Abstract
Both policy makers and health care regulators require measures to reveal instances of poor care. Both in-hospital mortality and mortality within 30 days of admission are commonly used measures to reveal the number of ‘excess deaths’. However, all models contain hidden assumptions which can invalidate the model. In this instance, all current approaches to hospital mortality make the assumption that there are no large spatiotemporal events capable of creating disparate effects in time and place in both all-cause mortality in general, and hospital mortality in specific. The study of international trends in deaths, does however, reveal one such large an unexplained effect against deaths, which also appears to simultaneously affect medical admissions, i.e. both the numerator and denominator in the models are subject to unexplained forces. The action of these forces upon all-cause mortality in England is demonstrated, along with evidence for spatiotemporal variation in age-standardised mortality within 30 days of admission for myocardial infarction across various locations and organisations in England. Cyclic and parallel movement in both all-cause mortality and SHMI are demonstrated in Milton Keynes, Ipswich and Northampton hospitals. An identical situation is demonstrated at the Wigan Royal Infirmary for HSMR. Finally the potential contribution of these events to the ‘excess’ deaths at the Mid Staffordshire hospital are investigated. It would appear that hospital mortality models do indeed contain a ‘fatal’ flaw which severely limits their ability to detect true instances of excess hospital mortality. Correspondence to: Rodney P Jones (PhD, ACMA, CGMA), Healthcare Analysis and Forecasting, Camberley, UK, Tel: +44(0)1276 21061; E-mail: hcaf_rod@yahoo.co.uk
Highlights
Following the Mid Staffordshire Hospital scandal in England, and the ensuing Francis report into standards of patient care [1], there has been considerable debate in the UK regarding the use of different measures of hospital mortality in order to detect hospitals which may have poor processes of care, leading to otherwise avoidable deaths.Two such measures are widely used, namely, hospital standardised mortality rate (HSMR) using a method developed by Dr Foster Intelligence [2], and the standardised hospital mortality indicator (SHMI) developed by the University of Sheffield at the request of the Department of Health [3,4]
When there is high seasonality in the underlying data, a running total effectively de-seasonalises the trend and has the advantage of minimising the higher Poisson variation seen in the smaller monthly figures
The point of initiation of the sudden step-up occurs at the foot of the ramp, while the slope of the ramp and the peak 12 months from the foot of the ramp both measure the magnitude of the step-change
Summary
Following the Mid Staffordshire Hospital scandal in England (where financial targets were given priority over patient care and safety), and the ensuing Francis report into standards of patient care [1], there has been considerable debate in the UK regarding the use of different measures of hospital mortality in order to detect hospitals which may have poor processes of care, leading to otherwise avoidable deaths Two such measures are widely used, namely, hospital standardised mortality rate (HSMR) using a method developed by Dr Foster Intelligence [2], and the standardised hospital mortality indicator (SHMI) developed by the University of Sheffield at the request of the Department of Health [3,4]. The Dutch study had the distinct advantage that the HSMR and SHMI equivalent calculations used an identical model
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Fractal Geometry and Nonlinear Analysis in Medicine and Biology
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
