The news doing the rounds on the Diabetes news circuit today is that even small reductions in blood glucose levels can result in huge cost reduction benefits to the NHS, as a result of the IMPACT2 study. What surprises me is that this is considered news. Surely it’s something we’ve known about since the DCCT study, many moons past? And £5.5bn. Now that’s a reasonably large number in relation to the annual costs of the NHS, but hang on a sec, that’s over 25 years. Something smells a little like spin here. Doesn’t the NHS spend £9bn a year on diabetes, or something?
The press release states the following:
“This new study shows that, if sustained, even modest improvement in blood glucose levels can provide significantly improved outcomes for the 3.5 million people diagnosed with diabetes in the UK. As a result, if applied to the whole diabetes population, almost a million serious medical complications, such as blindness, amputation and kidney failure could be avoided over 25 years, improving the lives of people with diabetes.Better management of blood glucose levels over the same period could equate to a cost avoidance of £5.5 billion for the NHS.”
Which is a valid and important argument. But I read it and I’m confused. You see, there’s this graph from one the many 1996 DCCT reports that, as far as I can see, says very much the same thing:
This graph tells us that if Hba1C is reduced from 9% to 8.5% (i.e. an average blood glucose reduction from 11.7mmol/l to 10.9mmol/l, less than 1 mmol/l), the relative risk of Nephropathy, for example, is reduced by something approaching 20%.
So why, 20 years after it was published, are we only just drawing conclusions that long term small reductions in blood glucose levels will have a significant effect on reducing complications of Diabetes. We’ve known it for years. We’ve been encouraging diabetics to reduce their Hba1C for years to reduce complications. By reducing complications we can easily infer that we reduce costs.
For the headlines to show that, well, what do you know, reducing average blood glucose levels reduces complications and therefore reduces costs to the NHS is a somewhat belated reinterpretation of the data to draw, well, the same conclusions really.
But lets go into the model in a bit more depth.
What we’re talking about is £5.5bn savings over a 25 year period. If we linearly interpolate this, it’s a saving of £220mn per year. Now the model isn’t quite linear, and is weighted towards the ten to twenty year period, so by doing this linear interpolation, I’m attributing more savings to the early and late years of the process, but that’s really an aside. And let’s also be clear, there is no inflation included in my calculation, but I also don’t see any attribution for it mentioned in the paper itself.
If we look at the last set of data on treating Type 2 diabetes alone, NICE states that the cost to the NHS in 2012 was £8.8bn and provides an estimate of £9.8bn in 2015. Let’s consider that. The overestimated savings for point in time “now” across Type 1 and Type 2 diabetics make up 4% of the total spent just on T2 treatments four years ago and 2.2% of that spent in 2015.
And Diabetes UK estimate that in roughly twenty years time, the cost of treating diabetes will be £16.9bn per annum, so our saving at this point is now 1.3% per annum on our linear model. Even if I was to compound that value up, it would work out at 2.8% of the annual budget in 20 years time.
Based on this, I’d like to understand what constitutes the huge saving in the headline, because if we linearly extrapolate between 2015 and 2040, we see that the total spend on diabetes is £359.3bn and the saving is £5.5bn, which works out at a 1.53% saving in the overall spend on diabetes in that 25 year period.
Now in my mind, this isn’t a massive number. While £5.5bn on its own looks fairly large, it’s really peanuts.
While I’ve little doubt that reducing complications will make the lives of many better, and reduce costs to the NHS, this feels like spin of the highest order. Let’s be clear here. That annual saving equates to a saving per Type 2 diabetic per year (assuming the numbers of T2s don’t increase) of £59. That’s not even 9 packs of the cheapest test strips available. So while it looks like a big number, in terms of clinical treatments that you can provide for this, you don’t get much.
If we roll back to the headline then, that making small reductions in blood glucose levels will have huge benefits, I think we need to be very clear. These won’t be huge cost benefits. They’ll improve the quality of life for a reasonable number, but in reality, the only way to make really big savings is to work out how to do the big stuff. How to avoid Type 2 altogether (in our carb obsessed society that shuns fats as a fuel source) and how to really reduce Hba1C levels to lower relative risk levels.
We don’t need to pretend about this, the small improvements help. But we can’t take our eyes off the fact that the only way we are going to really make a difference is to do the much, much bigger and consequently much, much harder stuff. And that’s where we really need to focus attention. This is but a side show.