IADR Abstract Archives
Reliability and Validity of ABO Discrepancy Index and US-PAR Index
Objectives: To assess the reliability and validity of the American Board of Orthodontics Discrepancy Index (DI) and the United States Weighted Peer Assessment Rating (US-PAR) Index for malocclusion severity evaluation.
Methods: A stratified random sample of 120 finished cases gathered from six university orthodontic centers by performing the Angle Classification as a stratified factor. Sixty-nine orthodontists rated malocclusion severity on a five-point scale by assessing a full set of pretreatment records (study casts, lateral cephalometric radiographs, panoramic radiographs, and facial photographs) on each case and listed at least three reasons that affected their assessment. Three calibrated raters assessed DI and US-PAR Index components using published guidelines. Cronbach’s generalizability method was performed to analyze inter-examiner reliability. Bivariate correlations were applied to compare clinician judgment and the DI or US-PAR Index.
Results: Excellent inter-examiner reliability of clinician judgment, DI, and US-PAR Index was demonstrated by Intra-class correlation coefficient (rho= 0.995, 0.990 and 0.964, respectively). DI and US-PAR Index showed similarly good correlation with clinician judgment (r=0.700 and 0.707, respectively). Variabilities among different Angle classifications were found: the DI showed the highest correlation with clinician judgment in Class II patients (r=0.780), whereas the US-PAR index showed the highest correlation with clinician judgment in Class III patients (r=0.710). Both indices demonstrated the lowest correlations with clinician judgment in Class I cases (r=0.518 for DI and r=0.588 for US-PAR Index).
Conclusions: With a strong inter-examiner agreement, the panel consensus was used for validating the DI and US-PAR Index for malocclusion severity evaluation. Overall, the DI and US-PAR index were reliable for measuring malocclusion severity with significantly variable weightings for different Angle classifications. Further modification of the indices on different Angle classification may be indicated.
Methods: A stratified random sample of 120 finished cases gathered from six university orthodontic centers by performing the Angle Classification as a stratified factor. Sixty-nine orthodontists rated malocclusion severity on a five-point scale by assessing a full set of pretreatment records (study casts, lateral cephalometric radiographs, panoramic radiographs, and facial photographs) on each case and listed at least three reasons that affected their assessment. Three calibrated raters assessed DI and US-PAR Index components using published guidelines. Cronbach’s generalizability method was performed to analyze inter-examiner reliability. Bivariate correlations were applied to compare clinician judgment and the DI or US-PAR Index.
Results: Excellent inter-examiner reliability of clinician judgment, DI, and US-PAR Index was demonstrated by Intra-class correlation coefficient (rho= 0.995, 0.990 and 0.964, respectively). DI and US-PAR Index showed similarly good correlation with clinician judgment (r=0.700 and 0.707, respectively). Variabilities among different Angle classifications were found: the DI showed the highest correlation with clinician judgment in Class II patients (r=0.780), whereas the US-PAR index showed the highest correlation with clinician judgment in Class III patients (r=0.710). Both indices demonstrated the lowest correlations with clinician judgment in Class I cases (r=0.518 for DI and r=0.588 for US-PAR Index).
Conclusions: With a strong inter-examiner agreement, the panel consensus was used for validating the DI and US-PAR Index for malocclusion severity evaluation. Overall, the DI and US-PAR index were reliable for measuring malocclusion severity with significantly variable weightings for different Angle classifications. Further modification of the indices on different Angle classification may be indicated.