All paediatricians, paediatric radiologists, paediatric pathologists, forensic pathologists, and many other specialists have to deal with cases of suspected child abuse, and in terms of the generation of complaints from families this is a high risk activity. Many complaints are devoid of merit, but in some cases a complaint is justified because of a faulty approach. This review draws attention to the avoidable pitfalls associated with report writing when child abuse is under consideration (see box 1).

The modern guidance on the duties of experts in cases of suspected child abuse stems in part from a care proceedings case heard in 1990. A 3 month old baby was admitted to hospital with serious injuries comprising subdural haemorrhages of more than one age, damage to the brain, multiple rib fractures, and multiple limb fractures. Despite the evidence pointing to child abuse, a number of experts from various disciplines offered a variety of unjustified or highly improbable alternative medical explanations (including abnormally fragile bones) for the injuries. The judge was highly critical of some of the expert reports and evidence given in court, and in his published judgment he gave clear advice concerning the duties of an expert.¹ This and subsequent advice from various authorities on the work of an expert^2–12 is summarised in box 2. In addition, a High Court judge of the Family Division has written a useful handbook providing simple and practical guidance on all aspects of the duties of experts in care proceedings.¹³

DIFFERENT LEGAL PROCESSES

Care proceedings in England, Wales, and Northern Ireland involve a two stage process. The court must be satisfied, firstly, that a child is suffering or is likely to suffer significant harm attributable to care given to the child or its absence. These are known as the threshold criteria. If they are met, the court goes on to consider whether a care order is warranted on welfare grounds. Sometimes these two aspects are dealt with at separate court hearings. The first hearing, sometimes referred to as a “finding of fact hearing”, is devoted to whether or not a child has been harmed, and if so by whom. If the finding of the court is that the child has been harmed, then following a period of further investigation, including a risk assessment, and maybe a psychological assessment (of parents and/or children) and/or a psychiatric assessment, a second hearing (often rather inelegantly called “disposal”) decides whether or not a final order should be made, and if so what long term plans should be made to safeguard and promote the child’s welfare. The input of paediatric reports is mainly into the first part of the process. Parties to the process are the local authority, the child (whose interests are represented by a children’s guardian), the parents, and possibly other relatives. The paramount concern is the welfare of the child.^13,14

In criminal proceedings, the focus is on the criminal prosecution of one or more individuals who are suspected of harming or neglecting a child.

In some cases, civil and criminal proceedings run more or less side by side, and this carries the potential, discussed later, of directly interfering with the work of a paediatrician whose task is to prepare a report.

PROTOCOL FOR CASE MANAGEMENT IN CARE PROCEEDINGS

To avoid undue delay in making decisions about the future of a child, on 1 November 2003 a protocol for judicial case management was introduced.^14,15 The aim of this protocol is to complete a case within a maximum of 40 weeks, and one effect of this is to put pressure on doctors to provide reports by a specific date. Appendix C to the protocol provides a short but helpful code of guidance for experts.

MOST READERS OF REPORTS ARE NOT DOCTORS

Because most readers of reports in cases of suspected abuse are not medically or scientifically trained, it is essential to avoid or to explain any jargon. It is often helpful to provide a glossary that lists and briefly explains any technical terms.

When writing reports be aware that non-medical readers of the report will not necessarily apply appropriate caution to medical opinions, for example the dating of bruises or fractures. It is worth bearing in mind that some take medical opinion as gospel, rather than appreciating the limits of medical science.

THE OVERRIDING DUTY OF THE DOCTOR IS TO THE COURT

Regardless of whether one is dealing with civil or criminal proceedings, the overriding duty of the doctor preparing a report is to the court, and not to any of the parties such as the prosecution, the defence, the local authority, the parents, or the child. This contrasts with the primary responsibility in clinical care, where the primary duty is to the child. The doctor who cannot come to terms with this fact may with the best of intentions sow the seeds for serious criticism as events unfold. The scenario is simple enough: a child is admitted with what is thought to be abuse; the paediatrician, convinced that the child needs to be protected from the presumed perpetrator, unwittingly or deliberately exaggerates items in the report, such as the likelihood of abuse, the degree of pain suffered, or the amount of force inflicted. The aim of the doctor is to do everything possible to ensure that the child is protected. This is how the caring and concerned doctor can take the first steps on a path that ultimately may lead to criticism. In the face of a child with worrying injuries, it is essential to resist a natural tendency to overstate or exaggerate aspects of a report.

