Data from Hugh Harvey - Curated by Marshall Pearce - Last updated 07 April 2017
If I were to come to your place of work, be it an office, a shop or clinic, how would you want me to measure how good you were at your job? It's a tough question, one that doesn't have a readily standardisable answer, especially between different specialities and professions. Yet for some bizarre reason our UK politicians seem to have an enormous hard-on for measuring the hell out of NHS practice, and then waving the numbers around as if they were as sacred as the Ten Commandments. This is what I call 'metrics madness', a condition that only accelerates the nearer we get to an election.
The NHS is an utterly vast organisation, in fact so vast it's the fifth largest employer in the world. The scope of practice is so huge that not one single person knows what it actually does (and that includes Jeremy Hunt), let alone how to measure its efficiency. How on earth do you judge a system so complex, so diverse, and so immeasurable? Well, I can tell you it's not by asking how people 'feel' about it, or how many weeks it takes for someone with a coldsore to see their GP.
"How satisfied are you with your diagnosis of leukaemia, sir? Very, mildly or not at all?”
Patient satisfaction is one of those buzzwords that just plain irritates health care providers, mainly because it's a metric that almost completely misses the point of what health care is about. Being tongue-in-cheek about it - how can you measure how satisfied someone is by the care they received when they've just been told they've got 6 months to live, or their child has leukaemia. "Oh, yes, very satisfied, thanks." It's almost an insult to even consider asking the question. Everyone knows that 'satisfaction' surveys are also incredibly biased. Only the very dissatisfied, bored or deliriously happy actually give feedback; the middle ground generally can't be bothered. Yet satisfaction scores are bizarrely at the centre of some governmental initiatives for rewarding hospitals based on these scores.
In a British Social Attitudes survey early this year it was found that people who had had no contact with the NHS were 11% more satisfied with the way the NHS was run this year than in 2010. Amazing! These people had literally nothing to do with the NHS, had had no treatment, nor had any of their friends or family, but for some reason they were more 'satisfied'. What possible use can this statistic provide, why were these people even asked their opinion, and how can anyone suggest that funding should be based on these results? It's utter metric madness. It literally makes no sense.
Well, guess what Jeremy Hunt said: “These results are a tribute to hardworking NHS staff who are unstinting in their efforts to ensure patients receive a world-class service (sic)". The mind boggles. (Jeremy, I'm sure if I measured satisfaction scores with your work recently…)
Elsewhere in the same report the authors stated that several 'negative media stories' were likely responsible for a decrease in satisfaction in some areas, which is entirely inconsistent with the point of measuring patient satisfaction - basically they've just admitted that satisfaction scores can be directly manipulated by the media, and are therefore not directly linked to actual performance. The politicians, in my view, just shot themselves in the foot.
Another example of sheer metric madness is the '4-hour waiting time' measurement of A&E. 4 hours is a seemingly completely arbitrary number plucked from the air that has no basis in the real-world, is not used by any other emergency healthcare system and has no statistical validation behind it. Why not 3 hours, or 12? Who chose 4 hours, and why? Even the media can't get enough of this 4-hour statistic - to the point where the government threatened a media blackout on even mentioning it when things really hit the fan in the 2013/14 winter.
There are claims that 'reports of adverse patient outcomes occur after 4-6 hours'. Really, what a shock! Who'd have thunk that waiting 6 hours with a haemorrhaging leg could be bad for you? It's obvious that different medical conditions require different levels of urgency and you can't simply measure all conditions under the same time metric. It would be much more shocking to see polytrauma patients waiting 17 hours in A&E than it would for those with a sore throat. Yet this is not the comparison being made.
The Nuffield Trust agrees with me, calling for an end to our 'fixation' on this wholly unhelpful metric. They expertly point out that no one has hit this target consistently since July 2013, and it doesn't look hopeful that anyone ever will again. Finally someone is making some sense! Or are they?... Unfortunately despite their best efforts at appearing rational they also suggest we should replace the 4 hour target with... you guessed it... patient satisfaction scores. Consider my face well and truly palmed in disbelief.
"What we really should be measuring is the proportion of patients who received appropriate treatment within a timeframe appropriate for their condition”
Got a sore thumb? That'll be appropriately dealt with in 4 days time. Meanwhile we're appropriately dealing with this politician with a gunshot wound to his left forefoot.
Dr Hugh Harvey FRCR
Academic Clinical Fellow in Imaging
The Royal Marsden and The Institute of Cancer Research