SECTION 5 – SHMI Forecasting
5.3 More to Explore
SHMI Quarterly Profiling and Forecasting: https://www.aquanw.nhs.uk/resources/shmi-quarterly-profiling-and-forecasting/29382
Appendix A: Differences between HSMR, RAMI and SHMI Mortality Indicator (SHMI) **
Observed All spells culminating in death at the end of the patient pathway, defined by specific diagnosis codes for the primary diagnosis of the spell:
uses 56 diagnosis groups which contribute to approx.
80% of in hospital deaths in England*
Total number of observed in-hospital deaths
Number of observed in-hospital deaths plus deaths out of hospital within 30 days of discharge
Expected Expected number of deaths Expected number of deaths Calculated using a 10 year data set (as of 2012) to get the risk estimate
Expected number of deaths Calculated using a 36 month
Admission method
Source of admission
Diagnosis sub-group
Co-morbidities based on
Clinical grouping (HRG)
Primary and secondary diagnosis
Primary and secondary Procedures
Hospital type
Admission method
Further detailed methodology information is included in CHKS products, or specific enquiries to CHKS www.chks.co.uk
Sex
Age group
Admission method
Co-morbidity
Year of dataset
Diagnosis group
Details of the categories above can be referenced from the methodology specification document at http://www.ic.nhs.uk/services/
summary-hospital-level-mortality-indicator-shmi
Exclusions Excludes day cases and regular attendees
Excludes mental illness, obstetrics, babies born in or out of hospital, day cases, and patients admitted as emergencies with a zero length of stay discharged alive and spells coded as palliative care (Z515)
Specialist, community, mental health and
All England provider trusts via SUS
Data attributed to all Trusts within a ‘super-spell’ of activity that ends in death
UK database of Trust data and HES
Data attributed to Trust in which patient died
All England non-specialist acute trusts except mental health, community and independent sector hospitals.
Data attributed to Trust in which patient died or was discharged from
*HSMR does not exclude 20% of deaths, it looks for the diagnosis groups that account for the majority of deaths, and the figure of 80% is quite variable dependent on the case mix of the trust. HSMR could just as easily cover 100% of activity. It covers 80% of activity mostly for historical reasons and the fact that you get little extra value from the other 20%.
** NHS Digital publishes the SHMI indicator as observed, expected, denominator, value, upper control limits, lower control limits and banding. The term numerator is not used in the publication.
Appendix B - Metadata Page i of iii AQuA Quarterly Mortality Report Issue 17 Version 1.0 22nd August 2017
Appendix B: Metadata
Resource type Title Description Coverage Numerator Denominator Date Source
Mortality Charts 1 &
9
Discharge Method = 4 All discharges Latest published SHMI (12 month period)
HED
Mortality Charts 2 &
18
Discharge Method = 4 All discharges 1.4.2009 – 31.03.2016
HED
Mortality Charts 3 &
10
Discharge Method = 4 OR Died within 30 days
All discharges Latest published SHMI (12 month period)
HED Split as per Appendix B.3 of the SHMI Indicator
Specification i.e. Elective = Admission Method 11, 12, 13
Acute [NEL] = 21, 22, 23, 24, 28, 31, 32, 81, 82, 83, 84, 89, 98
Mortality Chart 4a SHMI - SHA 136 SHMI Trusts
Observed deaths Expected deaths Latest published SHMI (12 month period)
NHS Digital
Mortality Chart 4b &
11
SHMI – Funnel Plot
136 SHMI Trusts
Observed deaths Expected deaths Latest published SHMI (12 month period)
NHS Digital
Mortality Charts 5 &
19
NW SHMI 22 Trusts in North West
Observed deaths Expected deaths Latest published SHMI (12 month period)
HED
Mortality Chart 6 Observed and Expected
Discharge Method = 4 Discharge Method = 4 plus deaths from the HES-ONS linked mortality data file
Latest published
Discharge Method = 4 Discharge Method = 4 plus deaths from the HES-ONS linked mortality data file
October 2009 – September 2016
HED
Appendix B - Metadata Page ii of iii AQuA Quarterly Mortality Report Issue 17 Version 1.0 22nd August 2017
Resource type Title Description Coverage Numerator Denominator Date Source
Clinical Coding Chart 12 &
21 position of any episode or Specialty Code 315 in any episode
All discharges Latest published SHMI (12 month period)
NHS Digital
Clinical Coding Chart 13 &
22 position of any episode or Specialty Code 315 in any episode (where Discharge Method = 4)
Discharge Method = 4 plus deaths from the HES-ONS linked mortality data file
Latest published SHMI (12 month period)
NHS Digital
Clinical Coding Charts 14
& 23 primary diagnosis of any episode. Admission Method = 21 – 28, 31, 32, 81 – 89, 98.
