Abstract
Category:
Hindfoot; Midfoot/Forefoot
Introduction/Purpose:
Bluman et al. flatfoot classification (2007) is based on the posterior tibial tendon (PTT) rupture leading to a chronological appearance of several foot deformities. Since then, several discordances have been noted in this concept. An expert consensus met recently to update it. Emphasis on posterior tibial tendon rupture was shifted to a non-chronological approach focusing on five different independent foot and ankle deformations and each deformation's flexibility or rigidity. This concept, named Progressive Collapsing Foot Deformity (PCFD), was approved with a strong consensus. The aim of this study was to compare Bluman et al. and PCFD classifications. We hypothesize that both classifications will be intra and interobserver reliable and that the PCFD classification will allow a better distribution of the different types of foot.
Methods:
We performed a retrospective IRB approved study including 92 flatfeet. Three fellowship trained foot and ankle surgeons studied clinical information and X-rays and classified them in Bluman and PCFD classifications. One performed a blinded second assessment. Bluman classification was analyzed one time as initially described and a second time after removing the Angle of Gissane sclerosis sign. Bluman stage I represents isolated PTT dysfunction, stage II and III Flexible and Rigid Hindfoot Valgus, and stage IV Tibio-Talar Valgus. These stages are progressive and don't allow any combinations. PCFD classifies Hindfoot valgus (A), Midfoot Abduction (B), Forefoot Varus (C), Peritalar Subluxation (D), and Tibiotalar Valgus (E). Combinations of these deformities is allowed. Each deformity can be Flexible (1) or Rigid (2). Interobserver and Intraobserver reliabilities were determined with respectively unweighted Fleiss' and Cohen's kappa values. Descriptive analysis was performed on the 276 readings to highlight discrepancies between classifications.
Results:
Inter and Intraobserver reliabilities were respectively moderate (K=0.55) and substantial (K=0.62) for Bluman and moderate (K=0.56) and very good (K=0.85) for PCFD. The 276 readings were spread into 10 subgroups in Bluman and 64 in PCFD. 2.9% of the flatfeet were classified Bluman stage I, 31.5% stage II, 43.8% stage III and 21.7% stage IV. Bluman stages II and III were mainly composed of PCFD 1ABC (respectively 40.2% and 28.1%). The most represented Bluman subgroup was IIIB (32.6%) whereas after removing the Angle of Gissane sclerosis sign from the classification it was the IIC subgroup (44.2%). PCFD A, B and C were mainly composed of Bluman subgroup IIIB (respectively 35.7%, 43.6% and 36.2%) and PCFD D and E of Bluman subgroup IVB (31.9% and 73.3%).
Conclusion:
Both classifications showed moderate reliabilities although the PCFD represented 6 times as many different choices by readers. Bluman stage I was rare, possibly because it is based on PTT dysfunction with little or no deformity. This entity is no longer considered in the PCFD. Bluman stage III, assumed to represent rigid hindfoot varus included numerous flexible PCFD. Main confusion could come from the Angle of Gissane sclerosis sign, which leads to Bluman stage III. This sign is however a sign of extraarticular sinus tarsi impingement which should not lead to a triple arthrodesis as recommended in the Bluman classification.
