Acute poisoning among children and adolescents in southern Norway

Material and method

In the period 1 August 2014–31 July 2015, all poisoning-related admissions to the Departments of Paediatric and Adolescent Medicine, Sørlandet Hospital Trust, Kristiansand and Arendal were prospectively recorded. The departments cover a paediatric population of approximately 65 000, and in 2015 there were a total of 3 980 hospitalisations of patients aged up to 19 years (not including admissions to neonatal intensive care). For each admission, the paediatrician completed a form containing questions on various factors (appendix). With the exception of the completion of this form, the patients in the study received standard care. Prior to and during the study period, all doctors in the departments received information about which patients should be included and how the form should be completed.

All forms were reviewed retrospectively. In several cases, information was missing and this was then retrieved from the patient’s admission record. A note was made if the information could not be found in either the form or the hospital medical record. To ensure that all hospitalisations were captured in the dataset, diagnostic code searches were performed on multiple occasions during and after the study period (diagnostic codes T4n–T50, F10–F19 and X6n). In those cases where patients who met the inclusion criteria were detected through diagnostic code searches, but where the form had not been completed on admission, the necessary information was retrieved from the patient’s hospital medical record. If a patient was hospitalised several times during the study period, only the first admission was included.

The form included questions on age and sex. We recorded the living arrangements of the child/adolescent under the following categories: nuclear family (living with both parents), separated family (parents divorced or living with a single parent), foster home, institution or living alone.

The substances consumed were recorded on the form under the categories: alcohol, narcotics, paracetamol and other. However, as medications other than paracetamol were taken in many cases, while processing the data we changed the categories to the following: medications (prescription and over-the-counter drugs), illicit drugs, alcohol, multiple drug intoxication (medications or illicit drugs combined with alcohol) and other/unknown (where the substance was difficult to categorise). In most cases, the doctor completing the form had recorded the substance ingested under ‘other’. When not specified in the form, this information was retrieved from the medical records.

The form did not survey how the adolescents had obtained access to the medications they took. We considered this to be of interest. Using the medical records, we therefore recorded whether the medications were for the adolescent’s own therapeutic use or whether they had been found in the home. If this was not recorded in the medical records, we searched the medication module to determine whether the medication had been prescribed to the adolescent. If this could not be established with certainty, it was recorded as unknown.

The form had three checkboxes for the intention behind the poisoning: ‘accidental’ in which the substance was ingested in error, for example, where young children found medicines in the bathroom; ‘self-harm’ where the poisoning was part of self-injurious behaviour, and ‘other’, which was intended to cover substance misuse-related poisonings and other hard-to-define situations. However, many of the patients with substance misuse-related poisonings were incorrectly placed in the ‘accidental’ group by the doctor completing the form, and the patients’ medical records were therefore reviewed retrospectively to ensure that patients had been placed in the correct groups. In the following, the categories ‘accidental’, ‘self-harm’, ‘substance misuse-related poisoning’ and ‘other’ are used.

The form recorded whether the adolescent had other signs of self-harm, such as scars as a result of cutting, and whether he or she had previously been subjected to physical violence or other forms of abuse. We also recorded whether assessment by a member of the Department of Child and Adolescent Psychiatry was requested during the admission and, if so, whether he or she considered the adolescent to be at increased risk of suicide. Information on which adolescents had been or were under treatment at the Department of Child and Adolescent Psychiatry was retrieved from the ‘previous illnesses’ section of the admission notes after the study period. Any follow-up after discharge was also recorded.

All data were collected and processed in SPSS version 23.

The project was endorsed by the management of the Division of Medicine and the Department of Child and Adolescent Psychiatry at Sørlandet Hospital Trust and was approved as a quality assurance project by the Norwegian Centre for Research Data. The project was approved without the requirement to obtain consent.

Results

Over a one-year period, we recorded 88 hospitalisations of 81 patients. The patients comprised two groups: 13 children (16 %) aged 0–8 years (mean 2.8 years, SD 2.8) and 68 adolescents (84 %) aged 13–18 years (mean 15.5 years, SD 1.5). Some adolescents were hospitalised several times over the course of the period. Only the first hospitalisation is included. The children’s group consisted solely of accidental poisonings with various substances, and no further data are presented for this group.

Key characteristics of the adolescent group are shown in Table 1. Girls were over-represented with 48 admissions (71 %). Medication positioning was the most frequent reason for admission (29 cases, 43 %), followed by alcohol poisoning (26 cases, 38 %). Alcohol was the most common substance among the boys (10 cases, 50 %).

Medications were involved in 32 (47 %) poisonings, exclusively in 29 cases and as part of multiple drug intoxication in 3 cases. Paracetamol was ingested in 16 (24 %) cases. Girls accounted for 24 out of 29 poisonings involving medications alone. Where medications were ingested, they were for the individual’s own therapeutic use in 14 (44 %) cases, while in 16 (50 %) cases the medications were found in the home.

Self-harm was reported as the intention behind the poisoning in 32 (47 %) cases. Substance misuse- related poisonings were responsible for 35 (52 %) hospitalisations.

