Urgent findings in CT chest examinations: Radiography and medical students' capabilities and learning perspectives - a preliminary study.

Abstract

Introduction: Radiography and medical students (RMS), upon graduation, require capabilities to provide life-saving care through identification and communication of urgent findings on radiological imaging. This preliminary study investigated RMS' ability to identify and categorise urgent findings on CT examinations. It also explored their experiences of image interpretation education.

Methods: A sequential explanatory mixed-methods study was employed. Participants were students who had recently completed Year-4 (radiography/medicine) or Year-2 (medicine-only) from three Australian universities. Urgent finding identification capabilities were assessed via a test-set of 10 CT examinations with a range of findings (normal, abnormal but non-urgent, abnormal but urgent) that was developed on a validated learning and assessment platform. Each case required selection of: normal versus abnormal; where abnormal, finding/s from a pre-defined list; urgency level; self-rated confidence. Learning experiences were then explored via a survey (Likert statements, free-text responses).

Results: Mean sensitivity, specificity and accuracy (with min‒max ranges), respectively, were: All participants [n = 30] 0.87 (0.60-1.0), 0.63 (0.40-1.0), 0.75 (0.50-1.0); 4th-year radiography [n = 17] 0.89 (0.6-1.0), 0.66 (0.4-1.0), 0.78 (0.60-1.0); 4th-year medicine [n = 8] 0.85 (0.6-1.0), 0.50 (0.4-0.6), 0.68 (0.50-0.8); 2nd-year medicine [n = 5] 0.80 (0.60-1.0), 0.75 (0.75-1.0), 0.78 (0.78-0.78). False positives were highest for cases with non-urgent abnormalities in all groups. Free-text responses revealed students' desire for dedicated urgent finding educational resources with high 'repetitive learning' effects.

Conclusions: RMS demonstrated considerable performance in identifying urgent abnormalities as a group, but individual capabilities varied from pass-level to perfect. Together, participants demonstrated limited ability to correctly classify non-urgent CT chest abnormalities.

Implications for practice: This study highlights opportunities for targeted urgent-finding resource development, focused on improving consistency within cohorts and reducing false positive rates.