Sunday, December 19, 2010

A Watchful Eye

            There are few areas within forensic pathology that generate more debate and controversy than the subject of inflicted pediatric head trauma. Often in these cases, the reporting history is scant and typically the baby is stated to have been ‘just found unresponsive’ or experienced an ‘unwitnessed accident’. In lieu of this lack of information, a set of diagnostic criteria were decided upon to make the diagnosis of inflicted head trauma more straightforward. Traditionally these criteria included: 1) subdural hematoma; 2) cerebral edema; and 3) retinal hemorrhages.
            A dogma was established that if a child had all three signs, the child was abused. As time progressed and more scientific studies were published this belief came under fire and it is now accepted that not every subdural hemorrhage is abusive and that not every swollen brain was the result of inflicted trauma. That left retinal hemorrhages, and physicians began to place great importance in the discovery of such lesions.
            More recently the dogma has been if a child has subdural hemorrhage, cerebral edema AND retinal hemorrhages, then the injuries were inflicted. The presence of retinal hemorrhages became the preverbal ‘last straw’ in such that if you had the former two injuries but still couldn’t decide if the injuries were accidental of inflicted, the presence of the latter could sway you, and you could call the case inflicted head trauma.
            I was reared in the generation of pathologists who believed this hypothesis to be true, and therefore so did I. During my years of working in forensic pathology I have come across a number of cases where a child has had subdural hemorrhage, cerebral edema and retinal hemorrhages in the absence of trauma. The histories were excellent and there was no evidence for abuse. I had to rethink my entire belief system.        
Photograph 1A
            Photograph 1A depicts pre-retinal (arrowhead), intra-retinal (arrow), and post-retinal (asterisk) hemorrhages that were multifocal within the retina in a child who died following febrile seizures (click photo to enlarge). Furthermore, these lesions were also located at the periphery and at the optic disc, locations that were thought to be classic for abusive head injury. Photograph 1B shows multiple retinal hemorrhages (arrows) at the optic disc in the same child, the optic nerve is identified with an asterisk.
Photograph 1B
            Now if you search the literature you will find many case reports describing non-traumatic retinal hemorrhages in children. What once was truth became a little foggy for me!
            Enter retinal folds. Focused now shifted to the presence of these tiny creases within the retina, called retina folds. Usually described clinically, they are difficult to see under the microscope because of artifacts introduced by cutting the eye. These little lesions became somewhat of a holy grail, if they were seen in the setting of the other criteria needed to diagnose inflicted head injury, then the case was abuse, no question. There are some physicians who will diagnose abuse with retinal folds alone, in the absence of subdural hemorrhage.
            With focus now shifted to the presence or absence of retinal folds, a few case reports were published that demonstrated such findings in non-inflicted head trauma. The debate raged on and the argument became that yes, you can have non-abusive retinal folds, however, you have to have a story of severe trauma. Case reports describing non-abusive retinal folds included car accidents, a 40-pound television falling on a child’s head and an adult falling on a child’s head.
            Retinal folds became associated with severe head injury, and in the absence of such a history, they were diagnostic of inflicted head injury. This became part of my practice. If I had a case where the history did not suggest severe head trauma, but was one of being just found unresponsive or having fallen off the couch and saw retinal folds in conjunction with subdural hemorrhage and cerebral edema, I would opine that the injuries were likely to be inflicted. Yes, retinal folds can be seen with accidental head injury, but for me, the story had to involve some very serious injuries such as a car accident.
            Photograph 2 demonstrates an artifactual retinal fold in a non-traumatic eye. Photograph 3 depicts a retinal fold in a case of inflicted head trauma. In photo 3 the retina appears to be torn off the eye and there is hemorrhage rising up into the fold. Also, notice the retinal hemorrhages adjacent to the fold. Compare this image to the artifactual fold, and notice the lack of these findings in the artifact.
Photograph 2
Photograph 3
            After finally feeling as though I had come to a common ground within myself regarding the diagnosis of inflicted head injury, I had a case where a young child died following a non-traumatic hemorrhage within the basal ganglia. Examination of the eyes demonstrated multifocal pre-retinal, intra-retinal, and post-retinal hemorrhages AND retinal folds (photograph 4). I didn’t know what to think, retinal folds in the absence of any history of trauma? Yes.
Photograph 4
            This led me to think hard and examine the literature further. I now think that retinal hemorrhages and folds are likely the result of sudden and severe increases in intercerebral pressure.
            The literature regarding inflicted head injury is extensive and contradictory. In a modern forensic world were evidence based medicine has become vital, what is the pathologist to do? The question one should ask themselves is not ‘are these injuries inflicted or accidental?’ but ‘could these particular injuries occur in this particular child under these particular circumstances?’ This is not to say that the pathologist should disregard the classical findings; yes, abused children have subdural hemorrhages, cerebral edema, and retinal hemorrhage/folds. But is every child who has this constellation of findings abused? No. I treat every pediatric case as though it is a homicide until I can prove to myself that it is not.
            I am sure there are people who would disagree with me. All I can say is take every case on it’s own merits and don’t go into the autopsy with the bias of having already made the diagnosis of abuse on just the findings described above.  There is no lesion that is pathognomonic for inflicted pediatric head injury. One must have a high index of suspicion, but keep an open mind. The autopsy is just a single piece of information; the results must be interpreted in the context of all other available data, including birth records, pediatric records, police reports and scene investigations.


  1. Fantastic inaugural blog post. Given the near ubiquity of retinal folds in eye specimens, can a pathologist give much weight to them other than in the context of a funduscopic exam? That raises a question: Is there a role for postmortem funduscopic exam?

  2. Thanks, Dr. Moore!
    You bring up an interesting point. I think even making the diagnosis of postmortem retinal folds on histology is tricky. I have tried performing fundoscopy in a few cases with some success. The postmortem interval is the real problem with fundoscopy, the cornea blurs very quickly. Here is a link to an abstract in Forensic Science International by Amberg and Pollak detail the endoscopic examination of the fundus!