NHS Hospitals with Doctors on the board of directors have better outcomes

NHS Hospitals with Doctors on the board of directors have better outcomes

Amanda Goodall has shown that hospitals with doctors as chief executives have 25% better clinical outcomes (statistically significant) in US hospitals. This is seen in other areas where 'expert leaders' have better outcomes. Kirkpatrick and Veronesi have looked at the board composition and found that in general boards of directors having more clinicians have lower HSMRs. Very specifically they found that boards of directors with more doctors in them clearly have a lower HSMR, higher CQC rating (actually their predecessor the healthcare commission's ratings) and higher patient satisfaction.

This prompted me to look at the Keogh 14 NHS Trusts that have been identified by the Department of Health and others as having problems mainly as a result of higher SHMI. The findings are of course compatible with the published research. 

All the 14 Keogh Trusts put together have only 3 doctors in their boards apart from their medical directors. Since medical directors on boards are statutory they are a common factor in all boards anyway.  So if medical directors are excluded from the calculations then the 

Keogh 14 trusts have 3 doctors (excluding Medical Directors) out of  184 board directors 1.63% of the board are doctors excluding MDs

Compare that to the 14 hospital trusts  with the lowest SHMIs (as of 2011) who have 15 doctors (excluding medical directors) out of 195 board directors 7.69% of the board are doctors excluding MDs

If we looked at HSMR (as of 2011) and compared high 14 and low 14 HSMR hospital trusts (the 14 is simply a number to match Keogh - there is no real logic or magic on the use of the number 14 here) a similar picture emerges:

NHS Hospitals with highest 14 HSMRs - 5 doctors (excluding Medical Directors)amongst 189 board directors  2.64% of the board are doctors excluding MDs

NHS Hospitals with lowest 14 HSMRs - 16 doctors (excluding Medical Directors) amongst 191 board directors. 8.37% of the board are doctors excluding the MDs

Medical directors as already mentioned are a mandatory appointment. Any other doctors appointed to the board is a sign of the value and recognition  by the trust and the appointment committees either on the basis of what the trust thinks that doctors bring to the table or as a recognition of research findings that expert led organisations do better. It is very clear that more doctors on the board of directors is associated with better outcomes.

It may not be politically correct to say so but it simply makes sense to appoint more doctors to the board of directors. 

What is important is that increasing the number of doctors in the board in the high SHMI or high HSMR hospitals must not be done as a matter of ticking the box - that will be very disrespectful to the concept. It should come out of a recognition of the value that the medical profession brings to the system as borne out by the findings above. 

It is also possible that when we cynically manipulate the undeserving into boards or when all boards have a higher number of doctors there will still be a difference between low and high performing hospitals. That is a different and new issue to be dealt with as it emerges. However in the meanwhile if as a result of increasing doctors in the boards we get better results we should respectfully and gratefully accept that.

It makes sense to have doctors on boards - let us do it.

Additional information added on 25 May 2015 - The difference in doctors in the board of directors between the Keogh 14 trusts and low SHMI trusts mentioned above is statistically significant with a p value 0.0081 (significant at p < 0.05 ) using a chi-square test

©M HEMADRI

Follow me on Twitter @HemadriTweets
 

Notes: 

The trust boards listed above were identified from their respective websites accessed on 21 September 2013

Amanda Goodall's article: http://www.amandagoodall.com/CESifo_DICE-Spring_2013_Goodall_FINAL.PDF

Kirkpatrick and Veronesi's article on Clinicians in Boards: http://www.cihm.leeds.ac.uk/new/wp-content/uploads/2012/05/Clinicians-and-Boards.pdf

2011 SHMI and HSMR lists: http://www.bbc.co.uk/news/health-15897345?sort2

2012 low 14 SHMI trusts list is quite similar to 2011. I used 2011 since it was easier to access.

There were 7 BME board directors in Keogh 14 trusts and 7 BME board directors in 14 lowest SHMI trusts  

Keogh 14 Trusts

Basildon and Thurrock

Blackpool

Buckinghamshire

Burton

Colchester

Dudley Group

East Lancashire

George Eliot

Medway

North Cumbria

NLG NHS

Sherwood Forest

Tameside

United Lincoln

14 Low SHMI Trusts (2011)

West Middlesex

North West London Hosp NHS FT

Sheffield

Ealing

James Paget

Chelsea & Westminster

Newham/Barts

Whittington

Cambridge

St Georges

Kingston

UCLH

Royal Free

Imperial

14 Highest HSMR trusts (2011)

Morecambe Bay

Isle of Wight

Hull & East Yorks

North Cumbria

George Eliot

Yeovil

Dartford & Gravesham

University Hosp of North Staffordshire

Northampton General Hospital

Dudley Group

NLG NHS

Shrewsbury and Telford

Medway

Sherwood Forest

14 lowest HSMR trusts (2011)

