Aim To determine antecedent patterns of healthcare use by children fatally or seriously harmed by maltreatment.
Methods The authors analysed recorded healthcare use for children who were the subject of a serious case review (SCR) in England in 2005–2007. The SCRs were initiated when a child under 18 years old died or was seriously harmed, maltreatment (abuse or neglect) was a factor, and there were lessons for interagency working. The authors analysed a purposive sample (N=40), similar in key demographics to all 189 SCRs in England in 2005–2007.
Results Children had extensive recorded contact with universal (N=34/40; 85%) and secondary (N=26/40; 65%) healthcare services and children's social care (N=21/40; 53%). Thirty-one children (78%) had recorded health visitor and/or school nurse contact. Fourteen children (35%) had missed appointments. Almost three-quarters (N=29) had complex family problems recorded (parental domestic violence, alcohol/drug and/or mental health problems). Data quality regarding healthcare use was poor, and the extent and type of ‘missing data’ varied by age.
Conclusions Complex paediatric and family problems and a high level of contact with services preceded serious adverse events. Universal health services are likely to be well placed for giving ongoing and family-orientated support to vulnerable families. The absence of standardised data collection and any control group limits how far the Biennial Analyses of SCRs can meet their stated objective of identifying national trends and patterns. Linking SCRs to healthcare databases would provide a control group, improve understanding of the population context and diminish demands for data from professionals delivering care.
Statistics from Altmetric.com
High-profile investigations into individual cases of child death from maltreatment in England have shown that children and their families had multiple antecedent contacts with health services.1 2 Current evidence suggests that, compared with controls, maltreated children have an approximately twofold increased risk of presenting once or more to emergency departments (EDs) and primary care services,3,–,8 are twice as likely to have inconsistent or incomplete primary care provision,5 9 10 are between one and three times more likely to be receiving healthcare services for developmental or speech delay, autism or cerebral palsy,11 12 and are between two and four and a half times more likely to ever use healthcare services for behavioural or psychological problems.6 11,–,13 Less is known about maltreated children's use of both health and social care services.
What is already known on this topic
▶ High-profile investigations into cases of child death following abuse and neglect have revealed multiple antecedent healthcare contacts.
▶ Information is lacking on whether these patterns apply to seriously maltreated children overall.
What this study adds
▶ Complex paediatric and family problems and a high level of contact with services preceded death or serious injury in children who were the subject of a serious case review.
▶ Poor data quality found in serious case reviews could be addressed by linkage to routine healthcare databases and would reduce demands on professionals.
We evaluated patterns of healthcare prior to a serious adverse event in a nationally representative case series of children who died or were seriously harmed due to maltreatment. The findings can be used to generate hypotheses about service needs and patterns of service use by severely maltreated children.
We investigated antecedent patterns of healthcare use and social care contact in children who died or suffered serious harm following maltreatment, using the national sample of children who were the subject of a serious case review (SCR) in England between 2005 and 2007.14,–,16 See figure 1 for methods and table 1 for details of variables extracted. We could not investigate healthcare contacts as a marker of a serious event, as SCRs do not include a comparison group
We extracted data from the overview reports (a part of the full SCR; see figure 1). We did not have access to children's health records or to more detailed information analysed locally. To preserve anonymity, we grouped children into age bands, ordered from youngest to oldest within the bands (table 1). Of 189 SCRs conducted between 2005 and 2007, 63 overview reports were supplied to the research team undertaking the Biennial Analyses (figure 1).14 The completeness of the 63 overview reports varied considerably. All analyses presented here are based on the intensive sample of 40 children (figure 1).
The ages of the 40 children in the study ranged from birth to 17 years old. Two-thirds (N=26; 65%) were 5 years or younger at the time of the incident (table 1). The four youngest children were less than 24 h old (N=2) or less than a week old (N=2). Half the children (N=20) were boys. Seventy per cent (N=28/40) died, and 30% (N=12/40) were seriously harmed. No children were related.
Universal (primary) healthcare services
Eighty-five per cent of children (N=34/40) were recorded as using at least one universal healthcare service (table 1, figure 2). However, 14 of 34 (42%) of children in contact with universal care services had some missed antenatal or immunisation/well-baby appointments. For 10% of children (N=4/40), the health visitor or school nurse was the only recorded universal healthcare service.
Reported health problems and use of secondary care services
Of 18 children (N=18/40; 45%) with a health problem recorded (table 1), one had only a ‘minor’ problem, eight had low birth weight, seven were admitted to the Special Care Baby Unit, and seven had neurological or behaviour problems. Three children had two or more of these problems.
