General Surgery Report Details Huge / Unbelievable Variations In NHS Procurement Prices

Some of the very best procurement people we know work in the NHS, and much has been done in recent years to develop a real procurement community in the sector. And yet, if the latest report from Professor Tim Briggs is to believed, there is still appalling overspend on medical items in too many hospitals.

However, that little word “if” is important in that last sentence. Because some of the data presented in the report issued last week looks dodgy to say the least.  The “General Surgery - GIRFT Programme National Specialty Report” comes from Briggs, an eminent surgeon, and has much interesting to say about the autonomy of surgeons and variation in the system. However, it is of course the section on procurement that interests us.

It gives a table (on page 24) which compares the prices for 17 items of surgical equipment across 15 trusts who participated in the study and site visits for the GIRFT ("get it right first time") team. That shows amazing cost variations. For instance:

Highest price Lowest
Small laparoscopic wound protector £126 £0.36
Laparoscopic harmonic shears £2473.71 £27.75
Size 5 port with trocar £220 £9

 

These variations are so great that frankly, we just cannot believe them. The only possibility if they were accurate would be some huge fraud on the NHS perpetrated by suppliers and buyers working together. The report then says “The variation was so startling that the team requested further data to confirm whether trusts were describing the same items. It was also suggested that pack size may be a factor, with trusts potentially benefiting from buying in bulk”.

But it does not go on to give the explanations that were obtained from the trusts. And surely, “pack size” might mean one quote is for a single item, when another could be for a pack of 12? In which case, that table is meaningless.

This is very frustrating. If these variations are real then we need to hear from the trust paying £2473 instead of £27.75. If they aren’t real, then they shouldn’t be published.

The follow-on table looks at just three trusts, and takes “pack size” into account, apparently. That shows in general more believable variations of perhaps 20% or so. But even here, there is a concern. One trust gives £89.43 for a single unit price for  “size 5 port with trocar” (for a pack of 6) – another claims to pay £14.40 per unit But are we sure the first aren’t quoting the cost of a unit of 6 – their price is suspiciously close to six times the other trust’s single unit cost?

The HCSA (Health Care Supply Association) issued a supportive Tweet last week – “HCSA welcomes Tim Briggs Report and shares concern at lack of price transparency and price variation- the NHS needs to do much better”.

Fair enough - but I would have liked to see a bit more challenge on the data front from HCSA. Not in a defensive spirit, just dodgy data does no-one any good, and this does make NHS procurement look silly. After all these years, are HCSA members still paying 300 times more than they should for a “small laparoscopic wound protector”?  If so, they should be drummed out of the HCSA and their jobs.

The other real irritation is having the data but no explanation as to the “why and who”? Isn’t it about time that the trusts were identified – by all means give them the chance to explain themselves, but we need to know what is happening here. Otherwise it is just endless "shock horror" bench-marking reports and we never seem to move forwards.

There are also strange goings on around the PPIB price benchmarking tool, which I’m told trusts are finding useful – yet there are arguments about funding with NHS Improvement and Department of Health wrangling over the cost. That is tiny compared to the potential for savings – something maybe for the new DH Chief Commercial Officer, Steve Oldfield, or new Commercial Director, Melinda Johnson, to sort out?

 

Voices (15)

  1. Sam Unkim:

    Anyone from one of the 15 Trusts like to comment ?

  2. Final Furlong:

    Here we go again. Price variation exists because so many trusts buy the same products from the same suppliers at different volumes, through different supply channels, against different contract durations, with different commercial terms, using different levels of procurement skill, supporting different levels of clinical subjectivity, and so on…. And to everyone’s shock, suppliers treat all of these customers differently. And then we analyse different data sets, using different spend analytics specialists, who look at the data in different ways, to generate a different answer each time. Same (old) story.

    However, before anyone publicly publishes any insights gleaned from any data, they need to check its validity beforehand. Some of the GIRFT examples just look impossible, but anything is possible in the NHS (given the dynamics above).

    Re: Simon’s comment, The Atlas of Variation was a great example of using accurate but incomplete data. Yes, that trust bought that product at that price from that supplier at that time and at that volume (but at the end of the year they received a 50% rebate which didn’t get captured). Great headlines. It didn’t make a jot of difference.

  3. John Warrington:

    I feel I should add a little context to this debate. GIRFT is about clinicians challenging clinicians to do better. As we all know, clinicians generally have a poor view of costs, if any view at all, so when the GIRFT clinical leaders ask their colleagues what they use and what they pay it should be no surprise that the answers are suspect. Most likely local clinicians will have asked their finance and procurement teams and then handed over raw information to the GIRFT team, and as someone pointed our earlier the team would have had little time to verify what they received.

    It would have been so much easier if the GIRFT team could have had access to a single national procurement dataset and then proactively engaged their colleagues in delivering better value for money, but of course no such dataset exists. PPIB is getting there but until the coding issue is sorted as Nicola points out, data will always be suspect.

    I would humbly suggest the NHS procurement community gets behind the GIRFT initiative and PPIB as this is the first time I have seen NHS clinicians collectively concerned about costs and they are going to need all the help they can get to address the variation that clearly exists if not the extremes highlighted in this report.

    By way of balance though, DH 2016-17 accounts show that apart from Pharma, trusts have finally managed to keep a lid on the year-on-year inflation we have seen over the last 5 years and I believe this is down to hard work of the NHS procurement community. But there is no time to rest on laurels, and the opportunity for procurement to work closely with clinical leaders at all levels is not one to be missed as they don’t come around too often!

