Hallux valgus and metatarsus adductus measurements: inter-reader reliability and correlations on radiographs and MRI

R. Dessouky, N. Heineman, L. Zhang, J. Hummel, J. Skweres, D. Wukich, A. Chhabra

Research output: Contribution to journalArticle

2 Scopus citations


Aim: To assess inter-reader reliability of metatarsus adductus (MA) using the traditional method and Engel's angle (EA) on radiography and magnetic resonance imaging (MRI) and assess correlations with hallux valgus (HV). Methods and materials: Ninety consecutive patients with radiographs and MRI of the foot were included. Two readers measured HV angle (HVA), traditional metatarsus adductus angle (MAA), and EA on radiographs and HVA and EA on MRI. Three- and two-way mixed model analyses were used for reader agreements. Ninety-five percent bootstrap confidence intervals were calculated. The linear mixed model was used for association between HVA and EA/MAA. Results: Mean age and male to female ratio was 54.2±15.4 and 0.4:1, respectively. Mean HVA and EA were 20.6±9.4 and 21.2±8, 21.2±8.3 and 22.4±7.5 on radiographs and MRI, respectively. Mean MAA was 18.5±5.7 on radiographs. Inter-reader agreement was good for EA (ICC=0.73, 0.6) and moderate for MAA (ICC=0.41). Positive correlations between HVA, MAA, and EA on radiographs and MRI were found, but none were statistically significant (p=0.44 and 0.87). Conclusion: Engel's angle is more reproducible. Although positive correlations exist between the degrees of HV and MA, they are not statistically significant.

Original languageEnglish (US)
JournalClinical Radiology
Publication statusAccepted/In press - Jan 1 2018


ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this