Article Text

The short-term cost of falls, poisonings and scalds occurring at home in children under 5 years old in England: multicentre longitudinal study
  1. N J Cooper1,
  2. D Kendrick2,
  3. C Timblin2,
  4. M Hayes3,
  5. G Majsak-Newman4,
  6. K Meteyard1,
  7. A Hawkins5,
  8. B Kay6
  1. 1Department of Health Sciences, University of Leicester, Leicester, UK
  2. 2Division of Primary Care, University Park, Nottingham, UK
  3. 3Child Accident Prevention Trust, London, UK
  4. 4Norfolk and Suffolk Primary and Community Care Research Office, Hosted by South Norfolk CCG, Norwich, UK
  5. 5Institute of Health & Society, Newcastle University, Newcastle upon Tyne, UK
  6. 6Emergency Department, Bristol Children's Hospital, Bristol, UK
  1. Correspondence to Professor Denise Kendrick, Division of Primary Care, Floor 13, Tower Building, University Park, Nottingham NG7 2RD, UK; denise.kendrick{at}


Background Childhood falls, poisonings and scalds, occurring predominantly in the home, are an important public health problem, yet there is limited evidence on the costs of these injuries to individuals and society.

Objectives To estimate National Health Service (NHS) and child and family costs of falls, poisonings and scalds.

Methods We undertook a multicentre longitudinal study of falls, poisonings and scalds in children under 5 years old, set in acute NHS Trusts across four UK study centres. Data from parental self-reported questionnaires on health service resource use, family costs and expenditure were combined with unit cost data from published sources to calculate average cost for participants and injury mechanism.

Results 344 parents completed resource use questionnaires until their child recovered from their injury or until 12 months, whichever came soonest. Most injuries were minor, with >95% recovering within 2 weeks, and 99% within 1 month of the injury. 61% emergency department (ED) attendees were not admitted, 35% admitted for ≤1 day and 4% admitted for ≥2 days. The typical healthcare cost of an admission for ≥2 days was estimated at £2000–3000, for an admission for ≤1 day was £700–1000 and for an ED attendance without admission was £100–180. Family costs were considerable and varied across injury mechanisms. Of all injuries, scalds accrued highest healthcare and family costs.

Conclusions Falls, poisonings and scalds incur considerable short-term healthcare and family costs. These data can inform injury prevention policy and commissioning of preventive services.

Statistics from

Request Permissions

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.


In England, unintentional injuries occurring in or around the home are a leading cause of preventable death and disability in children under 5 years old.1 Falls, poisoning and scalds are common in this age group, leading to 18 300, 5100 and 1420 admissions, respectively, in 2012/2013.2 In 2002 (the most recent year for which national surveillance data were collected), falls, poisoning and thermal injuries (data not presented separately for scalds) accounted for >280 000 emergency department (ED) attendances.3

In the UK, there is very limited evidence on the costs of unintentional home injuries to children under 5 years old. Based on the average cost of an ED attendance of £114,4 280 000 attendances would cost the National Health Service (NHS) nearly £32 million. The 24 820 hospital admissions would cost the NHS £19.1 million (based on an average of £586 per short stay case and £2461 for long stay cases).5 However, average costs cover all ages and all reasons for attendance or admission and may not accurately reflect costs of treating home injuries to children under 5 years old. In addition, they exclude costs of additional treatment by family doctors and other healthcare practitioners.6 ,7 They exclude indirect costs, such as lost output due to reduced productivity caused by injury, costs to the family or society and losses due to premature death. Information on the costs of injuries is important for prioritising spending on prevention, treatment and rehabilitation services and for economic evaluations of interventions.8–11 Several studies have attempted to quantify economic costs associated with unintentional injuries,5 ,12–17 but estimates are not always specific to children,5 ,13 ,16 ,17 or to children under 5 years old.18

Estimates of short-term and long-term cost of injuries, predominantly based on UK data, were highlighted in the Chief Medical Officer's 2012 Annual Report18 and a report by Public Health England.1 Both report high financial costs, with average hospital and other health service costs per case immediately after injury (across all injury types and all ages), being almost £2500,13 and note that lifetime cost of a childhood traumatic brain injury can be £4.95 million at 2012 prices.19 ,20 While there are estimates from other countries, these cannot be easily compared between countries and across time due to differences in healthcare systems, absence of standardised methodologies and different approaches used.21 This paper presents detailed information on direct and indirect costs resulting from unintentional home injuries in children under 5 years old in the UK who were participating in a series of case–control studies exploring modifiable risk factors for falls, poisonings and scalds.


Study design

Multicentre longitudinal study set in four acute NHS Trusts in Nottingham, Newcastle, Bristol and Norwich, UK.


