Number % Medication/vaccination error 10 5.1 Delayed/failed diagnosis of septicaemia 8 4.1 Delayed/failed diagnosis of meningitis 7 3.6 Extravasation 7 3.6 Delayed/failed diagnosis of unspecified sepsis 6 3.1 Delayed diagnosis of anorectal abnormality 6 3.1 Delayed/failed cardiological diagnosis 6 3.1 Delayed diagnosis of appendicitis 6 3.1 Misdiagnosis of epilepsy 6 3.1 Delayed diagnosis of a fracture 4 2.1

The other incidents included a delay or failure in diagnosing a number of conditions including brain tumours, tumour recurrence, testicular torsion, bowel perforation, shunt blockage, Turner’s syndrome and intussusception. Further events that led to successful litigation were gastrostomy-related errors, cold-light injuries and pressure sores.

The NHSLA were also able to provide a breakdown of the causes of the errors that led to litigation in these 195 cases. These are summarized in Table 1.2:

Table 1.2 Causes of incidents leading to successful litigation, 01/04/2005–31/03/2010

The NHSLA listed the outcomes that resulted in these 195 incidents. The most common outcome was death (74 out of the 195 cases), followed by unnecessary pain (35 cases), unnecessary surgery (16) and brain damage/developmental delay and scarring (both 12 cases); 5 cases resulted in amputations, 3 in visual problems and 2 in a perforated appendix.

The total cost of litigation in each of the 195 cases (including the damages and the costs) ranged from £600 to £3,044,943.

We also obtained information from NCAS which is a division of the National Patient Safety Agency. It provides advice, support and formal assessments of doctors working in the United Kingdom, the Isle of Man, Channel Islands and Gibraltar. The nonclinical reasons for referral to this service comprise predominantly behavioural and conduct issues such as poor team working and bullying. But we concentrated on those paediatric cases, where the doctor had been referred for assessment because of specific clinical concerns. Over the ten years from 2001 to 2011, there were 63 cases. The most common reasons for referral are shown in Table 1.3:

Table 1.3 Most common reasons for referral, 2001–2011

Common themes emerge from these studies (details of which appear in the references below). The most common error in the NHSLA cases was a failure to treat or diagnose an infection, particularly meningitis and septicaemia. The NHSLA and NCAS studies also showed failings in prescribing.

In contrast to the NHSLA, the NCAS study showed that the commonest reason for a complaint against a paediatrician was related to the diagnosis and management of child protection cases. This reflects the focus of NCAS referrals, which is different from that of the NHSLA; most NCAS referrals are made by NHS managers and doctors, rather than by patients.

The main causes of the errors in the NHSLA study were very much as expected, with a delay or failure in diagnosis and treatment being by far the commonest.

Most errors in clinical practice result in little or no harm. However, the NHSLA study shows that the commonest outcome in the cases analysed was death. Looking a little further down that list, the fourth most common outcome was brain damage. These observations reinforce the importance of trying to minimize errors in clinical practice. When things go wrong, they can go spectacularly and tragically wrong.

Pulling all this research together, we believe that there are certain keys areas where doctors would benefit from advice. We aim to provide such advice in the following sections. This will include advice on how to make the correct diagnosis promptly, avoiding prescribing errors, checking test results and acting on abnormal findings, avoiding errors in practical procedures, avoiding resuscitation errors and how to act in child protection cases. We shall start by looking at the patient consultation and how to identify the sick child.

The patient consultation (Cases 1, 2, 17, 24, 26, 36)

The patient consultation is, in a very practical sense, at the heart of the doctor–patient relationship. It gives an opportunity for face-to-face communication and for the doctor to build up a rapport with the patient and to win his trust. If handled correctly, at the end of the consultation, the doctor should be armed with the information that he needs to reach a diagnosis, or at least a differential diagnosis. And of all consultations, it is the first that is perhaps the most important; that first consultation will strongly influence all that follows. How should it be conducted?

A good history is essential in making a correct diagnosis. The history on its own leads to the diagnosis in over 60% of cases. Once a patient has given his account of the presenting complaint, the doctor should ask questions to clarify his understanding of the patient’s symptoms. All this is perhaps obvious, but it all comes down to communication. Good communication should aid diagnosis; poor communication will hamper it. Thus a doctor must listen carefully to patients and their parents. In the words of the nineteenth-century Canadian physician Sir William Osler, ‘Listen to the patient. He is telling you the diagnosis.’ This requires skill and patience and can be difficult if time is short.

