Shaking and Other Non-Accidental Head Injuries in Children

Drs Minns and Brown provide a broad review of inflicted head injuries (IHIs) from medical, epidemiological, legal, and social viewpoints. The most extensively treated topics, clinical (neurological) manifestations and biomechanics of infant and toddler head injury, consume the first third of the text. The text expands on the 2005 article by Dr Minns¹ on the same topic. In both, he postulates 4 different clinical presentations for IHI. Six percent of IHIs are attributed to a “hyperacute encephalopathic or cervico-medullary syndrome,” in which acute cervicomedullary injury results in immediate apnea, leading to death or severe hypoxic-ischemic injury. Here the pathologic studies of Geddes et al² are referenced. They observed children with IHI to have microscopic or gross evidence of injury at the cervicomedullary junction but little evidence of traumatic axonal injury. These observations were very helpful in redirecting our thinking to the role of such injuries and apnea as a cause of morbidity. Unfortunately, they jumped from there to totally unsupported conclusions that these injuries might result from minor traumatic events. Additionally, they postulated that the apnea with accompanying hypoxia might be the cause of the subdural and retinal hemorrhages seen in these children, instead of a direct result of the original trauma. They have subsequently recanted their “unified hypothesis,” but not until after it had severely damaged child protection.³ The second pattern is an “acute encephalopathic” one, with extensive symptoms of intracranial and retinal bleeding and cerebral dysfunction. Dr Minns placed 53% of the cases in this category and attributed it to shaking or repetitive whiplash injury, with or without accompanying impact. A third group of 21% of the victims had a “subacute nonencephalopathic” presentation, with intracranial bleeding and extracranial evidence of abuse but without the severe neurological symptoms. Finally, 20% had a “chronic extracerebral presentation” of enlarging head size from chronic subdurals and variable symptoms of increased intracranial pressure, but they lack other signs of abuse. The earlier mentioned figures are from a Scottish database, extensively reported in the text. Although these clinical formulations are presented as discrete scenarios, I feel they are more likely extremes, while most children have some blend of these patterns. For example, we often see children with chronic subdurals, picked up by increasing head sizes, who in retrospect have a postnatal jump in head size, simultaneous with a period of irritability and vomiting. Many, on skeletal survey, are also found to have healing rib or other fractures. It is naive to assume these children did not sustain severe head injuries with immediate concussion simply because no one is admitting it or no one recognized the cause of the symptoms at the time.

The Edinburgh, Scotland, group has not limited themselves to clinical characterization of IHI but also initiated biomechanical experimentation. Drs Minns and Morrison present their own dummy-shaking studies, as well as finite element modeling of infant whiplash injury. They, as everyone attempting biomechanical prediction of these injuries, are hampered by the lack of data on how the infant primate brain responds to repetitive shear forces in back-and-forth angular accelerations. They found that adults can shake 2-month-old dummy infants at 3 to 5 cycles/s, for 24 seconds. Heavier infant dummies are shaken at a similar rate for fewer cycles. Their finite element analysis suggests that these rates are in the range most likely to cause damage. Drs Minns and Morrison weigh in on the shake vs impact argument in favor of shaking alone causing some of IHIs. This opinion is based on confessed cases, their biomechanical model's support for the observed injury patterns, and the rare independently observed cases.

Fortes of the text include Dr Minns' viewpoint as a neurologist, offering differing ideas and literature than usually known to pediatricians. Many of the chapters have extensive charts of past literature to illustrate their messages. The chapter on the outcome and prognosis in IHI is a good example, providing the most extensive literature compilation I have seen. Most of the graphics, which are liberally included, are of very good quality and effectively illustrate the educational points. The text is multidisciplinary, providing good clinical, radiological, and pathological information. For example, Dr Jaspan's discussion of computed tomography findings and their evolution is very detailed and perceptive.

Although the text includes many non-UK references, the chapters and reference lists are weighted toward UK sources. As a result, some excellent studies are not included, for example, the North Carolina prospective, population-based incidence study by Keenan et al.⁴ Likewise, the Starling et al study⁵ of perpetrator confessions, which provides the strongest evidence to date that shaking alone can cause injury, is missed. Similarly, I would have liked more data on the difficulties interpreting mixed density subdurals on the initial computed tomograph.⁶ The biomechanical data also are heavily weighted to that from Edinburgh, providing little reference to others, including the recent studies of Raghupathi et al.⁷ Her group has not yet provided, but is closing in on, thresholds for repetitive whiplash shear thresholds for the infant animal brain.

This is not a text for armchair reading but is recommended for anyone actively caring for, evaluating, or testifying in these cases. It does supply some unique information and some alternative viewpoints to much of the literature.

Correspondence: Dr Feldman, 2101 E Yesler Way, Seattle, WA 98122 (kfeldman@u.washington.edu).

Financial Disclosure: None reported.