Physician-Estimated Depth as a Screening Tool for Computed Tomography Evaluation of Pectus Excavatum

Pectus excavatum (PE) is the most common congenital chest wall anomaly with a reported incidence of 1/300 to 1/400 live births and a male predominance. Preoperative evaluation of defect severity typically requires a calculation of the Haller index (HI) and/or correction index (CI) using computed tomography (CT) or x-rays. The purpose of this study was to determine whether physician-estimated depth (PED), a bedside screening tool, could be used to identify a subset of pediatric patients in whom CT was unnecessary.

After institutional review board approval (IRB #032018-091), we retrospectively reviewed all patients with a diagnosis of PE between 2009 and 2018 at our academic pediatric center. Demographic information including age, sex, and body mass index were abstracted. Imaging was reviewed to obtain HI and CI and to retrospectively calculate PED. The PED is calculated at the bedside by measuring the depth of the pectus at the site of greatest depression relative to a horizontal surface laid across the deformity. For this retrospective study, we calculated the CT-derived PED by measuring the depth from the horizontal on the respective CT images. Patients without imaging studies and patients with pectus carinatum, arcuatum, or mixed deformities were excluded from this study.

A total of 94 patients met inclusion criteria. Of these, 82% were male, with a median age of 15 y. Patients were further subdivided by BMI, with 46% of patients having a BMI of <18.5 kg/m² (i.e., underweight), whereas 54% of patients had a BMI of ≥18.5 kg/m². Using a threshold PED of 2 cm, patients with a BMI of <18.5 kg/m² had correct classification rates of 93% and 95% using PED relative to HI and CI, respectively. Patients with a BMI of ≥18.5 kg/m² had correct classification rates of 80% and 88% using PED relative to HI and CI, respectively, at the same 2 cm threshold.

PED is a viable screening tool for the preoperative evaluation of PE with a 2 cm threshold providing the combination of high sensitivity, specificity, and correct classification rates especially in underweight patients.