• regional reports

A recent development in Northern Ireland shows much promise for injury control in the province. A new electronic injury surveillance system has been introduced at the Royal Belfast Hospital for Sick Children, our only paediatric hospital. Information on the cause of injury is being collected on all children presenting to the accident and emergency department. The first six months' data has been examined for completeness and accuracy.

Of the 11 683 cases, 4567 (39%) were injuries and of these 3883 (85%) had complete coded injury surveillance data—external cause, location, activity, intent. The free text narrative was used to assess the accuracy of the coding of external cause of injury and 71% were deemed to be correctly coded. The areas of greatest confusion were foreign bodies being coded as poisonings, falls from a height being classed as low level falls, and swimming pool immersions which were in fact other injuries at a swimming pool. Over 200 cases that were coded as other or unspecified were capable of being allocated to other codes on the basis of the free text. Accuracy was greatest for burns, scalds, and poisoning by medication.

To determine the overall value of the information we developed a scoring system against which a 10% sample (457) was assessed. The system allocated points for the following pieces of information: the coded data fields of external cause, location, activity, intent (4 points); the free text—nature of complaint (1 point), specific details on location, activity and mechanism of injury (3), measurement in terms of height, quantity, volume etc (1), safety precautions (1). The highest score possible was 10 but our highest score was 8 (in 12 cases), followed by 66 cases at 7 points, and 211 cases at 6 points. If all four coded fields are correctly completed it is easy to get at least 4 points so this showed us that improvements need to be made to the quality of information in the free text field.

Some initial discoveries:

• Two thirds of the burns by touching a hot object were to boys and one third were by touching an iron.

• Scalds to children were evenly divided between boys and girls and half pulled something hot on to themselves.

• After paracetamol solution, white spirit and essential oils were the agents most commonly associated with poisonings.

Although this is not earth shattering for countries that have well established injury surveillance, it is the first time that we can produce detailed and specific information for those working in injury prevention in Northern Ireland.

Our short evaluation showed the need for “help” documents and a more structured training programme for the nurses responsible for completing the injury surveillance data. The technical problems with the system link to the length of the free text field, the need for mandatory fields, and some additions to the external cause of injury codes. These are in the process of being addressed with a training manual and comprehensive index being written. Feedback of some of the initial findings to the staff is being planned in order to highlight the importance of the data for injury control. When we come to undertake a second evaluation one measure against which we will review the quality of the data is the scoring system we devised. This work was undertaken by a medical student on a short studentship. Funding is being sought to enable on-going analysis and dissemination of the findings.