England's Latest Innovation Scorecard: A Patchy Picture

A common gripe from those in the pharma industry in England is that even when their products have been found to be cost effective by the National Institute for Health and Care Excellence (NICE) (not always easy to do these days), uptake in the NHS can be ‘low and slow’.  The latest innovation scorecard, just released from the Health and Social Care Information Centre (HSCIC) on  June 11th, 2013, still paints a patchy picture.

Measuring access to innovation
The innovation scorecard was one of the tangible outcomes from Innovation, Health and Wealth published back in December 2011.  It’s essentially an effort to increase the adoption and spread of innovation in the NHS.  Part of the problem is not knowing what happens to innovations when they are available for use across the NHS.  Knowing this could kick-start efforts to improve compliance with NICE Technology Appraisals (TAs).  That might particularly be the case where it’s shown that some parts of England are below either their peers (although the HSCIC says that the scorecard data isn’t really supposed to be used for benchmarking but it’s tempting to do so), or below expected use. But just as the HSCIC notes, there’s just no routine data to know.  This has led to a bespoke approach to tease out just what happens to NICE approved products across the NHS.  It’s not easy, or simple, hence the approach has evolved over time with this being the 2nd edition of the innovation scorecard, building up from the first innovation scorecard in January 2013. But even now, not all medicines in all TAs can be included.

The June 2013 innovation scorecard
The June 2013 innovation scorecard covers:

• 2011 data previously presented in Use of NICE appraised medicines in the NHS in 2010 and 2011. This compares expected to observed use for 7 areas

• 2012 data for Defined Daily Doses (DDDs) for 12 medicines in primary care from NHS Business Services Authority data

• 2012 data for medicines purchased in secondary care for 62 medicines in mg or units per 100,000 Finished Consultant Episodes (FCEs) (roughly speaking FCEs are completed stays in hospital), with data coming from the Commercial Medicines Unit at the Department of Health

• 2012 data for medicines purchased in secondary care for 8 medicines in mg per 100,000 FCE bed days, with data coming direct from the pharma companies

Variation
It’s pretty hard to interpret much from the scorecard to be honest, but it’s a valiant effort.  The most intuitive area is the comparison of observed uptake versus expected uptake; that’s probably also the area that is hardest to do too, because working out what was expected isn’t straightforward let alone knowing with confidence what was prescribed.  Part of the complexity is linked to the multiple indications for some drugs, and in others, focusing use on sub-groups of the patient population or perhaps even just interpreting the TA recommendations.  The scorecard shows:

• 4 medicines/groups of medicines where use is higher than expected overall at the England level (Varenicline (TA123), Osteoporosis (TA160, 161, 204), Statins (TA94), Ranibizumab (TA155))

• 3 medicines/groups of medicines where use is lower than expected overall at the England level (Acute Coronary Syndrome (TA47), Prucalopride (TA211), Febuxostat (TA164))

For comparisons of DDDs per 100,000 FCE the take away is simply that it varies, and it can vary a lot, across England.  For example, the biggest difference across Clinical Commissioning Groups (who are newly formed organizations responsible for spending circa £80 billion of the NHS budget) was close to 370,000 DDD per 100,000 population for Zopiclone (ranging from 58,023 to 427,988 DDD per 100,000 population).  That compares to the smallest range seen for Zalepon with a range from 0 to 265 DDD per 100,000 population.  What that really means though is difficult to determine; in essence the major source of variation accounted for by the scorecard is for population size and there are a myriad of other drivers for variation aside from that.

What next?
The HSCIC are keen to get feedback, particularly as the scorecard is classed as ‘experimental’ statistics. That just means that the approach to generating them are still under development and suggest that they are used with caution.  The HSCIC are inviting people to let them know what they think via a feedback form. They’re asking people to think about:

• Suitability of the data and any alternatives (though it’s hard to think what could have been missed with so many groups involved in the work, including pharma companies like GSK and Roche)

• Whether there should be an additional central data collection or if local reporting requirements could work just as well (and that’s a hard one to answer at a time when the ‘new’ NHS is bedding down)

Some might point out that perhaps efforts should be made to look more closely at more recent innovations, some of the drugs included in the scorecard have been in use for years.

Other key issues are more about answering the ‘so what’ question.  If a CCG, or hospital, isn’t in line with expectations, just what will happen? Will action be taken when they’re above as well as below expectations?  And crucially, what does this all mean for patients and their health, and for future innovation?

That there will be variation is inevitable, perhaps what we need to do is to move to a ‘comply or explain’ approach.  Given how much energy (on many people’s parts; companies, patient groups, clinicians, NICE itself) is expended on NICE TAs there really should be compliance with NICE TA recommendations and only ‘reasonable’ departures from guidance made, supported by a transparent rationale.  Especially if that captures the unexpected quirks which might actually need a change in the guidance; real life tends to be somewhat different to the evidence used in NICE appraisals after all.

Leela Barham is an independent health economist. You can contact her at leels@btinternet.com