Shaken Baby Syndrome and Other Mysteries
A letter submitted to the American Journal of Forensic Medicine and Pathology,
Spring 1998

John Plunkett, M.D.
Laboratory & Medical Education Director,
Regina Medical Center, and Coroner,
Minnesota Regional Coroner's Office Regina Medical Center
1175 Nininger Road Hastings, MN 55033

Source of Support: None

"The problem isn't with what we don't know, the problem is with what we do know that isn't so". - Will Rogers

"If the law has made you a witness remain a man of science. You have no victim to avenge, no guilty or innocent person to ruin or save. You must bear witness within the limits of science." - Paul H. Broussard Chair of Forensic Medicine Sarbonne, 1897

A surreal and disturbing electronic media event followed the guilty verdict in Commonwealth (Massachusetts) v. Woodward last fall. One of Ms Woodward's attorneys suggested that the "shaken infant syndrome" was not as well supported by scientific data as was generally thought and that the trial outcome might have been different if the evidence had been presented to a jury of neuroscientists and engineers. Surreal: Almost immediately a letter to the Editor cosigned by fifty child abuse "experts" (including at least one pathologist) was published on the Internet and sent to more than a dozen newspapers, including the New York Times, the Washington Post, the Chicago Tribune and the Los Angeles Times. The letter ridiculed the idea that the syndrome should be re-examined, saying that it was "a well characterized clinical and pathologic entity with diagnostic features...virtually unique to this type of injury". The signators reproved [defense witnesses] who would "challenge the specificity of these features [by] speculating on other causes in court" and stated that the "rebleed theory in infants is a courtroom diagnosis, not a medical diagnosis, and [that] the jury properly rejected it."(1)

 

Disturbing: Not a word of response from the forensic pathology community.

 

When Dr John Caffey, a pediatric radiologist, described the "shaken-whiplash syndrome"(2), he did so within the context of a battered child with multiple (multidirectional force) injuries, and postulated that the weak neck muscles and relatively large head size of an infant made him/her particularly susceptible to cerebral injuries from shaking. I suspect that Caffey and others evaluating head injuries in the '40s, '50s and '60s asked a number of caretakers if the infant had been "shaken" and were told "yes" in at least some cases. The caretakers were never asked about an "impact" because direct trauma was not part of the theory. Scientific theory was quickly accepted as scientific fact: Subdural hemorrhage and retinal hemorrhage in an unconscious or dead child is a shaken infant; there is no need to "prove otherwise"; only a fall from a two story building or a motor vehicle accident could cause such an injury, if it was not due to shaking. Studies critically evaluating the biomechanics of rotational brain injury and a subdural hematoma, available from experiments performed for (among others) the automotive industry (3-9) and the space program, were forgotten, not sought or ignored.

 

Many years after Caffey's publication Drs. Duhaime and Gennarelli(10 - 12) developed a model to demonstrate the particular suspectibility of the infant brain to shaking injury, but were unable to generate the required force unless the head was impacted against a solid surface. The impact deceleration forces were greater than the shaking deceleration forces by nearly two orders of magnitude, and Duhaime and Gennarelli concluded that "severe head injuries commonly diagnosed as shaking injuries require impact to occur and that shaking alone in an otherwise normal baby is unlikely to cause the shaken baby syndrome". This conclusion, from the first actual scientific study of "shaking" injuries, was quickly adopted by the child abuse professionals, who didn't miss a beat: Well certainly, the child is "frequently" thrown to the floor or into a wall once the adult grows tired of shaking the child. The "shaken-slammed infant syndrome" superceded the "shaken infant syndrome". However, scientific theories die slowly and the child abuse experts believed and continued to teach that pure impact rarely caused the syndrome, and that at least 1 - 2 violent "shakes" were necessary to "prime" the brain prior to impact.

