A comparison of sagittal MRI and lateral radiography in determining the Insall–Salvati ratio and diagnosing patella alta in the pediatric knee

Abstract

The Insall–Salvati ratio is a technique for determining patellar height that relies on bony landmarks. Magnetic resonance imaging (MRI) and plain radiography are used interchangeably to assess the Insall–Salvati ratio in the pediatric population despite the lack of validity in the literature. The purpose of this study was to investigate if the Insall–Salvati ratio and patella alta as determined on MRI are comparable to those determined on radiography in pediatric patients. We conducted a retrospective review of 49 pediatric patients (age range: 7.5–17.0 years) with unfused growth plates who underwent both knee MRI and lateral knee radiography. Measurements for calculating the Insall–Salvati ratio (the ratio of patella tendon length to patella length) were obtained by three observers. Data were analyzed using paired t-tests and Pearson’s correlation. A reliability assessment and inter-method agreements were performed. Patella alta was defined as an Insall–Salvati ratio\u2009>\u20091.2. Additional cutoffs of Insall–Salvati ratios\u2009>\u20091.3 and\u2009>\u20091.4 were also analyzed. There was no statistically significant difference between Insall–Salvati ratio as determined on MRI (mean: 1.20) and radiographs (mean: 1.25; P\u2009>\u20090.05). There was a strong correlation between Insall–Salvati ratio as determined on MRI and radiographs (Pearson’s r\u2009=\u20090.6) with moderate consistency (Cronbach’s alpha\u2009=\u20090.78). There was a good level of agreement between the diagnosis of patella alta on MRI and radiographs when defined as an Insall–Salvati ratio greater than 1.2 and 1.3 (Cohen’s kappa\u2009=\u20090.61). The results demonstrate a strong association between Insall–Salvati ratio and patella alta derived from MRI and radiographs in children ages 7.5 years and older.