And to continue with the theme of listening and understanding … Doctors often see patients who speak poor English. Where this gets in the way of understanding, the doctor should find a translator who can speak the language of the patient. If one is not on site, most hospitals have phone access to translators 24 hours a day. A judge is unlikely to excuse a mistake caused by a failure to use a translator.

After the history comes the examination. Examining any patient may be difficult, but examining a sick child even more so. A general examination should be conducted with a particular focus on the ailing system. A doctor should try to put the child at ease; it may be a good idea to give them a toy at the beginning of the consultation, if they are very young. Starting with the hand is often a nonthreatening way of commencing the examination and talking to the child may well help to calm him down. The examination may need to be opportunistic: for example, the doctor may have to wait until the child is quiet and calm, before he can listen to the precordium. If the child is very fractious, the doctor may wish to leave the child for a while, if the situation is not urgent, and to examine the child a little later, perhaps after a feed or after analgesia has had time to take effect.

There are key areas that should not be omitted in the course of particular examinations, such as the examination of the throat in a febrile child. If the doctor were to omit this, he could be committing a significant error; he may fail to diagnose tonsillitis; he may fail to find the source of the fever. If a doctor finds a child difficult to examine, a more experienced doctor with a better, more experienced examination technique should be asked to see the patient.

If the examination is limited one should make that clear in the notes, together with the reason why.

Once the history and examination have been completed it is important to make a diagnosis or to consider the differential diagnoses in order of probability. The doctor should list the investigations required to clarify the diagnosis and to provide further details about the illness. A management plan should then be constructed. Following this pathway encourages a doctor to rigorously analyse the ailment afflicting the patient.

The history, examination, diagnosis, investigations and management plan should be clearly documented. It is also very important to note negative findings in the history and examination. The investigation results should be obtained and documented promptly. The importance of clear and thorough documentation cannot be overemphasized.

The requirement to see all patients who present to the ED within four hours can sometimes cause doctors to prioritize patients inappropriately. Thus a doctor may find himself rushing the history and examination in order to meet this deadline. Errors may occur as a result. But it is no defence for a doctor to argue this in court. This, along with ergonomic factors such as the stress, tiredness and depression, is a system issue which may adversely affect the outcome of the patient consultation. If the doctor has concerns about the systems in place at his hospital, he should discuss them with his superiors.

A delay or failure in making a diagnosis and a delay or failure of treatment are the two commonest causes of errors. The reasons behind these delays and failures are many; they include poor paediatric training, a lack of knowledge, failure to recognize when help is required and when a more senior opinion is needed, and an inadequate hospital and departmental induction to the job.

Clinicians often form a hypothesis about the likeliest diagnosis early on in the consultation. A doctor should repeatedly question this hypothesis and reassess alternatives. He should also bear in mind the diagnoses that he cannot afford to miss. So when he is presented with a febrile child, meningitis should be on his list of differential diagnoses. Meningitis may be both clinically and statistically unlikely, but it should at least be considered, if only because it is a treatable condition with potentially catastrophic sequelae.

Doctors can also make what are termed cognitive errors when they are looking for a diagnosis. A detailed analysis of cognitive errors is outside the scope of this book and the reader is referred to the references at the end of this section. However, these are some examples:

• ‘Confirmation bias’, where information gathering is geared towards confirming rather than refuting the diagnosis. This may cause a doctor to overlook important alternative diagnoses. In the course of his examination and assessment of the child, the doctor should not overlook symptoms and signs that may be inconsistent with the initial diagnosis and may suggest other possible conditions.
  
• ‘Premature closure’, where a doctor makes what he considers to be a definitive diagnosis early on in his treatment of the patient and then fails to reconsider the diagnosis at a later stage even if the circumstances have altered.
  
• ‘Availability bias’, where a doctor gives too much weight to his own past experience and easily recalled examples, at the expense of other and often rarer diagnoses.

