The case against the Kent and Medway stroke review plans in which 3 HASUs (Hyper Acute Stroke Units) will serve the whole region (plus border areas).
Save Our NHS in Kent About Save Our NHS in Kent (SONIK)
18 February 2019
Save Our NHS in Kent is a community group that campaigns against any NHS changes it deems not to be in the public interest. We are affiliated to the national campaign groups Health Campaigns Together and Keep Our NHS Public. We have undertaken extensive research into the Kent and Medway stroke plans and found them to be highly problematic, both in the procedures used to reach decisions and in the plans themselves. This document will focus mainly on the difficulties inherent in the plan, were it to be implemented. Save Our NHS in Kent is of the opinion that most of the problems raised would be solved by shifting to a four HASU plan. We have designed this document to be as concise as possible, with more detailed information located in separate documents or online sources - these are signposted for your convenience. NB: This version of the report (version 2) contains changes to point A.5, to account for the recent publication of a new follow-up study on HASUs in London and Manchester.
Section A - Evidence of Lives Saved (London).
- The London HASU study, which is the sole source of evidence underpinning the idea that plans to centralise existing units into fewer HASUs will ‘save lives’, did NOT record deaths in ambulances as part of the study.1
- The London study showed an uplift in the number of lives saved prior to the period before centralised HASUs were introduced. However, the London hospitals were starting from a low base, as they were at that time underperforming compared with other areas.2
- In the London study, changes to dependency and disability outcomes were NOT MEASURED.3
- The London study was ‘pump-primed’, that is to say, financial incentives were provided to the hospitals involved to meet certain metrics, such as ‘door to needle time’, and ‘length of stay’. In Manchester, which was supposed to be the control area, these financial incentives did not apply.
6. It has been claimed that after the four year period of the London study was over, wards became overcrowded.4 A follow-up study, very recently published, found that the higher mortality rate and length of stay metrics were maintained in London after the study ended, but it still did not record deaths in ambulances or disability outcomes, the uplift was still very small compared with the population served, and over-capacity wards were not taken into account in the study either.5 The new study reiterates explicitly that the approach cannot be expected to work in non-urban areas: “The greater travel times in rural areas make centralisation challenging and may necessitate other solutions” and that it could negatively affect treatment in other hospitals in the area: “relatively little evidence exists to show whether centralising acute stroke care to a small number of high volume specialist centres produces better clinical outcomes at the system level across all stroke patients”. Peer reviewers have suggested, as they did at the time of the original study, that any improvements seen in London or Manchester are down to the reorganisation of care in the HASU stroke units, and are not due to centralisation.
7. PatientsintheLondonareawhereHASUswereimplementedbenefittedfromanupper limit journey time of 30 minutes, with most reaching a HASU within 16 minutes. This study is simply not applicable to Kent with its geographical spread; the centralisation of services proposed in Kent and Medway means an hour long journey for many.3 Therefore, the London study (“Impact of Centralising Acute Stroke Services in English Metropolitan areas on Mortality and Length of Stay: Difference-in Difference analysis” 2010) does not prove that there will be ‘lives saved’ or that there will be any improvement to death and disability outcomes in Kent and Medway.
Section B - Evidence Regarding Mortality Increase sand Health Disbenefits(Manchester, Northumbria, California etc).
- Other studies show centralisation hasn’t improved death and disability outcomes: The Northumbria study (2018) showed that where three stroke units were centralised into one HASU (at a cost of £95m), death and disability outcomes did not change at all. The only improvements to speed of treatment were to in-hospital timings; call-to- treatment time was not recorded in the study. There was no improvement shown to death and disability times despite the disruption and cost. There was no fall in death and disability measures either, presumably due to the fact that, unusually, there was no change to journey times for the population served (the three stroke units had been in three hospitals located within a few miles of each other in Newcastle).
- The Manchester trial, which was part of the London study showed no improvement in mortality. Disability outcomes were not measured.
- A study into the closure of A&E departments in England took place between 2009-11. It was designed to measure changes to patient outcomes as a result of improved care in fewer locations. The A&E reconfiguration came with the same promise that has been made in regard to the national stroke plans, ie that longer journeys are ‘offset’ by better quality treatment. However, the findings showed that no improvement to death and disability outcomes could be demonstrated, despite all the disruption and extra cost. An increase in the proportion of patients with an emergency condition who died went up by 2.3%.
- International studies show that being further from emergency care correlates greatly with increases in mortality rates for those with emergency, ie time sensitive, conditions. A California study from 2014 found that for patients with a greatly increased drive time (average 47 mins) the mortality rate shot up by 21%. Those with decreased journey times had a 17% lower chance of dying. (Srebotnjak et al).
- After a hospital or ED closure, patients suffering a myocardial infarction (MI), unintentional injury, stroke or sepsis have a greater risk of death if there is an increased distance to a hospital.
- Nicholl et al. found that, for serious emergencies, particularly respiratory problems, a longer journey distance to hospital appeared to be associated with an increased risk of mortality.
- The American Heart Association policy recommends that patients not bypass a closerPSC (primary stroke centre) in favour of an CSC (comprehensive stroke centre) ifsuch a diversion would add more than 15–20min of transport time. In other words,getting assessed and stabilised at a centre without top-level specialisms is preferable where additional journey times exceed 15-20 minutes. [“Mechanical Thrombectomy—A Brief Review of a Revolutionary new Treatment for Thromboembolic Stroke” (Pervinder Bhogal et al). [https:// doi.org/10.1007/s00062-018-0692-2]
- A BMJ Systemic Review concluded, in general terms about all health care: “In the debate between local versus centralised healthcare provision, 77% of the included studies showed evidence of an association between worse health outcomes the further a patient lived from the healthcare facilities they needed to attend.This was evident at all levels of geography—local level, interurban and intercountry level. A distance decay effect cannot be ruled out, and distance/travel time should be a consideration when configuring the locations of healthcare facilities and treatment options for patients.” [Kelly C, Hulme C, Farragher T, et al. “Are differences in travel time or distance to healthcare for adults in global north countries associated with an impact on health outcomes? A systematic review.” BMJ Open 2016;6:e013059. [doi: 10.1136/bmjopen-2016- 013059]
- Another study regarding centralisation of stroke care is expected to be completed in 2019; its findings ought to be taken into account before rushing ahead with the Kent & Medway Stroke Plan implementation. It is the Barcelona RACECAT trial which is an assessment of whether primary assessment and treatment at a local centre vs. direct transfer to a specialised centre is preferable in the treatment of stroke.
