# Improving Recognition and Reporting of Malignant Central Airways Obstruction on CT: The Need for Standardized Terminology and Structured Reporting

## TL;DR

This study shows that malignant central airway obstruction in lung cancer is often missed or poorly reported on CT scans, suggesting a need for standardized reporting methods.

## Contribution

The study identifies a lack of standardized terminology and structured reporting for malignant central airway obstruction in CT scans and proposes solutions to improve recognition.

## Key findings

- 13% of new lung cancer diagnoses had severe central airway obstruction on CT scans.
- Only 72% of these cases were reported in CT scan narratives, with inconsistent terminology used.
- The term 'central airways obstruction' was never used in any CT reports.

## Abstract

Background

Malignant central airways obstruction (CAO) is a clinically significant complication of lung cancer that can lead to severe morbidity if not promptly recognized and is frequently under-reported on CT.

Methodology

All new annual lung cancer baseline CT scans were retrospectively reviewed as part of an audit process in 2014, 2019, and 2020. CT scans were assessed by the interventional pulmonology team. Patients with radiological CAO, defined as obstruction >50% in the trachea, main bronchi, or bronchus intermedius, were assessed. Narrative reports were separated into the main text section, conclusion, and addendum sections. Key phrases describing CAO were then extracted for further analysis. Analysis was performed in R Studio, tidytext, and wordcloud packages.

Results

In total, 140/1,096 (13%) new diagnoses of lung cancer had severe CAO on admission. No difference was seen across the three time periods (45/342 (13%) in 2014, 43/370 (12%) in 2019, and 52/351 (15%) in 2020; P-valve = 0.448). Of these, 135/140 (96%) had available CT scan radiology reports. Patient performance status was >2 in 50/135 (37%).  The presence of CAO was reported on the index CT scan in 97/135 (72%) of cases. The dominant obstructive component was extrinsic in 70/135 (52%), with a median (interquartile range (IQR)) area loss due to obstruction of 86% (64%-100%). The words (including truncations) central, obstruction, and occlusion were used in 29%, 36%, and 16% of the main text reports, respectively. Extracted key phrases that described CAO contained a median of 16 (9.5-25) words. The phrase central airways obstruction, or CAO, was not used in any CT reports.

Conclusions

Unreported CAO on CT scans, with variable and non-standardized terminology when present, is a concern. Adopting a standardized, structured reporting approach and the refinement of trigger words may improve CAO reporting and recognition.