Hospital protocols and national and international protocols on different conditions provide information on the symptoms that need to be asked about, the signs that need to be checked, the differential diagnoses in a given set of circumstances and the appropriate treatment plan. Likewise, different courses such as the Advanced Life Support Group’s courses on Advanced Paediatric and Neonatal Life Support, and Child Protection provide very useful training in important areas of paediatrics.

As a very basic comment, it is training and experience that will lead to improved and more efficient history taking and examinations, to more accurate diagnosis and to better treatment.

Failure to identify a sick child (Case 31)

It is easy to identify a severely unwell child. The challenge is to spot the child who is not yet severely unwell, but may deteriorate rapidly, if he does not receive the right treatment. Such children will often present alongside hundreds of other children with self-limiting conditions. To make the task even harder, they may have only the subtlest of signs to indicate that they are individuals at high risk. Furthermore, it is well recognized that children’s physiology allows them to maintain essential physiological parameters within the normal range for far longer into the course of an illness than their adult equivalent. The Royal College of Paediatrics and Child Health has made a DVD called Spotting the Sick Child and NICE have issued guidelines on Feverish Illness in Children to help recognize such children.

When a sick child is not identified, this is usually because early warning signs of a critical illness were missed or ignored. Therefore when an essential physiological parameter (heart rate, blood pressure, capillary refill time, respiratory rate, SaO2 or GCS) is abnormal, this needs to be carefully explored to see if it is the first sign of an impending critical illness.

When cases involving sick children who were not correctly identified are reviewed, it is often found that the child had a single abnormal parameter (most commonly tachycardia) and that this was not acted upon. Reviewers generally conclude that the doctor did not act on this finding because:

• he attributed the abnormal parameter to some cause other than illness, for example the doctor considered that the child’s tachycardia was caused by distress or that a prolonged CRT was due to environmental cold; or
  
• he chose to ignore a single abnormal parameter because everything else was normal; or
  
• he failed to recognize that the parameter was abnormal, often because he did not know the normal ranges in childhood.

To try and prevent these common mistakes, Paediatric Early Warning (PEW) scores were developed. These provide age appropriate values for normal ranges with corresponding single and cumulative triggers for review by a senior doctor.

On occasion, a child can present to the ED critically ill after attending the ED just a few days earlier. It is natural to assume that ‘something was missed’ at the first attendance. In many cases this is true, but it is also true that children can deteriorate suddenly and rapidly. While all such cases should be explored to see if lessons can be learned, it can equally be that sometimes, when children are seen very early in the course of a critical illness, there are no early warning signs of a severe illness.

This only serves to reinforce the need to give patients clear ‘safety net’ advice, when they are discharged from medical care: that is advice about when they should reattend the ED or the GP, if the child fails to get better. Such advice should be given no matter how trivial the complaint may appear to be.

Inability to perform practical procedures competently

Following the guidance below should help avoid errors whilst carrying out practical procedures.

Practical paediatric procedures require good communication skills, manual dexterity, patience, a gentle touch and supervised practice. So it is important to be aware that skills acquired during training on adult patients may not translate directly to competence in the same procedures in a child. The doctor should be aware of the limits of his competence and should not exceed them without experienced supervision (GMC, Good Medical Practice, sections 3 and 12).

The choice of assistant is also important – an experienced assistant is much more likely to be able to help a child remain calm, to provide distraction and to keep the appropriate body parts still and in the correct orientation, for example during a LP.

The objective should be to perform the correct procedure on the correct patient, on the correct side, competently and with appropriate consent. Pay particular attention in the case of twins and where there are common surnames.

Give adequate analgesia or anaesthesia. Be aware of the dose limits for local anaesthetics in small children. If it is necessary to tap a hollow viscus or a fluid collection, an ultrasound examination may improve the chances of success, help the doctor’s understanding of the anatomy and reduce the risk of a ‘dry tap’.

Prepare the trolley and instrument pack. Ensure all the necessary instruments are present. Keep sharp instruments, needles and blades inside a tray to reduce the risk of a needlestick injury. Keep a count of needles and swabs.

Ensure that the patient is adequately monitored – a small child or neonate concealed under drapes may be rendered invisible and should be monitored with an ECG, oxygen saturation meter, and an apnoea alarm, with the limits appropriately set. An assistant should take responsibility for the safety of the child whilst the doctor performs the procedure. If indicated, the doctor should check that life support equipment is present and working, and that he knows how to use it if necessary.

