Linked HES-ONS mortality data
Using HES data alone to analyse deaths
Patient activity data in HES can be used to identify if a patient died in hospital. Deaths recorded in HES may be analysed by the main diagnosis for which the patient was being treated, however this data alone cannot be used to determine the underlying cause of death. For example, a patient admitted for a hernia operation (with a primary diagnosis of hernia) may die from an unrelated heart attack. HES data also has the limitation that it does not capture deaths happening outside hospital. Linking HES with the ONS deaths data addresses these issues.
ONS Mortality Data
The Office for National Statistics (ONS) collects information on cause of death from the death certificate. For registered deaths, the underlying cause of death is derived from the sequence of conditions leading directly to the death and is recorded on the death certificate. The death certificate also records a list of other conditions or diseases that the patient had at the time of death, which may or may not have directly contributed to the death.
HES-ONS linked mortality data
Linking ONS and HES creates a richer dataset which captures deaths of people who have attended or have been admitted or treated in hospitals, irrespective of whether they died in hospital or not. Combining the data in HES and ONS datasets provides additional valuable information, for example ONS provides cause of death which is not in HES. This data can be used for a wide range of analyses, eg long-term follow-up on survival rates.
Summary of deaths following admission or primary procedure
Using the linked HES-ONS mortality data, deaths both in and outside hospital following hospital admission or primary procedure have been calculated. Detailed data on aggregated counts of deaths within 30, 60 and 90 days by primary procedure and primary diagnosis is available for HES data years 2007-08, 2008-09 and 2009-10:
These tables contain data for patients who had been treated in hospital in this period and subsequently died within 30, 60 or 90 days of the admission/procedure. The procedures table contains aggregated counts by primary procedures and the diagnosis table contains aggregated counts by primary diagnosis at admission.
It is to be noted that high percentages/counts of deaths after admissions/procedures do not necessarily imply an issue. The numbers should be considered in the context of the total number of patients with the condition and the nature of the condition as the health of some patients prior to entering hospital and prognosis will be expected to be poor simply due the condition they are suffering from.
The unadjusted data can be used as a starting point for more detailed clinical analysis. It highlights the dataset as a valuable and rich source of information for gaining a greater insight into outcomes of patients. However, unadjusted data such as these are not recommended for comparing organisations' death rates eg some organisations treat sicker patients who are more likely to die whether or not they enter hospital. For this purpose, standardised data enables a fairer comparison. A review into the methodology for provider level mortality rates is available in the SHMI web page.
Accessing the data
To request data extracts and for further information, please visit the Bespoke Data Extract Services web page.
Provisional nature of ONS mortality data
The Office for National Statistics provides mortality data on a monthly basis which is provisional in nature. ONS provisional data has not been subject to full quality assurance and may not contain all deaths which were registered or which occurred during the period. Deaths which occurred in a given year may be registered in a subsequent year. The database remains open to accept these registrations so figures are subject to change. ONS also supplies a final annual refresh of mortality data for the year which is based on calendar year of death registration.
Efforts are continually being made to optimise the linkage algorithm due to the complexity of the process, and to accommodate the varying quality of data submissions to HES. Below is a list of recent modifications/enhancements made to the linkage algorithm. This may have an impact on data extracts of linked mortality data taken at different points in time. For more information on these and other historic changes, please refer to Appendix C of the HES Mortality Guide, available to download from this page.
|Change Implementation Date|
|1. Dealing with subsequent hospital activity after date of death January 2010|
|2. Inclusion of ONS mortality data for deaths registered in 1998, June 2010|
1999 and 2000 where underlying cause of death and cause of death
mentions are recorded in ICD-9
|3. Selection of best quality match when ONS death record matches August 2010|
to multiple HESIDs
|4. Processing poor quality data submissions for deaths in hospital November 2010|
The responsible statistician for this publication is Simon Rhea, HES/SUS Development - Section Head (email@example.com, 0845 300 6016).
The copyright to the information we are disclosing is held by the Health and Social Care Information Centre. The HSCIC has suspended the application of re-use licence fees as a consequence of government policy ('Making Public Data Public'), so you may re-use this information free of charge. Please ensure that the following copyright statement is included within your documents:
'Copyright © 2012 Re-used with the permission of the Health and Social Care Information Centre. All rights reserved.' Please also use the citation logo available for download from the HSCIC's website.