Health: Even the gentlest shake can kill

THERE ARE many things that put infants' lives at risk, such as congenital abnormalities, infection, or childhood cancer. Now there might be one more to add to the list: crying.

In a recent survey in Holland of 2,000 two-month-olds, mothers were asked what methods they used to soothe an inconsolable child. While most used traditional remedies such as rocking or a pacifier, 8 per cent admitted to slapping or shaking their child.

The dangers of shaking an infant were discussed in the trial of Louise Woodward, and here a registered childminder was convicted of the murder of a toddler in July, allegedly by shaking him. 50,000 cases of "Shaken Baby Syndrome" occur in the United States every year; one in four resulting in a fatality.

The controversy starts even with the name. Some doctors claim it should be called the "Shaken-Impact Syndrome", as in laboratory experiments it is impossible to reproduce the forces needed to cause the damage observed in real life, without striking the dummy's head on a hard surface. Others say these experiments are too simplistic, and ignore the fluidity of brain tissue in living babies. They cite the example of a "snowstorm" paperweight. Hitting it with a hammer may impart the maximum force, but shaking it gently a few times is what really gets things swirling about chaotically.

The principal brain injury caused by shaking is a "subdural haemorrhage", caused by the tearing of small veins inside the skull. Blood leaks into the subdural space, the dura being the outermost of the three layers of meninges, the membranes that cover the brain. Small bleeds may have very vague consequences, like poor feeding, vomiting or lethargy. Bigger ones compress the brain, and can be fatal; most non-fatal bleeds result in long-term disability. In the British Medical Journal last week, a team from the South West examined 33 cases of subdural haemorrhages that occurred in their area from 1993 to 1995. All were in children under two years of age. "In retrospect," the authors say, "27 cases (82 per cent) were highly suggestive of abuse." An alternate view was expressed in an editorial in a rival medical journal, The Lancet, in August. There it was suggested that recent publicised cases "should not lead to the premature diagnosis of `shaken baby syndrome' ". Two experts in the field responded to this editorial by saying it was "singularly unhelpful, [and] misrepresented current clinical thinking".

Despite these striking differences amongst the experts, there are some things about which everyone is in agreement. Firstly, shaking children cannot be regarded as a safe thing to do. Do anything with children that cry excessively - turn on some music, take them out in the pram or car, cuddle them, put them in their cots and go into another room - but never shake them. "Excessive crying" has been defined as crying for more than three out of 24 hours. One researcher put it more simply: "An infant cries excessively when the parents state it cries a lot."

The combination of a head relatively big for its body, weak neck muscles, and a pliable, mobile brain means that an infant withstands shaking far less well than an adult.

Another thing everyone agrees on is that subdural haemorrhage is not enough on its own to establish a diagnosis of child abuse. Some children with wider than usual subdural spaces can bleed into them after worryingly trivial falls. But a subdural haemorrhage should alert doctors to look for other signs of abuse, namely small fractures in the growth plates of other bones, fractures elsewhere of different ages, indicating abuse over a period of time, and the strongest sign that shaking or abuse was the true cause of the injury - bleeding into the membrane at the back of the eyeball; retinal haemorrhages.

The combination of subdural with retinal haemorrhages, along with X-ray signs of other injuries, is enough to confirm child abuse in the minds of most paediatricians. According to the BMJ study, which benefited from hindsight, physical abuse was very much under- diagnosed at the time in the group of children they studied. Opportunities to intervene and prevent further abuse were therefore missed.

The finer points of the mechanisms, pathology and nomenclature of Shaken Baby Syndrome will be debated for some time yet. But a vigorous shaking can do as much damage as dropping a child head first on to a concrete floor - something that all parents and those in charge of children should bear in mind, preferably before they end up in court.