 

If an impact is necessary to cause the "shaken-infant syndrome" injury, what force is required? The child abuse professionals tell us that it is equivalent to a fall from a two story building or a motor vehicle accident.(13 - 16) However, a fall from a two story building is almost exclusively translational force, even if the child is spinning, and is unlikely to cause a subdural unless there is a skull fracture with tearing of a dural sinus. A motor vehicle accident involves "low strain" rotational deceleration, again unlikely to cause a subdural although certainly capable of causing death(10) . (A major goal of automotive safety engineers is to increase deceleration time and distance, and is the reason children's safety seats are placed "backwards" in the car: Most motor vehicle accidents are associated with constant velocity or deceleration prior to impact, and the collision or pre-collision deceleration will push the infant's head into the padded seat, not away from it. ) There is no experimental data on immature skulls or brains, and we do not know the amount of force required to cause a subdural, retinal hemorrhage and brain injury in a child. However, adult data (human and other primates)(5 - 8, 10) and anecdotal evidence from children who have suffered clearly accidental head injuries from short distance falls(13, 16 - 18) indicate that an impact velocity of 15 feet per second, with a stopping time of 2 milliseconds and a rotational deceleration of 100 radian/second2 , will cause a subdural. This is equivalent to a gravity rotational fall of 3 feet with the head impacting a solid surface such as a linoleum floor, and rotating through a 60o arch during the 10 milliseconds or so prior to impact. (The impact velocity and distance may be even less if Aoki(17) and Howard's(18) evidence is sound and their conclusions correct.)

 

I do not understand the mechanism for or causes of re-bleeding in an infant subdural hematoma. A study by Gilliland(19) concluded that "some shaken infants have no impact", based on autopsy evaluations of 80 children with head trauma, 9 of whom were determined to have died from "shaking" alone. This study does not mention the ages of the children, the nature of the subdurals (acute, subacute or chronic), or the time from hospital admission to removal of respirator support. If an "impact" occurred 14 - 21 (and perhaps fewer) days prior to death, it would be invisible to the unaided and microscopic eye. I have personally reviewed more than thirty records of children, living and dead, with retinal hemorrhage, a subdural 7 days - several months old when first evaluated, and acute bleeding superimposed on the old hemorrhage, the overwhelming majority of whom had no physical or autopsy evidence for an "acute" impact, other injury, or other historic data to support "abuse" as the etiology of the rebleed. The "vast clinical experience" of the Internet physicians allowed them to conclude that each of these injuries and deaths was due to a non-impact shaking since "infants do not suffer massive head injury, show no significant symptoms for days, then suddenly collapse and die."(1) The plural of anecdote is not data; the sum of "vast clinical experience" may be knowledge but it is not science.

 

I do not understand the "retinal hemorrhage" litmus test for a shaken infant. No one knows what causes retinal hemorrhage, although it is highly correlated with rotational deceleration injury/subdural hemorrhage in children, but retinal hemorrhage indistinguishable from that found in rotational deceleration may be found in association with ruptured vascular malformations,(20 -23) arachnoid cysts(24 - 26) and CNS infections. The "pathognomic sign" has evolved from "retinal hemorrhage" to "flame-shaped retinal hemorrhage" to "multilayered flame-shaped retinal hemorrhage" to, most recently, "multilayered flame-shaped retinal hemorrhage with macular folds" with no data to support this progression other than the usual argument that the injury "must" have been caused by shaking. A respected forensic pathologist, recently retired, told me that "true pathognomonic signs are rare" and that "only academic, defined as 'little practical worth', physicians espouse this term." I am reminded of Abraham's negotiations with God regarding Sodom and Gomorrah: "Well, if not multilayered flame-shaped hemorrhage, how about multilayered flame-shaped hemorrhage with macular folds?" If the latest version of the "pathognomic sign" proves to be correct, it is still no more than a marker for a rotational deceleration injury, and does not tell us if the cause of the injury was a "shaken-slam" or a high-strain rotational fall.

 

Whether or not fatal head injury in infants is due to diffuse axonal injury is a mystery. The concept of "diffuse axonal injury" was developed almost 50 years ago(27) and helps to explain prolonged unconsciousness in adult head-injury accident victims who have no mass lesion, and to predict survivability from the injury.(28 - 30) The theory is relatively simple, although the injuries are "multifocal" rather than "diffuse", and experimental and clinical data upholds this model for understanding many adult head injuries on a macroscopic and microscopic structural level. However, all of the experimental data is on adults and "mature" primate (and other) brains, and there is no data on infant or immature primate brains. The theory, despite the lack of evidence that it can be extrapolated to children, has been embraced by the "shaken infant" experts as its own, and is used to conclude that "shaken infants" who die never have a lucid interval(31) , and that the cause of death in those who die shortly after the shaking is axonal injury. Neither conclusion is supported by available scientific data(32 - 33) . At least some children with head injury, rotational and otherwise, have a documented lucid interval prior to the development of symptoms, including those who subsequently die. Further, it is impossible to prove macroscopically or microscopically that axonal injury has occurred in children who die within 48 hours of an injury except for the rare child with pontine hemorrhage or a clearly pre-mortem tear in the corpus callosum. We simply do not know why some of these children die: It may be axonal injury, "malignant" cerebral edema, direct irritation of brain-stem breathing centers, or some other phenomenon we have not considered. The child-abuse catch phrases "immediately unconscious", "globally changed" and "unable to feed" make nice sound bites, but are scientifically indefensible and create an inescapable legal conclusion: The last person standing when the music stopped is the one who must have injured the child.

