At the core of this year’s planning guidance lay something rather unusual.
I think its intention was benign, but it has the potential to cause a lot of trouble, drive up waiting times and distort clinical practice right across the country.
This is it: “3.7: Commissioners and providers should plan on the basis that their RTT waiting list, measured as the number of patients on an incomplete pathway, will be no higher in March 2019 than in March 2018.”
The problems would unfold if the waiting list size were treated as a target (a phenomenon described nicely by Goodhart’s law). People look for creative ways to achieve targets, often with unintended consequences, and when it comes to achieving a waiting list target the shortcuts are (unfortunately) obvious and harmful.
Imagine you are responsible for a waiting list. It’s getting bigger and you’re taking flak for that. It also happens that you also have an all day operating list to fill.
You could book a highly complex patient who has already waited several months and will need at least seven hours in theatre followed by several days of high dependency care. That would reduce the waiting list by one patient. Or you could fill the session with straightforward day cases and reduce the waiting list by 14 patients.
With a target to meet and limited means of doing it, you make the obvious choice.
It has consequences. The waiting list size – an easy thing to measure – goes down. The case mix remaining on the waiting list – which is less obvious – gets heavier. The waiting time for straightforward day cases goes down; the waiting time for complex inpatients goes up.
You know it was the wrong thing to do. There is no clinical reason why straightforward day cases should jump the queue – and the heavy cases will have to be treated before long (though they may come to harm because of their unfairly extended wait). But these are all consequences of targeting the waiting list size.
This is also an example of what economics professor John Kay calls a tailgating strategy – driving a little closer to the car in front to shave a few seconds off your journey, in return for a slightly higher risk of dying in a horrible pile up. Other tailgating strategies include holding open staff vacancies, delaying maintenance and all the other ruses used to delay costs into the next financial year.
I don’t expect NHS England or NHS Improvement intend to delay complex cases, or even for the waiting list size to become a target. More likely, they understand waiting times are a function of two things – the size of the waiting list and the order in which patients are treated – and they simply want to control the former (via the planning process) without intending any detriment to the latter (as targeting the list size would do).
The 2018-19 year is young and there is plenty of time to stop these harms from happening. If my characterisation of NHSE/I’s intentions are correct then as a first step it would be helpful if they would clarify that (perhaps in the comments section below).
But the most important actions will happen at local level. In trusts and clinical commissioning groups, the focus should remain on limiting the waiting times of every patient, in every specialty and subspecialty, as the NHS constitution intended. The waiting list size is not an end in itself, but part of the means.
What would happen if waiting list size was a target?
If the waiting list size did become a widespread target, with the unintended consequences outlined above, how could we tell from the national statistics? I think several things would happen to the numbers that I keep an eye on:
- Waiting list management would deteriorate as a long waiting tail of complex cases developed, so the index of waiting list management would rise.
- The number of patients waiting several months would rise, while the number waiting over a year (which is a targeted priority for 2018-19) would remain more controlled, so the number of over 26 week waiters on the waiting list would rise.
- Waiting times would rise in specialties with heavier case mixes and fall in specialties with lighter case mixes, as the allocation of “extra” sessions changed. For instance, orthopaedics would probably rise and ophthalmology would fall.
- Waiting times would rise faster in those local services that already have long waits, so the distribution of waiting times would rise even more rapidly at the long waiting end.
- All of which would contribute to a general rise in how long 92 per cent of the waiting list has waited and (on the less helpful but more popular measure) on the percentage of the waiting list falling within 18 weeks.
Rob Findlay is director of the software company Gooroo, specialists in demand and capacity planning