Overview of available evidence

No suitable published literature was identified that addressed whether pregnancy is a direct result of child maltreatment. We did not search for epidemiological literature on teenage pregnancy.

Delphi consensus (see also Appendix C)

The GDG sought the opinions of the Delphi panel on statements about 16- and 17-year olds because of sensitivities around the age of consent. The following statements were included in the survey:

GDG considerations

It is the GDG's opinion that pregnancy in children as a direct result of sexual abuse falls within the legal framework outlined in the Sexual Offences Act 2003. Therefore, any pregnancy in a child younger than 13 years should be recognised to be a result of maltreatment. This still applies if two minors have engaged in sexual intercourse as it represents neglect by lack of supervision.

The GDG believes that the issues around consensual experimentation among 13- to 15-year-olds outlined in Home Office guidance should be taken into account when a young person of this age is pregnant: that guidance indicates that a pregnancy in this age group is not an immediate reason to suspect sexual abuse.

Despite the age of consent being 16 years in the UK, the GDG believes that healthcare professionals may observe circumstances around a pregnancy that should give rise to a suspicion of maltreatment. Namely, when there is a clear discrepancy in power, emotional maturity or mental capacity between the young woman and the putative father, or concern about incest or that the young person is being exploited.

The GDG sought the opinions of the Delphi panel on the recommendation about pregnancy in 16- and 17-year-olds (see above and Section C.2.4). The GDG accepted statements 25a and 26a from the Delphi survey. Although agreement was reached on Statement 25a, the GDG amended the definition of a ‘discrepancy in power, emotional maturity or mental capacity’ to provide examples that are meaningful for healthcare professionals. Based on the views of the Delphi panel, the GDG rejected its proposed statement about a concealed identity of the father. There was consensus within the GDG about the recommendation about children younger than 13 years and thus the views of the Delphi panel were not sought.

Overview of available evidence

A total of 201 articles (194 from the main search and seven from bibliographies) were identified and 60 articles were selected for detailed assessment. Of these, one systematic review and 11 additional studies have been included in the review.

Narrative summary

One systematic review (eight papers; search undertaken in 2002) assessed the initial diagnosis given when infants presented with an ALTE.⁴² The review included eight studies involving 643 infants seen in emergency departments or paediatric units. The study calculated that 0.6% to 0.8% of emergency admissions for infants were for ALTE. A total of 728 diagnoses covering 50 conditions were reported, of which 227 were gastro-oesophageal reflux disease (GORD), 169 were unknown, 83 were seizures, 58 were lower respiratory tract infection, 26 were ear, nose and throat (ENT) problems, 17 were breath-holding, 11 were metabolic disease, 11 were ingestion of toxins or drugs, eight were urinary tract infection (UTIs), six were cardiac problems, five were benign cause, and two were fabricated illness (0.3% of children). The study concluded that careful investigation of ALTE is needed because of many possible causes. [EL = 2+]

A prospective cohort study (n = 44184) undertaken in Austria investigated the epidemiology of ALTE.⁴³ The study identified 164 cases of ALTE, or 2.46 per 1000 live births. An underlying cause was identified in 91 of 164 cases (55%): 29% were respiratory, 22% were digestive (gastrointestinal) tract, 2% were congenital cardiac malformation, 1% were inborn metabolic errors and 1% were convulsions. The study made no conclusions in relation to child maltreatment. [EL = 3]

A prospective cohort study (n = 340) undertaken in Australia examined the diagnosis of ALTEs: 289 of 340 had a diagnosis of which 211 were GORD, 17 were airway pathology, 25 were fits/seizures, two were brain-stem tumours, two were hypoglycaemia, eight were respiratory syncytial virus, five were fabricated or induced illness (FII) (1.7% of those diagnosed, 1.5% of total) and 27 were abnormal pneumograms (11 with reflux).⁴⁴ Fifty-one had no abnormal finding. The study made no conclusions in relation to child maltreatment. [EL = 3]

A prospective case series (n = 128) from the USA of children younger than 24 months presenting at a single emergency department examined the diagnosis applied to cases of ALTE.⁴⁵ Of the 128 cases of ALTE, 51 were GORD, 38 were apnoea, 11 were choking episode, six were infection, five were bronchiolitis, five were upper respiratory infection, four were seizures, three were abuse (2.3% of total), three were swallowing disorder and two were breathing-holding spell. The study concluded that abuse was diagnosed in 2.3% of cases of ALTE and this should be considered in patients who present with ALTE. [EL = 3]