Chelsea & Westminster

Airedale

Kings College

Salford

Newham/Barts

Guys and St Thomas

Frimley Park

Whittington

Cambridge

St Georges

Kingston

UCLH

Royal Free

Imperial

High Mortality Hospitals Cannot Afford To Pay

In a previous post I showed that most high mortality trusts did not pay bank holiday extra rates/wages to staff for the Queen’s diamond jubilee bank holiday, while most low mortality trusts paid higher wages. 

 

A friend of mine who is an academic wrote back to me and said he could not resist doing a chi square on the numbers and found the p=0.01. I am no don to argue or explain stats but irrespective of statistical significance it is important to probe if there might be a deeper meaning or relevance. 

 

It is important to understand why the high mortality trusts did not pay higher holiday rates. Are they ‘mean spirited’ as the Unite Union portrayed them?

In my mind the underlying reasons are very simple and here it is:

QUALITY IS INVERSELY PROPORTIONAL TO COST 

 

And a high HSMR is broadly speaking poor quality care.

Financial reasons?

It might be something as simple as they had no money left to pay. Now that would be a perfectly reasonable assumption to make. Trusts get paid for activity, things like hernia repairs, aneurysm repairs, cardiac stenting, the kinds of things that you do to make patients get better. As far as I know the NHS tariff system through which the trusts get paid does not include things like deaths or complications. 

But in-hospital deaths are very costly; in-hospital complications are very costly. There is no mechanism for payment for that. So a hospital/trust which has high deaths and complications will obviously not have money to do anything else.

Well, it therefore might turn out that their inability to pay higher wages had no a financial reason at all; it may well be a by product of poor quality. High cost, deficits, losses are all a function of poor quality. 

 

If you pushed them they will come out with something like ‘in this financial climate we would like to channel all our sparse finances directly into patient care’ and you know what, they sure do; their patient care must cost excessive amounts of money due to higher rates of standardised mortality and higher complications.

Cultural reasons?

Perhaps they were unwilling to pay higher rates; management might not have felt the need to 'reward' staff who are unable to produce high quality measured in terms of mortality. Another reason might have been that the money might be better spent in a high mortality hospital in trying to reduce the mortality rather than paying more to staff when the law does not demand that you do so. These are a part of the mental make up and cultural reasons of management. They are right, well, partly right. It is also just possible that well rewarded staff might be motivated to engage in improvement. Works both ways but always difficult to decide which one is right for the given circumstances.

Finally, here is some speculation

But, why did some high mortality hospitals pay staff bank holiday wages? Surely the above arguments apply to them as well. Why did some low mortality hospitals not pay higher bank holiday wages? 

 

Now I am moving into speculation something which I try not to do too often. My gut feeling is that the high mortality hospitals who paid a higher wage are probably going to find reduced mortality soon or at the best they may continue to stay where they without slipping and getting any worse and the low mortality hospitals who did not pay may find their mortality going up or at the best they may stay where they are without getting any better

 

My speculation is an extension of my theory about money in hospitals, the trusts who are doing clinically well might have the spare cash to spend it on staff. If that was indeed the case, the staff deserve it.

©M HEMADRI 

Follow me on twitter @HemadriTweets

Any links between bank holiday pay and mortality?

The Queen's diamond jubilee celebrations are going on right now. The government declared a 'bank holiday' on Tuesday 5 June 2012 (http://www.direct.gov.uk/en/Nl1/Newsroom/DG_183806). Some of us will still be working over the celebration period to keep essential and emergency services going. This includes NHS staff. Individual NHS organisations can decide on whether they will treat this extra holiday as 'bank holiday' or 'public holiday'; they do not have to follow the government declaration of a 'bank holiday'. 'Bank holiday' attracts a higher rate of pay for those who work on that day along with some other terms advantageous to the employees. 'Public holiday' does not attract a higher rate of pay.

Unite Union surveyed their members and found that 113 NHS organisations were treating this as a 'public holiday' and hence no extra pay for staff. (http://www.unitetheunion.org/news__events/latest_news/_named-and-shamed__-_nhs_emplo.aspx) They have called this 'mean-spirited' and called their publication 'named and shamed'.

