Diagnostic Accuracy of Individual and Multimodal Pulp Sensibility Tests Compared With Radiographic Assessment for the Diagnosis of Pulpal Health: A Prospective Diagnostic Accuracy Study

Introduction: The diagnostic performance of individual pulp sensibility tests and multimodal diagnostic strategies for differentiating vital and necrotic pulps in teeth with endodontic-periodontal lesions remains uncertain. Therefore, this prospective diagnostic accuracy study evaluated and compared the diagnostic performance of individual pulp sensibility tests and multimodal diagnostic models with a composite reference standard. Materials and methods: Ninety-eight patients aged 18-65 years with endodontic-periodontal lesions underwent cold testing, heat testing, electric pulp testing, and pulse oximetry. The composite reference standard comprised radiographic assessment and intraoperative bleeding during the preparation of the access cavity. Diagnostic accuracy parameters such as sensitivity, specificity, predictive value, and area under the curve (AUC) were calculated for each test. Three multimodal models (parallel "OR," serial "AND," and weighted logistic regression) were developed and compared with the best-performing individual test using DeLong's test and McNemar's test. Results: A total of 98 patients were included in the study, comprising 52 (53.1%) males and 46 (46.9%) females, with 175 teeth evaluated. Pulse oximetry demonstrated the highest individual diagnostic accuracy (AUC = 0.905, sensitivity = 93.0%, specificity = 88.0%), followed by the electric pulp tester (AUC = 0.860), cold test (AUC = 0.825), and heat test (AUC = 0.750). The parallel model achieved the highest sensitivity (97.0%), whereas the serial model achieved the highest specificity (97.3%). The weighted logistic regression model provided the best overall performance (AUC = 0.933, sensitivity = 91.0%, specificity = 90.7%), significantly outperforming pulse oximetry (p = 0.032) and the parallel model (p < 0.001). Pulse oximetry emerged as the strongest independent predictor of pulp necrosis in multivariable analysis (adjusted OR = 11.13, p < 0.001). Conclusion: Pulse oximetry demonstrated the highest diagnostic accuracy among the individual tests evaluated for assessing pulpal health. Although the parallel and serial multimodal models achieved the highest sensitivity and specificity, respectively, the weighted logistic regression model showed the best overall diagnostic performance by integrating information from multiple tests. These findings suggest that a multimodal approach incorporating pulse oximetry, electric pulp testing, and thermal testing may improve diagnostic accuracy and facilitate more informed treatment planning in endodontic practice.