Good aseptic technique reduces the chance of infective complications. Empower the assistant to notice lapses and tell him to point out if gloves or instruments become contaminated. Clean the skin carefully, remembering that in infants, skin can be burned by alcoholic solutions, and that iodinated solutions can be absorbed transcutaneously.

Ensure that you understand the anatomy of the procedure. Practise using the instruments – most commonly used instruments are designed to be used by a right-handed operator and may not work optimally if used left-handed. Read the product inserts and instructions, especially if using a new type of catheter. Thin-walled catheters may become cracked if manipulated with metallic instruments. Examine distance markers carefully. The doctor should ensure that he knows what length of a tube or catheter will protrude from the patient at the end of the procedure.

When securing lines and catheters, ensure that any splints used maintain the body part in a position of function and that peripheral nerves and vessels are not compressed. Avoid encircling limbs with complete loops of tape in order to prevent a ligature effect and to reduce the risk of ischaemia.

At the end of the procedure, document the process carefully, retaining a note of the batch number and catheter type inserted. Obtain radiological confirmation of line positions if needed, record the findings and take reme-dial action if necessary. Remember to talk to the parents, and the child if old enough, at the end of the procedure. The doctor should explain what he has done, what should happen next, and make the parents aware of potential complications.

Failure to check test results or act on abnormal findings (Cases 1, 8, 21, 27, 35)

It is stating the obvious to say that if tests are requested, the results have to be looked at (ideally by the doctor who ordered them). They have to be considered with the care to be expected of a competent doctor.

Life at the coalface is always more complicated. Circumstances may intervene: the doctor who actually requested the test may finish his shift before the results are available, leaving another doctor to deal with them. The handover to that doctor may have been inadequate. Alternatively, he may simply be too busy to give them proper attention, with the result that an abnormality is overlooked. While those circumstances might point to weaknesses in the system, they do not absolve the responsible doctor (or his employing Trust) from a charge of negligence.

That is why it is so important that clear notes are made, pointing out the need to follow up the results of investigations.

When it comes to interpreting the results, however, the situation is different. A misinterpretation obviously might be negligent: the inexperienced doctor may not appreciate the significance of an abnormal finding, or the incompetent doctor may realize it but just not act on it. However, a competent doctor is allowed to make errors of judgement without necessarily being labelled negligent. So, in a subtle case, with confusing symptoms and signs, an expert may take the view that the notional competent doctor might well have made the same mistake as the one accused of negligence.

The subtle case, though, generally invites the objection that a second opinion should have been sought. Take the case of ‘The Neonate with Abnormal Movements’ (Case 21). The infant’s condition was rare and the results of the lumbar puncture were difficult to interpret. But the expert has commented that a microbiologist’s opinion should have been sought. He appears to think it likely that a microbiologist would have stimulated a new line of enquiry. Would the ordinary competent doctor have considered that line of enquiry?

Similarly, in the case of ‘The Boy with a Limp’ (Case 1) the inexperienced doctor who saw no abnormality on the X-ray is criticized by the expert for not getting a second opinion, in circumstances where there was no written report on the X-ray.

Another, not uncommon scenario is a fluctuating situation where an abnormal result comes and goes. At the time, a considered judgement may be made to ‘watch and wait’. Later, with the benefit of hindsight, an expert may be able to pinpoint exactly when that policy ceased to be appropriate, with a clarity which was unavailable at the time. But that is the benefit of hindsight.

Prescribing errors (Cases 7, 15, 30)

Prescribing errors are very common. In the year 2007–08 there were 10 041 medication errors in children in England and Wales, the highest incidence being in children aged 0–4 years.

Doctors should always refer to The British National Formulary for Children, which is updated annually. The age and weight should always be stated on the prescription and in some cases the height and surface area are also necessary. The large weight range, from <1 kg in the case of a premature neonate to >100 kg in an obese adolescent, increases the risk of a doctor making a medication error. Doctors should be aware that the responsibility for a mistake lies with the person who has signed the prescription. The following principles should be adhered to when prescribing medication:

• Write prescriptions legibly, ideally in capitals.
  