Half of the children (N=21/40; 53%) were receiving follow-up from secondary healthcare services at the time of the incident (table 1), of which three-quarters had recorded follow-up by a paediatrician (N=16/21; 76%). Of all 40 children, only 15% (N=6) were recorded as receiving follow-up from Child and Adolescent Mental Health Services. Half of the 40 children (N=22; 55%) had recorded admissions to hospital or Emergency Department attendances (table 1). Two-thirds of the children (N=26; 65%) had either recorded admissions or Emergency Department attendances or both (table 1).
Family factors and children's social care contact
Parental problems were common. Almost two-thirds of cases (N=29/40; 73%) had a concern recorded about complex family problems (table 1). These included: parental substance abuse (N=13/40; 33%), domestic violence (N=21/40; 53%) or parental mental health problems (N=25; 63%). These data reflect practitioner awareness or concern, not service use.
Over half the children were recorded as being known to children's social care (CSC) at the time of the incident (N=21/40; 53%) (table 1), and more than four out of five families were recorded as previously receiving follow-up from CSC (follow-up in child, sibling or parent lifetime N=31/40; 78%). Six children (N=6/40; 15%) were the subject of a Child Protection Plan at the time of the incident (table 1) and five were the subject of a court order (4 care orders, 1 contact order).
Service use across sectors
Half (N=11/21; 52%) of the children with CSC follow-up at the time of the incident had missed appointments or had a period of not being registered with a general practitioner (GP). Fourteen of these 21 children (67%) had a serious health problem and/or use of secondary healthcare services recorded and one further child had a minor health problem (table 1). Missed appointments and lapses in GP registration (N=7/19; 37%) and recorded health problems and/or secondary health care use (N=10/19; 52%) were also common among children without CSC follow-up at the time of the incident.
Only a minority of children (N=4/40; 10%) had no record of healthcare use in their lifetime or any recorded history of CSC follow-up for themselves, siblings or parents. Two were concealed pregnancies and two were adolescents.
The extent of missing data varied by the age of the child at the time of the incident. A third (N=5/14; 36%) of the children without recorded secondary care follow-up, hospital admissions or ED visits were less than 8 weeks old. Birth problems were more likely to be recorded for children under five (N=8/16 children under 5y had birth problems recorded) than over 5 years (N=1/14 children over 5y had birth problems recorded). Information about antenatal care and GP registration was also most likely to be recorded for children aged under 1 year and least likely to be recorded for children aged 11 years or older (table 1).
Most children who died or were seriously harmed due to child maltreatment had a record of previous contact with universal and/or secondary healthcare services and/or CSC services. Many children were in contact with multiple services, and few children were not known to any of these services. Most families' records indicated professional concerns about complex family problems. Health visitors and school nurses had contact with most children, including those with little other service use. GP registration was infrequently mentioned. Missed routine appointments with universal health services were common. These findings are likely to underestimate service use.
Comparison with existing literature
In the absence of a control group for SCR data, we contextualised our findings using indirect comparisons with the general child population, focusing on data from England or the UK (table 2). These indirect comparisons are limited by differences in populations, time at risk and outcome measures, and should be interpreted with caution.
Rates of universal health service use (antenatal care, immunisations, GP registration) appear higher in the general population than in children who were the subject of an SCR between 2005 and 2007 (table 2).17,–,19 In the general population, rates of partial immunisation and lack of antenatal care are higher in certain ethnic groups, young parents and/or families from deprived areas, probably due to barriers to accessing services.17 19 It is likely that the families of children who are the subject of a SCR experience similar barriers to receiving care (45% of the intensive sample of children who were the subject of a SCR were living in ‘poor conditions’(N=18/40) and 10% of all children who were the subject of a SCR 2005–2007 were born to ‘teenage mothers’ (N=18/189)).14 A recent systematic review found no strong evidence that maltreated children attend ED more than other children after socio-economic factors were taken into account.7
Between 13% and 31% of all women in the general population experience domestic violence in their lifetime,20 and almost 40% of mothers (and 20% of fathers) in the general population received a diagnosis or were treated for depression by the time their child was 12 years old21 (table 2). The higher levels of recorded complex family problems for families of children who were the subject of a SCR might be explained by hindsight (ie, concerns in light of the incident) and/or increased ascertainment given their degree of contact with services.
