Predictors of Intraoperative Difficulty and Postoperative Examination Abnormalities in 164 Orbital Operations

Abstract

Although orbital fractures are common, prediction of outcomes in orbital surgery can be quite challenging. We aim to identify predictors of intraoperative difficulty, operating time, and postoperative examination abnormalities in subjects undergoing post-traumatic orbital reconstructions. This is a retrospective cohort study of all consecutive orbital operations performed at a private, Level 1 trauma center in Portland, Oregon, USA over an 82-month period. All subjects that underwent exploration of the internal orbit for traumatic indications during the study period were included in the cohort. Four plating styles, surgical approach (transorbital vs transantral), days from injury to first surgery, fracture size (approximated as a rectangle using linear measurements from computed tomography scans), anteroposterior fracture position, and medial wall involvement were examined. The primary outcome variable was intraoperative difficulty (defined as requiring revision after intraoperative imaging or return to the operating room). Secondary outcome variables included operating time and postoperative examination abnormalities. Age and sex were included. χ2 and Regression analyses were performed using a significance level of P<.05. One hundred and sixty four orbital operations were performed (90 isolated injuries and 74 combined orbital/midface injuries) on 155 subjects (73% male, mean age 39.8years, standard deviation 16.7). In subjects with isolated orbital fractures, medial wall involvement was associated with intraoperative difficulty (P=.01). When using a transantral approach, intraoperative difficulty was more likely in more anterior fractures (P=.02). Plating style was associated with operating time (P=.03), with median times from 81 to 105minutes (range 21 to 248minutes). Postoperative examination abnormalities were more likely in the transorbital approach group (P=.01). Neither days to first surgery nor intraoperative difficulty were associated with postoperative examination abnormalities. Postoperative eyelid changes were seen in 13.6% of transorbital approaches and 0% of transantral approaches. Correction of gaze restriction and enophthalmos were more likely than correction of diplopia (P<.01). Medial wall involvement is associated with intraoperative difficulty in orbital surgery. Anteriorly positioned fractures are better treated transorbitally, while posterior fractures may be amenable to transantral repair, thus avoiding risk of lower eyelid changes.