Previous/other self-harm was recorded in 24 (35 %) cases, almost all of them girls. Previous poisonings were recorded in 13 adolescents (19 %).

In total, 37 (54 %) of the adolescents had completed or were currently receiving treatment at the Department of Child and Adolescent Psychiatry. This was the case for 28 (88 %) of the adolescents who reported self-harm as the intention behind their poisoning, and for eight (23 %) of those with substance misuse-related poisoning. Assistance in assessing suicide risk was requested in 37 (54 %) of the cases and always when the intention behind the poisoning was given as self-harm or a suicide attempt. A request was made in 28 (97 %) cases of medication poisoning and in three (12 %) cases of alcohol poisoning. In total, 15 (22 %) of the adolescents were considered to have an increased suicide risk, 11 (23 %) girls and four (20 %) boys.

Information on whether the adolescents had been subjected to physical violence or other forms of abuse was not recorded in 60 cases (88 %), either in the form or in the medical records.

A total of 37 adolescents (54 %) were followed up in the Department of Child and Adolescent Psychiatry following discharge from the Department of Paediatric and Adolescent Medicine, with 9 (12 %) patients acutely hospitalised. None of the discharge summaries mentioned follow-up by the patient’s general practitioner or other sectors of the primary healthcare service.

Discussion

Over the course of a year, 81 children and adolescents were hospitalised with poisoning. Poisonings in children were exclusively accidental in nature. In adolescents, ingestion of medications was more common in girls, whereas alcohol was the predominant cause of poisoning in boys. However, 33 % of the girls had also consumed alcohol. In common with several European studies, we found that paracetamol was the most commonly taken medication (4). Besides paracetamol, the adolescents mainly took medications that had been prescribed for their own therapeutic use. Eliminating access to means is important for preventing repeated self-harm, and suicide (2). Providing parents with guidance on the storage and administration of medications can be an important preventive measure.

Dysfunctional family relationships are a risk factor for self-harm behaviour (5). It is worth noting that 25 % of the adolescents lived outside the parental home, either alone or in an institution or foster home.

Almost half of all poisonings in the adolescent group were linked to self-harm, while around half of the hospitalisations involved substance misuse. Underlying mental illness is a known risk factor for self-harm behaviour in adolescents (5, 6). This is consistent with our dataset, in which 54 % of the adolescents had previously been or were currently under treatment at the Department of Child and Adolescent Psychiatry. In cases where self-harm was stated as the intention behind the poisoning, the Department of Child and Adolescent Psychiatry always provided assistance upon request. However, this was requested for only seven out of 37 (19 %) poisonings involving substance misuse. It may appear that consumption of alcohol and illicit drugs is not regarded as risk behaviour, but rather as normal behaviour in adolescents. This is despite the fact that, in 2009, Hodgins et al. showed that Swedish adolescents who sought help for substance misuse were at significantly increased risk of death, mental illness and further substance misuse (7). The group with substance misuse-related poisonings make up a large proportion of our dataset, and we believe that hospitalisations with acute poisoning provide a good opportunity to survey the drinking and substance use habits of adolescents, either while in hospital or at their general practice.

In our dataset, as many as 15 (22 %) of the adolescents were considered to have an elevated risk of suicide. Self-inflicted poisoning is not a common method of suicide among adolescents (2, 8, 9). However, in 2012, Tørmoen et al. studied a Norwegian adolescent population and found that one-third of adolescents who had self-harmed also reported previous suicide attempts (10). A study from Finland on suicide in children and adolescents found that 17 % of women who committed suicide had previously been hospitalised with self-inflicted poisoning (9). This underlines the importance of viewing self-inflicted poisoning as a risk factor for suicide.

Given that a history of abuse is a known trigger for risk behaviour in adolescence (11) and has been found to explain up to one-third of self-harm in women (12), we specifically wished to survey whether the adolescents had suffered physical or sexual abuse. Interestingly, in 88 % of cases, neither the medical records nor the form contained any information about this. It appears that this subject is not broached with adolescents, which is notable in light of a report by the Committee on Child Physical Abuse [Barnevoldsutvalget] of May 2017 (13), emphasising that all areas of the support services are doing too little to uncover abuse and neglect in children and adolescents. It could be argued that this topic falls within the remit of the Department of Child and Adolescent Psychiatry and thus does not belong in non-psychiatric medical records. However, studies have shown that too little information on this topic is gathered, even after long-term follow-up by the child and adolescent psychiatry services (14). That is not to say that paediatricians should explore these issues in all children and adolescents hospitalised with acute poisoning whatever the cause. But, in adolescents with serious risk behaviour, it may be worth considering whether this topic should be explored more routinely. If so, the most appropriate person or agency to ask these questions should be determined to ensure that already vulnerable young people do not fall between two stools.

Major review articles have highlighted a lack of knowledge regarding the type of follow-up that is most beneficial for this patient group (2). In our dataset, 47 % were discharged without any form of follow-up. In particular, 80 % of the group with substance misuse-related poisonings received no offer of follow-up. It is especially notable that the discharge summaries never mention general practitioners with respect to follow-up, despite the fact that general practitioners should be well-placed to play a key role in the further assessment and follow up of adolescents with risk behaviour.