THE MEDICAL CHRONOLOGY; A USEFUL TOOL

It can sometimes be helpful to one’s own thinking, and is useful to non-medical people involved, to prepare a medical chronology that lists exactly what happened and when, taking care to include information from typewritten sources (letters, discharge summaries), handwritten medical notes, handwritten nursing records (which often contain information not included elsewhere), and the results of investigations.

INTENT AND MOTIVATION

The binary terms “accidental” and “non-accidental” and “unintentional” and “intentional” are often employed. However, they mean different things, and both pairs have their drawbacks. There are two key points to be made. The first is that intent to harm a child is uncommon. In most cases of child abuse, the adult loses control of his or her reactions and injures a child. The deliberate and calculated act of setting out, in cold blood, to harm or injure a child is believed to be rare. It is therefore incorrect and needlessly inflammatory to label all cases of child abuse as “intentional”, a term that is often equated by lay people with the word “deliberate”. The second point is that no-one can look at an injury and know with certainty the intent of the person who caused the injury, although there are a few particularly extreme cases where it is only natural to wonder what was going through the mind of the person who caused the injuries. The key point is that the job of the paediatrician (or radiologist or pathologist) is to distinguish between accident or naturally occurring disease on the one hand, and abuse on the other. The paediatrician cannot know the mindset of the person believed to have caused the injury, and speculation on this aspect should be avoided. Once it has been established by the court that the case is one of abuse, then experts in social work, psychology, and psychiatry can become involved and examine motivation.

Paediatricians should not try to be mind readers. The implication is that the terms “intentional” and “unintentional” should be avoided in paediatric reports on the topic of suspected child abuse.

INTERVIEWING THE PARENTS OR CARERS

When a child is referred to a paediatrician because of a clinical problem, the paediatrician has three main tools: the history, physical examination of the child, and a study of the available data including investigation results and previous records. Most paediatricians would not dream of giving a clinical opinion without taking a history, and yet it is not uncommon for paediatricians to undertake a “paperwork exercise” of preparing a report without ever meeting the child or the family. While there is a desire to avoid needless repetition of physical examination of small children, particularly long after an injury has occurred, it is hard to justify avoiding speaking to the parents or carers.

It is likely that the paediatrician who has been asked to conduct a paper exercise will have available witness statements from the parents and/or carers and others, and there may also be a transcript of police interviews. There should also be copies of the medical records. Alas, none of these is a substitute for a properly taken history of ones own. However detailed are interviews by police or lawyers, neither have the medical knowledge, training, or expertise to take a medical history. Furthermore, the medical records all too often contain incomplete histories obtained by the admitting junior doctors. The fact is that all clinicians perform their clinical duties with the aid of various routines, and departure from these routines by the use of short-cuts may be associated with a greater risk of error. Interviewing parents or carers is time consuming, and may be inconvenient and cause delay, but preparing reports in cases of possible child abuse is not an activity where a doctor should contemplate second best.

A paediatrician who does not attempt to interview the parents risks being criticised for by-passing the usual routines and failing to consider all aspects of the case. Recent press publicity has indicated that parents may be exceptionally aggrieved by paediatricians willing to make a confident diagnosis of abuse without ever meeting the parents and giving them the opportunity to explain their side of the story.

Although civil and criminal proceedings arising out of the same injuries are separate processes, there are various legitimate ways that information can flow between the two. Thus it is that in some care proceedings, lawyers acting for parents may advise them against being interviewed, for fear of disclosures that could assist an ongoing police investigation or criminal prosecution. The practice of giving such advice has been “strongly deprecated” by a High Court judge of the Family Division, notwithstanding the fact that the parent sought to be interviewed was suspected of killing two of her children.¹⁶

Having mentioned the importance of interviewing carers, in the light of recommendation 65 of the Victoria Climbié Inquiry report,¹⁷ it is worth also remembering (in selected cases) the importance of considering interviewing the child.

