Accidental poisonings in children cause major illness, worry, sometimes even death and place significant demands on NHS resources; yet most are preventable. Poisonings are more common in poorer households but research shows that providing cupboard locks and home safety advice improves safe storage of poisons by families. However, child poverty has increased over the last 8 years whilst child health services promoting home safety (health visitors, school nurses, safety equipment schemes) have been cut. The impact of this on childhood poisonings, especially among poorer families is unknown.
There are no recent figures describing which medicines or substances are involved in childhood poisonings. In particular, we do not know if medicines involved in poisonings were prescribed to household members or bought over-the-counter.
We will use information from GP records, hospital admissions and deaths to explore which medicines or substances have been involved in poisonings in 0-11 year olds over the last 20 years, whether these are prescribed medicines and, where possible, whether household members, including children themselves, were prescribed these medicines. We will describe how the frequency of poisonings has varied from the poorest to the richest households over time to assess the impact of austerity measures.
The results will raise awareness amongst prescribers of the importance of giving safety advice to families of young children. Results could also inform revisions to appropriate/safe prescribing limits. Those commissioning and providing home safety services will be provided with real-world evidence of the impact of service cuts to inform future decisions around commissioning priorities.
Background: Poisonings among 0-11 year olds are an important cause of ill health but are potentially preventable. Up-to-date evidence of the exact substances involved, socioeconomic gradients and time trends, including any effect of the austerity agenda, are lacking. No studies have examined the relationship between substances involved in childhood poisonings and medications prescribed to household members. Updated population level data are therefore required to guide commissioning decisions, safe prescribing and safety advice offered to families.
Aims: 1.To estimate population level medically-attended poisoning incidence among 0-11 year olds in England from 1998-2018 by poisoning substance, sex, age, deprivation level, geographical area and changes over time. 2.To examine the relationship between medicinal poisonings in these children and medicines prescribed to household members by primary care.
Design and Setting: An open cohort study of all 0-11 year olds in the Clinical Practice Research Datalink (CPRD) with available linked Hospital Episode Statistics (HES) and Office for National Statistics (ONS) mortality data, plus nested case-control study.
Method: The cohort study will establish poisoning episodes by substance using ICD-10 (HES/ONS data) and/or Read (CPRD data) codes. Incidence rates and adjusted incidence rate ratios will be calculated using Poisson or negative binomial regression for poisonings in relation to substance, age, sex, deprivation, year and geographical area. A series of nested case-control studies will be constructed using cases (poisonings involving the 5 most common medicinal substances identified in the cohort study) and controls (no poisoning with the specified substance). Conditional logistic regression will be used to examine odds ratios for poisonings involving these substances and that specific medicine being prescribed to a household member.
Impact: Results will be used to inform prescribing practices and safety advice offered to families within primary care. They will also assist healthcare commissioners and individual practitioners make decisions about poisoning prevention interventions.
The outcome will be incident poisoning events during the study period, recorded in any one of: primary care data, HES admission data or ONS mortality data by the use of at least one relevant Read code or ICD-10 code (see provisional code list in appendix A). Such codes will be identified from either the Medcode variable within the Clinical file (CPRD data) or from HES admission (ICD variable within Diagnosis codes) or ONS cause of death codes. See section N for more details.
Other key variables will include: sex, age, socio-economic deprivation (using Index of Multiple Deprivation 2010 quintiles), geographical area (Region variable within the Practice file), year of event and relevant prescribed medication to any household member (this will require use of the Family Number variable in the Patient file, as well as prescription data within the Therapy file). Prescription data will only be extracted following completion of the cohort study. These data will then be used to define exposures in the nested case-control elements.
Edward Tyrrell - Chief Investigator - University of Nottingham
Edward Tyrrell - Corresponding Applicant - University of Nottingham
Denise Kendrick - Collaborator - University of Nottingham
Elizabeth Orton - Collaborator - University of Nottingham
Laila Tata - Collaborator - University of Nottingham
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation