Generating an annual estimate of the proportion of deaths thought to be due to problems in care for the NHS


To explore whether retrospective case record review (RCRR) can be used to create a national estimate of deaths thought to be due to problems in care. This will be based on the sampling of a proportion of deaths to build an annual indicator for the NHS.


Current measures of patient safety are limited as they do not provide a robust overarching measurement for the safety of healthcare services. Unlike hospital mortality rates (HSMR and SHMI) which are criticised for being indirect measures of the quality of care, RCRR is a more direct assessment of deaths thought to be due to problems in care.

Our proposed annual estimate will use external clinical reviewers to carry out RCRR in a representative sample of deaths. This will be distinct from the RCRR process carried out by providers themselves and will generate data only at a national level.

Our aim is to ensure that the data generated are robust and meaningful, to avoid fuelling the on-going debate about what statistics can and can’t tell us about mortality. It will also be vital to ensure the indicator is reproducible and able to detect change over time.

Latest news

  • Healthcare Quality Improvement Partnership (HQIP) will manage procurement of this work, which will likely commence by Q1 2016/17.