isolation, breaking with tradition, and the quality of teaching should also pay dividends in the development ofprimary care. but became more positive when they considered the implica- One way of doing this would be to set up academic practices tions of learning in the community. The authors say that with an additional partner to reduce service loads and "telling is selling"; discussion with interested groups tended contracts to provide both teaching and research.9 This would to increase support. also help to prevent the potential isolation of students outside One important barrier to any change of site for medical a hospital base but would again require funding additional to education is the cumbersome funding system, whereby that proposed at King's College. the complex funding provided jointly by the Department In all these developments there is a risk that primary of Health and the Higher Education Funding Council is care departments may be seen as competing with hospital divorced from educational accountability, which is itself specialties and basic sciences for both curricular time and unclear. In addition, the provision of the service increment funding. It is vital that all change is underpinned by a clear for teaching and research is governed by rules that do not educational philosophy related to the aims of the overall allow the increment to be transferred outside the hospital curriculum. Hospital specialists also serve the community, system. At present the Department of Health (through the but their views were not sought in the King's College project. family health services authorities) provides small fees for the Examples of good teaching in the present curriculum and teaching of undergraduates in general practice but makes no house officer posts should not be lost but reinforced. Com- formal provision for the supervision of house officers. The bined staff development sessions, such as those reported at report from King's College recommends that medical educa- King's College, could both reduce any tension and improve tion should be both funded and monitored for quality by the the confidence of staff working outside the hospital setting. Higher Education Funding Council. Medical schools would Finally, all change must be carefully evaluated. Evidence allocate funds to the various sites where students learn. based education is as important as evidence based medical Two further proposals should be considered. Firstly, practice. teachers need training to develop their abilities. This is a task JENNY FIELD already taken seriously in the education of postgraduate Senior lecturer trainees in general practice, and without it consistently high ANN-LOUISE KINMONTH quality teaching is unlikely to be maintained in any context.7 Professor Primary Medical Care, Because formal training in educational skills has been largely University ofSouthampton, absent in hospital settings, reorganisation of the present Southampton SO1 6ST funding system might not cover staff development and new ways to teach teachers might have to be devised. 1 Silver GA. Victim or villain. Lancet 1983;ii:960. Secondly, teaching and research in the community requires 2 General Medical Council. Tomorrow's doctors. London: GMC, 1993. 3 Seabrook M, Booton P, Evans T, eds. Widening the horizons of medical education. London: King's an infrastructure, including space, administration, and Fund, 1994. information technology. Teachers and researchers in general 4 Wharton M. Clinical training in the community: a holistic approach. BMJ 1995;310:407. 5 Harris CM, Dudley HAF, Jarman B, Kidner PH. Preregistration rotation including general practice practice need an ability to appraise evidence from scientific at St Mary's Hospital Medical School. BMJ 1985;290:1797-9. research and clinical examination. Because they share these 6 Wilton J. Preregistration house officers in general practice. BMJ 1995;310:369-72. 7 Entwistle N. The impact of teaching on learning outcomes in higher education: a literature review. needs and research is likely to be more effective when Sheffield: Committee ofVice Chancellors and Principals, 1992. 8 Haines A, Jones R. Implementing findings ofresearch. BMJ 1994;308:1488-92. clinicians and researchers share priorities and values,8 closer 9 Royal College of General Practitioners. Conference of academic organisations in general practice. links between clinical teachers, researchers, and clinicians Research andgeneralpractice. London: RCGP, 1994.

