Beyond the Chest X-Ray: Comparative Performance of Preoperative Tests for Phrenic Nerve Dysfunction in Brachial Plexus Injuries

Background: The phrenic nerve, arising from C3-C5 and innervating the diaphragm, is frequently affected in brachial plexus injuries (BPI). Given its role in respiration, preoperative assessment is essential to anticipate respiratory compromise in high-risk patients (e.g., elderly, obese, pediatric, smokers) and to determine its suitability as a donor nerve in selected cases. Traditional tests such as inspiratory-expiratory chest radiographs (CXR) are commonly used but have limited sensitivity. More recent modalities-including chest sonography, pulmonary function testing (PFT), and magnetic resonance neurography (MRN)-have been described, though their relative performance in BPI remains unclear. This study systematically compared these methods against intraoperative phrenic nerve stimulation as the reference standard. Methods: A retrospective review was performed of patients undergoing brachial plexus reconstruction between 2020 and 2024. Inclusion required documented intraoperative phrenic nerve stimulation and at least one preoperative diagnostic test. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) were calculated. McNemar's test assessed diagnostic agreement, DeLong's test compared AUCs, and logistic regression identified predictors of phrenic nerve injury. Results: Among 113 patients, chest sonography provided the highest diagnostic accuracy, with 100% sensitivity, 80% specificity, and an AUC of 0.900. CXR showed the greatest specificity (93%) but low sensitivity (44%), yielding an AUC of 0.687. PFT demonstrated moderate sensitivity (86%) and low specificity (58%), with an AUC of 0.718. MRN performed poorly: MRN of the injured C4 root showed 22.2% sensitivity, 84.4% specificity, and an AUC of 0.533, while MRN of the injured C5 root yielded high sensitivity (87%) but very low specificity (6.5%), with an AUC of 0.467. Conclusion: Chest sonography provides the most reliable assessment of phrenic nerve function in BPI. Although CXR offers high specificity, its low sensitivity limits its use as a standalone test. A combined approach with sonography and CXR may enhance diagnostic accuracy.