Email your comments

It is unlikely that any geriatrician would disagree with the entirely reasonable statement made by Damien Jenkinson and Roger Boyle when they say: “The Stroke Strategy for England brings tremendous opportunities for developing services for all people with stroke and at risk of stroke” (BGS May newsletter).

It is good to see that their efforts have moved stroke up the political and health agendae. Their article aimed to lessen anxieties about the overall potential detrimental impact on stroke services, of focusing preferentially on acute stroke care. It ended with the rallying call of, “This is our greatest opportunity to redesign the whole stroke care pathway and truly collaborative working has never been more important”, seemingly forgetting about the last “greatest opportunity” to redesign and improve stroke services just over half a decade ago in the form of chapter five of the 2001 Older People National Service Framework (NSF).²

The proposed stroke care pathway that the NSF policy hoped to deliver looks rather different from the 2007 National Stroke Strategy (NSS)³ “hub and spoke” vision of stroke care delivery that was originally outlined in Roger Boyle’s own December 2006 Mending Hearts and Brains report.⁴ The NSF was a partial success story as evidenced by the serial improvements that have been documented by the RCP (London) National Sentinel Stroke Audits.^5,6 Such changes, however, were difficult to achieve in many cases as unlike the coronary heart disease NSF or the cancer NSF, resources allocated to the Older People’s NSF were targeted by the government on priority sections of the NSF that did not include stroke.⁷

The annual cost of stroke is put at £7 billion.¹ .A relatively meagre £105 million⁸ of central funding has been allocated over three years to implement the stroke strategy. History may be about to repeat itself unless more centrally allocated resources are directed towards all aspects of the proposed stroke care pathway.

Controversial claims and evidence based policy
At the beginning of this year the Prime Minister gave a speech at King’s College London, setting out his agenda for the NHS.^9,10 He made a statement underlining the importance of evidence based change and in effect, evidence based policy making¹¹ and said “So my guarantee to you today is that our vision for change will be based on clinical evidence and the drive for a more preventative health service”. Does the National Stroke Strategy match up to that guarantee?

There are a number of claims and statements made by Damien Jenkinson and Roger Boyle in the May article that cast doubt on this guarantee in relation to the vision for stroke. They should not be allowed to pass unchallenged and can be divided into sections as outlined below, covering the evidence for comparative performance with European neighbours, the relative population effects of thrombolysis and stroke units, the under-selling of stroke unit effects, and the preferential funding within the stroke care pathway.

The evidence for the worst outcomes in Europe: findings of comparative international studies
“The participating hospitals were not intended to be representative of stroke care in the particular country but to give examples of the different ways in which stroke care may be provided across Europe”; “The residual differences in outcome were too large to be explained by variations in care and most likely reflect differences in unmeasured baseline factors”; “This leaves large variations unexplained and suggests systematic bias”. These three quotes - the first from the paper by Grieve¹², the second from Weir¹³, and the third from an editorial review of Gray’s paper¹⁴ by Dippel¹⁵ – hardly amount to a resounding vote of confidence in the certainty of findings that are said by both the Department of Health and the National Audit Office (NAO)¹⁶ to show that stroke care in the UK is the worst in Europe thereby justifying a policy change to the European model of focusing resources more heavily on acute aspects of care.

These three studies were never designed to look at the differences between European services; Weir et al’s paper, that was based on 15,116 patients enrolled in the International Stroke Trial, was liberally spread with sensible caveats about how the findings should be interpreted and used with a degree of caution; in the final paragraph of the discussion the authors stated that the study “demonstrates the potential limitations of using analyses of observational data to explain international differences in outcome after stroke.” With such limitations and stark warnings, can the Department of Health justify what has been admitted to as “limited data” in these papers being used as strong supporting evidence for their claims about the UK?

It seems as though the vision for change is “evidence-light” rather than “evidence-based” and one that would not seem to deliver on the Prime Minister’s January 2008 guarantee.

Levers for change and the evidence: relative population effects of thrombolysis and stroke units
Firm assertions made by the Government do not always turn out to be correct or to have had a particularly solid evidence base. Were weapons of mass destruction (WMD), the lever to invade Iraq and stimulate regime change, ever found?¹⁷ Will 10 per cent of stroke admissions ever be thrombolysed? Manipulating potential levers of change to support a particular strategy aim does not always guarantee a government and its ministers being bathed in glory or public gratitude.

