Preventing childhood scalds within the home: Overview of systematic reviews and a systematic review of primary studies

Objective: To synthesise and evaluate the evidence of the effectiveness of interventions to prevent scalds in children. Methods: An overview of systematic reviews (SR) and a SR of primary studies were per-formed evaluating interventions to prevent scalds in children. A comprehensive literature search was conducted covering various resources up to October 2012. Experimental and controlled observational studies reporting scald injuries, safety practices and safety equipment use were included. Results: Fourteen systematic reviews and 39 primary studies were included. There is little evidence that interventions are effective in reducing the incidence of scalds in children. More evidence was found that inventions are effective in promoting safe hot tap water temperature, especially when home safety education, home safety checks and discounted or free safety equipment including thermometers and thermostatic mixing valves were provided. No consistent evidence was found for the effectiveness of interventions on the safe handling of hot food or drinks nor improving kitchen safety practices. Conclusion: Education, safety checks along with thermometers or thermostatic mixing valves should be promoted to reduce tap water scalds. Further research is to evaluate the effectiveness of interventions on scald injuries and to disentangle the effects of multifaceted interventions on scald injuries and safety practices.


Introduction
Children are at particular risk of thermal injuries. Globally, thermal injuries are the 11th leading cause of death between the ages of 1 and 9 years and the fifth most common cause of non-fatal childhood injuries [1]. The majority of thermal injuries in the under-fives are scalds [2]. They are important as they can result in long term disability, have lasting psychological consequences and place a large burden on health care resources, with an estimated 19 million disability-adjusted life years lost each year [3]. The treatment of scalds is resource intensive. In the USA between 2003 and 2012, the average cost per hospital stay for scald injuries in the under-fives was between $40,000 and $50,000 [4]. The total cost of treating hot water tap scald injuries to children and adults in England and Wales in 2009 was estimated at £61 million [5].
Most scalds in the under-fives occur at home [2,6]. They are most commonly caused by hot liquids from cups or mugs, baths and kettles [8,9]. Bath water scalds are more likely to involve a greater body surface area especially in infants and toddlers and are more likely to undergo admission to hospital, transfer to specialist hospital or burns unit [8].
There are a number of systematic reviews that have synthesised the evidence on scald prevention interventions. However, most of them reviewed interventions to prevent a range of childhood injuries including scalds, some do not report conclusions specific to scald prevention and the remainder report conflicting conclusions [10][11][12][13][14][15]. One review [16] focussing on interventions specific to reducing thermal injuries in children concluded that there was a paucity of research studies to form an evidence base on the effectiveness of communitybased thermal injury prevention programmes. A meta-analysis for which the searches were undertaken in 2009 found home safety education, including the provision of safety equipment, was effective in increasing the proportion of families with a safe hot tap water temperature, but there was a lack of evidence that home safety interventions reduced thermal injury rates or helped families keep hot drinks out of the reach of children [14].
There is therefore a need to consolidate evidence across existing reviews and update the evidence with more recently published studies to inform policy, practice, and the design and implementation of scald prevention. Overviews that synthesise all available evidence on a topic are more accessible to decision makers than multiple systematic reviews and can avoid uncertainty created by conflicting conclusions from different reviews, which may vary in scope and quality [17]. Overviews are useful where, as is the case for programmes to prevent scalds, there are multiple interventions for the same condition or problem reported in separate systematic reviews [18]. This paper presents the findings from an overview of reviews of childhood scald prevention interventions and a systematic review of primary studies to enable the most up-to-date information on scalds prevention interventions to be evaluated.

Literature search
We searched Cochrane Central Register of Controlled Trials (CENTRAL), Cochrane database of systematic reviews, MEDLINE, Embase, CINAHL, ASSIA, PsycINFO and Web of Science from inception to October 2012. We also hand-searched the journal Injury Prevention (March 1995-August 2012), abstracts of World Conferences on Injury Prevention and Control , reference lists of included reviews and primary studies, and a range of websites and trial registers for potentially relevant studies. No language limitation was applied.

