A randomised controlled trial of general practitioner safety advice for families with children under 5 yearsBMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7144.1576 (Published 23 May 1998) Cite this as: BMJ 1998;316:1576
- a Colwick Vale Surgery, Colwick, Nottingham NG4 2DU
- b Division of General Practice, University of Nottingham Medical School, Queen's Medical Centre, Nottingham NG7 2UH
- Correspondence to: Dr Kendrick
- Accepted 22 April 1998
Objective: To assess effectiveness of general practitioner advice about child safety, and provision of low cost safety equipment to low income families, on use of safety equipment and safe practices at home.
Design: Randomised, unblinded, controlled trial with initial assessment and six week follow up by telephone survey. Twenty families from intervention and control groups were randomly selected for a home visit to assess validity of responses to second survey.
Setting: A general practice in Nottingham.
Subjects: 98% (165/169) of families with children aged under 5 years registered with the practice.
Interventions: General practitioner safety advice plus, for families receiving means tested state benefits, access to safety equipment at low cost. Control families received usual care.
Main outcome measures: Possession and use of safety equipment and safe practices at home.
Results: Before intervention, the two groups differed only in possession of fireguards. After intervention, significantly more families in intervention group used fireguards (relative risk 1.89, 95% confidence interval 1.18 to 2.94), smoke alarms (1.14, 1.04 to 1.25), socket covers (1.27, 1.10 to 1.48), locks on cupboards for storing cleaning materials (1.38, 1.02 to 1.88), and door slam devices (3.60, 2.17 to 5.97). Also, significantly more families in intervention group showed very safe practice in storage of sharp objects (1.98, 1.38 to 2.83), storage of medicines (1.15, 1.03 to 1.28), window safety (1.30, 1.06 to 1.58), fireplace safety (1.84, 1.34 to 2.54), socket safety (1.77, 1.37 to 2.28), smoke alarm safety (1.11, 1.01 to 1.22), and door slam safety (7.00, 3.15 to 15.6). Stratifying results by receipt of state benefits showed that intervention was at least as effective in families receiving benefits as others.
Conclusions: General practitioner advice, coupled with access to low cost equipment for low income families, increased use of safety equipment and other safe practices. These findings are encouraging for provision of injury prevention in primary care.
We assessed the effectiveness of general practitioner advice about child safety, and provision of low cost safety equipment to low income families, on safe practices at home
The intervention increased safe behaviour and use of safety equipment
The intervention was equally effective in families receiving means tested benefits as in those not receiving benefits
The effectiveness of this intervention should be evaluated over longer periods, in other practices, and when delivered by other members of the primary healthcare team
The Health of the Nation suggests that primary healthcare teams should provide safety advice to parents during child health surveillance programmes, advise on and provide access to safety equipment, check and advise on hazards in the home, provide advice on first aid, and advise the community on safety.1 Studies have suggested that a lack of time and expertise are often quoted as factors that limit the provision of injury prevention in primary care.2–10 Hence, it has been suggested that any initiative to be introduced into general practice must be quick and easy to carry out.11
Studies in the United States have shown that counselling by physicians improved safety behaviour and reduced hazards,12–15 and one small study showed a reduction in falls in infants.16 In addition, increasing access to safety equipment increased the installation of smoke alarms17 and socket covers but not cupboard locks, which were more difficult to install. 18 19
However, differences between the healthcare systems of the United Kingdom and the United States may limit the generalisability of these studies to UK settings. We therefore undertook this study to assess the effectiveness of counselling on injury prevention by a general practitioner in conjunction with access to low cost safety equipment for families on a low income in the United Kingdom. The study received approval from the ethics committee of Queen's Medical Centre.
Subjects and methods
The study population comprised the 169 families with children aged ≤5 years that were registered with a single handed general practice in an urban area of Nottingham. The 165 (98%) families that responded to a questionnaire on child safety practices were numbered from 1 to 165, and we used random number tables to allocate them, by number, to an intervention or a control group. We calculated that 73 families were required in each group, based on β=0.1, α=0.05, a baseline possession of safety equipment of 60%,20 and a difference of 25% in possession of safety equipment. The figure shows the flow of families through the trial.
We used a questionnaire to obtain information on families' use of safety equipment; storage of sharp objects, cleaning products, and medicines; risk factors for unintentional injury; and sociodemographic factors. The questions on risk and sociodemographic factors had previously been validated.21 The questionnaire was designed to be administered by telephone or postal survey. It was piloted in another general practice with a similar patient population, with 30 questionnaires administered by each method. No major changes were made to the questionnaire based on the pilot study.
The questionnaire was pre-coded. Each safety practice was assigned a category combining several aspects of safety. For example, for the storage of sharp objects, safe storage was defined as all sharp objects stored above adult eye level or always kept in cupboards or drawers that were always locked. Moderately safe storage was defined as some sharp objects stored below adult eye level in cupboards or drawers that were only sometimes locked or only some of which had locks. Unsafe storage was defined as some or all sharp objects always stored below adult eye level in cupboards or drawers that were not locked.
The questionnaire was administered at baseline and at follow up, six weeks after intervention, by telephone by the general practitioner (MC) or sent by post to those families without a telephone. Non-responders to the postal questionnaire were sent a reminder three weeks later.
The validity of the responses was assessed by home visits to a random sample of 10 families in each of the intervention and control groups two weeks after the second questionnaire. MC, who was blind to the responses on the questionnaire, made the home visits.