Participants were parents of injured children (cases) who were participating in one of five case–control studies. To be eligible for the case–control studies, the children had to be aged under 5 years old and attending an ED, minor injury unit (MIU) or admitted to hospital following a fall, poisoning or scald. Full methodological details for these studies have been published.22–25 Parents were recruited to the case–control studies either face-to-face at attendance or admission for their child's injury or were approached by post within 72 h of attendance or admission. Recruited parents were asked if they were interested in participating in one of three other child injury research projects nested within the case–control studies. Those expressing interest within 2 weeks of their child's injury were eligible for the costs of injury study and were preferentially entered into the study to minimise the time between injury and questionnaire completion, except where the other nested studies had not met recruitment targets. Parents were recruited between June 2010 and February 2013.

Data collection

Participants were given (if recruited face-to-face) or sent (if recruited by post) a study information leaflet, a consent form and a questionnaire to complete covering the first two weeks after the injury. Further questionnaires were sent at 1, 3 and 12 months post injury to those who had not fully recovered at previous time points. Non-responders received up to three reminders: a postal reminder containing the full questionnaire, a telephone reminder and a postal reminder containing a mini-questionnaire to determine whether the child was fully recovered. Questionnaires covered health service resource use, family costs and expenditure for the time period since the previous questionnaire was completed. Participants were sent £5 gift vouchers for use in local stores for each completed questionnaire returned.


Self-reported resource use data were combined with the unit cost data obtained from NHS Reference Costs 2012,4 Personal Social Services Research Unit costs of health and social care 20125 and the British National Formulary 2012,26 and summed together to obtain the average cost per participant. Unit costs and sources are shown in online supplementary tables S1 and S2. All costs were inflated to 2012 UK sterling.27 Participant data were included in all cost categories where resource use was reported. Not all participants reported information for all categories; therefore, average costs for each category were calculated using variable numbers of responders. Total average costs were calculated only for participants with complete data on all cost categories. Average costs for each type of injury were estimated.

The main analysis was a complete case analysis. Sensitivity analyses were undertaken for NHS costs, non-NHS costs and total costs to check the robustness of findings to missing data. Multiple imputation was undertaken assuming data were missing at random. The imputation model included all cost component variables (listed in tables 4 and 5), which sum together to produce the total overall cost. Due to non-normality of the cost component variables, predictive mean matching was used for the imputation. The imputation model also included the socioeconomic and injury characteristics listed in table 1. Fifty data sets were imputed and were combined using Rubin's rules.28

Table 1

Characteristics of study participants


Parents of 435 children were invited to participate, 351 (81%) agreed, of which 344 (98%) provided data on recovery from injury and were included in the analysis (figure 1). Seven participants who were known not to have fully recovered at 2 weeks but who were subsequently lost to follow-up were similar in terms of age, sex, injury mechanism, admission status and NHS costs to those not lost to follow-up and their characteristics are shown in online supplementary table S3. Complete data were available for NHS costs on 288 (84%) participants, for non-NHS costs on 314 (91%) participants and for combined NHS and non-NHS costs on 268 (78%) participants. Parents of 95% of children reported full recovery within 2 weeks and 99% (n=340) within 1 month of injury. Most (75%) injuries were falls, with 18% being poisonings and 7% scalds. The mean age was 23 months, and 49% were male. Participants were relatively disadvantaged with 43% receiving state benefits, 37% living in non-owner-occupied accommodation and 15% of households without any adults in paid work. Few children (8%) had long-term health conditions prior to the injury (table 1).

Figure 1

Recruitment and follow-up of study participants.

Table 2 shows NHS and non-NHS resource use. Most children (61%) attended ED or the MIU (hereafter referred to as ED) and were not admitted, 35% were admitted for 0–1 days and 4% were admitted for ≥2 days. Three-quarters (76%) received low-cost investigations and treatment in ED (category 1 investigations and category 1 or 2 treatments, see table 2 for examples). The proportions were similar across all injury mechanisms (ranging from 70% to 82%) but were higher for scalds (100%). Hospital admission was most common for poisoning (65%) and scalds (47%). Few used GP (General Practice) (7%), outpatient (8%) or health visiting services (5%). Only 7% were prescribed medication after their injury but more (46%) used over-the-counter medications. These were rarely used for poisonings (5%), with similar percentages across the other injury mechanisms (ranging from 33% to 52%). One-fifth (18%) purchased aids or appliances, most commonly items of safety equipment, after the injury. Few parents incurred travel costs (9%), 16% lost time from work, 18% used informal childcare for the injured child and 33% used formal childcare for other non-injured children in the family, most commonly for poisonings and scalds.

Table 2

NHS and non-NHS resource use reported by parents by injury mechanism

Table 3 shows NHS, non-NHS and total costs per child by admission status, length of stay and injury mechanism. As expected, the highest NHS costs were for children admitted for ≥2 days, followed by those admitted for 0–1 days. The very small number of children admitted for ≥2 days precludes comparisons of costs by injury mechanism. The mean costs for admissions for 0–1 days were similar for all types of falls and poisonings (range £720–747), but the cost of scalds (£1011) was considerably higher. Mean ED costs were very similar for all types of falls (range £115–127). They were lowest for poisonings (£97) and highest for scalds (£178).

