Validation and Reliability of the Progressive Collapsing Foot Deformity (PCFD) Classification

CATEGORY: Midfoot/Forefoot; Other INTRODUCTION/PURPOSE: Recently, a group of experts proposed Progressive Collapsing Foot Deformity (PCFD) as a new terminology to describe better the complexity of this pathology, as well as a new staging system based on rigidity and different patterns of deformity. The proposed system (Table 1) better incorporates recent data and understanding of the condition, better allowing for standardization of reporting, and guiding treatment decision-making to achieve optimal outcomes. The classification includes two Stages: flexible (1) and rigid (2) deformity each of which may or may not be associated with five subclasses of deformity including hindfoot valgus (A), midfoot abduction (B), forefoot supination (C), peritalar subluxation (D), and ankle valgus instability (E). This paper aims to determine the reliability and the validity of the PCFD classification system. METHODS: This is a survey-based study distributed through REDCap. Participants were provided with video and case examples explaining how the classification system is used. Three groups of surgeons (Group 1 fellows in training, Group 2 surgeons in practice for 1-4 years, and Group 3 in practice 5 years or more) classified 20 cases of PCFD. Correct response rates for the 20 cases, and subclasses and substages of each individual case were calculated. The mean and 95% confidence interval were used to describe normally distributed numerical data. Values of the above parameters were compared among the three different groups using Tukey-Kramer HSD test of Oneway ANOVA. P<0.05 was considered statistically significant. RESULTS: Ninety-four of 113 returned surveys were completed and included in the analysis. There were 18 in Group 1, 23 in Group 2, and 53 in Group 3, with significantly different overall correct diagnosis rates (75.8%, 51.7%, and 37.6% respectively). There was no significant difference among the three groups with regard to the correct response ratio of different Stages. The correct response rates for each Class of deformity were 96.62% (95% CI=94.02%, 99.23%) for Class A (hindfoot valgus); 77.66% (95% CI=75.06%, 80.26%) for Class E (ankle instability); 87.98% (95% CI=85.38%, 90.58%) for Class B (midfoot abduction), 87.99% (95% CI=85.39%, 90.60%) for Class C (forefoot supination) and 57.29% (95% CI=54.70%, 59.90%) for Class D (peritalar subluxation), with the correct response rates of Class A, B, C and E being significantly higher than that of Class D. CONCLUSION: Three groups of surgeons with varied experience demonstrated a higher accuracy of responses for hindfoot valgus, and ankle instability and a lower accuracy for midfoot abduction, forefoot supination and peritalar subluxation. The classification as presented to the reviewers provided sufficient information to guide accurate decision making for correct responses, however some classes of deformity, in particular peritalar subluxation may have been poorly defined in the clinical presentations and may require more clinical and radiographic information to guide decision making for correct classification.