• The generic names of drugs should be used whenever possible.
  
• Beware of drugs with similar sounding names, for example prostacyclin and prostaglandin.
  
• Always double-check the prescription. This is particularly important when using unfamiliar drugs, or familiar drugs in an unfamiliar setting. If calculations are involved, try to get a medical or nursing colleague to check the arithmetic. If someone questions the prescription, the prescription should be checked carefully. During working hours a pharmacist may be available to check the prescription.
  
• In larger children beware of calculating a dose, based on weight, that is greater than the dose one would use for an adult.
  
• The strength or quantity to be contained in tablets, liquids etc. should be stated (e.g. 125 mg/5 ml).
  
• Dose frequency, and also often specific times, should be stated and in the case of drugs to be taken ‘as required’ a minimum dose interval should be specified.
  
• Great care needs to be taken with the decimal point. The unnecessary use of decimal points should be avoided (e.g. 5 mg not 5.0 mg). The zero should be written in front of the decimal point where there is no other figure (e.g. 0.5 mg not .5 mg). To avoid confusion over the placing of decimal points amounts less than 1 milligram should be prescribed in micrograms (e.g. 500 micrograms not 0.5 mg).
  
• The correct units should always be used. Avoid abbreviations. In particular, do not abbreviate the terms micrograms and nanograms.
  
• Ensure that the correct route of administration has been prescribed. See, for example, the vincristine case above, when the chemotherapy was administered intrathecally instead of intravenously.
  
• Clear plans should be in place for the necessary monitoring of drugs. For example, gentamicin levels.
  
• Ask about allergies, and the nature of the reaction. Document the reply that the patient has given.
  
• Careful consideration should be given to the information that is provided to patients and their families concerning side effects. Some doctors believe that it is only necessary to tell patients of risks that have more than a 1% chance of occurring (the Electronic Medicines Compendium quantifies the incidence of side effects of many drugs). This is not an accurate reflection of the law (see Consent below). A doctor should always consider what would be reasonable to tell the patient and his family. For example, in many circumstances it is right to mention rare but potentially serious side effects, such as the risk of agranulocytosis when taking carbimazole.
  
• Doctors should ensure that the medication is not contraindicated. For example, gentamicin in renal failure.
  
• The doctor should ensure there are no interactions with other medicines that the patient is taking and warn the patient about possible interactions with over-the-counter medicines.
  
• Repeat prescriptions should be regularly reviewed to ensure that they are still necessary.
  
• The administration of medicines should be carefully documented to ensure that drugs are not given twice.
  
• Doctors should beware of vaccination errors. It is important to check that the correct vaccine is being given, that the dose is correct and that parental consent has been obtained.

Mistakes in the calculation of doses of opiates are a common serious prescription error. So, too, are errors in the prescription of anticonvulsants, cytotoxics, antibiotics and intravenous fluids. In time, computer packages and online prescribing should become available and facilitate the correct calculation of drug doses, help with drug interactions and diminish the incidence of medication errors. Safe prescribing modules are part of the undergraduate curriculum and are also incorporated into several foundation year training programmes.

Failures in resuscitation

Failure to resuscitate properly can be divided into clinical failings (e.g. failure to correctly treat airway, breathing or circulation issues) and nonclinical failings (e.g. failures in leadership or communication).

Clinical failings

If a patient’s airway is inadequately resuscitated or if he experiences breathing problems, this usually indicates that the paediatrician has not involved an anaesthetist early enough. However, this is rarely an issue, as most resuscitation teams will automatically contact an anaesthetist and seek their input at an early stage.

If a patient experiences circulatory problems, then this usually means that fluid resuscitation is either inadequate or excessive.

Inadequate volume resuscitation occurs particularly in:

• severe sepsis patients, who can require very large volumes of fluid: for example 120 ml/kg–200 ml/kg of volume is not uncommon;
  
• burns, where the modified Parkland Formula for paediatric burns (>10% body surface area), requires 2 ml/kg/% burn in the first 8 hours after the time of injury in addition to normal maintenance fluids.

Junior doctors who have not used such large volumes of fluid in practice can become anxious and hesitant. They may be afraid to administer fluid in such large quantities. But to delay the administration of such volumes would be a mistake.