Table 3

NHS, non-NHS and total costs by admission status, length of stay and mechanism of injury

Non-NHS costs followed a similar pattern to NHS costs and illustrate the financial burden that injuries place on families. They were highest for children admitted for ≥2 days, followed by those admitted for 0–1 days and lowest for those attending ED but not admitted. Mean costs were highest for scalds among those admitted for ≥2 days (£399) and those admitted for 0–1 days (£200). Mean costs for those attending ED but not admitted were highest for falls from furniture (£68) and scalds (£48).

Tables 4 and 5 show components of overall NHS and non-NHS costs. Table 4 shows ED costs were higher for children admitted to hospital for ≥2 days, followed by those admitted for 0–1 days and lowest for those attending ED but not admitted. ED costs were similar across injury mechanisms for those attending ED but not admitted and for those admitted for 0–1 days. Again the small numbers admitted for ≥2 days precludes comparisons between injury mechanisms. Table 5 shows costs for admissions far outweigh costs for other NHS services. For parents, the major costs were for informal childcare and time off work. These costs were highest for scalds (mean costs: £91 informal childcare; £78 time off work) and lowest for falls on one level (mean costs: £17 informal childcare; £9 time off work).

Table 4

Mean costs per child for emergency department (ED) services by admission status and injury mechanism

Table 5

Mean costs per child for NHS and non-NHS services by injury mechanism

NHS and non-NHS costs were not found to be correlated with deprivation, although there was some evidence that participants on benefits on average had lower non-NHS costs than those not on benefits (see online supplementary figures S1 and S2). NHS and non-NHS costs by injury mechanism were similar across study centres, except for Newcastle, which was difficult to compare due to the low number of participants recruited (N=19, table 1).

The multiple imputation model used the cost component-level data together with socioeconomic and injury data to estimate the total NHS and non-NHS costs for all participants (online supplementary table S4 reports the number of missing observations for each variable included in the imputation analysis). Findings from the multiple imputation analysis were comparable to the complete case analysis with the exception of mean total cost of falls on one level (£2022; SE £1177) and falls from furniture (£2448; SE £651) for those admitted for ≥2 days. This is likely to be due to the small number of children admitted for ≥2 days resulting in substantial uncertainty in cost estimates as portrayed by the large SEs.


Main findings

Our study provides new information on the costs to the NHS and families of unintentional home injuries in England in children under 5 years old. The majority of our study population were children attending ED (61%), with 35% admitted for ≤1 day and 4% admitted for ≥2 days. Most injuries were relatively minor, with >95% recovering within 2 weeks and 99% within 1 month of the injury; hence, our cost estimates predominantly represent the short-term costs of injuries. Despite this, the mean healthcare cost of an admission for ≥2 days was typically £2000–3000, for an admission for ≤1 day was £700–1000 and for an ED attendance without admission was £100–180. For admissions for ≤1 day and for ED attendances, healthcare costs were highest for scalds. Costs for admissions for >2 days were difficult to compare across injury mechanisms due to small numbers. Mean family costs were considerable (up to £400 for an injury requiring admission for ≥2 days, £200 for an injury requiring admission for ≤1 day and £70 for and ED attendance not requiring admission) and mainly comprised costs for childcare and for time off work. Family costs varied across injury mechanisms, being highest for scalds.

Strengths and limitations

This is the first study to our knowledge reporting parent and healthcare costs for a range of common injuries in children under 5 years old in England. Seventy-eight per cent of participants were included in the complete case analysis. Multiple imputation produced similar estimates to the complete case analysis, with the exception of costs of falls on one level and falls from furniture for children admitted for ≥2 days. The number of children admitted for ≥2 days was small, especially for each injury mechanism, giving rise to considerable uncertainty around cost estimates and possibly explaining the differences between the complete case and multiple imputation analyses.

The last data collection from the national Home Accident Surveillance System in 20023 reported 5% of children who attended ED following an injury were admitted. As admissions are over-represented in our study, we analysed costs for admitted and non-admitted children separately. Our large number of admissions for ≤1 day may partly reflect development of short-stay paediatric assessment units that assess, investigate, observe and treat more minor injuries, most commonly head injuries and ingestions.29–31 We also only studied five injury mechanisms, so findings are not generalisable beyond these injuries. Although differences in healthcare systems and costs can limit the generalisability of single-country economic studies, our finding that minor injuries result in substantial healthcare and family costs is likely to be similar in other high-income countries.