 

I am concerned by the reduction of the criteria for a diagnosis of the "shaken-slammed infant syndrome" to absurdity. I have recently reviewed a number of infant deaths in which the cause was stated to be "shaken-slammed infant syndrome" but in which the child had no impact injury, no subdural hemorrhage and no retinal hemorrhage. The death was attributed to rapid development of cerebral edema (none had uncal or cerebellar tonsillar grooving or herniation) and "microscopic" subarachnoid and parenchymal hemorrhage. Impact against a "soft" surface and/or Gilliland's data was used to support the conclusions. (Physics 101: An increased deceleration time or deceleration distance markedly decreases the likelihood of brain trauma, unless we are talking about a 40 mph motor vehicle accident or a fall from a two story building). The caretaker in each of these deaths was charged with murder. None of these children had any evidence for a head injury in the opinion of 3 - 5 reviewing pathologists, who thought that three of the deaths were due to SIDS, one to SIDS/overlaying and one to a viral encephalitis/Reye's Syndrome. It was only after the deaths were reviewed by a number of other pathologists, who were in unanimous agreement, that charges were dismissed.

 

Our obsession with the media, and our willingness to uncritically believe what they tell us, is a puzzle. I was called by a television "news magazine" producer last week, asking if I had seen infant death cases in which I thought "shaking" had been misdiagnosed. I told her that I had. She asked if I had watched the Massachusetts trial on television. I told her that I had not, but that I was familiar with the medical evidence. She confided to me that the defendant was not "likeable", and wondered if those charged in the cases I had reviewed were "middle-class" and "likeable". Likeable? Middle-class? Ted Bundy was likeable and middle-class! I understand the need for the media to "sell" their story, but is a "good" parent or caretaker capable of fatally injuring a child, intentionally or otherwise? Of course. May a "bad" parent or "unlikable" caretaker have a child die from natural causes or an old injury while in his/her custody? Of course. I remembered the Judy Collins' song "Dress Rehearsal Rag" (I am a child of the sixties): "It's down to this, it's down to this, and wasn't it a long way down, and wasn't it a strange way down?". Are these concepts so complicated that the public must be depend on impressions and ignore the science? Is "likability" the criterion for truth?

 

Finally, I do not understand the common practice of plea bargaining in these cases. The defendant is usually indicted for first- or second-degree murder, and in most states faces a life sentence if convicted since the death involved a particularly vulnerable infant. Capital charges are the rule, even when the apparent head injury is isolated. However, as the trial date approaches and the prosecutor realizes that there are problems with the time of the injury (relative to the defendant's access to the victim) or the nature of the injury itself, the bargaining process begins. The defendant is offered a plea to a reduced charge with a limited jail sentence, given a deadline for accepting the offer, and threatened with a recommendation for upward sentencing departure if there is a conviction on the original charges. I have personally seen three young defendants, each with young children, accept such offers in deaths where there was either no head injury, or an injury that was old and could not be dated. Two accepted a plea to manslaughter with a seven year sentence and one a plea to manslaughter with a three year sentence. This type of plea bargaining is particularly cynical and troubling. If the person is guilty as initially charged, society does not need that person on the streets for a long time. If the facts support that the person was guilty as pleaded, he/she should have been charged as such initially. It would certainly appear to both the casual and the interested observer that the prosecution's goal is not a search for the truth but the finality of a conviction.