In addition, most junior doctors are familiar with the APLS principle that after 40 ml/kg of volume resuscitation, it is best practice to electively intubate and ventilate a child with positive pressure to reduce the risk of pulmonary oedema. Applying this principle, a doctor may hesitate to give more than 40 ml/kg of volume without the child being intubated or at least being reviewed by an anaesthetist prior to the additional fluids being administered. This, too, can be a mistake and could cause an unnecessary delay in providing the correct treatment. If a child is severely shocked they are most likely to arrest at induction of anaesthesia, as most anaesthetic induction agents cause hypotension.

It is less usual for a doctor to administer excessive volume resuscitation. When it does happen, it is usually in cases where children need smaller volumes of resuscitation fluid or where the doctor mistakenly treats the child’s blood gas results and not the child.

Specific conditions, when smaller volumes should be administered, include:

• children with known or undiagnosed cardiac conditions (either structural defects or cardiomyopathy) who are suffering from cardiogenic shock;
  
• children in diabetic ketoacidosis who are particularly vulnerable to cerebral oedema following excessive fluid resuscitation;
  
• paediatric trauma patients, who are now managed with permissive hypotension; in such cases, overgenerous fluid resuscitation can exacerbate haemorrhage by preventing clot formation.

If he does not interpret blood gas results correctly, a doctor can mistakenly administer too much resuscitation fluid. Doctors need to be aware that the use of 0.9% saline for fluid resuscitation can itself cause a hyperchloraemic acidosis. Thus it is possible for a child’s condition to improve and for him to clinically respond to fluid resuscitation with normalized physiological parameters but for the blood gas levels to show a worsening acidosis. If the doctor fails to distinguish between a lactic acidosis (from poor perfusion) and an iatrogenic hyperchloraemic acidosis, then he can administer excessive and unnecessary fluid resuscitation, and cause iatrogenic pulmonary oedema.

Nonclinical failings

This topic is a very large one and cannot be covered in any detail in this section. Suffice to say that human interactions are as important as the clinical steps in a resuscitation.

The importance of these issues in resuscitation has been increasingly recognized and is now taught in APLS/PALS courses and in simulation training. Specific examples of nonclinical failures are discussed in some of the case histories (see Case 29).

Sources of error in child protection cases (Cases 6, 33)

Failure to recognize child abuse/maltreatment

GPs, paediatricians and ED doctors must be aware that child protection cases do not come with a label. They must be constantly alert to the possibility that children brought with illnesses, minor injuries or more serious injuries may have been abused. They must remember that domestic abuse, alcohol and substance misuse and mental health problems in parents/carers are major risk factors for child abuse.

There are several roles that a doctor may have in child protection cases, some of which may overlap. He may be called on to treat a child who has suffered physically from abuse; and/or he may be asked to examine a child to determine whether there has been abuse; and he may be called on to give evidence at court.

There are two main ways in which the courts become involved in child abuse cases. The suspected abuser may be put on trial in the criminal courts, accused of rape, indecent assault or grievous bodily harm or worse. The family courts may be called on to determine whether a child should be taken from his parents, and if so, who should look after him.

Doctors, especially paediatricians, are trained to ‘listen to the mother’ and to believe what she says about her child. In the vast majority of cases this is, of course, the right thing to do. The difficulty is spotting the parent or carer who is lying, covering up or deliberately misleading professionals. Doctors are not trained to do this and find it hard to think in this way. It is important to be able to ‘think the unthinkable’ and to ‘look behind what you are being told to what you are experiencing’.

Failure to act

If a doctor has a gut feeling that something is not right they should act on it. Doctors must know what to do and who to contact (e.g. the consultant paediatrician, a social worker or the police) if they have concerns about a child’s safety. If they have concerns about sexual abuse they must contact the appropriate consultant paediatrician immediately so that an urgent medical examination can be arranged if indicated.

Failure to document

History

The importance of taking an accurate and detailed history cannot be overemphasized. The history may be taken from more than one person, for example each parent or a parent and the child himself, and it must be absolutely clear who the history was taken from and whether they were present when a particular incident occurred or are giving someone else’s account as told to them. The details of the history should be so accurate that when the history is relayed to another professional they are able to ‘see’ what happened as though watching a video recording. Details should include the time and place of incidents, accidents, falls etc. as well as who was present, who was nearby (e.g. in another room) and exactly what happened. Doctors should not interview children without prior discussion with children’s social care and usually the police. However if a child says things spontaneously, these statements should be noted verbatim, if possible. It may be useful for the doctor to help a child by saying, ‘Can you tell me what happened?’, but leading questions such as, ‘Did your dad hit you?’ must always be avoided.