Study selection
We included systematic reviews, meta-analyses, randomised controlled trials (RCT), non-randomised controlled trials (NRCT), controlled before-after studies (CBA) and controlled observational studies (cohort and case-control studies) targeting children aged 0-19 and their families to prevent unintentional scalds. The outcomes of interest were unintentional scalds, hot tap water temperature, use of thermometers to test water temperature, lowering boiler thermostat settings, use of devices to limit hot tap water temperature, keeping hot drinks and food out of reach, and kitchen and cooking practices. Potential eligible primary studies were identified from included systematic reviews by scanning references and further eligible primary studies were identified from additional literature searches of electronic databases and other sources. Titles and abstracts of studies were screened for inclusion by two reviewers. Where there was uncertainty about inclusion from the title or abstract the full text paper was obtained. Disagreements between reviewers were resolved by consensus-forming discussions and referral to a third reviewer if necessary.

Assessment of risk of bias and data extraction
We assessed the risk of bias in included systematic reviews and meta-analyses using the Overview Quality Assessment Questionnaire (QQAQ) [19]. The risk of bias of randomised controlled trials, non-randomised controlled trials and controlled before-after studies was assessed with respect to random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias. The risk of bias in cohort and case-control studies was assessed using the Newcastle-Ottawa scale [20]. Data on study design, characteristics of participants (e.g. age, ethnicity, socio-economic group), intervention (content, setting, duration, intensity), and outcomes (injuries, possession or use of safety devices and safety practices) were extracted using separate standardised data extraction forms for reviews and primary studies.
Quality assessment and data extraction were conducted by two independent reviewers, with disagreements being resolved by consensus forming discussions and referring to a third reviewer if necessary.

Data synthesis
In view of the clinical heterogeneity between studies in terms of design, population, intervention and outcomes, data were synthesised narratively by types of outcomes including outcomes related to safe hot water temperature, safe handling of hot food and drinks such as keeping hot drinks and food out of reach of children, kitchen and cooking safety practices such as using cooker guards or keeping children out of kitchen and other outcomes related to scalds that could not be classified specifically.

3.1.
Study selection Fig. 1 shows the process of identification and selection of studies. Four meta-analyses (each of which also contained a narrative systematic review) and 10 systematic reviews and 39 primary studies were included in the overview. Of these primary studies, 34 were identified from published systematic reviews and meta-analyses and five were identified from the additional literature search ( Table 1). Tables of excluded studies are available from the authors on request.

Study characteristics
Characteristics of included reviews are shown in Table 2. One review focused on community-based programmes to prevent scalds [16], while the remainder covered a range of injury mechanisms including but not specific to scalds. Only one review drew conclusions specific to scalds prevention interventions [16]. Two meta-analyses combined effect sizes from studies reporting safe hot tap water temperature [11,14] and one combined effect sizes from studies reporting keeping hot food and drinks out of reach [14]. Four systematic reviews narratively synthesised the evidence on the effect of interventions on scald injuries [12,13,15,16,21] and three on safe hot water temperature [10,12,15,21]. Seven systematic reviews reviewed the effectiveness of interventions on prevention of child injuries including burns and scalds, but did not make conclusions specific to scalds prevention [22][23][24][25][26][27][28]. The 39 eligible primary studies included 26 RCTs, 3 NRCTs, 7  CBAs, 2 cohort studies and 1 case-control study. The characteristics of included primary studies are show in Table 3. Most of the included studies employed multifaceted interventions including home safety inspections, education or counselling, provision of educational materials and safety devices. Included studies less commonly reported multifaceted home visiting programmes aimed at improving a range of child and maternal health outcomes, community multimedia campaigns, scald prevention education delivered through lectures or workshops, in clinical consultations, via specially designed computer programmes or other online educational material.

3.3.
Risk of bias in reviews and in primary studies Assessment of risk of bias is shown in Table 2 for reviews and  Table 3 for primary studies. For reviews, OQAQ scores ranged from 1 to 7. For primary studies, 12 of the 26 RCTs (48%) had adequate allocation concealment, 10 (40%) had blinded outcome assessment and 14 (52%) followed up at least 80% of participants in each group. Of the nine NRCTs and CBAs, none had blinded outcome assessment, two (22%) followed up at least 80% of participants in each group and two (22%) had a balanced distribution of confounders between treatment groups.

Findings from included reviews and primary studies
Findings from included reviews are shown in Table 2 and from primary studies in Table 3.       3.5.