The intervention consisted of standardised advice and safety leaflets concerning smoke alarms, stair gates, fireguards, cupboard locks, covers for electric sockets, door slam devices, safe storage of medicines, sharp objects, and cleaning materials. Families receiving means tested state benefits were offered a smoke alarm for 50p and two window locks, three cupboard locks, six socket covers, or a door slam device for 20p, all available from the surgery at the time of the consultation. Stair gates and fireguards were offered at £5 per item via the health district's low cost scheme, which was available to families receiving benefits across Nottingham Health District (including control families) and was accessed via health visitors, with equipment being delivered to a local health centre for collection by parents.
The intervention took place during child health surveillance consultations or opportunistically during other consultations, or the family was asked to make an appointment specifically for the intervention. The control group received routine child health surveillance and routine consultations, but without the intervention. The mean length of consultation for safety advice was 20 minutes.
We analysed the data, on an intention to treat basis, using SPSS for Windows.22 The results are presented as relative risks (95% confidence intervals) of using safety equipment and behaving safely, and the number needed to treat to facilitate one family to use safety equipment or behave safely. We assessed the consistency between the responses to the questionnaire and observed safety practices by means of κ coefficients.23
The consistency of responses to the questionnaire and at the home visit was high: 21 questions showed complete agreement, with κ coefficients of 1; for five questions κ=0.75-0.99; for six questions κ=0.59-0.74; and for four questions κ<0.60. Two questions had almost complete agreement (95% in each case), but the κ coefficient was low because all but one of the responses were positive on the first questionnaire.24 The two remaining questions with low κ coefficients concerned the level at which sharp objects were stored in the kitchen (κ=0.49) and the use of socket covers on unused sockets (κ=0.33).
Table 1 shows the baseline characteristics of the study population. Thirty three per cent of the families in the intervention group and 35% of those in the control group reported that at least one of their children had had more than one attendance at the general practice or accident and emergency department for an injury.
After the intervention, families in the intervention group were more likely to use a range of items of safety equipment than were control families (table 2). A higher proportion of families in the intervention group were categorised as safe for their storage of sharp objects and medicines and for safety of windows, fireplaces, electric sockets, smoke alarms, and door slams (table 3), suggesting that even when these families did not obtain items of safety equipment, such as cupboard locks and window catches, they did change their safety behaviour.
After stratifying the results by receipt of means tested benefits, we found that, among those receiving benefits, a significantly higher proportion of families in the intervention group than controls were categorised as safe for five of the nine safety practices. Among those not receiving benefits, significantly more families in the intervention group were categorised as safe for three of the nine safety practices. This suggests that the intervention was equally, if not more, effective in the families receiving benefits.
The high response rate to the baseline questionnaire suggests the results of this study are generalisable to the practice population. The similarity of the study population to that of Nottingham in terms of sociodemographic factors25 suggests these results may be applicable to a wider population, although the lower proportion of families belonging to an ethnic minority in the study means that caution must be exercised in extrapolating the results of this study to minority groups.
As safety practices were self reported, it is possible that families receiving the intervention overreported safety practices to a greater degree than did control families, so overestimating the effect of the intervention.26 However, the high degree of consistency of responses to questionnaire and the safety practices observed on the home visit suggest that overreporting did not occur to any great degree and did not occur differentially in the intervention group.
The results from this small study suggest that general practitioners can increase safety practices through giving routine safety advice and providing low cost safety equipment. The short follow up period means we cannot draw conclusions about the long term effectiveness of such an intervention, and further studies are needed. Further evaluation is needed in other practices to see if our findings can be replicated elsewhere. Furthermore, the effectiveness and cost effectiveness of other members of the primary care team undertaking the same intervention programme requires evaluation. The short time scale and small sample size of this study precluded any assessment of reductions in frequency or severity of injury, but such evaluations are needed before new interventions are introduced into routine primary care.
Our finding that the intervention was at least equally effective in families receiving benefits is important as there is debate about the relative effectiveness of population versus targeted approaches to injury prevention in primary care. 21 27 28 This study used a population approach, with tailoring of the interventions to specific groups in order that families relying on state benefits were not disproportionately disadvantaged by taking part in the interventions. This is the first UK study to suggest that a population approach would be equally effective in different socioeconomic groups. It has been argued that the population approach would lead to widening inequalities in health, as interventions may be less effective in those most at risk.27 This study suggests this is not the case, but further work, with a larger sample from a larger number of practices, is needed to confirm this finding.
The safety leaflets used in this study included Play it Safe (Health Education Authority, 1996), Your Baby's Safety At Home and At Play (Mothercare, 1994), and Home Safety Checklist (Child Accident Prevention Trust, 1996).
DK coordinated the formulation of the primary study hypothesis, discussed core ideas, designed the protocol and questionnaire, and participated in data analysis and writing of the paper. MC initiated the idea for the study; discussed core ideas; carried out the literature search and pilot studies; modified the questionnaire; administered the questionnaire, intervention, and validation studies; entered and verified the data; and participated in data analysis and writing of the paper.
Funding: Nottingham Health Authority provided a grant of £500 for the purchase of safety equipment. This research was undertaken as part of a Masters Degree in Medical Science in Primary Health Care, in the Division of General Practice at Nottingham University.
Conflict of interest: None.