STUDYING THE MEDICAL RECORDS

It sometimes happens that one is invited to prepare a report on the basis of very limited information such as a telephone call, a few photographs, or a couple of witness statements. Requests to prepare reports in child protection cases without access to vital materials such as the child’s medical records should always be resisted; it is absolutely essential that the full medical records are sought.

EXAMINING THE PATIENT

The need for a child to be re-examined specifically for the purpose of preparing a report will depend on the circumstances. It is important to give early notice of a need to examine a child and the reasons for the examination, as children who are the subject of care proceedings cannot be examined without the consent of the court.

CLINICAL PHOTOGRAPHS

It is common for solicitors to fail to provide photographs, sometimes because they do not know of their existence. A study of the nursing section of the hospital medical records may uncover the fact that a hospital or police photographer visited, and one may have to ask for these photographs to be obtained. One should insist on good quality glossy prints of original photographs, enlarged if necessary, and not make do with laser prints or colour photocopies.

It is worth remembering that photographs, particularly those of very extensive injuries, have the potential to be highly prejudicial, which is why in many abuse cases photographs of the dead and injured child may well be withheld from the jury. Paediatricians are only human, and require considerable self-discipline to stand back and try to make an objective analysis of photographic evidence.

Photographs or video recordings of the home, the scene, and any domestic items should be viewed if they are available. They may be helpful in determining, for example, whether or not a particular fall was from a low level.

It is important to bear in mind that for a variety of reasons, photographs of suspected injuries can be misleading. Photographs may fail to show lesions that were undoubtedly present, either because the quality of the image is poor or because the lesion was particularly difficult to photograph. In some situations photographs can suggest the presence of lesions or injuries that do not exist, and in a recent case no less than five experts were seriously misled into incorrectly diagnosing abuse because two-dimensional colour photographs failed to accurately reflect a three-dimensional normal variant appearance.¹⁸

ESTIMATING THE AGE OF BRUISES

It is common to be asked to provide an opinion on the age of bruises. Textbooks have provided schedules of colour change of bruises over time. However, these schedules are unreliable,¹⁹ and should not be used. A few rather limited studies of the colour changes of bruises, some using colour photographs and some using direct observation, all indicate that attempts to age bruises based on their colour is fraught with difficulties.^20–23 The time course of the appearances of bruises may vary with the location, depth, extent, and nature of a bruise. The only established fact is that the presence of a yellow colour within a bruise indicates that it is at least 18 hours old.^20,21

It is perfectly reasonable for an experienced doctor who has examined the child to state that a purple bruise “looked fresh”, meaning it was unlikely to have been inflicted more than two days previously,¹⁹ but considerable caution is required when estimating the upper and lower age limits for bruising.

Histopathology of bruises is worth performing in some fatal cases. Histology can rule out skin lesions such as blue naevi and can confirm bruising in those with dark skin. Histology of a bruise may enable the pathologist to say it is very fresh (no cellular reaction), recent (infiltration with neutrophils), or more than approximately two days old (presence of haemosiderin).

BLOOD COAGULATION TESTS

A frequent error is to believe that a normal “clotting screen” (such as a full blood count, prothrombin time, and partial thromboplastin time) excludes the presence of a coagulation disorder. It does not. It only excludes the commonest conditions that may cause spontaneous bleeding or serious bleeding following trauma.²⁴ A “clotting screen” is no more than a screening test, and if there are pointers to a coagulation disorder (such as a history of a bruising or bleeding tendency in the patient or the family), referral to a haematologist and further investigation will be required. There are a number of coagulation disorders, individually but not collectively very rare, that may pass undetected by a “coagulation screen”. Just one example is the Hermansky-Pudlak syndrome, normally a mild disorder of platelet function, but which in one 7 week old infant was associated with both subdural and retinal haemorrhage.²⁵

THE INTERPRETATION OF RADIOGRAPHS

It is essential that radiographs are reported by radiologists who are familiar with the broad brush of children’s radiology, particularly in relation to the different patterns of injury in accidents. It is also essential that the radiologist has knowledge and experience of paediatric disease, its effects on the radiographic appearances, and the numerous radiological appearances which are normal variants. In practice this means that the films need to be seen by a paediatric radiologist. Copy films vary in quality, and numerous errors have been made as a result of relying on copy films rather than ensuring that the original films are supplied.