Children who were the subject of a SCR had high levels of contact with social care services (78% lifetime prevalence of index child, sibling or parent contact with social care services). This might be due to the criteria of ‘lessons for interagency working’ that were used to select children for SCR.16 Information on the proportion of children who have ever had contact with CSC is not available for the UK, but 4% of all children (aged 0–15 years) in Western Australia have had a child protection notification by age 10,22 and in Southern Australia 23% of white children by 16 years of age.23
There are parallels between our findings and those of the Confidential Enquiry into Maternal and Child Health (CEMACH) study of UK child fatalities.24 CEMACH noted that: universal (primary) healthcare services were in contact with many children (42%) in the 3 months prior to death but missed universal healthcare appointments were common and worrying (3%); and a high proportion of children (77%) had complex medical needs. As with our study, missing data on primary healthcare use was common (20%). The CEMACH study concludes that primary care is an important service for children who are at risk of death, based on their coverage of the child population, their role as the repository of information from across health services (eg, about injuries) and ability to provide continuity of care for children with chronic disease.
Strengths and weaknesses
We used the best available source of national data for England on deaths or serious harm related to child maltreatment. Limitations included the lack of standardised or systematic recording of information in the lengthy narratives and missing information (‘no mention’). Our findings are likely to underestimate need and service use. It was not possible to ascertain frequency or intensity of service use, appropriateness or quality of contacts with services, or whether parental problems were alleged or confirmed, ongoing or time-specific or recognised by services at the time of the incident. Without comparable population-based data, no inferences can be drawn about whether the findings of chronic healthcare problems and high levels of service use reflect a deprived environment or exposure to maltreatment or both. SCRs do not collect any measures of deprivation, and therefore indirect comparisons cannot be made between similarly deprived children in the general population.
Implications for practice
Our findings raise three hypotheses:
Insufficient attention is being given to the role of healthcare for children subject to serious maltreatment.
The picture of complex family problems and chronic health problems in the child indicate considerable ongoing healthcare need but there was a high level of ‘missing’ data about health and healthcare service use (no mention of) in the SCR overview reports.
A substantial proportion of children known to CSC were not registered with a GP and/or had missed appointments with universal services.
Seriously maltreated children might benefit from proactive and long-term engagement with universal healthcare services.
Eighty-five per cent of children were recorded as in contact with a universal healthcare service, while only half were being followed up by CSC at the time of the incident.
Health visitors and school nurses were in contact with children who had little other recorded service use.
There is unmet healthcare need in seriously maltreated children.
Many children had missed appointments with universal healthcare services and/or periods when they were not registered with a GP.
Despite higher-than-average healthcare need in seriously abused children, rates of complete use of universal healthcare services might be lower than average, and rates of hospital admissions and ED use do not appear higher in children who were the subject of an SCR.
Implications for SCRs
Given the potential for SCRs to impact negatively on child safeguarding (creating a risk adverse culture of blame and by diverting resources25 26), it is important that data generated by SCRs be robust and useful, and provide value for money. Our findings of poor quality and unsystematic collation of data in SCRs, and the variable judgements which cases should be the subject of a SCR, cast doubt on whether this approach (via the Biennial Analysis) can meet the SCR's secondary objective of identifying national trends and patterns in these children.15 These problems will not be solved by the policy requiring all deaths related to maltreatment to be subjected to SCR.15
To maximise value for money, it is essential that the complete SCR dataset reaches the research team conducting the biennial analysis. Computerised processes may help achieve completeness.
Service providers responsible for funding SCRs should consider linking data on maltreatment-related deaths to routinely available healthcare data and the child death overview processes. This would ensure systematic data capture to build chronologies of service use, would increase data completeness, accuracy and comparability between SCRs, and, most importantly, would allow comparisons of patterns of care with similar children who do not die. Such an approach would also release professional time to address lessons learnt locally.
Funding The Department of Health contributed to the project. JW's time was funded by a joint MRC and ESRC PhD studentship. This work was partly undertaken at Great Ormond Street Hospital/University College London, Institute of Child Health which received a proportion of funding from the Department of Health's National Institute of Health Research (‘Biomedical Research Centres’ funding). The Medical Research Council provides funds for the MRC Centre of Epidemiology for Child Health. The research also took place at the Centre for Research on the Child and Family at the University of East Anglia where MB, SB and PB are located. Their time was funded by the Department of Children Schools and Families.
Competing interests None.
Ethics approval Ethics approval was provided by the University of East Anglia Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.