Section C - Impacts as reported on by third parties (IIA, KHPO)
- The initial Independent Impact Assessment (Mott MacDonald) commissioned to form part of the PCBC (Pre Consultation Business Case) across five separate assessment points showed a negative impact for three of the data points, and neutral/no likely impact/improvement for the other two.
- The Kent Public Health Observatory produced an ‘Evidence Review’, again this was included in the PCBC documents. Their findings revealed a lack of appropriate evidence and concluded that money might be better spent on prevention.
- A second Independent Impact Assessment was produced at the time of the DMBC (Decision Making Business Case) in Sept 2018; it found 4 positive impacts and 8 negative impacts. The assessment warned that ‘capacity could become constrained’, which could have ‘a negative impact on the responsiveness, safety and quality of patient care’. It also found that only 61.8% of those from the most deprived quintile will be able to access stroke services within 30 minutes by blue lighted ambulance, compared to 69.6% of the population; access within this timeframe for the most deprived quintile would actually drop by 23% if the stroke proposals are implemented. Other factors are raised such as concerns over SECamb’s ability to cope; the likelihood of increased staff turnover and loss of current expertise; a reduction in patient choice; and travel times.
Section D - Parties objecting to the Plans
- Thanet District Council is ‘vehemently opposed’ to the stroke plans and the relocation of QEQM hospital’s stroke unit to Ashford, which is a one hour journey away. Medway Council is mounting a judicial review. MPs Craig MacKinlay, Rehman Chishti and Kelly Tolhurst have publicly criticised the plans. Cllrs Lesley Game, Barry Lewis, Karen Constantine, Ros Binks, Emma Dawson, Ida Linfield, David Wildey, Wendy Purdey, Teresa Murray, David Royle, Sam Bambridge, Michelle Fenner, Iris Johnston, Peter Campbell, Ian Venables, Carol Messenger, Trevor Shonk, Jenny Matterface and many more have voiced opposition to the plans.
- There has been a lack of supporting voices outside of the Clinical Senate (which is arguably biased as Dr Hargroves acted as chair whilst also being the Lead Clinician for the plans) and a small number of consultants and GPs. It has been claimed in public meetings that all the staff at QEQM hospital are happy with the changes, whilst SONIK knows from private conversations that this is far from true.
- In the public consultation meetings in Thanet, the audience was on every occasion very much opposed to the plans; the only voices in favour were the panel and occasionally doctors from out of area that appeared to be planted in the audience to back up the panel’s arguments. A show of hands was conducted at a few of the meetings, revealing in each case that the consultees were not at all in agreement with the panel. The public consultation report showed incredibly strong opposition to the longer journey times that would result from the implementation of this plan, with 66% of respondents to a telephone survey stating that they were concerned about this part of the plan (across the whole area). In Thanet that number rises to 78%. The findings from the phone surveys stated that there was ‘one main and repeatedly noted concern - the length of travel time to HASUs’. The report also showed that 25% of respondents were unhappy with how the consultation had been conducted. [“Public consultation on proposed changes to urgent stroke services, Research analysis report” Summer 2018, DJS]
- The consultation of staff members at the hospitals involved also showed strong objections, with nurses in particular raising numerous concerns, including the inability to continue as a specialist stroke nurse due to the impossible travel times that the reconfiguration of services would impose. Doctors and consultants weren’t included in the consultation.
- Dr Tony Martin, (previously Chair of Thanet CCG and voting member of the JCCCG for the Kent and Medway Stroke Review) conceded in a public meeting that the current evidence claiming that lives will be saved is not sufficient: He said: “We haven’t seen hard evidence relating to areas similar to Thanet.” [https://www.facebook.com/SaveOurNHSKent/ videos/1017914045014841/] He subsequently retired in May 2018.
- Campaign groups Save Our NHS in Kent, Southend Save Our NHS, 999 Call for the NHS and Dorset Defend Our NHS as well as national group Keep Our NHS Public have all expressed serious concerns over the plans to centralise stroke care nationally into HASUs, and the many closures that will result.
Section E - the case against having only three HASUs for Kent and Medway/Specialist staff shortages/ recruitment and retention
- Appendix M in the consultation supporting materials (PCBC) sets out the reasons for refusal of a four HASU option; namely that the number of consultants required would be too great. The document emphasises that to have a fourth HASU in operation would require “recruiting 14 additional stroke consultants to Kent and Medway” without mentioning the fact that recruitment of at least eight of those 14 is already necessary to have three HASUs in operation. So, the difference between three and four HASUs is 6 consultants. But then there is a disclaimer that “larger units may require additional consultants”, suggesting that the figure of eight extra consultants needed for the three HASU plan could potentially rise to 11, given that all three HASUs will be large. If it is likely that the three planned HASUs all need extra consultants, it leaves us with the question as to whether the option of an entire HASU has been removed simply for the lack of potentially only three consultants, when 8-11 new consultants are needed for the basic option of three HASUs.
- There are 2,500 strokes per year in Kent and Medway. That is already more than enough for four HASUs according to the guidelines often quoted by the STP in regard to the need for 600 strokes per year per HASU, or for 5 HASUS if you accept the current guideline of 500. If population growth is taken into account, then the number of strokes per year will rise too (the assumption that the number of strokes will remain flat and not increase in years to come has already been challenged by the JHOSC, resulting in the STP conceding that their calculations would have to be adjusted.)
