The presentation of deliberate self-harm (especially overdoses) to accident and emergency departments and subsequent management by child and adolescent mental health services is comparatively well documented (Reference BrentBrent, 1997; Royal College of Psychiatrists, 1998). However, it is now recognised that other ‘accident presentations’ are commonly associated with a range of psychosocial difficulties and “health-risky behaviour” (Reference MilgramMilgram, 1993; Reference RivaraRivara, 1995). For instance, in the US substance misuse (alcohol or drugs) has been reported as commonly associated in teenagers with presentations in accident and emergency departments or trauma admissions (e.g. Reference Loiselle, Baker and TempletonLoiselle, 1993; Reference Mannenbach, Hargarten and PhelanMannenbach, 1997; Reference Spain, Boaz and DavidsonSpain, 1997). Recommendations have been made to use screening more and subsequently refer those positive to appropriate services (Reference Maio, Portnoy and BlowMaio, 1994; Reference Buchfurer and RadeckiBuchfurer & Radecki, 1996).

In the UK drug and alcohol use and misuse has increased among young teenagers (Health Advisory Service, 1996; Reference Miller and PlantMiller & Plant, 1996; Reference ColemanColeman, 1997). A British inner city casualty audit (Reference ConnorConnor, 1997) revealed that about 200 under-17s were brought in under the influence of alcohol in one year. Anecdotally, the inner city casualty department of the hospital in which the child psychiatry department is located reported an increase of children attending in intoxicated states. However, very few had ever been referred to child psychiatry.

The study

Casualty notes were collected on all 11- to 16-year-olds who had presented to the accident and emergency department during the course of two non-consecutive weeks. The entry made by the casualty officers was analysed and attenders were classified into three groups of risk:

1. (a) A low-risk group, in which the history was internally consistent and compatible with the injury sustained, for example, a child brought in by a teacher with a cut finger, after attempting to ‘slam-dunk’ a basketball into the hoop.

2. (b) A high-risk group, where either there was evidence of risk-taking behaviour, such as an overdose, or the history did not explain the injury.

3. (c) An uncertain-risk group, where there was insufficient history to judge risk (this included four youngsters who did not wait to be seen).

A note was made of whether the possibility of drug or alcohol misuse had been enquired about.

The high- and low-risk groups were then compared against the following parameters, which were recorded on the casualty card:

1. (a) who they were accompanied by (family member/nobody or person other than family);

2. (b) gender;

3. (c) age;

4. (d) time of presentation (night/day);

5. (e) diagnosis (medical/trauma);

6. (f) outcome of attendance (discharged/admitted/‘other’).

Statistics

The χ² test was used to test associations of the binomial variables: risk, gender, who accompanied the patient, diagnosis, and outcome. The association of risk with night versus day was calculated using Fisher's exact test. The association of risk with age was calculated using the Mann—Whitney U-test and further logistic regression analysis.

Findings

A total of 130 presentations of 11- to 16-year-olds to the accident and emergency department was recorded over the two weeks under study. In none of the 130 cases was a record made of an enquiry into the possibility of drug or alcohol use, regardless of the circumstances of the presentation. Of the 130, 71 (55%) were considered to be low-risk, 22 (17%) high-risk and 37 (28%) uncertain-risk (the last group was excluded from the statistical analysis).

Four parameters were not significantly associated with risk: age, gender, diagnosis and time of presentation. For outcome a high-risk individual was significantly less likely to be discharged, admitted for non-psychosocial reasons or followed up in the fracture clinic (P=0.006) (see Table 1). Of the 10 ‘other’ recorded for the high-risk group, four were sent back to the GP (one unwanted pregnancy, one self-harm, one medically unexplained shortness of breath and one child with recurrent fainting episodes). There were five referrals to child psychiatry, some of whom were admitted, all having harmed themselves in some way. One child was referred to casualty review, having punched through a window. It is of interest that four attenders said that they were pregnant or possibly pregnant.

Table 1. Risk by outcome

Risk level	Discharged	Fracture clinic	Non-psychosocial admission	Other	Total
High risk	8 (36.3%)	1 (4.5%)	3 (13.6%)	10 (45.5%)	22
Low risk	43 (61.4%)	5 (7.1%)	14 (20.0%)	8 (11.4%)	70

Of the ‘other’ category in the low-risk group, five were referred back to the general practitioner (suture removal, chicken pox, conjunctivitis, asthma and laceration), two were referred to out-patients and one for casualty review, all with musculoskeletal injuries.

Of the ‘unascertained’ group four removed themselves before a full history or examination was taken. However, the presenting complaints were: one assault, one facial injury from a fight, one ‘inadvertently stabbed self’ and one rash — so that while the cases were not completely examined, risk looks high. Of these two were unaccompanied, two were with their mother.

When risk was compared with ‘accompanied by’, there was a significant result showing that risk was associated with not being accompanied by a family member on presentation to casualty (P=0.006), for example, a 15-year-old girl unaccompanied to casualty with a complaint of ‘Tippex in the eye’ raising the suspicion of solvent misuse (see Table 2).

Table 2. Risk by “accompanied by”

Risk level	Family member	Non-family member/nobody	Total
High-risk	12 (54.5%)	10 (45.5%)	22
Low-risk	59 (83.1%)	12 (16.9%)	71
Total	71 (76.3%)	22 (23.7%)	93

Comment

It is important to identify psychosocial risk contributors to the presentation of youngsters in accident and emergency departments and to recognise where drugs or alcohol are part of this. At least 17% of the 130 presentations to casualty of 11- to 16-year-olds were high risk in terms of evidence of behaviour or inconsistency of the injury with the history. We found that another indicator of difference between the groups was whether or not the child was accompanied by a member of the family, so that this is important to record. That the difference between the groups was recognised in some way by staff was reflected in the differences of outcome between high-risk and low-risk groups, with a higher proportion of the former being referred on to child psychiatry usually for deliberate self-harm or to the general practitioner. Only one case of self-harm out of six was not referred to child psychiatry. While there was no evidence in the casualty note system of referral on to social services there is a back-up notes review system provided by a liaison health visitor.

The specific problems of drugs and alcohol are still being overlooked despite evidence suggesting that there is increasing use of drugs and alcohol in the young, and that there is a strong likelihood of associated presentations in casualty. There may be a number of reasons for this. First of all, there may be pressures of time and that substance misuse is not even thought about. Second, if it is thought about and asked, what should the casualty doctor do with this information? If every child who experiments with drugs is referred, the child psychiatry services would be quickly swamped. Similarly, this creates a problem for the child psychiatry services, who do not necessarily have the expertise and experience in dealing with children with genuine addictions, or access to specialist services in the way adult psychiatrists do. Yet a casualty ‘crisis’ could be a good opportunity to pick up and intervene with those with substance misuse problems.

As the next stage of the audit we fed back the findings to a variety of groups — accident and emergency, the child psychiatry department, paediatricians. We developed brief advisory guidelines for casualty staff on how to assess and appropriately refer children who may have misused drugs or alcohol, and we have also produced a leaflet for young people and parents advising them about sources of help if there are concerns about the use of drugs or alcohol. Meanwhile, discussions with the addictions directorate and the local purchaser are ongoing and we hope to be able to provide better services for the target group of young people who misuse drugs and alcohol and have associated psycho-social problems.