Initial Assessment of the Persistent Dentoalveolar Pain Disorder Screening Instrument

Objectives: Persistent dentoalveolar pain disorder (PDAP) is a poorly understood chronic pain disorder that is likely to be heterogeneous with regards to its underlying pathophysiology. To aid in studying this disorder, a brief screening instrument (PDAP screener) is needed that can identify patients in clinical and general populations. Therefore, the objective is to develop the PDAP screener and examine its psychometric properties.
Methods: Previous qualitative research was used to generate nine PDAP-specific items for the instrument. The full instrument included five expert-derived non-PDAP items to improve specificity. Initial assessment was performed using a simple bipolar ordinal scoring system with convenience samples of patients: PDAP (n=22); three pain conditions commonly confused with PDAP (temporomandibular disorders [TMD, n=41], trigeminal neuralgia [TN, n=21], acute dentoalveolar inflammation [ADI, n=42]); and pain-free controls (n=21). Patients were diagnosed using published criteria by boarded clinicians. The following analyses were conducted with the screener: internal reliability; exploratory principal factor analysis to determine factor structure; discriminant function analysis to examine which items could discriminate between PDAP and non-PDAP groups; and receiver operating characteristic (ROC) curve analysis for sensitivity and specificity.
Results: The nine PDAP-specific items were examined as a stand-alone set of items: Cronbach’s alpha=0.83; factor analysis demonstrated only one factor and the loading ranged 0.37-0.76; ROC area=0.71 (95%CI: 0.60-0.81). The full 14-item instrument had the following psychometric properties: one discriminant function which correctly classified 83% of cases versus non-cases; ROC area=0.74 (95%CI: 0.63-0.86). The sensitivity of the full scale using the simple bipolar scoring system was 0.73 (95%CI: 0.50-0.89) and the specificity=0.64 (95%CI: 0.55-0.72). Sensitivity and specificity improved when the standardized canonical discriminant function coefficients were used as weightings for their respective item: sensitivity=0.82 (95%CI: 0.60-0.95) and specificity=0.82 (95%CI: 0.74-0.89).
Conclusions: Initial assessment in a limited sample suggests the 14-item instrument has acceptable psychometric properties to undertake further validation.