 

Forensic medicine must embrace evidence-based medicine, now. We need to abandon the term "shaken-slammed infant syndrome" and use an actual description of the injury mechanism, i.e., "rotational deceleration", in those children who have a subdural hematoma and definite evidence for an impact. We need to recognize that "retinal hemorrhage" regardless of its characteristics is at best an external marker for a "probable" head injury. We need to differentiate an "old" or "subacute" subdural from an acute bleed and admit that we know nothing about the actual cause(s) for a rebleed. We need to acknowledge that there are very limited data regarding a "lucid interval" in a child following rotational head injury, and that the concept of "diffuse" axonal injury may not explain why some children die very quickly while others appear to have a symptom-free or relatively symptom-free interval prior to death. We need to cautiously interpret a caretaker's story that is inconsistent with the physical findings, since the caretaker may not know the "true" history, especially if a lucid interval may occur.

 

We need to avoid basing decisions on the "likability" or previous history of a caretaker, unless the physical evidence unequivocally allows us to make the judgement independent of the history. We need to abandon simplistic, inaccurate and unproven phrases such as "globally changed", "immediately unconscious", "fall from a two-story building", and "force so great the perpetrator must have known that he/she would cause injury". (How does one ever know what another human being knew or intended?) Modern medicine has many examples of concepts and technologies accepted as proved and promoted as "standard of care", only subsequently to be demonstrated to have no value and wrong (routine use of fetal monitors, routine skull x-rays in children with head injuries, Pap smears in woman who have had had a hysterectomy for nonmalignant disease, indications for blood component therapy, etc.)(35) We need to differentiate between what we scientifically know to be true, and what we think or hope to be true. We don't need advocates in front of audiences, including juries, demonstrating the mechanism of shaking in a syndrome we do not understand. We don't need a third or fourth National "Shaken Baby Conference" to promote more unfounded theories for infant head injuries and to suggest prosecution methods to counter "untruth" defenses. The concept of a "shaken infant syndrome" deserves to be examined and re-examined, even when we think we finally have it "right".

 

Too many of my colleagues (and most other physicians and almost the entire general public) think our profession is the "whodunit" discipline. It is not. Forensic pathology is the "what happened" specialty. When our focus is on the "who" we forget the "what" or may consider it unimportant. Worse, we may alter our explanation/interpretation of the "what" to make it conform to our opinion of the "who". The need to consider alternative explanations ceases, doors to further inquiry close: Do not go beyond, you will find nothing there. Objectivity fails because we are forced to defend an advocacy role, be it for the state or for the defendant. We must not forget that our only responsibility is to bear witness within the limits of science...