- The rule of 500/600 stroke patients per year per HASU is itself merely a guideline. Inthe NHS document “Stroke Services: Configuration Decision Support Guide”, which is designed to aid local bodies such as CCGs and STPs in a sensible and effective implementation of the national Stroke Review plan, it states: “It is recognised that guidelines can never provide the answer for every situation and do not replace sound clinical judgement and good common sense. Clinical guidelines are only likely to be applicable to 80% of clinical situations, 80% of the time.” These guidelines are not set in stone, but that is the way they have been presented to us during the consultation period. In the case of Kent and Medway, where public opposition, dangerous journey times, an increasing population and poor road networks are significant contributing factors, then sticking rigidly to a three HASU plan is not advisable. The three HASU plan also necessitates (according to planning so far) the removal of acute stroke services from the communities that will benefit from them the most, ie the deprived communities of Medway and Thanet.
- The 500/600 rule comes from guidance that is itself not based on any specific calculations and is not evidence-based. Although we have been informed that 600 stroke patients per unit, per year is necessary to guarantee the continuing expertise of specialists, the figure of 600 to 1,500 comes from NHS England guidance from 2015, and the lower figure of 500 is suggested by The Royal College of Physicians. No research has been used to reach those numbers; they seem to be arbitrary figures. Preliminary consultation documents show that such figures are often required by insurers to guarantee independent practices; it is a concern that insurance considerations may have helped to shape 500/600 rule over and above safety requirements for all parts of the Kent and Medway population.
- The Clinical Senate document confirms that smaller units can work just as effectively as larger ones, but might affect ‘economies of scale’: “Some units of a smaller size have however demonstrated generally good quality outcomes and so this minimum number of 600 should not be seen as an absolute requirement..we recommend that commissioners and units aim for activity of at least 600 cases per annum to maximise the manpower, training, experience and potential quality gains enabled by larger units, and the likely financial economies of scale.” [p14, 4.1.3, Appendix J, South East Coast Clinical Senate Review of Case for Change.]
- Monitor have also found that there is no relationship between the size of hospitals andhigh level quality of care, and that with more general specialities there is also lessevidence for the benefit of centralisation as whole. “Monitor’s research has found no clear evidence that smaller acute hospitals performed any worse clinically than larger counterparts. The analysis of a variety of clinical measures found no systematic evidence of poorer quality in small hospitals and found only a limited effect of size on financial performance.” [https://www.gov.uk/government/consultations/challenges-facing-small-acute-nhs- hospitals]
- There are not enough specialist staff employed in Kent and Medway at the current time to staff the proposed three HASUs. As mentioned in item E.1, eight (or possibly 11) additional stroke consultants will be needed to implement the three HASU plan. The STP have explained that despite staff shortages nationally, they believe that the presence of the HASUs will be enough to prompt the appropriate clinicians to move to Kent and take up employment here. If that is true, then the presence of four HASUs could strengthen that ability to attract the staff required, especially if only a few extra are required (see point E.1).
- Guidance around how stroke consultants will work in the designated HASU/ASUs is still unclear, in terms of how full use is made of their working hours. Aspects of the PCBC emphasised the need to specialise (hence the 500/600 rule) but other aspects of the PCBC suggested that consultants would also be forced to generalise in order to service the HASU/ASU setup.
- The shortage of consultant staff to cover current services in Kent and Medway hospitals has still not been explicitly demonstrated. SONIK have requested data on this, but have been unsuccessful in attaining it from the STP. We also requested information that would back up the anecdotal evidence given at listening events, but have also not been supplied with it. During listening events, the panel tended to refer to national shortages of consultants and then raise anecdotal examples. No evidence of a regional shortage for stroke consultants was ever provided.
14. It has not been proven that the staffing problem couldn’t be solved with better incentives, and in addition many of the Kent hospitals currently have managed to employ their full quota of stroke staff. Surveys of managers in Kent hospitals on the topic of recruiting staff did show a unanimous conclusion that it is difficult to recruit; but also studies have shown that the area with least recruitment issues in the country is London, with the second least being Kent. British Medical Association data shows fill rates for specialist roles in the NHS broken down by region. In 2015, KSS region (Kent, Surrey and Sussex) is second best after London at filling roles. In 2016, KSS drops down to third place, but there's barely a difference between the top three, which are all close to 95%. [‘BMA Specialty Fill Rates by region’ 2013-16]. 15. SONIK have asked Kent and Medway STP about their efforts to recruit staff, and were informed that roles had been advertised extensively. However it appeared to be standard practice advertising of roles in the usual places, nothing that would constitute a special effort. A website was created in 2018 to promote east Kent as an area to live and work in for NHS staff. Although it is a pleasant website with positive staff testimonials, it does appear to merely be paying lip service to the matter. GP numbers are especially low in certain parts of Kent, and if (as we have been informed) stroke specialist recruitment is another area that is struggling to recruit to the extent that services must be cut, then why have the recommendations of the BMA not been adopted? A report (“The State of Medical Recruitment”, BMA, 2017) suggests that working conditions are the big factor that causes trained doctors to leave the NHS, and would be a big factor in persuading them to return. Suggestions include more affordable childcare that is flexible and available to cover out of hours work; shared parental and carers leave; deferred training posts; improvements to shift patterns; e-rostering; more flexibility for part time roles, and many other suggestions. Providing subsidised, convenient, out of hours childcare for the families of much-needed specialist staff could attract the consultants, therapists and nurses that we need, and would be considerably cheaper to provide that the costs of this reconfiguration; so why hasn’t it been trialled? The idea of cash incentives for medical professionals similar to ‘London weighting’ have been suggested to the STP, but we were told ‘that isn’t done’ and that it would be too expensive. Alternatives to the ‘one size fits all’ national plan are either unwelcome or simply haven’t been considered. 16. The Impact Assessment report from Sept 2018 concluded that the reconfiguration could result in the loss of specialist stroke staff. See point C.3. 17. It has been argued that the presence of a HASU will attract the best staff to the unit and to that hospital, bringing more staff to the area and counteracting the alleged recruitment problems that Kent hospital trusts face. But what if the staff simply migrate from other Kent hospitals.