Caution is required when interpreting routine clinical reports on radiographs. The original request form for the investigation will usually be very restricted (for example “? fits ? cause”) and often omit essential clinical information. The report provided by the radiologist at the time may not have been written with forensic investigations or legal proceedings in mind. The author of the report may need to be approached to clarify matters or provide further information.

FORCES NEEDED TO CAUSE INJURY

A 3 month old infant is found to have unexplained healing rib fractures and metaphyseal limb fractures, and one is asked to comment on the forces required to cause the injuries. One is well advised to exercise caution and avoid dogmatic statements about how flexible are the bones of babies and how enormous the forces must have been. The truth is that the requisite scientific studies, in which human infants are deliberately injured in different ways and with different degrees of violence in order to establish precisely the nature and extent of the forces required to produce injuries, have not been performed and never will be. The only exception to this is the unusual studies by Weber in which dead babies were dropped on to various surfaces from a height of 82 cm in order to study the genesis of skull fractures.^26–28 Biomechanical studies using theoretical constructs or various models cannot provide the required human data. The nature of child abuse is such that, unlike accidental injuries, the reliable accounts of independent witnesses are rarely available. Confessions of perpetrators may tend to understate the degree of force. The only really solid ground is that in an infant with healthy bones, normal handling and normal activities do not produce fractures, and domestic accidents (such as short falls) rarely produce significant injury. It is self evident that significant force must be needed to break a bone, but in the absence of any reliable hard scientific data it is wise to avoid overstating the amount of force that is likely to be involved.

PAIN RESULTING FROM FRACTURES

Fractures cause two kinds of pain. One is acute pain resulting from the forces applied to the bone and the pain resulting from the bone breaking. The other is ongoing pain occurring in the days and weeks after a fracture has occurred. The immense variability means that over-confident assertions are worth avoiding. While the occurrence of the fracture itself is certain to cause significant immediate pain, the way that this pain is communicated to carers or parents can vary between different children and at different ages.

In some cases the ongoing pain is obvious for all to see, and the limb of a child is manifestly not being used for some days after a fracture has occurred. However, ongoing pain after a fracture is highly variable, and, for example, in infancy, rib fractures and metaphyseal limb fractures often produce no detectable ongoing pain at all, presumably because the bone is stable and the periosteum often little disturbed. These injuries commonly pass undetected by clinicians examining the child and the nursing staff looking after the child. This is sometimes because the fractures are in a state of healing, and sometimes because something else is dominating the clinical picture. The point is that caution is required before concluding that a reasonable carer should have known that something was seriously amiss in a child with rib or metaphyseal limb fractures.

BIAS

It requires constant vigilance to avoid bias creeping into a report. On reading the medical records of a child with a suspected combination of abuse and neglect, one finds that the patient has failed to attend appointments in the hospital or the community, and accordingly (and quite innocently) one adds to the list of problems “three failed outpatient appointments”. What is at fault is that probably quite inadvertently one has made a point of including negative information without putting it into context, and without providing the available positive information. In fact, although our hypothetical patient missed three appointments, he or she kept 15 others. Furthermore, when the child was unwell the mother always made a point of seeking medical help. Finally, the data for the family have omitted the normative data. Many other babies in the neighbourhood may be missing just as many appointments; this is not an excuse for missing the appointments, but it is important contextual information.

Selective extraction of negative information is one of the most common faults in medical reports. Vigilance and a strong sense of fair play are needed to avoid this trap. There are few human beings for whom one can find nothing positive to say.

A doctor who is involved in new research should be conscious of the natural tendency to promote his or her own findings, and should make every attempt to avoid becoming subjectively biased. It is essential that an expert considers and mentions in a report all relevant material, including that which tends to throw some doubt on the expert’s conclusion.¹¹

