Visual diagnosis of pectus excavatum: An inter-observer and intra-observer agreement analysis

Abstract

Background/purpose: Among patients suspected of pectus excavatum, visual examination is a key aspect of diagnosis and, moreover, guides work-up and treatment strategy. This study evaluated the inter-observer and intra-observer agreement of visual examination and diagnosis of pectus excavatum among experts.

Methods: Three-dimensional surface images of consecutive patients suspected of pectus excavatum were reviewed in a multi-center setting. Interactive three-dimensional images were evaluated for the presence of pectus excavatum, asymmetry, flaring, depth of deformity, cranial onset, overall severity and morphological subtype through a questionnaire. Observers were blinded to all clinical patient information, completing the questionnaire twice per subject. Agreement was analyzed by kappa statistics.

Results: Fifty-eight subjects with a median age of 15.5 years (interquartile range: 14.1-18.2) were evaluated by 5 (cardio)thoracic surgeons. Pectus excavatum was visually diagnosed in 55% to 95% of cases by different surgeons, revealing considerable inter-observer differences (kappa: 0.50; 95%-confidence interval [CI]: 0.41-0.58). All other items demonstrated inter-observer kappa's of 0.25-0.37. Intra-observer analyses evaluating the presence of pectus excavatum demonstrated a kappa of 0.81 (95%-CI: 0.72-0.91), while all other items showed intra-observer kappa's of 0.36-0.68.

Conclusions: Visual examination and diagnosis of pectus excavatum yields considerable inter-observer and intra-observer disagreements. As this variation in judgement could impact work-up and treatment strategy, objective standardization is urged.

Levels of evidence: III.

Keywords: Diagnosis; Inter-observer agreement; Intra-observer agreement; Pectus excavatum; Visual assessment.