Examination

Weight, height and head circumference should be measured and plotted on growth charts in all cases. A full general physical examination should be performed. The examination should be conducted whenever possible in a private, well lit room. In cases of possible physical abuse two doctors should conduct the examination together if possible – to confirm findings and for one to examine, measure and describe lesions whilst the other documents them. This may not be possible in emergency situations but often the examination can be repeated later by a more senior doctor in the best possible environment. Each lesion should be drawn onto a body map and also described in detail, for example ‘on the left upper arm, just above the elbow on the posterior aspect of the arm, there is a blue bruise measuring 1.5 × 2 cms’. If there are only a few lesions, the descriptions can be written on the body maps, but if there are many it is better to number the lesions on the body maps and list them with their descriptions on a separate page. Each page must be signed and dated by the examining doctor(s).

Other documentation

Lord Laming in his report following the death of Victoria Climbie noted the importance of documentation of not just the history and examination, but also of the interaction between the child and the parents/carers, of things said by the child or of any unusual behaviour observed.

All discussions between professionals, for example doctors and social workers, must be documented whether these were in person or by phone. All referrals must be in writing. This may be in addition to a verbal referral which may be needed for expediency and to give more of a flavour of a case than may be possible in writing.

It exaggerates the true legal position, but it does no harm to state it: ‘If it’s not written down it didn’t happen!’ Good notes and records are key to any child protection case.

Failure to communicate

Regular communication with colleagues is of the utmost importance. Doctors dealing with a case of possible child abuse must seek senior advice at an early stage both to protect the child and themselves. When ‘handing over’ a possible child protection case it is vitally important to be as clear and accurate as possible. For example in the Victoria Climbie case it was said that marks on her body were not thought to be deliberately inflicted. At handover this became ‘no child protection concerns’ – clearly a statement with different connotations. When communicating with nonmedical professionals, for example social workers and police, doctors must ensure that they use words that can be clearly understood and that they are clear about the possible causes of the injuries. Doctors who have medical evidence to share should make every attempt to attend child protection conferences. Although written reports can be sent, it is much more likely that nonmedical participants in the conference will understand the implications of the medical findings, if the doctor is there to explain them, and if members of the conference appear to underestimate the seriousness of injuries, for example, the doctor should be able to point this out.

Summary of common errors

• Not being suspicious enough when things don’t add up – ‘Think the unthinkable.’
  
• Failure to recognize the impact of domestic violence on children in the household. These children are at high risk of deliberate harm, of getting caught up in violence and being injured inadvertently and of emotional harm from observing violence.
  
• Failure to put the child’s needs first. The best interests of a child and his parent(s) normally go hand in hand. But when there is any suspicion of child maltreatment the child’s interests are paramount, above those of the parent/carer.
  
• Not referring upwards (to a consultant paediatrician) at an early stage.
  
• Sloppy or inadequate history. The history should be good enough ‘to play the video’.
  
• Inadequate examination. Examination must be thorough and well documented, including clear body maps. Injuries must be described accurately. It is embarrassing to have to admit in court that a lesion shown on the left leg was actually on the right leg!
  
• Not admitting a child to hospital when abuse has not been excluded and it is not clear that home is a safe place.
  
• Not getting all previous health records (hospitals, GP, HV, school nurse).
  
• Not checking if the child is the subject of a CP plan.
  
• Failure to ensure chain of evidence. If swabs are taken from a child with a vaginal discharge where there is a possibility of sexual abuse it is essential to set up a chain of evidence. When we speak of a chain of evidence, we are referring to the integrity and history of physical evidence, from its collection to the time when it is produced at court. This is to ensure that the swab results are definitely the child’s. Thus the doctor who takes the vaginal swab must place it in a sealed, labelled package, with details of the patient, the nature of the swab, the date it was collected and who collected it. When it is handed to the next person in the chain on its journey to the microbiology lab, he should record details of the time and date of the handover, the person to whom it is given and so on. If someone forgets to record this information, the quality of the evidence is affected and can be challenged by the lawyers. In some cases the evidence will be rendered inadmissible.
  
