Readmission rates and HES

Readmission rates can help to monitor the success of the NHS in preventing, or reducing, unplanned readmissions to hospital.

There is no nationally agreed definition of a readmission. However, there are definitions for specific readmission measures in some clinical areas, such as those within the Compendium of Clinical and Health Indicators.

Using readmission rates and making comparisons

In HES, when considering readmission rates, we usually only take into consideration emergency readmissions (elective readmissions were always an intentional part of the treatment). However, we exclude emergency readmissions in cases where patients have cancer or mental health issues, as emergency readmissions in such instances are often considered to be a necessary part of the patient care.

Emergency readmission rates can be an effective measure of treatment. However, the entire patient pathway should be taken into consideration. Take a readmission to hospital with an infection following a hip replacement, for example. Was the infection as a direct consequence of the original procedure/interventions, or was it to do with the level of aftercare, or the patient's own actions?

To accurately compare organisations and make fair judgements on the basis of readmission rates it would be useful to determine what acceptable readmission rates would be in a range of situations. It would also be necessary to take into consideration a range of factors, including:

• The completeness and quality of coding of the records being examined
• Differences in coding practices between organisations
• The case-mix, socio-economic mix and demographics of the relevant local population
• Complications that occurred in hospital, rather than following discharge
• Patients discharging themselves against medical advice
• The levels and quality of primary care and community resources available
• The severity of the condition

The benefits of readmission rates

There can be a wide variation in readmission rates between similar NHS organisations. These variations can act as a trigger to look at practice within an organisation or geographical area. This could in turn help to prevent avoidable readmissions and lead to improved levels of care.

Can I get readmission figures from HES?

Readmission rates aren't routinely calculated in HES. This is mainly because it is not possible to say with absolute certainty that we have identified all potential readmissions, and that those that we have identified clinically relate to the original admission. There are several reasons for this, some of which are discussed below.

Patient identifier (PSEUDO_HESID)

In order to identify episodes in HES that relate to a single patient, a value (PSEUDO_HESID) is assigned to every episode. If the patient identifying information recorded on two episodes is thought to belong to the same patient then the same PSEUDO_HESID is recorded on both.

The PSEUDO_HESID relies on the completion and quality of data recorded in episodes. If records are incomplete then it may not be possible to identify records that belong to the same patient, making it impossible to identify all related readmissions.

Unfinished episodes

HES data is held as (financial) data years. Some episodes in HES will start in one data year, but not finish within the same year. Such episodes are known as unfinished episodes. Unfinished episodes don't generally contain all the information that a finished episode would, making it inadvisable to use them to calculate readmission rates.

To avoid using unfinished episodes, you should look at discharges up to and including a specific date. For example, for emergency readmissions within 28 days of discharge, you should look at discharges up to and including 4 March of each year. This will mean that any readmissions occurring within 28 days would fall within the end of the financial year being considered.

Duplicate records

The IC works closely with trusts to remove duplicate records from HES, but some may remain. Duplicate records may appear in HES. In order to calculate readmission rates, duplicate records must be identified and excluded from any analysis as they may look like same day readmissions.

Further information

The following emergency readmission rates, developed by NCHOD and based on HES data, are currently available from the NCHOD website:

• Emergency readmission to hospital within 28 days of discharge (data years 1998-99 to 2006-07) for patients aged:
• 0 - 15 years old
• 16 - 74 years old
• 75+ years old
• 16+ years old.
• Emergency readmissions to hospital within 28 days of discharge (data years 1998-99 to 2005-06) for patients who were originally admitted for:
• fractured proximal femur
• stroke
• hip replacment surgery
• hysterectomy.