My interest includes hospital mortality and I wanted to find out what the high mortality hospitals and low mortality hospitals did in terms of the bank/public holiday pay arrangements. I took the list of 21 low mortality hospitals and 19 high mortality hospital from drFoster's hospital guide and then cross checked with Unite's named and shamed list. The findings are interesting to put it mildly.

My findings are:

7 OUT OF 21 LOW MORTALITY HOSPITALS ARE IN UNITE'S LIST (suggesting that they are not pay bank holiday rates)

14 OUT OF 21 LOW MORTALITY HOSPITALS ARE NOT IN UNITE'S LIST (suggesting that they are paying bank holiday rates)

14 OUT OF 19 HIGH MORTALITY HOSPITALS ARE IN UNITE'S LIST (suggesting that they are not paying bank holiday rates)

5 OUT OF 19 HIGH MORTALITY HOSPITALS NOT IN UNITE'S LIST (suggesting that they are paying bank holiday rates)

Low Mortality Hospitals (as per dr Foster)	Unite's named-and-shamed list
Barnet and Chase Farm Hospitals NHS Trust	In Unites' list
Barts and the London NHS Trust	In Unites' list
Cambridge University Hospitals NHS	In Unites' list
Chelsea and Westminster Hospital NHS	Not in Unite list
Epsom and St Helier University Hospitals	Not in Unite list
Frimley Park Hospital NHS Foundation Trust	Not in Unite list
Guy’s and St Thomas’ NHS Foundation Trust	Not in Unite list
Imperial College Healthcare NHS Trust†	Not in Unite list
King’s College Hospital NHS	In Unites' list
Kingston Hospital NHS Trust†	Not in Unite list
Newham University Hospital NHS Trust†	Not in Unite list
North West London Hospitals NHS Trust	Not in Unite list
Royal Devon and Exeter	Not in Unite list
Royal Free Hampstead NHS Trust	Not in Unite list
Sheffield Teaching Hospitals NHS	In Unites' list
South London Healthcare NHS Trust†	Not in Unite list
St George’s Healthcare NHS Trust	Not in Unite list
The Whittington Hospital NHS Trust†	In Unites' list
University College London Hospitals	Not in Unite list
University Hospitals Bristol	In Unites' list
West Suffolk Hospitals NHS Trust	Not in Unite list
High mortality hospitals (as per dr Foster)	Unite's named-and-shamed list
Blackpool Teaching Hospitals NHS	Not in Unite list
Buckinghamshire Healthcare NHS Trust	Not in Unite list
Burton Hospitals NHS Foundation Trust	Not in Unite list
Dartford and Gravesham NHS Trust	In Unites' list
George Eliot Hospital NHS Trust	Not in Unite list
Hull and East Yorkshire Hospitals NHS Trust	In Unites' list
Isle of Wight NHS Primary Care Trust	Not in Unite list
Medway NHS Foundation Trust	In Unites' list
Mid Cheshire Hospitals	In Unites' list
North Cumbria University Hospitals	In Unites' list
Northampton General Hospital NHS Trust	In Unites' list
Northern Lincolnshire and Goole Hospitals	In Unites' list
Shrewsbury and Telford Hospital	In Unites' list
The Dudley Group of Hospitals	In Unites' list
The Royal Wolverhampton	In Unites' list
United Lincolnshire Hospitals	In Unites' list
University Hospitals of Morecambe Bay	In Unites' list
Worcestershire Acute Hospitals	In Unites' list
York Teaching Hospital	In Unites' list

Caution: This is a write up based on information that is publicly available so far. This analysis may not be accurate. We can find out the correct situation only when either NHS employers or individual trusts tell us whether they have paid extra treating it as a bank holiday or not paid extra treating it as a public holiday. So further enquiry and analysis would be needed to validate this. What is presented here is a mere observation and does not suggest cause and effect.

On the basis of current information (this may change when we have accurate information) it seems like there may be attitudes and cultures of organisations, management and staff, playing a bigger part in mortality and morbidity than we previously have assumed.

Please let me know if there are any factual errors in the above and I am only too willing to correct them.

©M HEMADRI 

Follow me on twitter @HemadriTweets

Mortality 1998 & Now - What can we learn?

Mortality 1998 and now: What can we learn?

I found these 4 pages full of 30 day hospital mortality rates from 1998 – 1999 which you find below the write up (you may have to print it, if you want to look at it as you go along). Those were the days before HSMR (though retrospective calculations should be available). Some powerful details are jumping out.

