The radio-opacity of fishbones.

Plain radiographs are often taken to localize fishbones which lodge in the pharynx. In this study a pig's neck was used to determine the radio-opacity of 14 different species of fishbones. Radiographs were taken with bones placed in up to four of the common sites of impaction and assessed on two occasions independently by two previously uninvolved ENT consultants. The visibility of the fishbones was scored. Inter and intra-observer agreement for species and site was determined. There was excellent interobserver agreement at both sessions for species (rs1 = 0.96, P < 0.0001, rs2 = 0.95, P < 0.0001) and site (rs1 = 0.98, P < 0.05, rs2 = 0.95, P congruent to 0.05). Intraobserver agreement for species of fish were similarly significant but not for site. The most opaque bones were cod, haddock, grey mullet and lemon sole. The site most often detected was the laryngopharynx and that most often missed was the tonsil. This study is clinically relevant because it is useful to know which species of fishbone are radio-opaque when clinical examination fails to reveal a foreign body and may clarify the decision to proceed to endoscopy.