A prospective case series (n = 157) from the UK of children (aged 1 week to 96 months) presenting once or more in one hospital setting examined the diagnosis applied to cases of ALTE.⁴⁶ The study reported that, of the 157 reported cases, 80 had no diagnosis. Of those diagnosed, two had disturbances in skin perfusion, seven had fabricated illness (9% of those diagnosed and 4% of the total), 18 had suffered suffocation (23% of those diagnosed and 11.5% of the total), 40 had hypoxaemic events with no evidence of suffocation or epilepsy, and ten had hypoxaemia induced by epilepsy. The study concluded that identification of mechanisms is essential to the appropriate management of infants with ALTEs. [EL = 3]

A prospective case series (n = 243) of infants younger than 12 months admitted to one tertiary unit in the USA examined the diagnosis given to cases of ALTE.⁴⁷ The study found 35 different causes for ALTE. Of the total cases, 80 were caused by infection, 69 were gastrointestinal, 32 were neurological including six (2.5% of total) abusive head injuries within this group, seven were airway obstruction, six were congenital or birth-related problems, 39 had an unknown cause, three were breath-holding spells, two were periodic breathing and one was vasovagal response. The study concluded that a wide spectrum of diseases and disorders can precipitate an ALTE. In relation to maltreatment, the study concluded that ‘Among them, abusive head injury, a recently recognized cause, occurs frequently enough to obligate its inclusion in the differential diagnosis.’ [EL = 3]

A retrospective case series (n = 60) from the USA examined the diagnosis applied to infants with ALTE.⁴⁸ The study setting was a single emergency medical service (EMS) over a 12 month period. The study found that 60 (7.5%) of 804 infants encountered met the criteria for ALTE (absence of breath, colour change, change in muscle tone). The diagnoses applied to these cases were: 20 (33%) had no diagnosis, seven (12%) were pneumonia or bronchiolitis, six (10%) were GORD, five (8%) were seizures, four (7%) were sepsis, four (7%) were upper respiratory infection, three (5%) were apnoea episodes, two (3%) were intracranial haemorrhage, two (3%) left against advice, one (2%) was bacterial meningitis, one (2%) was dehydration and one (2%) was severe anaemia. Furthermore, 35% of the 60 infants had been diagnosed with underlying conditions. The study reported one case of intracranial injury caused by maltreatment, but highlighted that in 20 cases no diagnosis was made and in two cases the parents left against medical advice. The study concluded that ‘An apparent life-threatening event in an infant can present without signs of acute illness and is commonly encountered in the EMS setting. It is often associated with significant medical conditions, and EMS personnel should be aware of the clinical importance of an apparent life-threatening event. Infants meeting criteria for an apparent life-threatening event should receive a timely and thorough medical evaluation.’ [EL = 3]

A retrospective case series (n = 73) of infants (mean age 7.4 weeks) who were seen at a single apnoea programme in the USA reported that 47 infants had negative investigation, 17 had recurrent events but no diagnosis, five had respiratory infection, two had GORD, one had pallid syncope and one had tracheal stenosis.⁴⁹ [EL = 3]

A retrospective partially controlled case study (n = 85) from the UK compared the medical and family history of maltreated children (30 of 39 children with maltreatment confirmed by covert videoing) and non-maltreated children (46 children with confirmed respiratory disease or epilepsy) presenting with ALTE.⁵⁰ The mean age of maltreated children when they first presented with ALTE was 3.6 months. The study found that in the 41 siblings of the maltreatment group there were 12 unexpected deaths compared with one unexpected death among the 52 siblings of the control group (P < 0.0001). [EL = 2−]

A survey of 11 apnoea monitoring programmes and four apnoea monitoring device vendors in the USA examined reports of infant deaths.⁵¹ Over a 5 year period, 1841 children were monitored. There were 25 reported deaths in this group: 13 due to SIDS, four due to non-accidental trauma (0.2% of total), six due to sudden unexpected death at home, one due to subarachnoid haemorrhage and one caused by cardiac disease. The study reported no specific conclusions relating to maltreatment. [EL = 3]

A retrospective case series (n = 28) from the USA of children who suffered proven non-accidental head injury examined their presentation and outcome.⁵² Of the children examined, only three were aged I year or over. The results showed that 16 of 28 presented with apnoea. Of those who presented with apnoea, 57% had a history of apnoea and 71% had previous seizures within 24 hours. The study found that 12 were left with severe disability, four died, one was in a vegetative state and seven survived. The authors concluded that trauma-induced apnoea is more important to outcome than the mechanism of injury. [EL = 3]