PARENTING ISSUES

Instructions to experts rarely invite conclusions about the quality of parenting, but often include a catch-all instruction inviting comment on any issues not covered in the instructions. The doctor preparing a report notices in the papers that one or more previous children, siblings of the child under consideration, had to be placed in foster care a number of times. In addition, the mother has failed to take the child to a number of appointments, and there are suggestions that the children have not been fed regularly, are filthy dirty, and are understimulated. After careful study, the doctor becomes anxious about this mother’s abilities to look after children. The actual evidence of abuse is far from clear, but the doctor is seriously worried about the welfare of the child. The pitfall for the caring doctor, fearful that in the absence of evidence of abusive injuries the child is going to be left in the care of this apparently inadequate mother, is to try to compensate for the lack of hard evidence of abuse by over-stressing the general concerns. This is the beginning of a slippery path that leads to the final but wholly inappropriate conclusion, along the lines of “this mother is totally unfit ever to have the care of children again”. However experienced, doctors should never make recommendations of this sort. This particularly applies to experts parachuted into the case who have never met the family or visited the home! Decisions about the child’s future are for the court to determine, after appropriate input from those with the relevant skills. Paediatricians must remember they are not social workers, clinical psychologists, or child psychiatrists.

It is perfectly valid, indeed quite proper, for a doctor to highlight areas of concern. But however grim the circumstances, and however valid the criticisms of the child’s care, a parent is likely to react with great bitterness when sweeping recommendations are made by doctors, particularly if the doctor has not met the family, or visited the home, or been involved in the care of the children in recent years. Furthermore, courts may accord an exaggerated respect to a doctor’s views about non-medical matters, making it all the more important to avoid excessive comments about poor parenting.

USING A PROFILE OF A CARER TO MAKE A DIAGNOSIS

A 4 month old infant is found to have multiple rib fractures; the radiological appearances (different stages of healing) suggest at least three separate episodes of injury, accompanied by two metaphyseal fractures of the long bones and a human bite mark. The carers are unable to offer any explanation as to how the injuries might have occurred. The fact that the mother’s current partner is not the father of the child, was himself in care in his own childhood, and has a record of violence, is of no value in proving that the diagnosis is one of abuse, a diagnosis which must be based solely on the features of the injuries and the lack of any explanation.

It is well recognised that stepfathers can sometimes harm a child, but the fact that a child with an injury has been in the care of a stepfather is not an indicator that injuries are the result of abuse. As with cigarette smoking and lung cancer, being a stepfather, having been in care in one’s childhood, and a previous criminal record for violence are risk factors for abuse. Risk factors cannot be used to make a diagnosis, but they can help explain why a condition has occurred, and they may inform a risk assessment of future harm.

These simple facts are sometimes overlooked in cases of Munchausen syndrome by proxy (the current fashion is to call this “fabricated and induced illness”). Instead of focusing on the primary issue, such as “has this child been suffocated, causing attacks of apnoea?”, or “has the history of haematuria been fabricated?”, the mistake is to attempt to make the diagnosis by constructing a personality profile of an alleged perpetrator and then show that the adult carer in question falls into the described pattern. This is a serious error.

BE PREPARED TO DISSECT AND ANALYSE INDIVIDUAL COMPONENTS

When making a diagnosis, the treating paediatrician may look at the overall picture, but those investigating a child with multiple injuries (such as the police, or social services) will almost certainly consider each injury separately, as will the courts. To anticipate this, it is advisable in a report to discuss the nature and possible causes of each injury, one by one. The evidence for abuse may be weak or non-existent for individual lesions, but the overall pattern may nevertheless point to non-accidental injury, maybe because of the extent or distribution of the injuries.

Another reason for considering each injury separately is that in some criminal cases one will find that for legal reasons information about certain injuries may be inadmissable during a court hearing. In this situation the ability to consider an injury independently of other injuries may be essential.

YOUR AREA OF EXPERTISE

It is a mistake to comment on an area that is outside one’s expertise. An expert should have sufficient practical experience in the area.¹¹ This means, for example, that a histopathologist whose expertise is in conducting postmortem examinations and dealing with the examination of tissue samples is not in a position to comment on the symptoms, diagnosis, treatment, or prognosis of living children. Lawyers may have a poor understanding of the different areas of medical expertise, and a question not in one’s domain should be firmly redirected. This can be difficult if the relevant expert (such as a paediatric haematologist) has not been instructed. The temptation to try to be helpful and “have a stab” at the question must be resisted.

QUOTING REFERENCES TO THE MEDICAL LITERATURE

It may be relevant to refer to the medical literature. For example, when asked to provide an estimate of the age of some bruising, it may be helpful to refer to the published literature on the subject, if only to show how little objective data is available. What is not acceptable is to selectively provide a few references that bolster the point of view that one wishes to advance, while ignoring all material that points in other directions. The volume of medical literature is such that one can usually find some articles to support almost any point of view, however outrageous. The court needs a balanced assessment of the possibilities.