Number of episodes Latest FY for which data has been published
HED
Clinical Coding Charts 15
& 24 primary diagnosis of the first episode. Admission Method = 21 – 28, 31, 32, 81 – 89, 98
Number of first episodes [i.e. Spells]
Latest FY for which data has been published
HED
Clinical Coding Charts 16
& 25 primary diagnosis of last episode. Admission Method = 21 – 28, 31, 32, 81 – 89, 98 (where Discharge Method = 4)
Number of last episodes [i.e. Spells]
Discharge Method = 4
Latest FY for which data has been published
HED
Clinical Coding Chart 17 &
26
Number of first episodes [i.e. Spells]
Latest published SHMI (12 month period)
HED
Appendix B - Metadata Page iii of iii AQuA Quarterly Mortality Report Issue 17 Version 1.0 22nd August 2017
Resource type Title Description Coverage Numerator Denominator Date Source
Sepsis Mortality Chart 27 Septicaemia CCS Group 2
Expected deaths for CCS Group 2
Latest published SHMI (12 month period)
HED
Sepsis Mortality Chart 28 Urinary Tract Infection CCS
Expected deaths for CCS Group 101
Latest published SHMI (12 month period)
HED
Sepsis Mortality Chart 29 Biliary Tract Disease CCS
Expected deaths for CCS Group 92
Latest published SHMI (12 month period)
HED
Sepsis Mortality Chart 30 Septicaemia CCS Group 2 VLAD
Individual Trust Cumulative Expected minus Observed mortality outcome based on SHMI.
March 2016 – February 2017
HED
Sepsis Mortality Chart 31 Urinary Tract Infection CCS Group 101 VLAD
Individual Trust Cumulative Expected minus Observed mortality outcome based on SHMI.
March 2016 – February 2017
HED
Sepsis Mortality Chart 32 Biliary Tract Disease CCS VLAD
Individual Trust Cumulative Expected minus Observed mortality outcome based on SHMI.
March 2016 – February 2017
HED
SHMI Forecasting
Chart 33 Likely impact of
‘losing’ a
Percentage relative impact predicted following ‘loss’ of quarter from SHMI
Latest published SHMI (12 month period)
HED
1 See Appendix D.1 of SHMI Methodology
2 This most closely reflects the episodes that are used in the SHMI calculation. Only a small proportion of second episodes are used [i.e. where the primary diagnosis of the first episode is an “R” code and the second episode has a primary diagnosis other than an “R” code].
Appendix C – Trust Codes Page i of i AQuA Quarterly Mortality Report Issue 17 Version 1.0 22nd August 2017
Appendix C: Trust Codes and Names
Trust Code Trust Name
REM Aintree University Hospital NHS Foundation Trust RXL Blackpool Teaching Hospitals NHS Foundation Trust RMC Royal Bolton NHS Foundation Trust
RW3 Central Manchester University Hospitals NHS Foundation Trust RJR Countess of Chester Hospital NHS Foundation Trust
RJN East Cheshire NHS Trust
RXR East Lancashire Hospitals NHS Trust
RXN Lancashire Teaching Hospitals NHS Foundation Trust RBT Mid Cheshire Hospitals NHS Foundation Trust
RNL North Cumbria University Hospitals NHS Trust RW6 Pennine Acute Hospitals NHS Trust
RQ6 Royal Liverpool and Broadgreen University Hospitals NHS Trust RM3 Salford Royal NHS Foundation Trust
RVY Southport and Ormskirk Hospital NHS Trust RBN St Helens and Knowsley Hospitals NHS Trust RWJ Stockport NHS Foundation Trust
RMP Tameside and Glossop Integrated Care NHS Foundation Trust RM2 University Hospital of South Manchester NHS Foundation Trust RTX University Hospitals of Morecambe Bay NHS Foundation Trust RWW Warrington and Halton Hospitals NHS Foundation Trust
RBL Wirral University Teaching Hospital NHS Foundation Trust RRF Wrightington, Wigan and Leigh NHS Foundation Trust