References:
1. Amaya M, Bechtel K, etal.(50 signators). Shaken Baby Syndrome and the Death of Matthew Eappen. Internet: http://silcon.com/~ptave/shaken.htm. November 11, 1997.
2. Caffey J. On The Theory and Practice of Shaking Infants. Its Potential Residual Effects of Permanent Brain Damage and Mental Retardation. Am J Dis Child 1972; 124:161-69.
3. Gurdjian ES, Webster JE, Lissner HR. Observations on Prediction of Fracture Site in Head Injury. Radiology 1953; 60:226-235.
4. Evans FG, Lissner HR, Lebow M. The Relation Of Energy, Velocity And Acceleration To Skull Deformation And Fracture. Surg Gyn Ob 1958; 107:593-601.
5. Gurdjian ES, Roberts CL, Thomas LM. Tolerance Curves Of Acceleration And Intercranial Pressure And Protective Index In Experimental Head Injury. J Trauma 1966; 6:600-604.
6. Gurdjian ES, Hodgson VR, etal. Impact Head Injury: Mechanisms and Prevention. Accident Pathology: Proceedings of an International Conference. Washington D.C. US Government 1968: 140-143
7. Nahum AM, Gatts JD. Impact Tolerance of the Skull and Face. Proceedings of 12th Stapp Car Crash Conference, P - 26, Paper 680785. New York: Society of Automotive Engineers, Inc., 1968.
8. Gurdjian ES. Impact Head Injury: Mechanistic, Clinical And Preventive Correlation. Springfield IL, Charles C Thomas. 1975.
9. Melvin JW, Evans FG. A Strain Energy Approach to the Mechanics of Skull Fracture. Proceedings of the 15th Stapp Car Crash Conference. P-30, paper 710871. New York: Society of Automotive Engineers, Inc, 1972.
10. Gennarelli TA, Thibault LE. Biomechanics of Acute Subdural Hematoma. J Trauma 1982; 22:680-686.
11 Duhaime A-C, Gennarelli TA, et al.. The shaken baby syndrome: A clinical, pathological, and biomechanical study. J Neurosurg 1987; 66:409-415.
12. Duhaime A-C, Alario AJ, et al. Head Injury in Very Young Children: Mechanisms, Injury Types and Ophthalmologic Findings in 100 Hospitalized Patients Younger Than Two Years of Age. Pediatrics 1992; 90:179-185.
13. Hall JR, Reyes HM, et al. The Mortality of Childhood Falls. J Trauma 1989; 29:1273-75.
14. Chadwick DL, Chin S, et al. Deaths from Falls in Children: How Far is Fatal? J Trauma 1991; 31:1353-1355.
15. Williams RA. Injuries in Infants and Small Children Resulting From Witnessed and Corroborated Free Falls. J Trauma 1991; 31:1350-52.
16. Reiber GD. Fatal Falls in Childhood: How Far Must Children Fall to Sustain Fatal Head Injury? Report of Cases and Review of the Literature. Am J Forens Med Pathol 1993; 14:201-207.
17. Aoki N, Masuzawa H. Infantile acute subdural hematoma: Clinical analysis of 26 cases. J Neurosurg 1984; 61:273-280.
18. Howard MA, Bell BA, Uttley D. The pathophysiology of infant subdural haematomas. Br J Neurosurg 1993; 7:355-365.
19. Gilliland MGF, Folberg R. Shaken Babies - Some Have No Impact Injuries. J Forens Sci 1996; 41:114-116.
20. McLellan NH, Prasad R, Punt J. Spontaneous subhyaloid and retinal hemorrhages in an infant. Arch Dis Child 1986; 61:1130-2.
21. Kanaan I, Lasjaunias P, Coates R. The spectrum of intracranial aneurysm in pediatrics. Minim Invas Neurosurg 1995; 38:1-9.
22. Tekkok IH, Ventureyra EC. Spontaneous intracranial hemorrhage of structural origin during the first year of life. Childs Nerv Sys 1997; 13:154-64.
23. Jay, V (Toronto), Personal communications.
24. Aicardi J, Bauman F. Supratentorial extracerebral cysts in infants and children. J Neurol 1975; 38:57-68.
25. Smith RA, Smith WA. Arachnoid Cysts of the Middle Cranial Fossa. Surg Neurol 1976; 5:246-52.
26. Munk PL, Robertson WD, Durity FA. Middle Fossa Arachnoid Cyst and Subdural Hematoma: CT Studies. J Comp Assist Tomo 1988; 12:1073-75.
27. Smith, SJ. Diffuse Degeneration of the Cerebral White Matter in Severe Dementia Following Head Injury. J Neurol Neurosurg Psychiatry 1956; 19:163-85.
28. Gennarelli TA, Thibault LE, et al. Diffuse Axonal Injury and Traumatic Coma in the Primate. Ann Neurol 1982; 12:564-574.
29. Adams JH, Graham DI, et al. Diffuse Axonal Injury Due to Nonmissile Head Injury in Humans: An Analysis of 45 Cases. Ann Neurol 1982; 12:557-563.
30. Blumbergs PC, Jones NR, North JB. Diffuse axonal injury in head trauma. J Neurol Neurosurg Psychiatry 1989; 52:838-841.
31. Willman KY, Bank DE, etal. Restricting The Time Of Injury In Fatal Inflicted Head Injuries. Ch Abuse Negl 1997; 21:929-940.
32. Nashelsky MB, Dix JD. The Time Interval Between Lethal Infant Shaking and Onset of Symptoms - A Review of the Shaken Baby Syndrome Literature. Am J Forens Med Pathol 1995; 16:154-7.
33. Plunkett J. Restricting The Time Of Injury In Fatal Inflicted Head Injuries. (Letter) Ch. Abuse Negl 1998 (In Press).
34. Fallows JM. Breaking the News: How the Media Undermines American Democracy. New York. Pantheon Books. 1996.
35. Vertosick FT. First Do No Harm. Discover 1998; 19-7: 106-111.

 

 

 

TOP

 

 

 

www.slimeylimeyjustice.org