Abstract
Purpose :
Non-accidental trauma (NAT) examinations in children are not entirely benign and preclude neurological examination. The majority of ophthalmic exams routinely ordered for patients suspected of NAT are negative. Our study elucidates clinical and imaging factors that correlate to retinal findings to increase the yield of positive exams and decrease the burden of potentially unnecessary exams, validating previous studies and introducing new screening variables.
Methods :
Our study is a retrospective chart review from May 2014 - August 2021 at a level-1 trauma center. 274 patients met the inclusion criteria of: 1) children </= 36-months-old, 2) concern for NAT, and 3) an ophthalmology consultation was placed. Through univariate and multivariate logistic regression, our study produces a screening algorithm for ophthalmic workup in NAT.
Results :
Due to a strong association of “abnormal neuroimaging” and retinal hemorrhage (RH) (OR of 170 (95% CI (10.245, >999.999)), multivariate logistic regression controlling for “abnormal neuroimaging” was performed to analyze all variables associated with RH. 1 or more abnormal neuroimaging finding had a statistically significant association with retinal hemorrhages and produced the strongest association with a univariate OR of 170. The multivariate model (p-value<0.0001 with a c-statistic of 0.980) proposes using the following variables for predicting retinal hemorrhage on exam: Abnormal neuroimaging, Glasgow coma score (GSC) < 15, altered mental status on examination, seizure activity, vomiting, burising, scalp hematoma/swelling, and skull fractures.
Conclusions :
If neuroimaging is obtained in the evaluation of suspected NAT, the yield of positive ophthalmologic findings would increase if certain non-ocular findings are present. If absent, ophthalmologic exams may not be necessary.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.