Sara R Teixeira1, Vitor F Dalto2, Daniel A Maranho3, Orlando S Zoghbi-Neto4, José B Volpon5, Marcello H Nogueira-Barbosa6. 1. Division of Radiology, Department of Internal Medicine, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: steixeira@hcrp.usp.br. 2. Division of Radiology, Department of Internal Medicine, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: fdalto@gmail.com. 3. Division of Pediatric Orthopaedics, Department of Biomechanics, Medicine, and Rehabilitation of the Locomotor System, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: dacmaranho@gmail.com. 4. Division of Radiology, Department of Internal Medicine, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: zoghbi47@gmail.com. 5. Division of Pediatric Orthopaedics, Department of Biomechanics, Medicine, and Rehabilitation of the Locomotor System, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: jbvolpon@fmrp.usp.br. 6. Division of Radiology, Department of Internal Medicine, Ribeirao Preto Medical School, University of Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, Sao Paulo 14049-090, Brazil. Electronic address: marcello@fmrp.usp.br.
Abstract
PURPOSES: To evaluate whether the pubo-femoral distance (PFD) can be used as an accurate screening test to diagnose developmental dysplasia of the hip (DDH) in an at-risk population compared with the Graf method. Second, to determine whether PFD assessment is feasible and reproducible regardless of the observer's experience. MATERIALS AND METHODS: IRB approved this retrospective single-institution study. Written informed consent was waived. Between January 2010 and March 2012, 116 neonates at risk for DDH were included. Infants' hips were distributed into two groups according to recommendation for treatment: non-dysplastic (ND; Graf I/IIA; 211 hips; 69 females/37 males) and dysplastic hip (DH; Graf IIB/IIC/III/D/IV; 21 hips; 8 females/3 males). One resident and one experienced radiologist reviewed ultrasonography images performed in the fourth week. To compare the groups, Student's t and Mann-Whitney tests for normally and non-normally distributed covariates were performed. Accuracy of PFD to diagnose DDH was calculated. Intraclass correlation coefficient (ICC) was calculated to assess inter-observer agreement. RESULTS: Mean PFDs of ND group were 3.09mm at neutral position and 3.64mm with the hip flexed. Mean PFDs of DH group were 6.29mm and 7.59mm, respectively. Sensitivity, specificity, and accuracy of PFD were 94.4%, 93.4%, and 97.2% (cut-off=4.6mm) at neutral position and 94.4%, 89.0%, and 95.5% (cut-off=4.9mm) with hip flexed. ICCs were 0.852 and 0.864, respectively. CONCLUSIONS: PFD is comparable with Graf method, enabling physicians to differentiate patients who should undergo treatment from those who should not. PFD can be used as a screening tool for diagnosing DDH with high accuracy, even by inexperienced radiologists.
PURPOSES: To evaluate whether the pubo-femoral distance (PFD) can be used as an accurate screening test to diagnose developmental dysplasia of the hip (DDH) in an at-risk population compared with the Graf method. Second, to determine whether PFD assessment is feasible and reproducible regardless of the observer's experience. MATERIALS AND METHODS: IRB approved this retrospective single-institution study. Written informed consent was waived. Between January 2010 and March 2012, 116 neonates at risk for DDH were included. Infants' hips were distributed into two groups according to recommendation for treatment: non-dysplastic (ND; Graf I/IIA; 211 hips; 69 females/37 males) and dysplastic hip (DH; Graf IIB/IIC/III/D/IV; 21 hips; 8 females/3 males). One resident and one experienced radiologist reviewed ultrasonography images performed in the fourth week. To compare the groups, Student's t and Mann-Whitney tests for normally and non-normally distributed covariates were performed. Accuracy of PFD to diagnose DDH was calculated. Intraclass correlation coefficient (ICC) was calculated to assess inter-observer agreement. RESULTS: Mean PFDs of ND group were 3.09mm at neutral position and 3.64mm with the hip flexed. Mean PFDs of DH group were 6.29mm and 7.59mm, respectively. Sensitivity, specificity, and accuracy of PFD were 94.4%, 93.4%, and 97.2% (cut-off=4.6mm) at neutral position and 94.4%, 89.0%, and 95.5% (cut-off=4.9mm) with hip flexed. ICCs were 0.852 and 0.864, respectively. CONCLUSIONS: PFD is comparable with Graf method, enabling physicians to differentiate patients who should undergo treatment from those who should not. PFD can be used as a screening tool for diagnosing DDH with high accuracy, even by inexperienced radiologists.