A survey of 51 of 127 (n = 20 090) apnoea monitoring programmes in the USA investigated the prevalence of FII. The results showed that 54 (0.25% of total) cases of FII were reported. The mean age of infants with this diagnosis was 8.2 weeks. Detailed information on 32 of these cases showed that 18 were re-hospitalised between one and four times, 13 were re-hospitalised five or more times and one was unknown. The study concluded that FII presents as unexplained multiple, serious apnoea events occurring in the presence of only one person (not witnessed).⁵³ [EL = 3]

Evidence statement

Evidence from one systematic review, six prospective case series, three retrospective case series and two surveys were included in the review. The evidence shows that ALTEs account for 0.25% to 0.8% of emergency hospital attendances. Studies showed that infections, gastrointestinal problems, seizures and ‘unknown’ causes were the most common diagnosis applied, accounting for 545 of 728 diagnoses in the systematic review. The evidence shows that maltreatment is diagnosed in 0% to 15.5% of cases, but these figures were dependent on the aim of study, date of study, patient population and the investigations undertaken. One survey of apnoea monitoring programmes showed that 18 of 32 (56%) of infants who were subject to FII were readmitted to hospital on multiple occasions.

GDG considerations

There are many causes of ALTEs and the literature suggests that an ALTE due to maltreatment is rare. However, the high number of children with unknown diagnosis represents a potentially hidden population of maltreated children. The GDG found no clear evidence on the significance of multiple ALTE presentations in an individual child. Drawing on their collective clinical experience, the GDG believes that multiple ALTE presentations in the absence of a medical cause indicates a reason to be increasingly concerned about maltreatment. The GDG believes that a child who has had repeated ALTEs that have only been witnessed by one parent or carer can be at risk of serious harm and therefore this presentation represents a reason to suspect maltreatment.

The systematic review of nosebleeds in infants shows that nosebleeds can occur in cases of deliberate suffocation. The GDG believes that a nosebleed in an infant in conjunction with an ALTE should prompt investigations into the cause of these events.

There was consensus within the GDG about the recommendations in this section and thus the views of the Delphi panel were not sought.

Overview of available evidence

No relevant evidence that fulfilled the inclusion criteria was identified.

GDG considerations

The GDG's opinion is that the clinical signs and symptoms of poisoning do not differ between accidental and intentional poisoning and therefore concluded that it is of utmost importance to identify indicators relating to the circumstances and context of the poisoning incident.

The result of a biomedical investigation in a child may reveal that the child has hypernatraemia. This can be a result of, for example, diarrhoea and vomiting, over-concentrated preparations of formula feeds or deliberate salt poisoning. The GDG believes that the cause of hypernatraemic dehydration should be elicited and maltreatment excluded if appropriate and therefore recommends that maltreatment be considered if a child has hypernatraemia.

There was consensus within the GDG about the recommendations in this section and thus the views of the Delphi panel were not sought.

Overview of available evidence

One case series was identified.

Narrative summary

A case series (n = 205, ages younger than 19 years) sought to improve the understanding and recognition of inflicted paediatric submersion in children who sustained submersion injury and were hospitalised or autopsied.⁵⁶ All events were categorised as either having been inflicted or unintentional through a review of abstracted case scenarios by two paediatricians using pre-established criteria. Sixteen submersions were judged to have been inflicted and 186 as having been unintentional. Two cases were confirmed as having been intentional submersions. In the inflicted submersion group, all children were younger than 5 years.

By comparing these two groups, it was found that submersions were four times more likely to occur in bathtubs than in other sites (RR 4.14; 95% CI 2.35 to 7.29 according to our own calculations from published data; the given RR was 6.28; 95% CI 2.51 to 15.69) The data published showed that nine of 16 bathtub submersions were inflicted and 25 of 184 were unintentional.

There were no differences found between inflicted and unintentional submersions in the duration of submersion. In general, the numbers in the inflicted group were very small and therefore differences between the groups are difficult to verify. Only two cases were confirmed as being inflicted. The authors concluded that unexplained physical injuries, developmental implausibility or changing history are the main features for the recognition of inflicted submersion. [EL = 3]

Evidence statement

One study suggests that it is difficult to distinguish inflicted from unintentional submersions.

