Editor’s blog Monday 26 April 2010: Bovvered about bonuses, and measuring up the coding creep conundrum

The Lib Dems have given their FOI stuff on managers’ bonuses to The Observer, leading to “>this not-wildly-shocking story about managers’ bonuses.

There is a debate to be had about incentivisation and unintended consequences. It would have to involve the degree to which public sector managers can control the inputs and outputs of their work in any meaningful way. If the debate were a rational one, it would probably not make bonuses very large if it concluded they were vital (and there should logacally also be maluses, or forefeits for poor performance). Incentives should also only be payable after a suitable time lag.

This is not that debate. It is shrewd, streetwise politics by the Lib Dems, who are on the record planning to address senior managers’ pay: it gives them a nice lead in the ‘demonising managers’ steeplechase.

Some of the story is particularly silly. Heart of England FT’s deputy CE Beccy Fenton’s £32,000 ‘bonus’ looks in truth more like a cut of the £1 million she earned for the Trust from consultancy. And quite a small cut at that: less than 3%.

Meanwhile, the Royal College of Nursing bids to be taken unseriously, in its pre-conference publicity for an online poll of 287 nurses’ opinions about the NHS.

This story suggests that “nine in 10 of those responding to the poll said patient care was compromised at least several times a month. Nearly a third said it happened on most shifts. Most hospital wards are operating with an average of 13% fewer staff than they officially need.”

You have to assume that if the RCN can’t get more than 287 people to fill out its survey, they are in serious trouble.

The RCN probably employs more nurses than that.

In his coverage of this story, Jeremy Laurance of The Indpendent observtes that the Conservatives have been at the old FOI lotion too: “>”Figures obtained by the Conservatives under the Freedom of Information Act, published yesterday, showed 24 trusts were planning a net cut of 651 doctors over four years and 26 trusts were planning to cut 2,050 nurses”.

This derives from the Sunday Times article, revealing the FOI findings in more detail, and associated feature on consultants' pay for overtime. Good work, comrades!

Meanwhile, the very smart Nigel Hawkes of Straight Statistics writes in the BMJ about the topical controversy over how quality of NHS providers is measured. You will remember that we touched on this last week.

Hawkes questions whether patient coding making hospitals appear better than they are in his feature ‘Patient coding and the ratings game’

The leaping-off point is the disparity between the HSMR methodology used by Dr Foster to rate Mid-Staffs as a high-performing hospital – the ninth-safest in England, and one of the five most improved - and the actually quality of its care, which the Care Quality Commission termed “appalling”.

Hawkes’s piece is partly inspired by Dr Foster rivals CHKS, whose recent research has shown “that the number of codes has been creeping up over recent years, yet crude death rates in English hospitals show virtually no change over the past five years. There is also a big variation in the average number of diagnostic codes per patient from hospital to hospital”.

He suggests that one possible explanation could be the increasing proportion of patients classified as needing palliative care. The next section is technical: I shall quote from the BMJ press release to avoid confusion.

“CHKS figures show the number of deaths coded Z51.5 (the code used for palliative care) was under 400 a month in 2004, but had reached more than 1,800 a month by June 2009. 

Patients coded Z51.5 are assumed to have come into hospital to die, so performance calculations make allowance for that. This means that a few heavy users of the Z51.5 code could have reduced their HSMRs from 110 (above average) to 90 (below average) simply by increasing the frequency of use of the palliative care code”

Hawkes is clear that he has no proof this is happening, but writes, “there is no question that the situation is open to manipulation by trusts.”



This follows the BMJ’s publication of this feature by Richard Lilford and Peter Pronovost, which proposed that ‘Using death rates to judge hospital performance “a bad idea”’, and the accopmpanying editorial by Nick Black, calling for HSMRs to be abandoned and replaced by better and more specialist metrics.

There is clear evidence that metrics need to progress. HSMRs do not tell us everything; they do not tell us nothing. There are better measures to be made, no doubt; equally, there are too many NHS organisations, providers and commissioners, who are not running on what imperfect data we have. As the data used gets refined (always easier when it is used often), so closer focus on the data-illiterate needs to ensue.

Meanwhile, the Sainsbury Centre for Mental Health has an excellent new publication on its website on how to put recovery work at the centre of mental health programmes.

Implementing Recovery: A methodology for organisational change aims to help mental health services measure how well-placed they are to support the people who use them to build the lives they want, settiing out the three stages a service needs to reach on ten key challenges to become fully focused on recovery. It allows mental health services, their users and their commissioners to judge how well they are doing in meeting each of the ten challenges.

Don’t let mental health be your Cinderella, will you? Her sisters are none too fun. Go read, and then act on the reading.