In order to arrive at a population effect of 1.3% with thrombolysis - based on an absolute risk reduction of 13.1% with the treatment - it would imply that Damien Jenkinson, Roger Boyle and the government’s health ministers who are promoting this policy change to focus on acute stroke believe that 10% of all stroke patients are eligible for thrombolysis, not just 10% of ischaemic strokes. Do they mean that in England thrombolysis will be given to 10% of all 110,000 stroke patients – 11,000 people? If so, that is double Boehringer-Ingelheim’s estimate given to NICE during the appraisal of Alteplase; the drug company’s paper estimated that at a 10% rate in 2011 there would be 5,512 doses of the Alteplase delivered to ischaemic stroke patients in England and Wales.¹⁸ Even if the Department of Health is referring to only 10% of the 73,000 stroke admissions in England – a statistic reported by Health Minister, Ann Keen MP, in an October 2007 written answer to a question in Parliament¹⁹ - there would still be 32% more patients being treated at the 7,300 Department estimate than the 5,512 estimated by Boehringer-Ingelheim for both England and Wales. Who is more likely to have looked very closely at the details and come up with a realistic and achievable view of the extent of use of a treatment: (a) the drug company manufacturing the product that needs to forecast potential drug production requirements and the profits produced for its shareholders or (b) the government buying it for use in the NHS?

Similarly, these 7,300 or 11,000 figures are at odds with the Department’s own excellent Impact Assessment²⁰ document figures for the use and effects of thrombolysis. Damien Jenkinson and Roger Boyle indicated the absolute benefit with thrombolysis to be 13.1% meaning that for every 1,000 patients treated 131 will regain independence, who would not otherwise have done so. Table 6 of the Impact Assessment document shows that with thrombolysis the expected base case rate of recovery to independence is 549 with a range between 307 and 792. Working backwards from those figures with the 131 / 1000 estimate of recovery, the number of doses of Alteplase that would have to be given would be 4,191 to achieve the base rate recovery figure of 549. The range of use of Alteplase would be 2,344 to 6,046 doses each year.

At the Department’s estimate of 7,300 doses the number of patients recovering independence would be 956, a figure close to the 1,000 people recovering with thrombolysis mentioned by Gordon Brown in his January 2008 speech, that is also the figure given in the main body of the NSS. The Department of Health has declined to explain why the thrombolysis outcome numbers in the NSS do not match with those estimates produced by UCLAN’s up to date economic modelling work in the Impact Assessment.

The Department’s 5.6% v 1.3%, four times more effective population effect claim undersells the relative benefits of organised stroke unit care in the pathway when compared with thrombolysis; 5,512 of 110,000 patients treated with thrombolysis is a 5% not 10% population treatment rate. Using the 13.1% absolute risk reduction with thrombolysis given in their article produces an overall population effect for thrombolysis of 0.65%. Using the 5.6% population effect of stroke units given in their article means that organised stroke unit care is around 8 – 9 times more effective than thrombolysis on a population basis.

Underselling the effects of stroke units in the NSS
The benefits of stroke units in terms of outcomes are well recognised. On page 33 of the 2007 NAO publication “Joining forces to deliver improved stroke care”¹⁶ it stated: “The Stroke Unit Trialists’ Collaboration showed that by providing stroke unit care rather than general medical care, five out of every 100 people are able to return home independently, two less require institutional care, and three less are likely to die.” This in effect is stating that the number needed to treat (NNT) to avoid one death is 33 (100 / 3) and the NNT to enable one person to regain independence is 20 (100 / 5). These figures are echoed in paragraph 76, page 25 of the Department’s own Impact Assessment20 where it indicates that for every 33 patients treated an extra one survives and for every 20 patients treated an extra one is discharged independent. These Cochrane derived figures are accepted by recognised experts in stroke rehabilitation as being the expected outcomes with stroke unit care.²¹

What the above statements in the NAO and Department of Health documents quite clearly demonstrate is that for any given additional population of patients going through organised stroke unit care there will always be a larger number of people who regain independence than who will avoid death. They show that for every 100 additional patients accessing organised stroke unit care, 5 regain independence and 3 avoid death; for a 1,000 population the numbers would be 50 and 30; for 11,000 patients - that is 10% of England’s total stroke population - the numbers would be 550 independent and 330 deaths avoided.

It is baffling, given the apparent simplicity of all of the above, how six eminent stroke experts and one stroke tsar can appear to be so bad at numbers and detail in this particular case to cause the Secretary of State for Health to publish the National Stroke Strategy policy document containing a claim about the outcome with stroke unit care that clearly lacks face validity, has been given an incorrect reference, and contains figures that do not even accord with those produced by the Department’s own latest economic modelling work as published in the tables of the Impact Assessment. The minister’s policy document rationale for stroke units is given on page 31 and states, “If timely access to stroke units was increased to 75 per cent of stroke patients, this would prevent over 500 deaths per year and result in over 200 more independent individuals.” More deaths prevented than individuals returned to independence; is that what the evidence says should be expected from stroke unit intervention? The quoted figures look remarkably similar to those produced by the NAO in its 2005 stroke report; the estimates from economic research modelling work were 550 deaths avoided and 205 people returned to independence if the number of stroke patients accessing stroke unit care increased by 25% from 50% to 75%.²²

Does the Secretary of State for Health, who is ultimately responsible and accountable for the strategy and policy document, have any explanations that ought to be shared with the public and those commissioning stroke services as to why the potential impact of stroke unit care has apparently been undersold? Is this politics driving strategy and priorities within the strategy or is the evidence base driving strategy priorities?