The shaken infant syndrome Parents and other carers need to know ofits dangers

In children less than 1 year, non-accidental injury is the suggest abuse-signs include irritability, lethargy, vomiting, commonest cause of serious head injury'-much of it result- convulsions, apnoea, shock, and fluctuating consciousness. ing from shaking and impact. Last week saw the launch of The history may be confusing ifthe adult bringing the child to a leaflet, "Handle with Care," produced by the National medical attention has no knowledge that the child has been Society for the Prevention of Cruelty to Children and the shaken or withholds information. Shaking injuries that Department of Health, in response to increasing professional produce subtle clinical changes may never be brought to concern over the dangers ofshaking babies. medical attention. There may be a delay in seeking medical John Caffey first drew attention to the association of help because the perpetrator believes that the shaken child is fractures of the long, bones and subdural haematomas, and he asleep rather than unconscious. After a variable time, the implicated whiplash injury as the cause of the intracranial infant will develop signs of cerebral irritation, cerebral trauma.2 It is accepted that shaking alone can cause the brain oedema, or intracranial haemorrhage.4 Acute deterioration, damage,' and it is now recognised that the infant's head also convulsions, or respiratory or circulatory arrest may follow. undergoes rotational forces as well as whiplashing during Although clinical examination of the infant may reveal shaking.4 Injuries may result from single or multiple episodes bruising or other evidence of neglect in addition to the ofshaking. neurological signs, it often shows nothing unusual. Retinal The severity of the shaking force is such that shaking haemorrhages are present in between 50% and 80% of injuries cannot occur in any form ofplayful activity (as Caffey patients,6 and when other causes have been excluded are originally suggested2). A recent description of the act of virtually pathognomonic of child abuse. Ideally, the fundi shaking states that it is so violent that neutral observers would should be examined by an ophthalmologist who frequently recognise it as dangerous.' examines children. If the cerebrospinal fluid is examined The clinical presentation of a shaking injury may not frank haemorrhage or xanthochromia indicate acute or recent

