment effect in either of the two trials.
Conclusions: All short forms were satisfactory but those derived by the item impact method were more sensitive and offer advantages in detecting impacts of dentine hypersensitivity. The 15 item version is most sensitive and increases participant burden only marginally. The original long form of DHEQ remains the gold standard for use in trials.