18. Whether or not the additional staff numbers required for the three HASU plan in Kent are achievable has been questioned at numerous points in the process. The large numbers of additional staff promised helps to promote this plan as a major improvement to services; but what if the staff increases are simply unrealistic and the region ends up with a smaller number of units, but no, or very few additional staff? Then Kent and Medway will, in effect, have agreed to a reduction in service without the benefits. That is not what CCG leaders or councillors believe they are agreeing to. 19. Recruitment difficulties for specialised staff in the Kent and Medway catchment area has been heavily relied on in terms of the argument for three HASUs and no more. The only data publicly available regarding recruitment of specialists by region comes from the British Medical Association, which does appear to tell a very different story. The ‘fill rates’ for Kent, Surrey and Sussex are much better than in all other areas of the country except London. Fill rates refers to the percentage of advertised positions being filled, ie the rate of successful hires of new members of staff. The ‘fill rates for Neurology in KSS were 100% for all four years shown. The fill rates for the geriatric medicine category were: 2013, 85%; 2015, 85%; 2015, 100% and 2016 also 100%. According to this data, no attempts were made to hire in the stroke medicine category, but this may be due to stroke staff being categorised under the ‘geriatric’ category. These fill rates refer only to specialist staff, and the KSS area fill rates are quite respectable. The national fill rates for stroke, geriatric medicine and neurology also aren’t as bad as we have been led to believe. Dr David Hargroves, the Stroke Lead for the Stroke Review in this area claimed at listening events that “40% of stroke consultant vacancies are left unfilled”; SONIK asked the STP to provide a source for this data, and we were referred to the SSNAP data in the PCBC, which merely states that 40% of sites have ‘any unfilled stroke consultant posts’. This will be a much lower figure in terms of overall unfilled posts; it tells us that 60% of sites have 100% of staffrequired, and the remaining 40% of sites have a shortage of some extent; it may relate to small shortages such as part-times roles or a fraction of a WTE (whole time equivalent) for all we know, therefore reliance on this data from SSNAP is problematic. The BMA data on unfilled posts shows fill data both as a percentage and in terms of actual posts, and it shows the following: fill rates nationally for Neurology were 100%, 95%, 95%, 95%. The fill rate for the geriatric medicine category were: 2013, 79%; 2014, 75%; 2015, 90% and 2016, 83%. The stroke medicine category showed no posts advertised for all years apart from 2014 where the fill rate was 40%; however this statistic has a very low number of actual posts behind it - 5 were advertised, and 2 filled. Again, it is likely that the true numbers for stroke recruitment have been categorised under geriatric medicine. Whilst all these figures show that there is a shortage of specialists in stroke medicine, they do not demonstrate a shortage as dramatic as claimed by Dr Hargroves. [‘Specialty Fill rates by Region 2013-16’, BMA]. There is no valid argument as to why there can’t be four or more HASUs. The recruitment argument has not been proven, and is an unconstitutional argument that puts claims of finite resources over patient need. The 500/600 rules has no basis in evidence. Practical solutions to staffing shortages haven’t been attempted. The extent of national and local shortages in stroke medicine has not been sufficiently established, and may have been exaggerated.
Section F - Journey Times
- Journey times for patients in Thanet (currently 141,000 people, using a conservative estimate) will increase by approximately 566%. The majority of residents in this densely populated area (the 2nd most densely populated in Kent after dartford) live in the conurbations of Margate, Cliftonville, Ramsgate and Broadstairs. These are the current distances from QEQM (Queen Elizabeth the Queen Mother Hospital): Margate (5 mins), Cliftonville (9 mins), Ramsgate (12 mins) and Broadstairs (11 minutes). If this is averaged to 9 minutes, as a crude estimate of the average current journey time to hospital in an ambulance, then the increase to one hour amounts to a 51 minutes additional journey time, which is a 566% increase. This will put Thanet residents at clear risk of the dangers described and backed up by the evidence outlined in points B.3 -8.
- Medway residents living in more densely populated areas will see an increase from an average of 9 minutes to 30 minutes to their next nearest hospital where a HASU will be located, ie Maidstone hospital. Whilst this will be a journey that remains within a safe travel time of 30 minutes, it will result in a deprived area with a high incidence of strokes losing its acute care, and could mean the gradual loss of services and downgrading of a hospital that serves one of our poorer communities as specified in Section M. Tunbridge Wells residents will be experiencing an increase from 9 minutes to 34 minutes. This could put residents in Medway and Tunbridge Wells at risk from ‘distance decay’ as outlined in point B.8.
- The argument that “closer isn’t always better” or that “a shorter journey to a hospital without a hyper acute stroke unit can be worse for stroke patients than a longer journey to a hyper acute stroke unit” is one which can only be heeded if a) journey are not too much longer and remain within a safe limit and b) the improvements promised at a hyper acute unit are in fact substantial, evidence-based and relate to actual patient outcomes. See sections A and B for considerable evidence as to why this isn’t the case for the Kent and Medway Stroke Review.
- The “closer isn’t always better” argument also relies on offsetting of outcomes. The argument is that the detrimental effects on some patients due to longer journeys (much longer, in the case of Thanet) will be offset by the improvements in the HASU. If we want a high quality national health service fit to take us into the future and compete with health services across the globe, then shouldn’t we be aiming for a stroke service that doesn’t have to minimise the benefits of an improved service to patients due to the detrimental effect of longer journey times. Is this why the centralisations tested so far seem to struggle to show an uplift in death and disability outcomes? (see B1-3)
- In some cases, stroke patients are taken to hospitals such as Kings College Hospital in London for stroke treatment. This has been used in some cases to justify the “closer isn’t always better” argument; it should be noted that patients being taken to London are either extremely severe cases that are taken by air ambulance, or they are cases where the patient has first been assessed and stabilised at a local hospital before being transferred to london for specialist treatment. This is very different from the Kent & Medway stroke plan whose aim is to take ALL suspected stroke patients to a further away hospital in the first instance, with no primary assessment or stabilisation at the nearest hospital.