• Poor communication between colleagues at handovers and ward rounds.
  
• Poor communication between doctors, social workers and police – ‘not speaking the same language’. Doctors may often think that they have fully explained the medical findings, but actually these may not have been clearly understood by nonhealth professionals.
  
• Inadequate documentation – especially unusual interactions between parents and a child or unusual behaviour (e.g. sexualized) in the child.
  
• Not documenting all discussions, including telephone calls with other agencies.
  
• Being drawn by social workers or the police to give a definitive opinion on the cause or age of injuries when this is not possible. All the available evidence shows that bruises cannot be dated. Some fractures can be dated but only approximately – this should only be done by a consultant radiologist.

References and further reading

BMJ Group (2011–12) BNF for Children. www.bnfc.org

Davis T (2011) Paediatric prescribing errors. Arch Dis Child 96: 489–91.

Del Mar C, Doust J, Glaszious P (2006) Clinical Thinking: Evidence, Communication and Decision Making. Oxford: Blackwell Publishing (especially Chapters 1 and 4).

Department for Children, Schools and Families (DCSF) (2006) What to Do If You Are Worried a Child is Being Abused. www.education.gov.uk/publications/standard/publicationdetail/page1/dfes-04320-2006

Elstein A, Schwarz (2002) A Clinical problem solving and diagnostic decision making: selective review of the cognitive literature. BMJ 324: 729–32.

GMC (2007) 0–18 Years: Guidance for all doctors. http://www.gmc-uk.org/guidance/ethical_guidance/children_guidance_index.asp

GMC (2012) Good Medical Practice. www.gmc-uk.org/guidance/good_medical_practice.asp

Hampton JR, Harrison MJG, Mitchell JRA et al. (1975) Relative contribution of history taking, physical examination and laboratory investigation to diagnosis and management of medical outpatients. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1673456/pdf/brmedj01449-0038.pdf

Laming (2003) The Victoria Climbie Inquiry: Report of an Inquiry by Lord Laming (Jan. 2003). http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4008654

Maguire S, Mann MK, Sibert J, Kemp A (2005) Can you age bruises accurately in children? Archives of Disease in Childhood 90: 187–9.

Markert RJ, Haist SA, Hillson SD et al. (2004) Comparative value of clinical information in making a diagnosis. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1395780/

NICE (2009) NICE Clinical Guideline 89. When to Suspect Child Maltreatment. http://guidance.nice.org.uk/CG89

NPSA (2009) Review of Patient Safety for Children and Young People. http://www.nrls.npsa.nhs.uk/resources/?EntryId45=59864

Spotting the Sick Child (2011) Information on the accurate assessment of sick children from the Department of Health. www.spottingthesickchild.com

Vincent C (2010) Patient Safety, 2nd edn. Chichester: Wiley-Blackwell and BMJ books.

References and further reading specific to section on Person-centred paediatric errors and their causes

Advanced Life Support Group (2011) http://www.alsg.org/uk/

Carroll AE, Buddenbaum JL (2007) Malpractice claims involving pediatricians: epidemiology and etiology. Pediatrics 120: 10–17.

Marcovitch H (2011) When are paediatricians negligent? Arch Dis Child 96: 117–20.

Najaf-Zadeh A, Dubos F, Pruvost I et al. (2011) Epidemiology and aetiology of paediatric malpractice claims in France. Arch Dis Child 96: 127–30.

Raine JE (2011) An analysis of successful litigation claims in children in England. Arch Dis Child 96: 838–40.

Raine JE, Scarrott D (2012) Which clinical errors lead to the referral of paediatricians to the National Clinical Assessment Service? In press. Published online by The European Journal of Paediatrics.

Royal College of Paediatrics and Child Health (2012) http://www.rcpch.ac.uk/training-examinations-professional-development/professional-development-training/safeguarding-childr

1. In this book the general use of the personal pronoun (e.g. ‘he’, ‘him’, ‘his’) indicates, in all but specific cases, both male and female doctors, patients, etc.