This study is one of the largest single-institution reviews of OGIs, examining over 600 affected patients and analyzing the factors impacting the number of ophthalmic office visits and secondary surgeries within one year of the initial injury. These values have not been previously quantified and associated with injury characteristics in OGIs.

Assessment of clinical characteristics associated with greater healthcare use found that patients with poor presenting BCVA, such as LP/HM and CF or 20/400-20/200, had greater office visits and secondary surgeries. This trend held true in analysis of the common secondary surgeries excluding enucleation (Table 3). Although worse presenting VA is associated with poor final BCVA, other evaluations have found that BCVA typically stabilizes within the first three months of the initial injury, with a final BCVA improved about 3 lines of vision at one year [13]. Although the risk of significant visual impairment and legal blindness is high after OGI [13], any interventions that can preserve vision may significantly improve quality of life. Therefore, greater utilization of healthcare among patients with poor vision may reflect efforts to prevent progression to NLP. Similarly, zone of injury was a predictor for office visits, with the highest mean visits for zone I injuries (mean 5.7±4.9 vs 4.4±3.8 [zone II] vs 4.6±4.4 [zone III]). Zone I injuries have a better visual potential than zones II and III [14], likely due to decreased likelihood of posterior segment involvement. Although the structural integrity of the globe can be restored, damage to the retina and optic nerve may be irreparable [2, 15].

The type of trauma was also a predictor of office visits, with the most office visits for penetrating injuries (mean 5.9±4.9 [penetrating] vs 4.8±4.5 [blunt] vs 4.8±3.6 [non-traumatic]). Other examinations of OGIs found that penetrating injuries often have better prognosis than blunt injuries. Penetrating trauma tends to create a distinct entry wound with less risk of posterior segment involvement, while blunt trauma creates more complex ruptures with more risk of posterior segment involvement [16, 17]. Therefore, given higher potential for visual recovery in penetrating injuries, this etiology was also statistically predictive for more visits and secondary surgeries (Table 2). Conversely, eyes with poor prognosis had less utilization. For instance, BCVA was predictive of enucleation, with the highest mean among NLP patients (Table 3). This may be attributed to the lack of visual potential and risk of sympathetic ophthalmia. Patients presenting with BCVAs better than NLP may also undergo enucleations for deteriorating BCVA or other complications following OGI repair. Other predictors of enucleation included orbital floor fracture, eyelid laceration, and non-traumatic etiology. Among traumatic OGIs, orbital floor fractures and eyelid lacerations represent injury to the ocular adnexa. Both have been previously correlated with greater visual loss and the need for enucleation due to their association with corneoscleral injury and posterior globe involvement [15, 17, 18]. Non-traumatic OGIs comprised less than 2% of cases analyzed, which included cases of spontaneous perforations resulting from infectious or inflammatory conditions. Other examinations of non-traumatic OGIs have also associated these cases with similar causes [2]. Common etiologies such as infection, which may seed throughout the eye and cause rapid deterioration, may require aggressive intervention and eventual enucleation.

As expected, findings on presentation that are indicative for additional intervention, such as retinal detachment, vitreous hemorrhage, and traumatic cataracts, were also associated with more office visits [19]. Predictors for specific types of surgery were as anticipated (Table 3): endophthalmitis, retinal detachment, and vitreous hemorrhage were predictive for vitrectomy; presence of traumatic cataract was associated cataract extraction surgery; non-traumatic injury, surgical wound dehiscence, and endophthalmitis were predictive for corneal transplant. [19] The presence retinal detachment was statistically associated with need for corneal transplant although likely not clinically relevant. The type of injury was also a predictive factor for the total secondary surgeries, with the highest among non-traumatic OGIs (mean 0.8±1.0 vs 0.5±0.8). Due to the more chronic nature of these conditions (i.e. recurrent or nonhealing corneal ulcers), rather than an acute localized trauma, a longitudinal multi-faceted approach is required in the management of these patients.

This study has limitations related to scope and the nature of the methodology. The sample focused only on OGI repairs, therefore excluding non-operable cases. In addition, since the data is derived from a single urban Level I Trauma Center in New Jersey, the patterns observed may not reflect those observed in rural settings, community hospitals, or other geographic regions. Because of the retrospective design, the completeness and accuracy of data were dependent on the documentation available on the Epic EHR. Data collection was performed using emergency department notes, ophthalmology consultation, outpatient, and inpatient notes, and operative reports to develop a comprehensive dataset and minimize inconsistencies. Additionally, the focus of this study was on ophthalmology-specific follow-up and secondary surgeries. In cases of polytrauma or other complex injuries, multiple specialties would have been involved in patient care, potentially affecting overall healthcare use. This multidisciplinary involvement was not included in this analysis, although it may reflect in the high number of imaging studies reported. Resource utilization was quantified based on the total office visits and surgeries documented within the Epic EHR. However, it is possible that some patients received ophthalmic follow-up care outside of our institution, such as with an ocularist after enucleation. This data may not have been captured in the dataset.