McKinsey save the NHS

McKinsey in shock discovery that if you put things in numerical order half of them will fall below the median. What tautology will they demonstrate next?

Via DNUK again - here's a terrifying presentation from McKinsey*:

In February 2009 McKinsey was instructed by the Department to provide advice on how
commissioners might achieve world class NHS productivity to inform the second year of the
world class commissioning assurance system and future commissioner development. The
advice from McKinsey, in the form of the following slides, was provided in March 2009.

I recommend having a detailed read to see what a bunch of 20-something Oxford graduates with no experience of the health service can come up with. Basically if everything was done cheaper and more efficiently it'd all be fine - who'd have thought?

In slide 17 we see how if we just reduced all clinical staffing levels to the median we could save a few billion - no mention of whether staffing levels have an effect on quality of care (another race to the bottom) - indeed they later go on to argue against any attempts to ensure minimum staffing ratios:

"Some Royal Colleges are recommending introduction of mandatory staffing ratios on safety
grounds that will lead to increases in staff required above the activity growth e.g ratio of
1/28 per midwife 

Certain service reviews are also recommending more staff is required e.g. stroke, children"

But they argue we should "Limit introduction of mandatory staffing ratios" to save money - they can't be arguing for minimum staffing ratios for any reason, I'm sure it'll all be fine.

Highlights include slide 28 where they show how you could cut 'bed days' by 10% in mental health if only length of stay could be reduced to the median in the poor performers. I wonder whether those trusts with longer lengths of stays admit more severe patients and don't admit less severe patients, perhaps managing them in the community, and thus have already saved money over the other providers who are admitting less severe patients but getting lower average lengths of stay? Well I'm afraid McKinsey can't tell us because they are analysing data in the same way you would expect someone who has no idea of the context and a money-making axe to grind to do.

Presumably we could save money by abolishing all those specialist tertiary and quaternary hospitals with their high risk procedures because DGHs get better results for the same procedures. If only these specialist centres could be as good as those small local hospitals, the difference in performance couldn't possibly be because all the really difficult ones get sent to the specialist centres - how are we supposed to show that on our pie charts?

On slide 53 we can see some of the "procedures with limited clinical benefit" which are "relatively ineffective" or "potentially cosmetic" that we can 'de-commission' to save nearly £1 billion. Femoral hernia repair for instance, as someone on DNUK points out, the risk of strangulation is 45% at 21months in a femoral hernia, this is not a 'cosmetic' procedure!

A particular highlight is slide 60 where we find out that US physicians who do imaging (e.g. x-rays) in their offices do more investigations than those who need to refer to a radiologist. Presumably this means that we are going to massively increase the training places and consultant jobs for radiologists so we can save a few quid on chest x-rays by running each request past a radiologist first?

On slide 64 we can see that as digoxin in heart failure doesn't increase life expectancy only improve symptoms we shouldn't fund it at all. Move over NICE and the controversial QALY, if it doesn't stop you dying it isn't worth a penny.

As someone on DNUK points out, all their reasoning is based on using publicly available data to rank everyone, and then claim that £X million can be saved by making the 'worst performers' as good as the best performers. Conceptually difficult ideas such as figuring out if the factors making worse performers are actually amenable to intervention (e.g. you aren't going to make the people in rural areas all close together and nice and efficient for home visits like in a city however many graphs you draw) are just so much irrelevant detail.

I wonder how much we could have saved by getting some people who know what they're talking about to do a review instead of McKinsey and their one-size-fits-all Panglossian musings?

* This isn't new, for instance, the Ferret Fancier reported on it last year, but I think it is timely to revisit what facile idea people like this (yes Lansley, I'm looking at you)have about 'reforming' the NHS.