Role of High-Resolution Computed Tomography in Double-Lumen Tube Selection for Patients Undergoing Minimally Invasive Coronary Bypass.

Background

No consensus guideline exists for double-lumen tube (DLT) size selection, which conventionally relies on height, sex, and chest radiography. Computed tomography (CT) has been proposed to refine sizing, particularly in women and patients of small stature. We examined whether high-resolution CT (HRCT) measurement of the trachea and main bronchi improves DLT size selection beyond a height- and sex-based estimate in patients undergoing minimally invasive coronary artery bypass (MICS-CABG). The aim of this study was to determine whether HRCT measurement of the trachea and main bronchi improves DLT size selection beyond a height- and sex-based estimate in patients undergoing MICS-CABG.

Methods

A total of 140 patients were retrospectively analyzed who had undergone MICS-CABG and had HRCT within 30 days of surgery. Tracheal anteroposterior and transverse diameters and left and right main bronchus diameters were measured in lung, mediastinal, and bronchial windows (mean of six readings by two observers). The DLT size used was related to each parameter through Spearman correlation; the agreement between the height/sex estimate and the size used was determined using Cohen's kappa; discrimination for a larger tube (≥39 Fr) was evaluated through receiver operating characteristic analysis; and the incremental value of imaging was assessed using leave-one-out cross-validation.

Results

In total, 82.1% of patients were male (mean age 59.7 ± 10 years; height 171.7 ± 7.3 cm). The height- and sex-based estimates correlated most strongly with the size used (ρ = 0.885, p < 0.001), followed by height alone (ρ = 0.844, p < 0.001); the best imaging measure, the lung-window left main bronchus diameter, was weaker (ρ = 0.503, p < 0.001), and body mass index showed no association (ρ = -0.008, p = 0.926). The estimate matched the size used exactly in 85.7% of patients and was within one size for all patients (linear-weighted κ = 0.831). Height best discriminated a larger tube (AUC = 0.969). Adding the bronchial diameter to height and sex did not improve cross-validated prediction (exact match 80.7% with and without; AUC 0.972 versus 0.973). The mediastinal window yielded systematically larger calibers than the lung and bronchial windows.

Conclusions

HRCT airway measurement did not improve the prediction of the double-lumen tube size selected in routine practice beyond a simple height- and sex-based formula. Because the reference standard was the size clinicians actually placed rather than an independently validated optimum, these data argue against routine preoperative HRCT undertaken solely to predict tube size; they do not exclude a benefit of HRCT when selection is judged against clinical outcomes. CT remains valuable when abnormal airway anatomy is suspected.