RESUMO
OBJECTIVES: Plantar fasciitis (PF) is the most common cause of heel pain. Though PF is self-limited, it can develop into chronic pain and thus treatment is needed. Early and accurate prognostic assessment of patients with PF is critically important for selecting the optimal treatment pathway. Nevertheless, there is no scoring system to determine the severity of PF and no prognostic model in choosing between conservative or surgical treatment. The study aimed to develop a novel scoring system to evaluate the severity of plantar fasciitis and predict the prognosis of conservative treatment. METHODS: Data of consecutive patients treated from 2014 to 2018 were retrospectively collected. One hundred and eighty patients were eligible for the study. The demographics and clinical characteristics served as independent variables. The least follow-up time was 6 months. A minimal reduction of 60% in the visual analog scale (VAS) score from baseline was considered as minimal clinically important difference (MCID). Those factors significantly associated with achieving MCID in univariate analyses were further analyzed by multivariate logistic regression. A novel scoring system was developed using the best available literature and expert-opinion consensus. Inter-observer reliability and intra-observer reproducibility were evaluated. The appropriate cut-off points for the novel score system were obtained using receiver operating characteristic (ROC) curves. RESULTS: The system score = VAS (0-3 point = 1; 3.1-7 point = 3; 7.1-10 point = 5) + duration of symptoms (<6 months = 1; ≥1 6 months = 2) + ability to walk without pain (>1 h = 1; ≤1 h = 4) + heel spur in X-ray (No = 0; Yes = 2) + high intensity zone (HIZ) in MRI (No = 0; Yes = 2). The total score was divided in four categories of severity: mild (2-4 points), moderate (5-8 points), severe (9-12 points), and critical (13-15 points). Inter-observer agreement with a value of 0.84 was considered as perfect reliability. Intra-observer reproducibility with a value of 0.92 was considered as perfect reproducibility. The optimum cut-off value was 10 points. The sensitivity of predictive factors was 86.37%, 84.21%, 91.22%, 84.12%, and 89.32%, respectively; the specificity was 64.21%, 53.27%, 67.76%, 62.37%, and 79.58%, respectively; the area under curve was 0.75, 0.71, 0.72, 0.87, and 0.77, respectively. The Hosmer-Lemeshow test showed a good fitting of the score system with an overall accuracy of 90.6%. CONCLUSIONS: Based on prognostic factors, the present study establishes a novel scoring system which is highly comprehensible, reliable, and reproducible. This score system can be used to identify the severity of plantar fasciitis and predict the prognosis of conservative treatment accurately. The application of this scoring system in clinical settings can significantly improve the decision-making process.