Objectives:Treatment of borderline acetabular dysplasia is controversial. The existing literature lacks direct comparisons of different treatment approaches and focuses on lateral center edge angle (LCEA), failing to account for other important diagnostic characteristics. The purpose of this study was (1) to determine the most important characteristics in determining hip instability in this population, and (2) to develop a nomogram for clinical use and calculation of the Borderline Hip Instability Score (BHIS), and (3) to externally validate the BHIS in a multicenter prospective cohort of patients with borderline acetabular dysplasia.Methods:The current study included two parts. In Part 1, this study utilized a retrospective cohort study of 186 hips (178 patients) undergoing surgical treatment in setting of borderline acetabular dysplasia (LCEA 20°-25°) from a single surgeon experienced in arthroscopic and open techniques. Patients were excluded if over 40 years of age, Tonnis grade ≥2, prior ipsilateral surgery, or residual pediatric or neuromuscular disease. Multivariate analysis determined characteristics associated with presence of instability (treated with PAO +/- hip arthroscopy) or absence of instability (treated with isolated hip arthroscopy) based on clinical diagnosis of the single surgeon. During the study period, 39.8% of the cohort underwent PAO. Multivariate analysis with bootstrapping was performed and results were transformed into a nomogram and BHIS (higher score representing more instability). In Part 2, the BHIS was externally validated in a cohort of 114 patients with borderline acetabular dysplasia enrolled in a multicenter prospective cohort study across 10 other surgeons (with varied treatment approaches from arthroscopy to open procedures).Results:In Part 1, the most parsimonious and best fit model included 4 variables associated with instability: acetabular inclination (AI), anterior center edge angle (ACEA), maximum alpha angle, and internal rotation in 90 degrees of flexion (IRF). Odds ratio estimates and 95% confidence limits were 1.50 (1.28-1.76), 0.92 (0.86-0.99), 0.94 (0.90-0.98), and 1.11 (1.07-1.17), respectively. Notably, sex and LCEA were not significant predictors. The BHIS demonstrated excellent predictive (discriminatory) ability with c-statistic=0.89. Mean BHIS in the population was 50.0 (instability 57.7 ±7.9 vs. non-instability 44.8±7.3, p<0.001). BHIS demonstrated excellent predictive (discriminatory) ability with c-statistic=0.89. In Part 2, BHIS maintained excellent c-statistic=0.92 in external validation. Mean BHIS in this cohort was 53.9 (instability 66.5±11.5 vs. non-instability 43.0±10.8, p<0.001).Conclusion:In patients with borderline acetabular dysplasia, AI, ACEA, maximum alpha angle, and IRF were key factors in diagnosing significant instability treated with PAO. The BHIS effectively quantifies relative role of each factor and characterizes aspects of instability compared to the mean (BHIS=50) in this population. The BHIS score allowed for good differentiation of patients with and without instability in the development cohort, as well as the external validation cohort. Use of the BHIS score may facilitate efficient clinical characterization of important patient characteristics in the setting of borderline acetabular dysplasia.