DATA STATUS

This look at the data has some limitations in the sense that a snapshot of 1998 – 1999 is being compared to the most recent three years rather than for the whole period 1998 to 2011. So it is not a true performance over time that I advocate. Having said that, I feel that the lessons are still valuable based on the logic that if you are in the same place or in a worse place than in 1998 there has been no improvement either relative or absolute which is in general true (though in a purely technical sense that may not always strictly be the case for some hospitals). Also, I have only looked at some highlighting examples rather than a detailed research type of analysis so there may be other good examples that I have not looked at (possibly some place like Harrogate perhaps).

The current status was ascertained from dr Foster’s website in end of Dec 2011 and early Jan 2012.


Remaining where they were

Page one you will find Scunthorpe in the first section/small-medium hospitals rating at the higher end and it remains in the higher end.

East Yorkshire Hospitals rate is high (page 1). On page two you will find Hull Royal rates are high. They are now the HEY NHS Trust and are high currently.

Medway, Dartford and Gravesham remain at the higher end (currently in the highest 10 mortality)

Basildon and Mid Staffs remain at the higher end (page 1)

University Hospital Birmingham rates are at the higher end then as they are now (page 3)

Interestingly Bolton more or less remain where they were – nearer to the higher end of the spectrum.

Interestingly again, Luton more or less remain where they were - in the middle.

On page three under acute teaching hospitals; you will find that Chelsea & Westminster are low in 1998 as they are now.

Moving to a high mortality

North East Lincs (DPoW) seems to have moved from a low mortality to current high mortality.

There are some hospitals like George Elliot and Morecambe Bay who seem to have moved from somewhere in the middle to current higher rates of mortality.

Moving to a low mortality

In page one and page two you will find Birmingham Heartlands, Solihull and Good Hope (on page one) all of which form the HEFT, having a high mortality in 1998 but now have below 100 HSMR over 3 years and could be sited as a case of consistent improvement over time. They still not a ‘low’ mortality hospital though.

Bradford has moved from the higher end to very low mortality.

On page three you will find Kings and St Georges at the very high end in 1998 but are now very clearly in the lower end over a three year period. Page 4 has Airedale which was towards the higher end at that time, now for a good many years they are a low mortality hospital.




 
LESSONS (My personal interpretations)

As a generalisation, many hospitals tend to remain where they are. This is not unusual. I think I have already written about how it is human nature and natural physiology to maintain status quo. The difficulty is when a high mortality hospital remains high where status quo is not a good situation to be in.

It is possible to move from low to high mortality.

It is possible to move from high to low mortality (Bradford, Airedale).

The popular examples may not always be the really the improved ones (Bolton and Luton are popular).

Though mortality is a good headline important measure it is well linked up with good performances in other areas of safety and quality.

As an aside,
It seems to me that (once service configuration changes are explained) most of the improvement is related to the change and improving cultures of the organisation who act in a pro-active (rather than reactive way) - I have not provided reasoning and justification for my assertion here.

From a general reading around this I find that hospitals tend to take the CQC more seriously than dr Foster (or other companies analysing and reporting on mortality) – again I have not provided reasoning and justification for my assertion here.


Whom should we learn from?

If we do the learning circuits we will hear Luton. We will also hear often from Bolton especially about the lean systems. Till recently George Eliot was also doing the rounds talking about improvement. I have attended many of these presentations and have found them very useful from an educational and knowledge perspective.

From a political perspective it is relevant to learn from the experiences of University Hospitals Birmingham who ably withstood the bad press that they had in 2011 (whose CE Julia Moore was recently became Dame Julia Moore) and HEFT whose CE Dr Mark Goldman was seen to leave HEFT following bad press at Solihull (http://www.birminghampost.net/news/west-midlands-health-news/2011/07/20/former-heartlands-hospital-boss-lands-new-nhs-role-in-worcestershire-65233-29083826/ ; but Dr Goldman was reported as having ‘turned around’ hospitals and as being ‘wise’ in the Worcester area http://www.worcesterstandard.co.uk/2011/08/04/story-Wise-figure-takes-charge-of-hospitals-13829.html . It is probable that the cultures in these two Birmingham trusts are very different even though geographically they are not apart from each other. Political learning is very important as it is about perceptions, how people see them and how to handle them.

From a persistent long term improvement perspective we certainly need to learn more from HEFT, Kings, St Georges, Bradford and Airedale.

I am sure that there are many examples in each category that I have not looked into who are equally improved (or not as the case may be); hence the above is just a sample of what attracted my attention and not a thorough analysis.

This is certainly interesting but you ought to tell me if it provides any more insight.

©M HEMADRI 

Follow me on twitter @HemadriTweets