Guide to working with data and statistics

graphsGetting started with child accident data

You’ll need data on childhood accidents for different reasons – to make the case for including accident prevention in local priorities or investing in prevention programmes, to design effective plans to tackle serious injuries or to evaluate programmes. But getting to grips with data can feel like a challenging task, so we’ve developed this short guide to help you get started. Read on to find out more about the data you should be looking for and what it can tell you.

How to use this guide

National data can provide a useful starting point for your work on child accident prevention. That’s because your local pattern of deaths and serious injuries is unlikely to differ significantly from the national picture. See our data and statistics sources tool for useful links.

With accidental injury a leading cause of death, acquired disability and hospital admission for children, it’s far better to make a start on tackling known major causes of serious preventable injury, and improve your local data as you go along, than never start accident prevention work because your local data isn’t yet good enough.

As you do more accident prevention work, you’ll want access to local data to help you undertake more detailed planning, commissioning and programme delivery for your child accident prevention work. Read our guide for helpful pointers to local data sources and useful local contacts.

1. Focus on serious, preventable injuries

Before you get started, it’s worth remembering that many minor accidents that are very common in childhood are an inevitable part of growing up and having a healthy, active life. While it’s useful to have a broad understanding of all types of accidental injury from minor to very severe, the main aim of your data analysis should be to learn more about serious and preventable accidents which cause severe injury, disability and death.

2. Examine hospital admission figures

Hospital admissions data is a good place to begin, as it represents the more serious end of the accident spectrum. Hospital admissions have an important role in the new Public Health Outcomes Framework which includes an indicator for hospital admissions caused by unintentional and deliberate injury in under 18s.

To identify your accident prevention priorities you’ll need to dig deeper into hospital admissions data for unintentional injuries, ideally with the help of a public health data analyst. Examining data for a period of several years will give you a clear idea of trends and it will also give you an idea of priorities. One of the first things you can do is analyse admissions by age group and sex, as certain types of accident are closely linked to different ages and stages of child development. CAPT’s forthcoming guide, Accidents and child development, provides a helpful overview of the link between child development and childhood accidents.

Within every injury category there’ll be varying degrees of severity and preventability. For example, within falls there’ll be a large number of more minor injuries caused by things such as slips and trips or falls from playground equipment. But there’ll also be a smaller number of very serious – and highly preventable – injuries, for example from young children falling down stairs or out of windows.

3. Widen your analysis to other types of data

Although hospital admissions data can provide good insight into the more serious accidents happening in your area, road casualty and fire service data can all be useful sources of further information and intelligence about injuries to children. Contact your local road safety team to find out more about the circumstances in which children are killed or seriously injured in road accidents, and your local fire and rescue service for information about house fires involving children. 

4. Gather information on fatalities

Information from your local Child Death Overview Panel (CDOP) may be able to tell you about the circumstances surrounding childhood deaths from accidents. When you’re looking at data on child deaths from accidents within one area, the number of deaths from causes other than road traffic accidents are likely to be quite low. This can make it difficult to know whether particular deaths were ‘freak’ accidents or whether they are in fact typical for certain ages and stages of development.

ONS child mortality statistics can provide a helpful overview of causes of childhood deaths at different ages. CAPT’s forthcoming guide, Accidents and child development, also sets out the links between child development and childhood accidents. Sharing information with other CDOPs can be a good way to work around this issue.

You can also use the practitioner support section of the CAPT website to learn about the biggest safety risks for children of different ages.

5. Be aware that low numbers can conceal high costs

The most common types of accident aren’t necessarily the ones that incur the highest costs. For example, the number of admissions for head injuries is likely to be quite low, but these injuries are expensive for the NHS and highly traumatic for children and their families.

Average length of hospital stay is one useful indicator of the cost of an admission – longer stays are likely to reflect more complex injuries and higher costs. You can also look at our articles on the costs of childhood accidents. Understanding more about accidents with particularly high costs attached to them can help you to build a strong case for investment in prevention.

A data analyst may be able to extract admissions data involving admissions of more than three days which can be an approximate measure of severity and cost.

6. Use rates rather than numbers for benchmarking

When comparing your data with that of other areas or for England as a whole, look at rates of hospital admission (for example, hospital admissions per 100,000 population of children and young people) rather than at straightforward numbers. Working with rates gets around the fact that areas with high numbers of children and young people are likely to have a higher number of injuries by default.

7. Explore geographical trends

Exploring demographic information and geographical trends within your local area will help you to identify which populations have the greatest need, what interventions may be needed and which partners you may be able to work with.

You can work with a public health data analyst to rank your area’s wards by rate of hospital admission. As there are close links between accidental childhood injury and deprivation, it’s also useful to map accidental injuries against the relative deprivation level for each ward. IDACI child deprivation scores show the percentage of children aged 0-15 living in income deprived families, and data is available down to Middle Layer Super Output Areas within local authorities.

You can use your local knowledge of social, economic, cultural and environmental factors to determine why childhood accidents are more common in certain areas. Read about how NHS Wakefield used Google Maps to find out why accidents involving trampolines were on the rise.

8. Stay up-to-date with emerging issues

Even if you’re very familiar with child accident data for your area, it’s a good idea to keep informed about emerging causes of unintentional injury in children. If you are aware of newer risks such as strangulation from blind cords or suffocation from nappy sacks, you will be able to keep a watching eye out for these kinds of accidents.

For the latest news on emerging safety risks, see

9. Understand and address limitations with your data

It’s likely that you’ll come across issues which compromise the availability and reliability of data on accidental injuries among children and young people in your area. Common problems include inconsistencies in the way different organisations collect and code data, and challenges with sharing information in a timely manner. You may also find that your data on hospital admissions and A&E attendances includes a large number of accidental injuries coded under unspecified ‘other’ causes.

Reducing the proportion of injuries recorded with unknown causes will greatly improve the quality of your data and its usefulness for planning prevention work. Similarly, establishing a standardised approach to data collection, coding and sharing will give you with much more robust information to work with in the future.

To address issues relating to data, you’ll need to engage those involved in data collection and coding across different organisations. Demonstrating the ways in which you use the data they provide and how it has a real impact on the prevention of accidents can be a powerful incentive to produce detailed and consistent coding. You should aim to provide regular feedback to data stakeholders, to share insights that you’ve gained from the data and the decisions you’ve taken as a result. Read about how NHS Wakefield improved the quality of data coding for accidental injuries.

More information

Make sure you use the following contacts to help you in your work:

  • Public health analyst based in your public health team – local hospital admissions, including breakdowns by age and cause and possibly local A&E and mortality statistics.
  • Fire and rescue service – house fires involving children in your area.
  • Road safety team – children killed and seriously injured in road accidents
  • Child death overview panel – local child deaths with modifiable factors.

Use our data and statistics sources tool to search for national data on child accidents:

Updated March 2012

Updated July 2012