- SECamb’s Senior Operations Manager James Pavey has stated that the one hour ambulance journey from Thanet to William Harvey Hospital in Ashford is not a worry due to the fact that a two hour call to needle time has been deemed safe in recent (by the senate?). SONIK’s calculations reveal that call to needle time for some patients will inevitably exceed the two hour window quoted by Mr Pavey and the consultation documents, due to ambulance response times and loading. Strokes are categorised by SECamb as ‘category 2’, meaning that ambulances will arrive in 18 minutes on average, but the service anticipates that for 10% of patients the wait could be over 40 minutes. If we assume a worst case scenario where the response time is 40 minutes +15 minutes loading the patient + one hour in the ambulance, then the call to hospital door time is 1 hour 55 minutes. Some patients, the SECamb target states, could be MORE than 40 minutes. The aim is for all stroke patients to be treated within 30 minutes of admission to the hospital. This takes the call to treatment time up to a potential 2 hours and 25 minutes, and that is not taking into account those cases where the ambulance response exceed 40 minutes. There have been multiple reports in recent years to show that SECamb is not coping well. It should be the door to needle time of 30 minutes and the 2 hour call to treatment time are both ‘aspirations’; they are not upper limits or even targets that must be hit. SONIK asked what the consequences would be for the trusts if these aspirations are not fulfilled - there will be none. [http://www.secamb.nhs.uk/about_us/our_performance/response_time_targets.aspx]
- SONIK have asked the STP repeatedly to provide examples of any regions in the UK where HASUs have been implemented in such a way that large populations are left outside of the 45 minute zone. Nothing has been supplied to us. Looking at areas such as Northumbria, we see a HASU implementation where some residents are left with journey times above 45 minutes, but it is a tiny proportion of the population served, and those residents (living in very remote areas) had the same long journey times prior to the reconfiguration. Having no evidence to the contrary, SONIK has concluded that nowhere in England has a HASU reconfiguration taken place that leaves such a significant proportion of a population outside the 45 minute zone in terms of travel for emergency stroke care.
Section G - Facts about the centralisation so far
- Some trauma and some heart treatments (pPCI) have already been centralised without consultation in east Kent, where patients are now taken to William Harvey rather than QEQM or the K&C in Canterbury. It is claimed that this has been successful, but no data has ever been publicised to show evidence of improved patient outcomes since the relocation of these services occurred.
- HASUs are not yet prevalent, as has been implied and even claimed explicitly in some listening events and news reports. Glenn Douglas recently expressed his ‘embarrassment’ that Kent and Medway does not yet have a HASU, implying that we are falling behind the rest of the country (Stroke JHOSC meeting, Feb 1st 2019). This misleading device was also deployed in a number of listening events, where it was claimed that 50% of regions had adopted the HASU model, and that Kent was ‘behind the curve’. Concerned about the misrepresentation of the facts, SONIK wrote to Patricia Davies (Senior Responsible Officer for the Stroke Review at that time) and she confirmed in an email that the figure of 50% was in fact erroneous and should not have been used. Only a few areas have switched to HASUs so far, many of which are newly implemented. The HASU plan is in early stages of rollout (in terms of actual delivery of functioning centres), and the data on patient outcomes (as opposed to in- hospital timing metrics and length of stay) have only come from London, Manchester and Northumbria. It is wrong to imply that Kent is falling behind or that the nation plan to implement centralisation along the lines of the HASU model is at an advanced stage.
- Nowhere in England has a HASU reconfiguration taken place that leaves such a significant proportion of a population outside the 45 minute zone (in terms of travel times to a HASU) as in the Kent & Medway plans (see F.7).
Section H - Loss of 23 Beds by 2021There are concerns that the number of beds in the Kent area for stroke will decrease as a result of the plans. As a result of this reconfiguration we are due to see a reduction in available beds for stroke from 154 to 129 by 2021. This is in spite of a stated panel concern indicating that they do not believe, as has been stated, that stroke rates and prevalence will remain static despite a growing and ageing population, as has been argued during the consultation. The DMBC proposes to tackle this issue only through "prevention"; encouraging individuals not to smoke, or alter their lifestyle. The beds restructuring has been pinned entirely on the supposed power of preventative factors, in spite of the difficulty of successfully implementing such initiatives in areas of deprivation and the likelihood of an increasingly ageing and expanding population. The definition of beds is deliberately opaque. The loss of two beds by 2021 is made clear; but as there is currently "access" to 152 beds, and by 2021 there will be 129, this does amount to a loss of 23 beds, despite the ‘ringfencing’ arguments put forward by the STP.[Kent & Medway Stroke DMBC pages 18,131,133].
Section I - Thrombectomy
- Introducing the advanced prodecure Mechanical Thrombectomy to Kent has been given as one of the reasons to proceed with the plans. However, it must be noted that in the consultation materials, Thrombectomy is not a guaranteed part of the plans; it is a potential extra that may or may not happen.
- The number of patients eligible for Thrombectomy is very small.
- Increasing the number of eligible patients that can access Thrombectomy is widelyseen to be beneficial, and should be aimed for. However, it is not necessary to centralise Kent’s stroke units to do so. Stroke patients eligible for Thrombectomy can be stabilised at their nearest acute stroke unit with thrombolysis (clot-busing drugs), and once that happens the patient can be transported to a specialist centre where Thrombectomy can be done; once thrombolysed, there is a six hour window during which the patient remains eligible for Thrombectomy, which is ample time for a patient from anywhere in Kent to reach London, where it is currently offered. [“Mechanical Thrombectomy—A Brief Review of a Revolutionary new Treatment for Thromboembolic Stroke” (Pervinder Bhogal et al). [https://doi.org/10.1007/s00062-018-0692-2]
Section J - Rehabilitation
- Adequate rehabilitation will be essential to good patient flow in the new reconfigured service. It is also constitutes a large part of NHS stroke care overall, and has been identified as one of the areas most in need to improvement, with stroke survivors describing being left without adequate care following a stroke. It has been a surprise to many, including the members of the Stroke JHOSC, that it has been left until such a late stage to finalise rehabilitation plans which ought to have been at the core of planning and also ought to have been included in the public consultation. The most recent version of the DMBC dated January 19 states “Work has therefore taken place to develop plans for comprehensive and equitable community rehabilitation services, which will be delivered locally and will support the implementation of HASUs. This is being progressed by a Rehabilitation Working Group and will be led by a clinical lead currently being identified. The group report to the Stroke Clinical Reference Group, as shown in Section 1.3.2. This group includes a range of people from across health and social care plus patient representatives. It is expected that a rehabilitation business case will be presented to CCGs in spring 2019 to ensure standardisation of provision across the K&M.” The public were asked to comment on the plans without rehabilitation details, and the Stroke Joint Committee of Clinical Commissioning Groups voted in favour of the pans without details of the rehabilitation element being finished.