Incidence of scalds
Six reviews reported interventions to prevent scalds from two primary studies [29,30]. No meta-analyses reported the effect of interventions on the incidence of scalds (Table 1). The first study [30], an RCT, reported significantly fewer self-reported scald injuries (validated against hospital and insurance records) two years after a school-based education programme in the intervention group (0.31%) than the control group (0.93%) ( p < 0.05). The second study, a CBA, found a reduction in the number of scalds, particularly scalds from hot tap water and from hot cooking liquids being pulled from cooker tops, in the intervention areas over a 12 year period, but does not present similar data for the control area or the statistical significance of these findings [29].
Eleven studies reported significant effects favouring the intervention group for one or more outcomes related to safe hot tap water temperature including families having a safe hot water temperature, checking hot water temperature, and using engineering equipment to control hot water temperature (Table 3). This included nine RCTs [34,37,44,45,47,49,56,58,59], one CBA [43] and one cohort study [52]. Six studies reported significantly more families in the intervention than control group had a safe hot tap water temperature [34,37,43,44,47,49,59]. Five studies reported significantly more families in the intervention than control group checked or tested their hot tap water temperature [45,52,58], including one RCT specified using water temperature cards [49] and another using thermometers [56]. A cohort study found significantly more families exposed to the intervention lowered their hot water temperature than those not exposed to the intervention [52]. One RCT found significantly more families in the intervention than control group used spout covers for bath taps [56]. However, one CBA evaluating home safety checks, education and provision of bath water USPSTF [13] Narrative review b u r n s 4 1 ( 2 0 1 5 ) 9 0 7 -9 2 4 b u r n s 4 1 ( 2 0 1 5 ) 9 0 7 -9 2 4  No p values reported for any outcomes Gielen [35] RCT    Hot water temperature Lowered temperature on water heater I = 645 (54.25%), C = 516 (56.3%), p = 0.82 b u r n s 4 1 ( 2 0 1 5 ) 9 0 7 -9 2 4 Hot water temperature Tested hot water temperature I1 = 0 (0%), I2 = 0 (0%), I3 = 1 (4%), C1 = 2 (7%), C 2 = 0 (0%), C 3 = 0 (0%); only within group pre-post comparison p values reported   thermometers found significantly fewer families in the intervention group had a hot tap water temperature less than or equal to 52 8C than in the control group [42]. Most primary studies reporting significant effects on outcomes related to safe hot tap water temperature (including families having a safe hot tap water temperature, checking hot tap water temperature and using engineering equipment to control hot tap water temperature) employed multifaceted interventions. Three RCTs and one CBA provided safety education, a home safety assessment and safety equipment [37,43,44,49]. Two RCTs provided safety education and thermometers for checking water temperature [45,56]. One RCT provided education and thermostatic mixing valves fitted by qualified plumbers [34]. Two RCTs delivered educational lectures [47,59]. One RCT compared education plus supplying thermometers to supplying thermometers alone [58]. One cohort study compared families exposed to a multi-media scald prevention campaign with unexposed families [52].
Eighteen primary studies did not find a significant effect of interventions on outcomes related to safe hot tap water temperature including families having a safe hot water temperature, checking hot water temperature and using engineering equipment to control hot water temperature. These including 11 RCTs [31,35,36,38,40,42,48,50,53,55,57], two NRCTs [41,46], three CBAs [39,53,54], one cohort study [33] and one case-control study [32]. These studies evaluated integrated or individual interventions including home visits, home safety checks, counselling, safety education and offering safety devices.

3.7.
Safe handling of hot drinks and food Three systematic reviews and one meta-analysis looked into the effect of interventions on safe handling of hot drinks and food from seven primary studies [40,41,46,49,52,60,61]. Two more primary studies were identified through additional literature search [62,63] ( Table 1). The meta-analysis estimated the pooled odds ratio for the effect of home safety education on keeping hot food and drinks out of reach; it failed to find a significant effect of the intervention (OR 0.95, 95% CI 0.61, 1.48) [14].  Of the nine studies, one RCT evaluated the effectiveness of education plus home safety assessments [60]. It found that significantly more families in the intervention group tested the temperature of food prepared in a microwave oven than the control families. The remaining eight studies (see Table 3) evaluating a range of interventions, including home safety education, tailored safety advice, home safety assessments, provision of discounted or free home safety equipment and exposure to Safe Kids Week champion, found no significant differences between the intervention and control groups. These included three RCTs [40,49,61], three NRCTs [41,46,62] and one CBA [63] and one cohort study [52].