Preferential funding in the stroke care pathway
So what will the £32 million of extra central funding going to NHS demonstration sites be spent on? Will stroke units have a fair share to bolster the therapist deficits that the Department has identified? The answer judging from the details in the Impact Assessment is that all of the extra money is to be used to fund the “transient ischaemic attack” recommendation and the “brain imaging and thrombolysis” recommendation in table 1 of the report on page 13 that together at full implementation would cost £16.2 and £8.9 million a year respectively, giving a total of £25.1 million. The fast implementation choice for the strategy outlined on page 18 indicates that in year 1 the required funding would be 25% of the final amount, in year 2 it would be 43%, and in year 3 further increased to 60%. The total cost for funding of these two recommendations for the first three years of implementation of the Strategy would be £25.1 x (0.25+0.43+0.6) million which is £32,128 million.

It would be interesting to see how the Department justifies this pattern of spending as meaning “divided to support the whole pathway” or how it is going to achieve “equitable development of the entire stroke pathway” if additional central funding is not being made available for stroke unit care or early supported discharge schemes as appears to be the case.

Is history about to repeat itself?
As Damien Jenkinson and Roger Boyle stated, this is a tremendous opportunity to develop services for people with stroke and at risk of stroke. All geriatricians are likely to be very supportive of an evidence based approach that is compatible with the Prime Minister’s January 2008 guarantee.^23,24 However, no amount of service redesign and collaborative working will overcome a deficit in central funding and there is a real danger here that the raised expectations of both the public and those working in stroke services may once again not be fully met, as happened following the Older People NSF in 2001.

If the Department of Health is not just interested in allaying concerns but is serious about developing stroke services to match what it perceives to be the best in Europe and elsewhere in the world, the Department’s ministers need to look again very closely and very soon at how this policy is assisted in its implementation through extra central funding for all aspects of the pathway. This should not become another missed opportunity through lack of appropriate central funding for stroke services.

Nigel Dudley
Consultant in Elderly /Stroke Medicine
St James’s University Hospital, Leeds

References

1 Jenkinson D, Boyle R. Implementing the stroke strategy. BGS Newsletter, Issue 16, May 2008: 1 – 3
2 Department of Health (2001). National Service Framework for Older People. Department of Health
3 Department of Health (2007). National Stroke Strategy. Department of Health
4 Department of Health (2006). Mending hearts and brains. Department of Health
5 Gulland A. Hospitals struggle to cope with stroke care, says royal college. BMJ 2002;325:179
6 Shannon C. Stroke services have improved but remain overstretched. BMJ 2004;329:476
7 BBC News. Elderly to be prioritised in reforms. Thursday, 27 July 2000. http://news.bbc.co.uk/2/hi/in_depth/health/2000/nhs_reform/830309.stm (accessed 13/06/2008)
8 Short R. UK government to spend £105m extra on stroke services. BMJ 2007;335:1231
9 Prime Minister’s speech on the National Health Service, 7 January 2008.
10 Ham C. Gordon Brown’s agenda for the NHS. BMJ 336:53 – 4
11 Muir Gray JA. Evidence based policy making. BMJ 2004;329:988 – 9
12 Grieve R, Hutton J, Bhalla A et al. A Comparison of Costs and Survival of Hospital-Admitted Stroke Patients Across Europe. Stroke 2001;32:1684 – 1691
13 Weir NU, Sandercock PAG, Lewis SC et al. Variations Between Countries in Outcome After Stroke in the International Stroke Trial (IST). Stroke 2001;32:1370 – 1377
14 Gray LJ, Sprigg N, Bath PMW et al. Significant variation in mortality and functional outcome after acute ischaemic stroke between western countries: data from the tinzaparin in acute ischaemic stroke trial (TAIST). J Neurol Neurosurg Psychiatry 2006;77:327 – 333
15 Dippel DWJ. National variations in mortality and functional outcome: should we be worried? J Neurol Neurosurg Psychiatry 2006;77:288
16 National Audit Office (2007). Joining Forces to Deliver Improved Stroke Care. NAO
17 Prime Minister’s Iraq statement to Parliament, 24 September 2002.
www.number-10.gov.uk/output/Page1727.asp (accessed 24/4/2008)
18 Manufacturer Submission. www.nice.org.uk/nicemedia/pdf/StokeAteplFADMS.pdf (accessed 16/06/2008)
19 Strokes. Written Hansard, 30 October 2007: Column 1248W
20 Department of Health (2007). Impact Assessment. A new ambition for stroke. Department of Health
21 Young J, Forster A. Rehabilitation after stroke. BMJ 2007;334:86 - 90
22 National Audit Office (2005). Reducing Brain Damage: faster access to better stroke care. NAO
23 Craig GM. Involving users in developing health services. BMJ 2008;336:286 – 7
24 Rodgers H, Thomson R. Functional status and long term outcome of stroke. BMJ 2008;336:337 – 8

BGS Newsletter, Aug 2008
Issue 17 ISSN 1748-6343 17