Enophthalmos Following Orbital Fracture Repair: Identifying Risk Factors and Development of A Risk Calculator
Bashar Hassan*1, Seray Er3, Joshua Yoon4, Eric Resnick3, Cynthia T. Yusuf2, Fan Liang2, Richard Redett2, Paul Manson2, Robin Yang2, Michael Grant1
1Plastic and Reconstructive Surgery, R Adams Cowley Shock Trauma Center, Baltimore, MD; 2Plastic and Reconstructive Surgery, Johns Hopkins, Baltimore, MD; 3School of Medicine, University of Maryland, Baltimore, MD; 4Surgery, George Washington University, Washington,
Enophthalmos is an undesired outcome reported in 8-72% of cases following orbital fracture repair (OFR). Risk factors for postoperative enophthalmos remain unclear. We conducted the largest study to determine these risk factors and develop and validate a risk calculator for the prediction of this outcome.
A retrospective review of trauma patients who underwent OFR was conducted at two centers (2015-2019). Excluded were patients <18 years old and those with postoperative follow-up <2 weeks. Our primary outcome was incidence of postoperative enophthalmos at ≥2 weeks following OFR. The risk model for the prediction of postoperative enophthalmos was developed using 70% of the study population and validated using the remaining 30%. The C-statistic (area under the ROC curve) and Hosmer-Lemeshow tests were used to assess the risk model accuracy.
Of 253 patients, the median (IQR) age was 36.1 (27.8-50.7) years and follow-up was 79.0 (39.5-157.0) days. Postoperative enophthalmos was seen in n=21 (8.3%) patients. Our risk model for the prediction of postoperative enophthalmos included the following risk factors: older age, preoperative enophthalmos, medial wall fractures, near-total wall defects, and delayed OFR >1 week from injury. There was an incremental increase in the odds of postoperative enophthalmos with greater duration from injury to OFR. Patients with delayed OFR >30 days from injury were at 11.6 times the odds of postoperative enophthalmos compared to those who had OFR within 1 week from injury (aOR [95% CI] 11.59 [1.28->100]). Our predictive model had very good discrimination (C-statistic=84.5%), calibration (P=0.83), and validation (C-statistic=86.1%) (Figure 1). For a risk cut-off of 9.5%, the sensitivity and negative predictive value were 85% and 99%.
We designed the first validated risk calculator that can rule out enophthalmos following OFR in adults. Our risk calculator enables surgeons to preoperatively predict the absence of postoperative enophthalmos (for any risk <9.5%) in 99% of cases. This is useful for patient counseling and surgical decision making.
Back to 2023 Abstracts