3.8.
Kitchen and cooking safety practices Nine reviews reported the effectiveness of interventions on kitchen and cooking safety practices from 6 primary studies (Table 1) [41,52,56,60,64,65]. No meta-analyses reported pooled odds ratios related to kitchen and cooking practices. Two primary studies investigating interventions on kitchen and cooking safety practices were identified through additional literature search (Table 1) [32,62]. Two of the eight primary studies found significant effect of interventions. One RCT evaluating home safety education and home safety assessments reported that families in the intervention group were significantly more likely to have ''childproofed'' electrical heating devices in the kitchen (e.g. boiler, rice cooker) [60]. One NRCT evaluating home safety education, home safety assessments and burn and scald prevention workshops found that the intervention group were significantly more likely than the control group to have a ''child-protected'' cooker (not defined), and to have removed objects that a child could use to climb on to reach the sink [62]. However, the other six studies (Table 3) reporting on a variety of interventions including home safety education, home safety assessments, media campaigns, and free home safety equipment did not find any significant differences between the intervention and control groups in promoting kitchen and cooking safety practices. One RCT [65] evaluating the effectiveness of a school-based injury prevention programme found no significant differences between the practices of children in the intervention and control groups when cooking without an adult present. Another RCT [44] evaluating home safety education, home safety assessments and discount vouchers for safety equipment found no significant effect on keeping heating devices out of reach of children or for the use of stove guards. An RCT [56] assessing the effectiveness of an emergency department based home safety intervention found no significant effect on cooking on the back burners of cookers or turning pan handles towards the back of the cooker. An NRCT [41] evaluating providing tailored home safety education found no significant effect on keeping children away from the cooker or oven or on turning pan handles away from the edge of the cooker. One cohort study [52] evaluating Safe Kids Week 2001 found no significant differences between families who had been exposed to a media campaign on scald and burn prevention and controls for kitchen and cooking safety practices including cooking on the back burners of the cooker, keeping children out of the kitchen when cooking, turning pot handles to the back of the cooker and removing dangling cords of heating devices. A case-control study [32] investigating hazards in the homes of children who had presented with injuries from falls, burns, scalds, ingestions or choking found that no significant differences between cases and controls for having a cooker guard or not having dangling cords of heating devices.

3.9.
Other scald-related outcomes Eight reviews reported other scald-related outcomes such as burn safety scores which comprised a range of burn prevention behaviours such as pot handles left facing the edge of stove, not drinking tea/coffee or eating hot food when a child is on someone's lap, putting cool water in first when running a bath, or in some studies, undefined scald-related safety practices and undefined use of safety devices. No meta-analyses reported pooled odds ratios for any other scald-related outcomes. Four primary studies reported other scald-related outcomes. Two RCTs found significant effects on intervention groups from home safety education, home safety assessments and free home safety equipment on the burn safety scores (representing safer burn prevention practices) than the control groups [56,66]. One RCT found significantly more families in the intervention group made their homes safer after a television campaign, home safety advice, a home safety assessment check and advice on welfare benefits available to purchase safety equipment and local availability of equipment [64]. One CBA found no significant effect of a multi-faceted campaign (Hot Water Burns Like Fire) aimed at reducing the occurrence of scalds in children aged 0-4 years on scald prevention behaviours [51].

