Time series analyses

Hospital Episode Statistics (HES) inpatient data is available from 1989-90 onwards. During this time there have been ongoing improvements in data quality and coverage, which particularly affect earlier data years. As well as this, there have been a number of changes to the classifications used within HES records.

Changes have also been made to the organisation of the NHS. For example, two or more trusts may have merged to form one new trust, or a trust may have split into two parts, which may then each have gone on to merge with other trusts.

These changes play a big part in the accuracy and validity of time series analyses. It is therefore important to fully explore the possible issues before making conclusions about changes in year-on-year data.

There are articles on HESonline that offer information about what you should consider when planning time series analyses. These include:

• Changes in clinical coding
• Organisational changes over time.

If you have suggestions of further topics that would be of use, please contact us.