AbstractThe aim of this study was to estimate the sensitivity and specificity of palpation relative to necropsy for detection of keel bone fractures (KBF) in broiler breeders. In addition, the study investigated observer reliability of palpation. Four commercial breeder flocks (Ross 308 n = 2, Ranger Gold n = 1, Hubbard JA 757, n = 1) were included in the study. A total of 400 hens and 60 roosters were examined after culling or slaughter, at end of lay (63 weeks). Keel bones were first palpated independently by two raters with a varying degree of experience in palpation of keel bones in laying hens. After palpation, a necropsy with investigation of the keel bone was conducted. The prevalence of KBF at end of lay for hens, as diagnosed by necropsy, varied from 14 to 58%, while only one in 60 assessed roosters had KBF. Palpation had poor sensitivity, with differences between raters ranging from 56% to 36% when hybrid is not considered. The specificity of palpation for all hybrids together was 85% and 88%, for the two raters respectively. The two raters had moderate agreement, Gwet's agreement coefficient (95% confidence interval): 0.60 (0.52-0.69). In conclusion, palpation has poor diagnostic accuracy for detection of KBF in broiler breeders, and other methods, like necropsy, must be applied for reliable assessments. KBF appears to be a prevalent condition in different hybrids of broiler breeder hens, but not in roosters.