Discussion
This overview synthesised the largest number of primary studies evaluating child scald prevention interventions to date. Eligible studies were identified from comprehensive searches of published reviews, electronic databases, conference abstracts and other sources minimising the potential for publication and reporting bias. Rigorous procedures were used for study selection, quality assessment and data extraction. Our overview incorporated evidence from a spectrum of study designs including RCTs, NRCTs, CBAs, cohort studies and a case-control study to ensure maximum ascertainment of evidence in the field. There was little evidence of the effect of scald prevention interventions on the incidence of scalds. We were able to find only two studies reporting scald occurrence, one of which reported a significant reduction in the incidence of scalds following a primary school-based injury prevention programme targeting school children and parents [30]. The second reported a reduction in the incidence of scalds following a community burn prevention programme comprising home safety education, home safety assessments, the promotion and installation of cooker guards and lowering tap water thermostat settings [29]. However, the statistical significance of the reduction in scalds was not reported.
There was more evidence that home safety interventions are effective in promoting safe hot tap water temperature with two meta-analyses and 11 primary studies reporting significant effects favouring the intervention group. Most studies with significant effects provided home safety education, home safety assessments and discounted or free safety equipment including thermometers and thermostatic mixing valves. We did not find any consistent evidence that home safety interventions were effective in promoting the safe handling of hot food or drinks, or kitchen and cooking safety practices, but the number of studies reporting these outcomes was small. In addition, there was wide variation and a lack of standardisation in the tools used to measure these outcomes, which hampered evidence synthesis in general and metaanalysis in particular.
There are several limitations of the review. First, there was considerable heterogeneity in the content of interventions of included studies and most studies used multifaceted interventions, hence it was not possible to attribute treatment effects to specific components of interventions. Care needs to be taken in interpreting the effects of interventions on hot tap water temperature due to the varying definitions of a ''safe'' temperature used by different studies and some studies not providing the definition they used. In addition, the temperature defined as ''safe'' has reduced over time, with more recent studies using a lower temperature than older studies. Consequently it is possible that the interventions in our review may not reduce hot tap water temperatures to levels that would now be considered sufficient to substantially reduce the risk of scalds. There was also considerable variation in study populations across included studies, making it difficult to ascertain if interventions would benefit specific groups of children or families to a greater degree. The vast majority of included studies were undertaken in high income countries, limiting the generalizability of our findings to low and middle income countries. The risk of bias varied across studies, but up to half of the RCTs had adequate allocation concealment, blinding of outcome assessment and follow up of at least 80% of participants in each group. For the NRCTs and CBAs, none had blinded outcome assessment, and only one in five had follow up of at least 80% of participants in each group or balance of confounding factors between groups.
The new evidence we found was consistent with the findings from the two published meta-analyses [11,14] and from the published narrative systematic reviews [10,12,15,21] which found home safety interventions were effective in promoting a safe hot tap water temperature. Our findings were also consistent with the previous meta-analysis and many systematic reviews that failed to find evidence that home safety interventions improved other scald prevention practices or reduced the incidence of scalds.
Our finding that most studies which were effective in promoting a safe hot tap water temperature included home safety education, home safety assessments and free or discounted safety equipment differed from that of the review by Pearson and colleagues [27]. This review focussed on home safety assessments, with or without the provision of safety equipment. Since publication of that review, two new studies have demonstrated significant effects favouring the intervention group [34,37], both of which provided free home safety equipment. In addition, our review included a wider range of interventions and these differences may partly account for the apparent inconsistency in our findings.
Although this review focussed on interventions that could be delivered in health and social care settings, other engineering or legislative approaches may be beneficial in reducing scalds. A recent trial evaluating thermostatic control of social housing estate boiler houses with daily sterilisation demonstrated significant reductions in hot tap water temperature [67]. Legislative changes such as those requiring new boiler thermostats to be set at lower temperatures or requiring thermostatic mixing valves in domestic settings are likely to be cost-effective. An economic analysis of one of the trials included in this overview found home safety education plus fitting of thermostatic mixing valves as part of bathroom refurbishment of social housing stock saved £1.41 ($2.35, s1.70) for every £1 ($1.65, s1.20) spent [68]. A recent Canadian study evaluating legislation to lower thermostat settings on domestic hot water heaters accompanied by yearly educational information provided to utility company customers estimated cost savings of C$531 per scald averted [69]. It is therefore important that scald prevention strategies encompass other engineering and legislative approaches as well as educational ones.
The paucity of evidence we found highlights the need for research to investigate the effect of interventions on reducing the incidence of childhood scalds in the home, the safe handling of food and drinks, and safe kitchen and cooking practices. Researchers should use existing validated tools to measure these outcomes wherever possible to facilitate evidence synthesis and meta-analysis. In terms of helping households to have a ''safe'' hot tap water temperature, further analyses are required to disentangle the effects of providing home safety education, thermometers, home safety assessments and thermostatic mixing valves. Network metaanalysis has previously been used to good effect in synthesising the evidence for smoke alarms [70] and is likely to be helpful in this situation. Providers of child health and social care should provide education to reduce tap water scalds, along with thermometers or thermostatic mixing valves. Public health policy-makers and practitioners should develop and implement scald prevention strategies that encompass legislative, engineering and educational approaches to reduce scalds risk.