Trusts identified as performing poorly in the Dr Foster Hospital Guide 2012 have described the indicators used as “flawed” and questioned the organisation’s ability to assess new integrated care trusts.
Published today, the hospital guide shows that mortality rates across the NHS continue to improve, but 12 trusts are highlighted as causes for concern.
Dr Foster identifies that trusts are under increasing financial pressure and says that high levels of occupancy at the majority of hospitals make it harder to provide a safe and effective service.
Many hospitals are 95% full in winter time and nearly 29% of hospital beds are occupied by patients whose hospital stay might have been avoided. The report says a lack of integration with social and community care is often the problem.
Dr Foster takes into account four ways of measuring mortality; the hospital standardised mortality ratio, which measures deaths while in hospital; the summary hospital-level mortality indicator, which takes into account deaths both in and out of hospital; deaths after surgery; and deaths from conditions that patients would normally survive.
Buckinghamshire Healthcare NHS Trust is amongst the 12 worst performers with higher than expected rates in two out of four indicators.
It was also identified - along with Oxford University Hospitals NHS Trust, Royal Cornwall Hospitals NHS Trust, Royal Bournemouth and Christchurch Hospitals NHS Foundation Trust and Mid Yorkshire Hospitals NHS Trust - as having worryingly high mortality rates in the weekends.
The guide says the HSMR drops when trusts have more senior doctors onsite and senior staff numbers per bed have increased at a number of trusts over the past year.
Buckinghamshire said it was “disappointed” to be described as having a higher than expected mortality rate and believes it is not the case.
Medical Director Graz Luzzi said: “This is a hospital guide, but our care is also delivered at home, in the community and in a hospice.
“To be effective it is important to compare like with like. Hospice beds are obviously not the same, and the mortality rates between the two will be hugely different. Most hospitals in England don’t run hospices, but new integrated acute and community health care trusts do.”
Of the trust’s 700 beds, 100 are classed as community and palliative care beds.
“Since integration with the community hospitals and hospice, the trust questions whether the HSMR provides an accurate way of measuring its mortality rate,” Luzzi added.
“Dr Foster has acknowledged our concerns about applying the HSMR to integrated organisations by allowing the trust to include a comment explaining our integrated remit.
"But I question whether the information, without adjustment for the factors outlined above, should have been included at all. The Trust has done so much to improve patient safety and the overall patient experience, seeing the mortality rate drop year on year over the last three years.”
University Hospitals Birmingham NHS Foundation Trust is also identified as a poor performer.
Medical director Dr Dave Rosser said the HSMR is widely considered to be unsuitable for use as a comparative indicator between hospitals.
The new SHMI was introduced in October 2011 with a statement saying that “it should not be used on its own to judge or rank hospitals” and the mortality indicator relating to conditions of low clinical risk is “deeply flawed,” he added.
“To illustrate, one of the patients identified by Dr Foster to be in this category was a patient admitted into one of UHB’s specialist services with a condition called toxic epidermal necrolysis, which is known to have a mortality in excess of 50%.”
Dr Rosser said that if the current methodology is applied to the data two years ago, the trust would not have been classed as an outlier and if the methodology used two years ago was applied to the current data, the trust would not be an outlier this year.
All five hospitals identified as having lower than expected outcomes on three out of four measures are in London. These are: Barnet and Chase Farm Hospitals NHS Trust; Chelsea and Westminster Hospital NHS Foundation Trust; Guy’s and St Thomas’ NHS Foundation Trust; Imperial College Healthcare NHS Trust; and the Royal Free London NHS Foundation Trust.
The 12 trusts with higher than expected rates in two out of four indicators.
• Aintree University Hospital NHS Foundation Trust
• Blackpool Teaching Hospitals NHS Foundation Trust
• Buckinghamshire Healthcare NHS Trust
• George Eliot Hospital NHS Trust
• Hull and East Yorkshire Hospitals NHS Trust
• Medway NHS Foundation Trust
• North Cumbria University Hospitals NHS Trust
• Northern Lincolnshire and Goole Hospitals NHS Foundation Trust
• United Lincolnshire Hospitals NHS Trust
• University Hospitals Birmingham NHS Foundation Trust
• Walsall Healthcare NHS Trust
• Western Sussex Hospitals NHS Trust
© 2012 EHealth Media.
Morbidity data being analysed for mortality!Alison Bird 106 weeks ago
Dr Foster data is based on morbidity coding data, undertaken in hospitals to record what patients were treated/investigated for, not what they died from!
Mortality coding is undertaken by ONS (Office for National Statistics)!
Definitely Apples and Pears!!
We should sort out how we categorise the fruitPhilC273 106 weeks ago
It is fairly certain that there are some apples and pears being compared. It is also fairly certain that some Trusts are viewed more unfavourably than they should through these indicators.
However, we do need to have measures of performance and we do need to have a method that spots the developing Mid-Staffs type situation early.
Just disparaging or ignoring the indicators can%u219t be the answer. A few years ago A&E data was in a very poor state. Requiring it in SUS has led to some improvements. Surely what we need in the informatics community is a good discussion on how to make these indicators more representative and how we judge good and bad performance. Publishing them should at least drive a debate.
The biggest threat we have to health analytics is just dismissing something like this because it isn%u219t perfect.
re: Apples with Pearsmrtablet 106 weeks ago
Also I upcoded an unknown proportion of my apples to oranges, while you mistook green tomatoes for apples.
Plus I failed to ascertain the same proportion of pears that you did.
Finally your fruit farm is located in the Arctic circle while mine is in Nigeria.