Virtually all (99%) participants in our study fully recovered from their injuries within 1 month of the injury, so we have not been able to study the longer-term costs of injuries. In addition, our study included only 10 children admitted for ≥2 days, 8 who had ≥2 outpatient visits post injury and only 2 who had a further hospital admission for the same injury. Consequently, our estimates will underestimate the total costs of injuries in young children, particularly the cost of relatively rare, but resource-intensive injuries such as head injuries or complex fractures resulting from falls or scalds requiring longer-term care.

Our study participants were recruited from five case–control studies. We collected anonymised data on age and sex of children attending ED who were not recruited to the case–control studies and found only one significant difference in age between those recruited and those not recruited to one study (stair falls study; 19% recruited aged 0–12 months vs 15% of non-recruited; p<0.001). No significant sex differences were found across the five studies between recruited and non-recruited children. Therefore, our study population is probably representative of ED attenders in terms of age and sex, but representativeness in terms of other socio-demographic characteristics is unknown. Participants may have been more interested in child safety and may have been more likely to incur costs (eg, purchase of home safety equipment after an injury) than non-participants. However, as the costs of home safety equipment were small, this is unlikely to have had a major impact on our findings.

Comparison with other studies

It is difficult to compare our findings with other studies because of differences in study populations or healthcare services, the inclusion only of healthcare costs, specific types of (usually) more severe injuries or those with long-term consequences or different age ranges.14 ,17 ,18 ,20 ,21 ,32 ,33 One study reported an uncomplicated hot drink scald resulting in healthcare costs of £1850,34 a figure similar to our maximum cost for scalds for admission for ≤1 day. We have been unable to find UK studies reporting the costs to families in the under-fives with which to compare our findings.

An alternative approach to assessing the costs of injuries would have been to estimate the burden of childhood injuries by measuring what society would be prepared to pay in order to avoid childhood injuries (willingness to pay). This methodology has been used to estimate the cost of road casualties in the UK since 1988. Although the concept is appealing, well-documented practical difficulties include formulation of the questions asked and interpretation of the responses given.35 Despite this, Walter36 applied the road casualty methodology to estimate costs of home accidents, estimating the average cost of a non-fatal hospital-treated casualty aged under 5 years to be £11 338 (updated to 2012 prices).

Implications for research and practice

Our study, and the lack of other published studies, highlights the importance of measuring the cost of both major and minor childhood injuries to inform evidence-based policymaking for injury prevention and commissioning of preventive services. To enable calculation of the costs of injuries in children aged under 5 years old on a national level, larger studies are needed that cover all injury mechanisms that are representative of ED attenders and hospital admissions in the UK. It is important not only to measure the costs of injuries, but also to undertake economic evaluations of interventions to prevent child injuries in the home. Future costing studies and economic evaluations of interventions should measure the financial burden to families resulting from unintentional injuries in childhood as these can be substantial and explore economic impact by family characteristics.

What is already known on the subject

  • Falls, poisoning and scalds are among the most common injuries in 0–4 year olds, resulting in use of healthcare services in England.

  • There is little data on the healthcare or family costs of these common and mainly minor injuries in England.

What this study adds

  • Mean short-term healthcare costs were typically £2000–3000 for admissions for ≥2 days and £700–1000 for admissions for ≤1 day.

  • Mean short-term healthcare costs were typically £100–180 for emergency department attendances.

  • Mean short-term family costs were typically £100–400 for admissions for ≥2 days, £40–200 for admissions for ≤1 day and £20–70 for emergency department attendances.

Australia's National Firearms Agreement

I have probably written about this previously, but this example is so important it warrants repetition. In 1996, an assailant in Tasmania (Australia) used an assault rifle to murder 35 people and wound 19 others. Unlike the present situation in the USA, Australia's political parties quickly agreed to implement a National Firearms Agreement (NFA). Simon Chapman, the former editor of our sister journal, Tobacco Control, played a major role in bringing this about. Recently, JAMA published his account of the political response and the gratifying reduction in gun deaths that followed. In the USA, the firearm homicide rate in 2013 was 23 times higher than the rate Chapman et al report for Australia. The lesson is so obvious that it cannot be ignored.

Half of some air bags rupture

New data revealed that Takata air bags in certain Honda and Acura vehicles have a 50% chance of rupturing in a crash. This prompted two US senators to ask Honda to issue a ‘do not drive’ order to owners of these cars. They also asked that the cars be taken off the road while the defect is being repaired.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors All authors interpreted the data, critically revised the manuscript and approved the final version. NJC, DK and MH designed the study. CT, BK and AH collected data. NJC and KM analysed the data. NJC, DK, CT, MH and GM-N drafted sections of the manuscript.

  • Funding This paper presents independent research funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research Programme (RP-PG-0407-10231).

  • Competing interests None declared.

  • Patient consent Informed consent was assumed through return of completed study questionnaires.

  • Ethics approval Nottingham research ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.