- The Stroke Association report “The State of the Nation” makes clear that strokes are expected to continue rising until 2035, that strokes that happen earlier in life tend to be more severe, and that patients who suffer stroke earlier in life are more at risk of further strokes. The rate of first time stroke is due to rise by 59% over the next 20 years. Survival rates and therefore rehabilitation needs are due to rise by 123% in the same period. Given the deprived areas of Thanet and Medway will have a higher number of younger patients suffering a stroke and will also have considerably reduced access to reaching a stroke unit in a timely way, there is a serious concern that the rehabilitation plans will not be sufficient to cope with the numbers required, and that some of the most unwell patients will be put under the care of rehab units that are far removed from all of the HASU/ASU units where all the specialists will be located.
- The DMBC mentions that rehabilitation is already provided in the area by various bodies, including Virgin Health (p19). This raises concerns expressed by SONIK last year that the stroke plans may result in far more need for stroke out of hospital care, and that it might be provided by private companies whose record in delivering quality healthcare to date has been poor. We are concerned that these plans may be producing a business opportunity for private providers at the expense of good quality patient outcomes. Private providers will not be subject to the same accountability and transparency rules, for example, they will not be obliged to respond to FOI requests about the services they provide. In addition, social care is means tested, whereas NHS care is not. This increase in out of hospital care for stroke survivors could result in considerable costs being borne by the stroke patients and their families.
- Another concern is whether a rehabilitation ward will be situated at the QEQM in Thanet or not. It was widely covered in the press and a Thanet MP also spoke of rehabilitation ward at QEQM back in September/November 2018. However, the latest version of the DMBC does not confirm any locations, and suggests that rehab wards might be best located next to HASU/ASUs so that staff can be shared. If the rehab plan has changed, the STP ought to be making that completely clear, as a key concern for Thanet residents was the long journey times and expensive cost of journey to WHH hospital from Thanet for those wanting to visit relatives.
Section K - Reports of poor performance in hospitals currentlyThe criticisms of Kent’s existing stroke units come from SSNAP audit data. These are presented in colourful tables in the PCBC, and in particular in one table used for presentation to the public, which is Figure 7 from Appendix M, “Performance against Targets”. In this table, all of Kent’s District General Hospitals are shown with their % performance against 12 selected standards. The table is coloured in red green and amber, but the majority of fields are coloured red indicating what appears to be huge failings in Kent hospitals. 1. The ‘standards’ the hospitals are being measured by are guidelines designed to improve care; while these may be important, they are not patient outcome statistics. At no point have the mortality rates or disability outcomes of patients in these stroke units been disclosed to us. Should we not be looking at that data as well, and at improvements in mortality and disability outcomes over time in each unit? It might be more beneficial to see data that recorded severity level of stroke at admission and then measured that against outcomes in order to assess the performance and safety of a given stroke unit.
- Should we be closing units based on their inability to meet standards? Applyingstandards and setting targets can be beneficial in terms of identifying areas for improvement. The table may show a case for making improvements; should it be used to justify closures? The STP argue that improvements can only be made by pooling resources as staff are ‘spread too thinly’. Please see section E for our arguments against that reasoning.
- There are 21 SSNAP ‘standards’ which stroke units nationally are measured against. This table (Figure 7) has picked just 12 of them.
- The discussion has been reframed away from how well patients are being treated and what their outcomes are, to a discussion purely about performance against targets. Anyone who has ever worked in a target-driven environment will know that targets can improve performance, but can also encourage reprioritisation of objectives in a way that may not be beneficial, and can sometimes give a false picture of performance against the underlying, crucial objective - which would be in this case to save lives and prevent as much impairment as possible.
- How much do we really know about what lies behind the score for each of these standards? “Direct Admission” is a measure of ‘patients admitted directly onto a specialist stroke unit within four hours’. Every hospital in Kent is below the national standard. The national standard itself is only 57%. this was presented in listening events as a dreadful scenario where stroke patients are left languishing in A&E, ‘sometimes for four hours’. We know that the A&E department at QEQM has a separate room for stroke patients containing 3-4 beds; stroke patients are sometimes taken there and treated before being taken to the acute stroke ward, which could account for the the 49% figure against this measure for that hospital. It should be noted that stroke consultants and stroke nurses see patients in that room, and thrombolysis is administered there. This is quite a different story to that presented by Kent and Medway STP. We must ask ourselves if the other measures in this table are not quite what they seem; and if they constitute good reasons for closing the service down.
- As mentioned above, “Direct Admission” is only 57% nationally. All of kent’s hospitals are lower than this, and consequently marked in red.
Section L - Pressure on ambulance service1. During the consultation, the STP stated that that an extra £1m per year will be allocated to SECamb so they can handle the additional stress on the ambulance service as a result of so many additional journeys. SONIK asked SECamb about their annual running costs, which were £211 million for the year 2017/18. By our calculations, the budget increase for SECamb for the area affected by the Stroke reconfiguration is a tiny 0.98% increase. It is not clear if this is in addition to ordinary annual increases to budget or not. When the DMBC was published, it emerged that the annual increase had dropped by half to £500,000 per year (a budget increase of 0.47%). It is hard to see how this can be anything close to adequate.7
- The Independent Impact Assessment that was commissioned as part of the DMBCnoted that resource for longer journeys plus an increased number of transfers for patients already being treated in hospitals, remarking that “this will have a negative impact on the capacity of the ambulance service in terms of ambulance and paramedic resources”. As the ‘mitigations’ designed to counteract the capacity problems were not outlined prior to the impact assessment, it could not judge whether they were sufficient or not.
- SECamb is already overstretched and is still is special measures despite a big increase in funding.
- SECamb was rated as ‘requiring improvement’ in November 2018, improving on its previous ‘Inadequate’ score.
