SCIENTIFIC SECTION An evaluation of the Pi analysis in the assessment of anteroposterior jaw relationship Santosh Kumar, Ashima Valiathan, Pawan Gautam, Kalyana Chakravarthy and Priyanka Jayaswal Manipal College of Dental Sciences, Manipal, Manipal University Manipal, Karnataka, India Objective: This study has investigated two new cephalometric variables, the Pi angle and Pi linear in the evaluation of anteroposterior skeletal discrepancy. Design: Retrospective cross-sectional study. Setting: Manipal College of Dental Sciences, Manipal, India Subjects and method: A sample of 155 subjects (mean age 19.7 years) were subdivided into skeletal class I, II and III groups based upon ANB angle. Descriptive data were calculated for each variable and group. Receiver operating characteristics curves were used to examine sensitivity and specificity of the Pi angle in the discrimination between different skeletal groups. Correlation coefficients were obtained for each of the parameters to compare their relationship with other parameters in the class I group. Coefficient of determination, regression coefficient, regression equation and standard error of estimate were also calculated from the parameters showing significant correlation with the Pi angle. Results: Mean values for the Pi angle in skeletal class I, II and III subjects were 3.40 (¡2.04), 8.94 (¡3.16) and 23.57 (¡1.61) degrees, respectively. For the Pi linear they were 3.40 (¡2.20), 8.90 (¡3.56) and 23.30 (¡2.30) mm for class I, II and III subjects, respectively. Receiver operating characteristic curves showed that a Pi angle greater than 5 degrees had 89% sensitivity and 82% specificity for discriminating a skeletal class II group from class I. A Pi angle of less than 1.3 degrees had 100% sensitivity and 84% specificity in discriminating skeletal class III groups from class I. The overall accuracy for discriminating class II groups from class I was 85% and for class III from class I, 90%. Thus, a cut-off point between class I and II groups could be considered a Pi angle of approximately 5 degrees and between class I and class III, approximately 1.3 degrees. There were no statistically significant correlations found between Pi angle and ANB (0.07), Beta angle (20.04) and WITS analysis (0.19). The highest level of correlation was obtained for the Pi angle and Pi linear (0.96). Conclusion: The anglar and linear components of the Pi analysis are a suitable method for assessing anteroposterior jaw discrepancy in daily clinical practice. Key words: Anteroposterior discrepancy, cephalometrics, Pi analysis Received 15 July 2011; accepted 28 June 2012 Introduction An accurate anteroposterior (AP) measurement of the jaw relationship is critically important in orthodontic diagnosis and treatment planning. Since Broadbent’s introduction of the cephalostat in 1931, a number of geometrical parameters, such as the ANB angle, 1 WITS analysis, 2 AF-BF, 3 APDI, 4 and Beta angle 5 have been defined and used effectively for the evaluation of AP discrepancies affecting the apical bases of the jaws. However, none of the parameters currently defined give a definitive picture of the AP relationship. Any cephalometric analysis based on either angular or linear measurements has obvious shortcomings, which have been comprehensively discussed in the literature. 6 In particular, various cranial reference planes, such as the Frankfort horizontal and Sella-Nasion line have been used in the determination of jaw dysplasia; however, measurements related to the cranium do not provide a wholly reliable estimation of the AP jaw relationship within the dentofacial complex. 7 The line of reference from which the AP jaw relationship is evaluated, should Journal of Orthodontics, Vol. 39, 2012, 262–269 Address for correspondence: Santosh Kumar, Assistant Professor, Manipal College of Dental Sciences, Manipal University Manipal, Karnataka, India. Email: drsantoshortho@gmail.com # 2012 British Orthodontic Society DOI 10.1179/1465312512Z.00000000039