It is important to bear in mind that a high proportion of the medical literature on child abuse is deeply flawed. Much of the published material is contained in case series, and these always tend to over-represent particularly severe cases. It is unsafe to generalise from this kind of uncontrolled and often highly selected data.

NEW MEDICAL THEORIES

The place for new medical theories is medical journals and scientific conferences. Everyone agrees that court is the very worst possible place in which to float new theories. Doctors who use a court case of suspected abuse as a test-bed for a new unproven theory do the court a disservice and are inviting complaints from colleagues.

READILY ADMIT UNCERTAINTY

A number of publications on the subject of child abuse divide a case series (say of toddlers with fractured femur) into two categories, those who were abused and those who were not. Such publications are to be distrusted, as they fail to acknowledge the existence of a large category of “uncertain” cases in whom one never really knows whether or not the injury was the result of accident or abuse. The reality is that there are many cases in which doctors can do little more than voice suspicion or anxiety. When writing a report, one may feel (or actually be placed) under considerable pressure to express an undue degree of boldness and confidence, for fear that to do otherwise may weaken the case for the local authority or the prosecution. Regardless of pressure, never be afraid to say that one is simply not sure. “Unexplained bruising” or “unexplained bruising ? accidental ?non-accidental” are acceptable terms that plainly indicate a genuine diagnostic difficulty.

EVERYONE WILL READ YOUR REPORT

In clinical practice one may be accustomed (rightly or wrongly) to saying or writing things that are not shared with the parents. Bear in mind that whatever you write in a report for court will be disclosed to the parents, and in a criminal case to the person who is accused. In the case of care proceedings, this applies not only to one’s report but to anything else that one writes or says to any of the parties. It is good practice to assume that one will be asked to justify anything that one has said or written. In a report one should make a point of providing the reasons for having reached particular conclusions.

POINT OUT ANY LIMITATIONS OF YOUR REPORT

Through no fault of one’s own, there may be all sorts of limitations to a report. Certain papers may have been unavailable. Other reports may be awaited. One may have been unable to interview the parents. The photographs or radiographs supplied may have only been copies rather than originals. Be frank about any limitations.

DISAGREEMENT; MAINTAINING A PROFESSIONAL APPROACH

When dealing with a case of suspected abuse, one may be confronted by the report of a colleague with which one profoundly disagrees. Anger or bewilderment are common reactions. Here is some advice:

• Legitimate disagreement can occur, a legacy of the lack of objective controlled studies of what happens when someone injures a child.

• However strongly one may feel about a case (and maybe particularly if one feels strongly about a case), do not reject the alternative viewpoint out of hand, but give it the most careful consideration, and maintain a balanced and professional approach. It does no harm to ask oneself whether it is just possible that there is at least some merit in the opinion that has been advanced.

• Never ever consider a child protection case in terms of winning or losing. That is for advocates, not doctors. The challenge for the doctor writing a report is not to help win a case, but to do a careful, thorough, and honest piece of work to the very best of one’s ability.

• If, on reflection, one realises that one has arrived at an incorrect conclusion, it is a strength and not a weakness to readily acknowledge this. It is a common experience at experts’ meetings (a topic that is discussed elsewhere²⁹) that opinions change when one has a chance to better understand the reasoning of a colleague, or when one learns of new facts of which one was unaware. Experts who change their opinions for good reason on receipt of fresh information are respected by the court rather than criticised. However, if one changes one’s opinion, one should always explain the reasons for the change.¹³

• All doctors make mistakes. The most serious error is to refuse to admit one has made a mistake, even when it is pointed out.

• Remember (and take comfort from the fact) that the ultimate responsibility for making decisions rests with the court.

CONCLUSIONS

The task of distinguishing between natural disease or accident on the one hand, and abuse on the other, is often difficult. However skilfully the situation is handled by the doctors involved, there will always be a risk that parents or carers will react badly on learning that a diagnosis of abuse is under consideration. However, reports prepared for court hearings will inevitably be exposed to the very closest scrutiny, and it is worth being extra careful and cautious when preparing such reports.