The Dr Foster/Imperial College London methodology for the Hospital Standardised Mortality Ratio (HSMR) is published here in full. The statistical methods and factors used in the model have been included in all publications of HSMRs since 2001.
The methodology, including underlying log-odds, has been published on the NHS Choices website since 2009 when the Department of Health decided to publish HSMRs - an indication that HSMRs had become accepted as a valuable indicator of the quality of care offered by hospitals throughout England.
HSMRs should not be used in isolation. They provide an indicator of where a problem might exist and should be used as a trigger to investigation. Issues such as coding mistakes and under-reporting of co-morbidities (underlying health conditions) by hospitals can lead to high HSMRs, but where these have been ruled out it is important to note there may be a problem with the quality of care delivered by that organisation.
You can download the full methodology here
The Regression Coefficients (log-odds) can be found here
HSMR toolkit - a guide to HSMRs - Download as a pdf
Dr Foster Limited, trading as Dr Foster Intelligence. Registered Company Number 3812015