Presenter Status
Resident/Ph.D/Post graduate (> 1 month of dedicated research time)
Abstract Type
Research
Primary Mentor
Lisa Berglund, M.D.
Start Date
10-5-2023 11:30 AM
End Date
10-5-2023 1:30 PM
Presentation Type
Abstract
Description
Background: There is no consensus in the literature on the risk of developing avascular necrosis (AVN) after unstable slipped capital femoral epiphysis (SCFE). In recent years, a number of studies have sought to better characterize risk factors for AVN in this population. However, to the best of our knowledge, there is no study that uses predictive modeling to assess the chance of developing AVN based on different patient-centric factors.
Objectives/Goal: Our goal is to assess whether or not previously-published risk factors for AVN in unstable SCFE such as duration of symptoms, reduction type, type of surgery, timing of surgery, and severity of slip angle, are predictive of whether or not a patient will develop AVN.
Methods/Design: This was a 10-year chart review of patients with unstable SCFE. All patients had at least 12 months final follow-up. Receiver operator characteristic (ROC) area under the curve (AUC) analysis was performed to assess the predictive probability of developing AVN for all variables collected. An 8-stage hierarchical multivariable logistic regression was performed to control and adjust for variability between patients. Having no reduction in the operating room was identified as a confounder, and ROC AUC analysis was repeated after controlling for this variable.
Results: Our total cohort included 81 patients (61.7% male) with average age of 12.7±1.7 years, average BMI of 28.5±6.4, and average BMI percentile of 89.8±19.6. ROC analysis demonstrated an AUC of 0.73 (CI 0.62-0.85, p=0.006) for preoperative slip angle and 0.78 (CI 0.65-0.90, p=0.001) for postoperative slip angle reduction in predicting postoperative development of AVN. Logistic curves were created based on AUC analysis to determine the risk of AVN based on preoperative slip angle and slip angle correction (Figure 1). Eighteen (22.2%) patients developed AVN and seventeen (20.9%) eventually developed contralateral SCFE. Between the patients who developed AVN and those who didn’t, AVN patients were significantly older (13.1±1.6 vs. 12.5±1.7, p=0.033), had a significantly greater preoperative slip angle (54.6±18.3 degrees vs. 44.4±17.1 degrees, p=0.006), and were more likely to have had slip progression after initial fixation (p=0.004). The Odds Ratio (OR) of developing AVN after slip progression was 8 (CI 2.4- 26.9) and Relative Risk (RR) was 4.1 (CI 1.9-8.6).
Conclusions: This study contains one of the largest cohorts of patients with unstable slips analyzed to date. Using ROC AUC analysis, preoperative slip angle and postoperative slip angle correction are acceptable predictors of developing AVN. Using these logistic curves, surgeons and clinicians can estimate the percent chance that AVN will occur based on preoperative slip angle and slip angle reduction. AVN may be more likely to develop in older children who have higher slip angles. Additionally, slip progression is a significant risk factor for development of AVN.
MeSH Keywords
Slipped Capital Femoral Epiphysis; Avascular Necrosis
Included in
Predicting Avascular Necrosis after Unstable Slipped Capital Femoral Epiphysis
Background: There is no consensus in the literature on the risk of developing avascular necrosis (AVN) after unstable slipped capital femoral epiphysis (SCFE). In recent years, a number of studies have sought to better characterize risk factors for AVN in this population. However, to the best of our knowledge, there is no study that uses predictive modeling to assess the chance of developing AVN based on different patient-centric factors.
Objectives/Goal: Our goal is to assess whether or not previously-published risk factors for AVN in unstable SCFE such as duration of symptoms, reduction type, type of surgery, timing of surgery, and severity of slip angle, are predictive of whether or not a patient will develop AVN.
Methods/Design: This was a 10-year chart review of patients with unstable SCFE. All patients had at least 12 months final follow-up. Receiver operator characteristic (ROC) area under the curve (AUC) analysis was performed to assess the predictive probability of developing AVN for all variables collected. An 8-stage hierarchical multivariable logistic regression was performed to control and adjust for variability between patients. Having no reduction in the operating room was identified as a confounder, and ROC AUC analysis was repeated after controlling for this variable.
Results: Our total cohort included 81 patients (61.7% male) with average age of 12.7±1.7 years, average BMI of 28.5±6.4, and average BMI percentile of 89.8±19.6. ROC analysis demonstrated an AUC of 0.73 (CI 0.62-0.85, p=0.006) for preoperative slip angle and 0.78 (CI 0.65-0.90, p=0.001) for postoperative slip angle reduction in predicting postoperative development of AVN. Logistic curves were created based on AUC analysis to determine the risk of AVN based on preoperative slip angle and slip angle correction (Figure 1). Eighteen (22.2%) patients developed AVN and seventeen (20.9%) eventually developed contralateral SCFE. Between the patients who developed AVN and those who didn’t, AVN patients were significantly older (13.1±1.6 vs. 12.5±1.7, p=0.033), had a significantly greater preoperative slip angle (54.6±18.3 degrees vs. 44.4±17.1 degrees, p=0.006), and were more likely to have had slip progression after initial fixation (p=0.004). The Odds Ratio (OR) of developing AVN after slip progression was 8 (CI 2.4- 26.9) and Relative Risk (RR) was 4.1 (CI 1.9-8.6).
Conclusions: This study contains one of the largest cohorts of patients with unstable slips analyzed to date. Using ROC AUC analysis, preoperative slip angle and postoperative slip angle correction are acceptable predictors of developing AVN. Using these logistic curves, surgeons and clinicians can estimate the percent chance that AVN will occur based on preoperative slip angle and slip angle reduction. AVN may be more likely to develop in older children who have higher slip angles. Additionally, slip progression is a significant risk factor for development of AVN.
