The FT is shocked today. It is reporting that:
Taxpayers pay up to five times more for National Health Service operations in some hospitals than others, according to a Financial Times analysis that may fuel the healthcare funding debate.
The differences are so great that the NHS could save about £100m a year — equivalent to approximately 4,000 nursing salaries — across just three operations if it were to raise the performance of less efficient hospitals to the level of the top 25 per cent, according to an analysis of official data.
I would have thought that after the Piketty fiasco someone at the FT would have decided that it might be in their best interests to be a little more circumspect about their ability with numbers, but that is not apparently the case.
The FT says it has excluded the top and bottom 5% of numbers from their samples to allow for differences in 'cost allocations' but then says that the vast majority of cost difference depends upon how long people stay in hospital after treatment. I am dubious in both cases.
First, as an accountant I know only too well that there is no true cost of almost anything. That is particularly true in a complex organisation like a hospital. Everything in such cases comes down to the decisions made on how to allocate costs to activities. You could load costs onto the operating theatre. You could, just as easily, load costs onto bed occupancy. And you could, alternatively, decide that patient throughput was your cost allocation driver and in that case A & E and outpatient clinics would suddenly become very expensive. That's all because the variable costs of undertaking any activity in a hospital, whilst not insignificant, are always going to be only a part of the total, with that total being dominated in many cases by the cost of running the infrastructure of the hospital.
What is that infrastructure? Let's start with the heavy burden of administration that marketisation has imposed upon the NHS. And let's not ignore the costs of PFI: much of the variation noted here could come down precisely to that one issue. The cost variation could also depend almost entirely upon the decision that the government has made to allocate funds to the hospital, or not. Well funded hospitals have higher costs because they will spend the money that they are given to deliver potentially better outcomes but that is not noted in this survey.
Nor is something else of considerable significance: where these operations take place will have an impact upon their cost. In wealthy areas it is easy to discharge patients into the community because care is available. In other areas that is not possible. And some areas do, of course, have what are called ' skewed elderly' populations. Think of seaside towns with large retirement populations. In those places people cannot be discharged as quickly as they are in locations where the average age of the person having the operation is lower.
Maybe the FT took all these factors into account, but there is no indication of it in this report. What it looks like is another case of FT tabloid style journalism when the first number that fell off a spreadsheet is taken as a fact without considering any of the context that might explain it. If this paper wants to be taken seriously it has to be better than that.
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
I wanted to have a look at the data myself to see what the FT was doing here but no link is provided anywhere in the article to the data source used. And this was the paper that last week was calling for data transparency in the wake of Chris Giles’s criticism of Thomas Piketty. You couldn’t make it up.
One thing that immediately strikes me: has anyone tried to do the same analysis with private hospitals? There might well be as large, or greater, cost variations in the private sector, depending how the costs are measured. Also, the methodology for allocating the cost of a single operation is very difficult: how much of the fixed costs of running a hospital (infrastructure, medical equipment, admin and management staff, utility bills etc.) should be included? Very far from straightforward.
Neat question re private hospitals
There the massive interest charges paid offshore may well impact on cost!
This lack of reference does rather suggest the FT can’t even begin to justify its argument. It might as well be pulling figures out of the air, putting it politely 🙂