Muir Gray’s paper of the week: NHS England’s plan to pull the plug on ineffective procedures
Bottom line, chosen by Muir from the paper
NHS England proposed last week to stop or reduce routine commissioning of 17 interventions, including surgery for snoring, back injections, and knee arthroscopy for osteoarthritis, in favour of “less invasive, safer treatments that are just as effective”
NHS England’s states its aims as a “hierarchy of goals”: to reduce avoidable harm to patients, to save professionals’ time, to help clinicians keep practice in line with changing evidence, to create “headroom for innovation,” and to maximise value and avoid waste for patients and taxpayers
Implications for value improvement
The importance of this paper is that it separates out interventions into two different classes. Firstly there are interventions for which there is no evidence of effectiveness and these are obviously of zero value. Perhaps more interesting are the other interventions and it is good that the BMJ have put them into different colour categories. These other interventions are effective, so to some individuals they are of high value. But what is needed is to identify which individuals who will receive high value. Donabedian in 1980 pointed out that as you put more resources into service, for example resources to do more operations, you start by treating people who are most severely affected, so the benefit increases fast at first and then flattens off – the Law of Diminishing Returns. Unfortunately the harm goes up in a straight line and if you subtract harm from benefit then you get the graph shown below, the classic graph:
Now from an individual point of view when you start off with an intervention you only offer it to people who are really going to benefit, so the benefit to harm ratio is very high. As you do more interventions you start offering it to people who are less severely affected and for these people the maximum possible benefit is less, but the probability and size of harm is the same. You may reach a point where it’s futile – does more harm than good – and here is the graph from the NHS Atlas of Variation showing this:
So, this is a key paper with key concepts because we have to consider value from both the population and a personalised point of view.