344 BMJ VOLUME 310 11 FEBRUARY 1995 haemorrhage, but this investigation is not done when there lethargy but no lacerations, cerebral infarction, dr severely are concerns about raised intracranial pressure. Where safe, a raised intracranial pressure had subtle neurological sequelae subdural tap should be done as the analysis of cerebrospinal or persistent convulsions.'3 fluid is helpful in the dating ofinjuries. Who shakes their child? The shaking may represent a During the shaking episode the infant is often held by the response to tension and frustration generated by the infant's thorax4 and the compression forces on the ribs may result in incessant crying, which may be exacerbated by ignorance of fractures. Alternatively, the child may be held by the appropriate infant care. People experiencing stress may be shoulders and upper arms or feet.7 The squeezing and violent more prone to impulsive and aggressive behaviour.'4 movement associated with the shaking may cause the typical Until now social and medical effort has focused mainly fracture patterns of child abuse.8 Brain damage is often on diagnosing and treating the shaken infant, with little multifactorial with direct shaking injury to the brain being emphasis on prevention. Studies in the United States have compounded by hypoxic and ischaemic injury, infarction, shown that although between a quarter and a halfofthe public coning, and the pressure effects of subdural haematomas and did not know of the dangers of shaking infants, they retained impact trauma.7 information given in an awareness campaign.'5 Computed tomography in suspected cases ofshaking injury We do not know how many infants are shaken in Britain is essential. Subdural haematomas, sometimes of different each year, but over 100 deaths occur from child abuse or ages; subarachnoid blood; intracerebral and intraventricular neglect and in many of these children death is due to brain bleeding; cerebral oedema; diffuse loss of differentiation injury. Increased public awareness of the dangers of shaking between grey and white matter; and cerebral laceration and should reduce this number and reduce the disabilities of contusional tears may all be found in shaken infants.9 Skull survivors. radiology is essential as fractures may be missed on computed HELEN CARTY tomography.10 Magnetic resonance imaging supplements Consultant radiologist computed tomography by showing small subdural collections JANE RATCLIFFE Consultant in paediatric intensive care or identifying subdural haematomas of different ages and Alder Hey Children's Hospital, shaking injuries that are not visible on computed tomo- Liverpool L12 2AP graphy."' Owing to its availability computed tomography remains the primary imaging procedure. More recently, Drs Carty and Ratcliffe have been involved in the campaign to increase public high resolution ultrasonography performed through a patent awareness ofthe dangers ofshaking babies. fontanelle has been shown to be very sensitive in showing 1 Billmire ME, Myers PA. Serious head injury in infants: accident or abuse? Pediatrics 1985;75: 340-2. s shearing injuries and subdural and subarachnoid fluid.12 Late 2 Caffey J. On the theory and practice of shaking infants: its potential residual effects of permanent results of brain include brain damage and mental retardation. AmJDis Child 1972;124:161-9. injury multicystic encephalomalacia, 3 Hadley MN, Sonntag VKH, Rekate HL, Murphy A. The infant whiplash-shake injury syndrome: obstructive or communicating hydrocephalus, cerebral a clinical and pathological study. Neurosurgery 1989;24:536-40. 4 Kleinman P. Diagnostic imaging in infant abuse. AmJRoentgenol 1990;155:703-12. atrophy, infarctions, and gliosis. 5 Committee on Child Abuse and Neglect, American Academy of Pediatrics. Shaken baby syndrome: The most severe shaking injuries tend to occur in younger inflicted cerebral trauma. Pediatrics 1993;92:872-5. 6 Levin AV. Ocular manifestations of child abuse. Ophthalmology Clinics of North America infants as the head is relatively large in relation to the body. As 1990;3:249-64. body weight increases and neck muscles strengthen the 7 Brown JK, Minns RA. Non-accidental head injury, with particular reference to whiplash shaking injury and medico-legal aspects. Dev Med Child Neurol 1993;35:849-69. incidence ofbrain damage due to shaking falls and is rare after 8 Carty HML. Fractures caused by child abuse.JBoneyointSurgBr 1993;75:849-57. the second year oflife. 9 Han BK, Towbin RB, De Courten-Myers, McLaurin RL, Ball WS. Reversal sign on CT: effect of anoxic/ischemic cerebral injury in children. AmJRoentgenol 1990;154:361-8. The outcome depends on the severity of the shaking injury; 10 Saulsbury FT, Alford BA. Intracranial bleeding from child abuse. The value of skull radiographs. PediatrRadiol 1982;12:175-8. morbidity and mortality are high when the infant is comatose 11 Sato Y, Yuh WT, Smith WL, Alexander RC, Kao SC, Ellerbroek CJ. Head injury in child abuse: on presentation.35 Long term sequelae include profound evaluation with MR imaging. Radiology 1989;173:6537. 12 Jaspan T, Narborough G, Punt JAG, Lowe J. Cerebral contusional tears as a marker of child mental retardation, spasticity, motor dysfunction, blindness, abuse-detection by cranial sonography. PediatrRadiol 1992;22:23745. convulsions, and hydrocephalus.5 In one series of children 13 Sinai SH, Ball MR. Head trauma due to child abuse: serial computerised tomography in diagnosis and management. South MedJ3 1987;80:1505-12. shaken to unconsciousness 60% died or had profound 14 Dykes U. The whiplash shaken infant syndrome: what has been leamed. Child Abuse Negl mental or severe motor 1986;10:21 1-21. retardation, spastic quadriplegia, 15 Showers J. "Don't shake the baby": the effectiveness of a prevention program. Child Abuse Negl dysfunction. Others who had convulsions, irritability, or 1992;16:1 1-8.

New treatments for multiple sclerosis May delay deterioration

Encouraging progress has been made in the development of Neurologists held last October. Because much uncertainty treatments that reduce disease activity in multiple sclerosis. exists about the potential role of these very expensive The relation between disability and such activity (as reflected treatments it is important that the present position is clearly in the relapse rate or the appearance of new lesions on understood. In the review that follows, the figures for the two magnetic resonance imaging) is complex, but recent evidence recent reports have been derived from information provided suggests that it may be possible to delay, though not yet by the sponsors of the trials2 3; the data on which the abolish, the progress ofirrecoverable neurological deficit. conclusions of these two studies are based have yet to be The results of three large randomised controlled trials in subjected to peer review. ambulant patients with little or moderate disability have been In a trial of interferon beta-lb 372 patients with relapsing announced during the past 18 months. Only one has been and remitting disease were allocated to placebo or a low dose fully reported in the medical literature,' but two others (1 6 MIU) or high dose (8 MIU) ofthe drug, which was given were widely discussed at a joint meeting of the American by subcutaneous injection on alternate days. After two years Neurological Association and the Association of British the relapse rate (the primary end point) was significantly

BMJ VOLUME 310 11FEBRUARY1995 345