Abstract
Purpose :
Shaken baby syndrome (SBS), a subset of abusive head trauma, results from non-accidental violent head-shaking and has an estimated annual rate of 600-1400 cases. Collective lifetime costs for survivors are estimated to exceed $13.5 billion. Typical triad of SBS includes subdural hematoma, retinal hemorrhages and encephalopathy but the presence of this triad is currently disputed. The purpose of this study is to describe ophthalmic manifestations and associated neurotrauma in suspected SBS.
Methods :
We retrospectively surveyed the National Trauma Data Bank (2008-2014) for patients <3 yrs old admitted with the ICD9CM code 995.55 for suspected SBS. Statistical analysis was performed with SPSS software. Significance was set at p<0.05.
Results :
347 (13.9%) of 2495 patients who were <3yrs old, admitted with abusive head trauma and ocular injuries resulted from suspected SBS. Most were <1yr old (87.9%) and male (54.2%). Common eye injuries were retinal hemorrhages (30.5%), eye/adnexa contusion (14.7%) and retinal edema (10.7%). Common neurotrauma were subdural (75.5%), subarachnoid (23.9%) and intracerebral hemorrhage (10.4%). Mean (SD) injury severity score was severe, 20.2(8.2) and Glasgow coma score was moderate, 9.2(12.8). Eye/adnexa contusion (OR=4.06, 95%CI=1.89-8.63; p<0.001) and retinal edema (OR=5.27, 95%CI=2.56-12.79; p<0.001) had the greatest association with ICH than other neurotrauma. Retinal hemorrhages were not significantly associated with one type of neurotrauma over others.
Conclusions :
Our study supports previous studies that have shown that retinal and subdural hemorrhages were the most common ocular injury and neurotrauma in SBS, respectively. However, we did not find a significant propensity for their concurrence. We also found that retinal edema was significantly associated with both intracerebral and subdural hemorrhages. Overall, the data was reported to the National Trauma Database by ED or trauma staff so the diagnosis for SBS was likely suspected at the initial encounter and may not have been confirmed. This may explain lower than expected incidence of retinal hemorrhages in this cohort compared to what has been previously reported for SBS. Overall, data with confirmed cases of SBS including completed ophthalmic exams will provide more insight into the utility of the SBS triad for diagnosis.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.