GDG considerations

The GDG believes that a non-fatal submersion injury due to maltreatment can be caused by deliberate submersion or can occur as a result of lack of supervision. The account of the incident is key in determining the probability that maltreatment has occurred and suspicion should be raised when the account is inconsistent with the injuries. A drowning incident could also give reason to suspect maltreatment but unexpected child deaths are addressed by processes that are beyond the scope of this guidance.

There was consensus within the GDG about the recommendation in this section and thus the views of the Delphi panel were not sought.

Overview of available evidence

A systematic review and a comparative study were identified that considered repeated healthcare use as a sign of maltreatment.

Narrative summary

A systematic review that searched for studies that reported repeat attendances at accident and emergency departments (A&E) for injury in physically abused and non-abused injured children attending A&E found no relevant studies.⁷ Three studies were identified but excluded because of the way in which abused children were identified. Using a data set on injured children admitted for suspected physical abuse and a separate data set on re-attendance at hospital for injuries regardless of abuse status (both from the UK), estimates of re-attendance were calculated. Of 108 children attending A&E with an injury due to suspected abuse, 22 re-attended at least once with an injury. In a database of injured children regardless of abuse status, between 20% and 49% of pre-school injured children re-attended A&E with an injury within 12 months of the initial visit; 13% to 21% had at least three injury-related visits in a year. [EL = 2+]

A longitudinal study from the USA was identified that aimed to determine whether injury-related emergency department visits among children younger than 5 years were associated with child maltreatment reports.⁵⁷ During one calendar year, there were 56 364 injury visits by 50 068 children. Sexual assault cases were excluded from the study. The relative risk of having a substantiated report of physical abuse or neglect was 2.5 (95% CI 2.1 to 2.9) when children attended for two different injuries compared with one. For children with three injuries, the relative risk was 2.3 (95% CI 1.5 to 3.6) and for children with four or more injuries, the relative risk was 4.7 (95% CI 2.4 to 9.2). [EL = 2+]

Evidence statement

According to the systematic review, there is no UK-based published study that addresses the rate of previous attendance at A&E departments for injury in physically abused children in comparison with non-abused children. A recent US longitudinal data linkage study found a strong link between repeated attendance and substantiated maltreatment, suggesting that there is an increased tendency for children who have been maltreated to have sought medical opinion more often than non-abused children. Indirect data from the UK suggest that it is not uncommon for pre-school children to re-attend at A&E in a 12 month period irrespective of abuse status.

Delphi consensus (see also Appendix C)

The GDG sought the opinions of the Delphi panel on this topic. The following statements were drafted:

GDG considerations

Several studies of children who have sustained abusive fractures, thermal injury, inflicted head trauma and sexual abuse (see Section 4.1.8 on fractures, Section 4.1.5 on thermal injuries, Section 4.1.9 on intracranial injuries and Section 4.2 on anogenital signs, symptoms and infections) suggest that these maltreatments are repeated or ongoing. It is therefore likely that frequent presentation with injury is suggestive of child abuse.

The GDG considered that data from other countries could not be extrapolated directly to the UK population of children and young people. This is based on the fact that non-UK-based studies were conducted in health service settings with configurations and support infrastructures different to those found in the NHS. However, the relevant data were discussed by the GDG and used to inform their consensus-based recommendation.

The GDG believes that there are many innocent reasons why children may re-attend, so frequent re-attendance should not prompt an immediate suspicion of maltreatment without an examination of the circumstances.

The GDG sought the opinions of the Delphi panel on this recommendation and sufficient agreement was reached (see above and Section C.2.9).

Overview of available evidence

A number of systematic reviews were identified that brought together case reports of MSBP.

Narrative summary

A systematic review was identified that synthesised data on 451 cases of MSBP found in the literature between 1972 and 1999.⁵⁸ This review was an update of a paper published in 1987 that included 117 cases.⁵⁹ The mean age at diagnosis was 48.6 months (range 0–204 months) (n = 404) and 52% of cases were male. The estimated time between onset and diagnosis was 21.8 months (range 0–195 months) (n = 201). In 78.5% of cases, the perpetrator was the mother and in 6.7% of cases it was the father. Within the reports, children had, on average, three medical problems reported (range up to 19 per child). The most commonly reported symptom was apnoea (26.8% of case reports), followed by diarrhoea (24.6%) and seizures (17.5%). Seventy-six other symptoms were recorded in this case series and included behaviour (not defined), asthma, allergy, fevers, unspecified pain, infection and bleeding. Symptoms were induced in 57.2% of cases, and nearly half of these were induced while the child was in hospital. While the synthesis of information in this review is of high quality, reporting bias in case reports must be considered. [EL = 2++]

