Retinal hemorrhages common in shaken baby syndrome

Study, however, finds many non-ophthalmologists miss hemorrhages.

Issue: September 1, 2000

ByBob Kronemyer

MILWAUKEE, Wis. — Although retinal hemorrhages often accompany shaken baby syndrome (SBS), about 30% of patients have their hemorrhages undetected by non-ophthalmologists, according to the results of a retrospective, noncomparative case series that was recently published in Ophthalmology.

“The problem is that these kids present to an emergency-room doctor who can’t find the retinal hemorrhages,” said first author Jane D. Kivlin, MD, a professor of ophthalmology at the Medical College of Wisconsin here. “But having an ophthalmologist examine every child is simply not practical.” The authors state that the non-ophthalmologists’ difficulty in detecting these hemorrhages might be an important limiting factor in identifying shaken babies so they can be protected from further abuse. Educating emergency room doctors to detect retinal hemorrhages is one viable solution.

In the study, researchers examined the records of 123 children under 3 years old who experienced subdural hematomas from abuse. All patients were admitted to The Children’s Hospital of Wisconsin from January 1987 through December 1998 for their initial evaluation. Clinical features of eye examinations during admission and after discharge, as well as histopathologic observations of patients who died, were retrieved from medical records and statistically analyzed. The main outcome measures were visual and pupillary responses at initial examination, fundus findings, final vision, neurologic outcome of survivors and death.

Overall, 90% of the children had ophthalmic assessments. Retinal hemorrhages were detected in 83% of the examined children. “In previously published series, this percentage has ranged from 50% to 100%,” said Dr. Kivlin, a pediatric ophthalmologist. “However, this percentage depends on how other studies have defined their patients. For example, some studies required that patients have retinal hemorrhages to enroll.” In the current study, retinal hemorrhages were bilateral in 85% of affected patients. On the other hand, dome-shaped hemorrhages under the internal limiting membrane were not common.

Encouraging visual acuity results

---Retinal hemorrhage in a shaken baby.

Visual outcome at the last outpatient visit (mean follow-up 21 months) was assessed for 78% of the survivors (68 of 87 children). However, visual acuity outcomes were broadly defined because 40% of these children were assessed by non-ophthalmologists. Still, “more than one-half of the patients were known to see well at their last visit, while only one-fifth had some visual impairment,” Dr. Kivlin said. “I did not expect these encouraging results. I thought the percentage of good vision would be less because SBS is often associated with brain damage, death and blindness.” Non-seizure patients also were more likely to see well.

Vision attentiveness improved markedly in 10 patients who had poor vision before sedation. Nine of these patients improved to normal for their age, with a visual recovery time of from 12 hours to 4 weeks.

Vision linked to survival

The study found a strong association between vision response at initial examination and survival. “How the babies saw when they came in and how their pupils responded correlated very highly with whether a patient survived,” Dr. Kivlin said. “If a baby could see and the pupils responded, then he or she had a much better chance of surviving.” In contrast, poor visual response, poor pupillary response and retinal hemorrhage correlated strongly with death. In total, 29% (36 patients) died of their head injuries. “In general, about one-third of patients with SBS die,” Dr. Kivlin said. Two children died after discharge from the hospital (one from a chronic vegetative state and the other from enlarging subdural hematoma). A third deceased child had pigmented retinal scars from previous abuse, “a condition not previously observed histopathologically to our knowledge,” Dr. Kivlin said.

Most of the seven children who had circular retinal folds also died. Furthermore, children who did not have seizures were far more likely to die than those who experienced seizures. Retinal hemorrhages also correlated with death. A child was five times more likely to die if the ophthalmologist detected any retinal hemorrhages shortly after injury. “However, about five patients presented late, meaning they survived abuse. These survivors probably had retinal hemorrhages that disappeared before they were examined. So this data is slightly skewed,” Dr. Kivlin said.

Asymmetry of retinal hemorrhages was noted more often in the autopsy eyes compared with the patient’s clinical findings. “In the autopsy cases, hemorrhages occurred throughout the retina versus the posterior retina on clinical exam,” Dr. Kivlin said.

One finding that perplexed the investigators was an increase in retinal hemorrhages with increased age. Specifically, 100% of infants 25 to 36 months old had retinal hemorrhages, compared with 66% for infants 0 to 6 months old, 82% for infants 7 to 12 months old and 88% for infants 13 to 24 months old. “This increase was the opposite that was found in previous autopsy series,” Dr. Kivlin said. “The greater force required to shake a larger child may contribute to retinal hemorrhages, but a corresponding increase in mortality was not found.”

Quality of life

Quality of life for survivors of SBS varies greatly. “Some of these kids are really devastated. They are retarded, they are blind, they have seizures, they are confined to wheelchairs,” Dr. Kivlin said. “Others do fine.” Regardless, “child abuse has been estimated to be a big cause of mental retardation, where you don’t identify any other reason.”

Dr. Kivlin encourages ophthalmologists to work with survivors of SBS in improving vision to the fullest possible extent and offer low-vision rehabilitation services for those who need it. “All physicians tend to shy away from abused kids because it is so painful,” she said. In particular, “ophthalmologists as a group seem to be reluctant to suggest abuse. But we shouldn’t be in denial about abuse. Helping these children is very rewarding. Ophthalmologists can make a difference in these kids’ lives afterward by helping them to see as well as they can.”

For Your Information:

Jane D. Kivlin, MD, can be reached at the Medical College of Wisconsin Eye Institute, 925 N. 87th St., Milwaukee, WI 53226-4812; (414) 266-2020; fax: (414) 456-6300.

Reference:

Kivlin JD, Simons KB, Lazoritz S, et al. Shaken baby syndrome. Ophthalmology. 2000;107:1246-1254.