    1. Final Furlong:

      Can’t argue with that!

  4. Nicola Hall:

    These types of reported variation will always be questioned until we have a fully adopted data standard for comparison reporting. We have been involved in many inventory management projects , untangling the data is always one of the largest challenges. Generally the unit description is in the description, so per one is not the unit price. Implementing core systems is the key, forcing out these anomalies. The systems that’s are currently in place simply don’t give the visibility that procurement need. We need to just get on with the GS1 program at the coal faceto have a sets of data that can be compared properly. I doubt the mass fraud theory, I do believe the true picture can’t be analysed until we all have an equal base to start from

    1. Peter Smith:

      thanks for all the comments on this. I wasn’t seriously suggesting fraud by the way, that was just really making the point that I couldn’t believe the huge variations quoted. What annoys me somewhat is that people publish data that is clearly not meaningful – you wonder why? Just to get headlines? It really doesn’t help. We do need better NHS data but if every Trust employed a really good head of function and invested in some decent technology then the benchmarking wouldn’t matter so much to be honest.

      1. Sunshine Act 2010 - Effects on Physicians (NEJM 2013):

        Although … there is a well know legitimate point about the close relationship between Clinicians and highly profitable Medtech companies, hence the US Physician Payment Sunshine Act 2010, and some of the aspects of the recent UK Health Sector Medical Supplies (Costs) Act 2017 ( it would be good to get your thoughts on the later sometime ).

        Some would say that a lack of evidence led Clinical product governance, inhibited by close relationships with companies, has significantly contributed to escalating; product fragmentation & complexity, the demonstrable lack of price transparency, inefficient supply chains, inhibited innovation, and worst of all, increasing recalls and adverse product outcomes.

        The current data shows all of this, it’s just difficult to analyse as all big transactional datasets are. As you rightly point out, investment in NHS Spend Analytics capability & technology, have been consistently done on a shoestring by the Trusts and DH, often without due credence to the complexities associated with the task. I suspect GIRFT struggled with this data and product complexity combined with time pressures more than any deliberate headline grabbing.

        After all, legitimate variation and inefficiencies do exist in the supply chain and are a big sensational value in any case, with patients being effected at the end of it too!

        I don’t think that is the real problem though. It’s that once analysis is done accurately, even published in the likes of the Carter report in key medical categories like Primary Hips and Cardiac Stents, there is no clear mechanism for improved Clinical product governance or supply chain transformation to action the findings, so nothing moves on and the NHS England price variation report groundhog day rolls round every year, and every year, a fresh set of PIP, ASR, Hernia Mesh issues arise effecting patients.

        Hopefully GIRFT, as probably the best shot in a generation of national clinical: procurement alignment, as well as other initiatives and changes in the system will enable these mechanisms to emerge so that some of the serious problems in the Medtech supply chain can begin to be improved.

        If the mechanism does emerge then I disagree with you last point.

        Data and analysis for governance, compliance, benchmark and other things will be amplified rather than matter less.

  5. ScottP:

    As Simon says, we’ve know for years that there is legitimate variety and variation in product usage and price which belies underlying acute supply chain inefficiency.

    In this case, for harmonic shears, after the data has been reviewed further and things like pack sizes are considered, the avg price is likely to be around £480 with 21% legitimate variation from £330 to £510 for the same items. Of course price variability only tells a part of the story where inventories, support and transaction costs are also all inflated and also highly variable across the system.

    Aggregated; Invoice, PO and NHSSC data are an essential source of information when handled correctly and shed particular light on the inefficiencies of the Medical Technology supply chain where PO compliance is upwards of 85%.

    They are the only source for price and product usage governance and working towards better data aggregation, data quality and analysis in the future should be considered essential for NHS procurement, particularly in times of significant transformation.

    Supply Chain analytics and benchmarking are a complex exercise in healthcare but as the Carter review highlights, can be done well or poorly.

    1. bitter and twisted:

      17 items is a suspiciously un-round number, isnt it?

  6. bitter and twisted:

    The lesson is: NHS procurement data is an absolute pile of cack, and its amazing people dare to publish reports based on it.

    1. Mark Lainchbury:

      Just back from annual leave & can’t have that.

      47million of the 66m NHS PO lines on PPIB have been created from NHS-SC’s single national catalogue & are very consistent & very reportable.

      1. Bill Atthetill:

        Welcome back Mark!

        Does this include all related rebate data?

        1. Mark Lainchbury:

          That’s a politics problem, not a data problem.

          The NHS chose to scrap NHS Supplies’s absolute price guarantee (remember the blank “price challenge form” in front of every paper catalogue) just prior to its being outsourced to DHL.

          The NHS then choose to start giving rebates, regional & local prices and NOT make these variations transparent ( & reportable ) to all Trusts.

          1. Bill Atthetill:

            I have heard that one surgical glove manufacturer has some 30 individual rebate agreements with trusts. The bigger the trust, the bigger the rebate, one assumes. Isn’t this still a data problem (taking aside that it may be political problem too)?

  7. Simon Walsh:

    Thanks Peter- very useful analysis.
    You are right about the data though we ( HCSA) judged that there was a bigger issue of price variation and the Report should be a catalyst for such a debate.
    The Atlas of Variation taught us that you very easy become bogged down in data quality and lose sight of the big question.
    Thoughts welcome.

Discuss this:

Your email address will not be published. Required fields are marked *