Physician Estimated Depth: Can It Reduce Unnecessary CT Scan Evaluation of Pectus Excavatum?

BACKGROUND: The severity of pectus excavatum is determined by computed tomography (CT) derived indices such as the Haller Index (HI) and the Correction Index (CI). Physician estimated depth (PED) as an alternative measurement may preclude for CT. We retrospectively evaluated PED as a screening tool to identify surgical candidates. METHODS: Patients with a diagnosis of pectus excavatum between 1/1/2009 and 3/30/2018 were extracted from the electronic health record for review. Patients without available imaging were excluded. HI and CI were calculated from CT images. CT derived measurements acted as an approximation of PED. Using ROC analysis, we estimated the optimal PED cut-off for identifying surgical candidates according to an HI ≥ 3.25. RESULTS: A total of 94 patients were identified and all met inclusion criteria. Patients were predominantly males (82%) with a median age of 15 (IQR=14-16). Almost half (46%) were underweight (BMI < 18.5). The median HI was 4.1 (IQR=3.7-5.1) with 89% ≥ 3.25. The median CI was 39 (IQR=29-47). CI was highly correlated with HI (r=0.77, area under the curve, AUC=0.88). A CI of 27 was best at correctly classifying HI above/below 3.25 (sensitivity, Se=89%, specificity, Sp=90%). Median PED was 2.5 (IQR=2.1-3.0). Overall, a PED of 2cm correctly classified 86% of HI above/below 3.25 (Se=88%, Sp=70%, AUC=0.84). Among underweight patients, ROC parameters improved (93% correctly classified, Se=95%, Sp=75%, AUC=0.92). PED was highly correlated with CI (r=0.77, AUC=0.94). Among underweight patients, a PED of 2cm correctly classified 95% of CI above/below 27 (Se=97%, Sp=80%, AUC=95%). CONCLUSION: PED over 2cm can accurately identify patients who require CT imaging and pectus correction. Our findings show that PED may be employed easily in the clinic as a screening tool, thereby minimizing unnecessary CT scans. A prospective evaluation of PED is underway at our center.