A second systematic review searched for cases of MSBP that occurred outside the main countries where it is known to be well documented (the UK, the USA, Canada, Australia and New Zealand).⁶⁰ In 59 articles from 24 countries, 122 cases were identified. Some of these also appeared in the review cited above. The mother was the perpetrator in 86% of cases, the father in 4%, a spouse unrelated to the child in 4% and the grandmother in 2% (n = 93). The majority of children were aged between 3 and 13 years (52%) with 26% younger than 3 years and 12% aged 13 years or over; 9% were adults (n = 76). Male children comprised 54% of cases (n = 81). Counts were not given on the different presentations but the authors commented on similarities in distribution with other systematic reviews. A dissimilarity in the prevalence of induced apnoea was noted. [EL = 2++]

A narrative systematic review⁶¹ summarised the two articles above and added information from a study by Folks⁶² in which two patterns of presentation were identified: apnoea, seizures and cyanosis or diarrhoea, and vomiting, nausea and bone and joint problems. The most common forms of assault were suffocation, giving drugs and poisoning. The authors also noted the wide variety in fabricated illnesses. Histories of multiple hospitalisations and repeated medical investigations were also mentioned in cases of FII. [EL = 2++]

A study from the Netherlands identified cases of MSBP from the literature in paediatric gastroenterology patients.⁶³ The authors gave details of the mechanisms of fabrication and the medical investigations that were undertaken. No data were presented on the cases identified. [EL = 2−]

One study sought out cases of MSBP in children aged 6 years or over.⁶⁴ The authors identified nine cases from their clinic over a 2 year period (2001–2003) and 42 from the literature (1966–2002) and the oldest patient was 17 years (mean age across both groups was 9.3 years) (n = 41 as data were only available on 32 cases from the literature). False reporting occurred in all of the clinic cases (n = 9) and in 62% of the literature sample. [EL = 2−] Many of the cases from the literature are addressed in the systematic review discussed at the start of this section.⁵⁸

A retrospective chart review of 24 years detailed presenting complaints and associated falsified or induced conditions in cases of paediatric condition falsification.⁶⁵ Comparisons were made between cases where there was a history of allergy, asthma, sinopulmonary infections, ENT surgery or drug sensitivity (n = 71) and other cases of paediatric condition falsification (n = 33). Presenting features were asthma, sinopulmonary disease or hearing loss (14 children), CNS disease/seizure (23), apnoea (17), gastrointestinal symptoms (15), other infections (eight), failure to thrive (five), sexual abuse (two), immune dysfunction (one) and other (three). Associated falsified or induced conditions included haematological bleeding, infections, vomiting, diarrhoea, failure to thrive, apnoea, seizures and a number of others. [EL= 3]

Evidence statement

Studies that bring together reported cases of FII suggest that the most common presentations are apnoea, diarrhoea and seizures. Males are no more likely than females to be subject to this type of maltreatment and the perpetrator is the mother in most cases.

GDG considerations

The complexity of FII suggests that a case is unlikely to cause suspicion on first presentation to a healthcare professional as the histories that perpetrators provide are often intricate, knowledgeable and believable. Common methods of inducing illness are smothering and poisoning but any symptom or sign can represent FII. The GDG's clinical experience suggests that FII may only be diagnosed once there has been recognition that there are inconsistencies in the history, presentations and assessment findings. The GDG found descriptions of the indicators of FII made in Working Together to Safeguard Children¹ and its supplementary guidance⁶⁶ to be good representations and has adapted them for use here.

There was consensus within the GDG about the recommendations in this section and thus the views of the Delphi panel were not sought.

GDG considerations

A literature search was not conducted in this area as an evidence base in the medical literature was not expected. Poor school attendance or persistent lateness may constitute neglect of the child's education due to parental/carer failure to ensure that their child attends school. The stated reason for the poor attendance may be ill health and this may or may not be valid. The GDG believes that, in some circumstances, these absences may be due to fabricated illness and may go unnoticed by the school as ill health is an accepted reason for absence. The GDG notes that this is an uncommon occurrence but maltreatment should be excluded in these circumstances.

There was consensus within the GDG about the recommendation in this section and thus the views of the Delphi panel were not sought.