Is a Three-component Video-based Version of the Foot Posture Index Valid for Assessing Pediatric Patients With Orthopaedic and Neurologic Foot Conditions?

Background: The Foot Posture Index-6 (FPI6) is an assessment of foot position that can be useful for patients with orthopaedic complaints. The FPI6 rates six components of foot position from -2 to +2, resulting in a total score on a continuum between -12 (severe cavus or supination) to +12 (severe planus or pronation). The subscores are ratings made by the examiner and are subjective assessments of deformity severity. The FPI6 requires palpation of bony structures around the foot and therefore must be administered live during physical examination. Because it is sometimes impractical to perform these assessments live, such as for retrospective research, a valid and reliable video-based tool would be very useful.

Questions/purposes: This study examines a version of the FPI using three of the original six components to determine: (1) Are scores from the three-component version of the FPI (FPI3) associated with those from the original six-component version (FPI6)? (2) Is the three-component FPI3 as reliable as the original six-component FPI6? (3) Are FPI3 assessments done retrospectively from video as reliable as those done live?

Methods: A retrospective group of 155 participants (106 males; mean age 13 ± 4 years) was studied. All had undergone gait analysis including videotaping and in-person assessment using the FPI6. Ratings for three components (calcaneus inversion/eversion, medial arch congruence, and forefoot abduction/adduction) were extracted yielding an FPI3 score ranging from -6 to +6. The other three components of the FPI6 (talar head palpation, curves above and below the lateral malleolus, talonavicular joint bulge) were excluded from the FPI3. FPI6 and FPI3 scores and side-to-side asymmetry were compared for all participants and for diagnosis subgroups (cerebral palsy and Charcot-Marie-Tooth disease) using a Pearson correlation. Agreement for foot posture categorization between the FPI6 and FPI3 was assessed using weighted kappa. Intra- and interrater reliability of live and video-based assessments for the FPI3 and its components were examined using intraclass correlation coefficients (ICCs) and Bland-Altman analysis.

Results: Scores from the FPI3 and FPI6 are highly associated with each other, suggesting the FPI3 is an adequate substitute for the FPI6. FPI6 and FPI3 scores (r = 0.98) and asymmetry (r = 0.96) were highly correlated overall and within the cerebral palsy (r = 0.98 for scores; r = 0.98 for asymmetry) and Charcot-Marie-Tooth (r = 0.96 for scores; r = 0.90 for asymmetry) subgroups (all p < 0.001). Agreement between the FPI6 and FPI3 was high for foot posture categorization (weighted agreement = 95%, weighted κ = 0.88; p < 0.001). Interrater reliability for live ratings was similar for FPI3 and FPI6 and high for both measures (ICC = 0.95 for FPI6 and 0.94 for FPI3; both p < 0.001). High reliability was seen in video versus live ratings for the FPI3 total score and each of its components regardless of whether they were performed by the same (ICC = 0.98) or different (ICC = 0.97) raters (both p < 0.001), and interrater reliability remained high when the FPI3 was scored from video recordings (ICC = 0.96; p < 0.001).

Conclusion: The FPI3 is valid and reliable when done live or from video or by the same or different examiners. It is suitable for retrospective and multicenter research studies, provided videos are done using standardized protocols. Further research is recommended investigating possible ceiling and floor effects in patients with pathologic conditions.Level of Evidence Level III, diagnostic study.