Clinical photographs as teaching aids: how reliable are they?

A study to investigate the value and reliability of clinical photographs as teaching aids was undertaken. Twenty colour photographs were taken using the StarMed video-otoscopic system. The pictures, which were a mixture of normal and abnormal ears, were shown to 21 experienced otolaryngologists from the UK and Canada. These clinicians were asked to identify the abnormality if any. The median score for correctly identified pictures was 15 (range 12-18). This score was identical for both the UK and Canadian subgroups. Although the abnormalities were consistently well recognized with an average correct identification rate of 90 per cent (range 67-100 per cent), the 'normals' were recognized significantly less well at only 41 per cent (range 5-71 per cent) (chi-squared = 110.6; 1 df; p < 0.001). This result is probably due to failure of the camera to capture the huge variation and subtleties in the range of normal, and the clinicians' natural inclination to identify pathology, when in doubt. We would conclude that as long as this failing is recognized, clinical photographs, and specifically those from the video-otoscope, represent a useful and reliable teaching tool.