- In order to cope, SECamb outsources much of its 999 responses to private operators, and has also resorted to buying second hand ambulances.8
- In 2017, it was reported that SECamb were consistently failing to meet the national target for 999 call-outs. It was named as the worst performing ambulance service inresponding to 999 call-outs. [https://www.itv.com/news/meridian/story/2017-07-20/exclusive-report-on-worst-performing-ambulance-service/]
- SECamb has a deficit of £7.1 million. [Chief Executive AMM Presentation 2017, SECamb, page 3]
Section M - The long term affect on non-HASU hospitalsHospital departments and whole hospitals can be closed down if they are deemed ‘unsafe’. If recruitment of the right staff becomes too difficult due to constant undermining of a given hospital, through downgrades, closures and the constant transferral of patients to other hospitals, then that hospital begins to be at severe risk of closure without consultation. The hospitals that appear to be at most risk of eventual closure in the Kent and Medway area are: • QEQM in Margate (410 beds) • K&C in Canterbury (287 beds) • WH in Ashford (476 beds) (with K&C and WH in direct competition with each other)• Medway Maritime Hospital in Gillingham (588 beds) • MTW in Tunbridge Wells (512 beds) We heard Dr Mike Beckett tell us at the Minster listening event that ‘the future of healthcare is not in District General Hospitals’. Many NHS England planning documents including the Keogh report, which forms the basis of the current review of acute care, argue that DGHs are a thing of the past, and should be replaced with fewer ‘acute
hospitals’. Removing stroke services is the beginning of that process. [Review into the quality of care and treatment provided by 14 hospital trusts in England: overview report, Professor Sir Bruce Keogh,16 July 2013]. In an article entitled ‘Is it the End for District General Hospitals?’ from 2007, a description is given for what would need to happen before the overall number of hospitals in an are could be reduced. It reads like a blueprint for what has been happening in parts of Kent and Medway: "If the DGH is under threat, most vulnerable will be the smaller ones unable to provide a full range of services and without tertiary provision such as in cardiac care and neurosurgery....If we are to see the demise of smaller DGHs there will have to be either a large reduction in unscheduled admissions, significant reduction in lengths of stay or more alternative providers keen to enter the market....There have been inexorable reductions in length of stay since the inception of the NHS and these have been more significant in recent years. They will get shorter still through better integration of local health and social services, but is there any prospect of further large-scale opportunities? New technologies like emergency percutaneous cardiac intervention or thrombolysis for stroke may drive care to larger DGHs and tertiary centres but patients suitable for these procedures are likely to remain only a small fraction, less than 10 per cent of myocardial infarcts and strokes. And these tertiary centres will be looking for quick transfers back to the local hospital for those not ready for home, while there is not yet much appetite among private providers for less-predictable parts of the market that make up many of the average DGH admissions.” [https://www.kingsfund.org.uk/publications/articles/it-end-district- general-hospitals] Anyone who has been paying attention the the east Kent healthcare plans and the Kent and Medway Stroke plans will recognise that the elements listed above are exactly what the current STP plans involve. The removal of pPCI (percutaneous cardiac intervention) has already been removed from QEQM and Kent and Canterbury Hospitals, signalling that they are hospitals at higher risk of eventual closure. William Harvey hospital is one of the larger hospitals in Kent in terms of beds, and has been selected as a site for a HASU, so might seem to be lees at risk of cuts, closures or eventual closure; however, the Medway Maritime Hospital which has the largest number of beds in the area, and offers pPCI and a range of specialisms, was recently announced as not being selected as one of the sites for a HASU. The news shocked many who had seen the hospitals as one of the most obvious choices. This goes to show that in the quest to centralise care, any hospital may find itself suddenly consigned to the second league and potentially be allowed to deteriorate thereafter. The diminution of the K&C hospital in Canterbury over many years is a perfect example of how services can be removed little by little, resulting in a hospital that offers almost no acute care at all, no maternity, and is shrinking whilst the population is growing. One cut leads to another, the ability to recruit is affected, and more services are removed. In June 2017, junior doctors at the K&C were removed and transferred to the William Harvey Hospital in Ashford and the QEQM in Margate. The reason given was that “the lack of consultant cover meant that trainees were not adequately supported”. Shortly after, the Urgent Care Centre in Canterbury was closed ‘temporarily’, and has yet to reopen.
Announcements of unit closures and/or the commencement of consultation processes in themselves can severely affect a hospital’s ability to retain and recruit staff. In January 2018 the options for the Kent and Medway Stroke plans were formally announced, with QEQM and Canterbury excluded from the options for a HASU siting. By summer of that year, documents were circulated to the stroke staff at QEQM hospital explaining employment options, namely, move to Ashford or make other arrangements. The consultation was not yet complete, the final decision was due in early 2019. From that point, specialist staff have left the unit taking their years of experience and knowledge with them. According to a source in the hospital, only one specialist stroke nurse is willing to make the transition to Ashford. A document produced by The Surgical Forum of Great Britain and Ireland in October 2014 calls for a reduction in the number of hospitals in the United Kingdom. The report ‘The future of District General Hospitals’ is based on opinions expressed at a meeting attended by 20 clinicians and managers held by the forum. These are some of the statements from the report: “The United Kingdom and Ireland currently have too many hospitals attempting to provide both emergency and a wide range of elective care. ... The Forum was of the view that the NHS does not have too many beds, but that their function can often be redefined.” “The Forum were agreed that it would be appropriate for clinicians to voice concerns about the viability or otherwise of small hospitals, and make recommendations to politicians as to which hospitals should be reconfigured, closed, or have their function significantly changed.” “The Forum was unanimous in deploring what it termed ‘vested interest groups’ in attempting to keep unsafe or non-viable local services open where this is counter to achieving good clinical care.” These statements show that smaller hospitals are at risk as opinion amongst some parts of the NHS hierarchy are very much poised to close or redesignate the purpose of many hospitals (it should be stated that some contradictory viewpoints were expressed during the meeting). It must be remembered that at no point has the public been asked, at election time or otherwise, if it is in agreement with a reduction in the number of district general hospitals. In fact, it has not been a part of public discourse at all, as the topic has been largely kept out of the main news stories and has not been the topic of in-depth parliamentary debate. Apart from health campaigners, a number of politicians and the NHS transformation managers, no one knows about the direction of travel towards the demise of district general hospitals, and it looks set to stay that way.
