Low interobserver agreement among expert surgeons using a common morphological classification for pectus excavatum.

BACKGROUND

Since the initial publication of a visual classification of pectus excavatum in 2006-later revised in 2013-its use has gained widespread international popularity. This system categorizes pectus excavatum into three morphological types: cup-shaped, saucer-shaped, and trench-shaped. However, concerns remain regarding its reproducibility. We therefore conducted a two-stage evaluation of anonymized chest wall photographs: first within our surgical team, and subsequently with a panel of internationally recognized chest wall surgeons, to assess interobserver consistency.

METHODS

Comparative study assessing interobserver agreement among six thoracic surgeons specialized in chest wall malformations, using a blinded form with standardized medical photographs of 100 pectus excavatum patients. In a second phase, a simplified version of the form with ten patients was assessed by 17 members of the Executive Board of the Chest Wall International Group. Levels of agreement were analyzed. Further evaluation included the assessment of the results in relation to thoracic indices.

RESULTS

In the first phase, complete agreement among all six surgeons was observed in only 30 % of the cases. High agreement (≥80 % observers) was achieved in 49 %. Cases with high agreement had a significantly lower Titanic Index (64.1 ± 22.4 % vs. 78.1 ± 15.5 %, p < 0.0001), suggesting a relationship between localized deformities and greater classification consistency. In the second phase, consensus among international experts was even lower, with only 6 % achieving complete agreement.

CONCLUSIONS

Our findings highlight substantial interobserver variability in the classical pectus excavatum visual classification, even among experts. Relying solely on a classification based on surface anatomy may be inadequate for guiding treatment decisions, underscoring the need for objective metrics. Employing quantitative imaging indices such as the Titanic Index could enhance reproducibility and support more consistent clinical decision-making.

TYPE OF STUDY: Retrospective comparative study.

LEVEL OF EVIDENCE: III.