The forum’s concern that there are ‘too many hospitals’ but not ‘too many beds’ suggests that they favour reconfigurations that centralise hospitals into fewer locations. There is clear concern in the forum’s document and even more so in Bruce Keogh’s review that ‘politicians’ and ‘vested interest groups’ (campaigners) will obstruct the planning and raise objections, which is unwelcome. The NHS constitution does require that the public are involved in reconfigurations, and although consultations are taking place they should be better publicised and not treated as a PR exercise to facilitate an already agreed outcome. It is also not sufficient to merely have localised consultations, when these plans are intended for national implementation. National consultations should be widely publicised in mainstream media and all implications of the changes should be made clear. It should be mandatory for all consultations to include medical evidence that does not support the proposal as well as evidence that does, and with the same prominence. The initiative to dramatically reshape the NHS should not be devolved to the ‘transformation’ managers at local level; it requires far more scrutiny and comparison with overseas models, it should involve parliament and has to involve the general public who are the users and the owners of the service.
Section N - Financial Considerations7. It is stated in the DMBC that stroke services in Kent & Medway are running at a £7.5 million loss currently. It is curious that this is raised at all, or being referred to as a matter for consideration in decision making, given that NHS services are not required to make a profit, or to ‘break even’. The NHS is expected to provide a service according to need, not according to cost. Businesses are required to think about profit and loss, but the NHS is not. Public money is spent on getting people well so that they can continue to live healthy lives and continue to contribute to society. Stating that a service is running ‘at a loss’ or at ‘a deficit’ implies that there is a requirement to clear that deficit. It is well known that NHS spending boosts the wider economy in excess of the money spent - this is referred to as the ‘fiscal multiplier’. It is estimated that the NHS fiscal multiplier is between two and four [King’s Fund, Tackling poverty: Making more of the NHS in England, Jabal and Buck]. The OECD states that “the health system contributes to economic performance. It is a major employer – it accounts for nearly one in every ten jobs in OECD countries; health spending helps stabilise the economy in times of crisis, and it is a contributor to the productive capacity of OECD economies” [https:// www.oecd.org/health/2010-ministerial/healthsystemprioritieswhenmoneyistight.htm]. Health care spending plays a role in supporting health, well-being and an individual’s productive capacity, and therefore has knock-on effects for other parts of government spending. The Work Foundation reports that “in 2009, in the region of 11,000 people in England and Wales were enabled to return to work by hip replacement surgery, saving the UK welfare system £37.4m each year of their working lives” [‘Adding Value: The Economic and Societal Benefits of Medical Technology’, Bevan, Zheltoukhova and McGee]. We do not know the exact fiscal multiplier for acute stroke care in the NHS, but perhaps the STP or NHSE should be required to produce an estimate, and state the figure alongside every mention of services running at a ‘deficit/loss’.
- It is the role of the Kent and Medway STP to find ways to reduce the ‘deficits’ of the CCGs that fall into its catchment area. This is what is behind the ‘transformation’ plans recommended by the STP; the changes are billed as ‘improvements’, but what we see in this particular plan is a proposal to centralise care into fewer units (half the current number); a promise to increase the overall number of staff (which may not be achievable); a reduction in the number of beds; a vague possibility (not included in this plan, but hinted at) to provide thrombectomy at a later stage; and the promise of improved in-hospital timings. The in-hospital timings will be offset by longer journey times for most patients (in some cases, the longer journey will outweigh time saved in hospital), and it should be noted that improvements to patient outcomes cannot legitimately be promised or guaranteed (see sections A and B) - and patient outcomes are the metrics that really matter, ie whether patients are more likely to survive, and to survive unimpaired. Documents from the STP show that the intention is to ‘achieve financial balance’ for health and social care in Kent and Medway. In 2016, the ‘deficit’ was £109 million, projected to rise to £486 million by 2020/21 - so you can see that the job of the Kent and Medway STP is to cut services to a substantial degree in order to meet its expectation of saving a large projected figure of close to half a billion pounds. [page 6, ’Transforming Health and Social care in Kent and Medway: Sustainability and Transformation Plan’ 21st October 2016]
- The KHPO recommended that money would be better spent on prevention. With the threat of trained specialists and consultants continuing to leaving the NHS, it stands to reason that the NHS ought to listen to the needs of those struggling medics (see E.15), and fund initiatives that might help them to remain in the NHS, rather than lose that valuable expertise and experience which requires years and considerable financial outlay to replace. Perhaps NHS funds would be better spent on evidence-based prevention plans devised by the experts in Public Health departments, and on trialling schemes that attract and support medical professionals, enabling them to remain in the NHS, and giving them a reason to build a life in Kent.
Section A: 1. From the London Study - ‘Impact of centralising acute stroke services in English metropolitan areas on mortality and length of hospital stay: difference-in-differences analysis’ http://dx.doi.org/10.1136/ bmj.g4757, see page 8, ’Strengths and Weaknesses’ (excerpt below). 2. “Improvements in mortality in London were higher than the trend in improvement elsewhere, though London had previously lagged behind” [p16, ‘Kent and Medway Stroke Services, Evidence Review of Hyperacute Stroke Units’, Kent & Medway Public Health Observatory, Varshney et al] 3. ‘The Centralisation Myth’ and ‘Distance Does Matter’ - https://bit.ly/2QcXnd1 4. This was confirmed in an email from Dr Steve Morris, one of the study’s co-authors. He said “HASUs and SUs in London were offered a special tariff to cover the cost of the care they provided, but to obtain that tariff they were required to meet certain performance standards, for example in terms of staffing and provision of evidence-based care. Here is a link to a 2014 document describing the process (this is in the public domain):” http://www.londonscn.nhs.uk/wp-content/uploads/2015/01/Stroke-acute-commissioning-and-tariff-guidance-2014.pdf 5. ‘Impact and sustainability of centralising acute stroke services in English metropolitan areas: retrospective analysis of hospital episode statistics and stroke national audit data’, published 25.01.2019. The study and peer review document can be found here: https://www.bmj.com/content/364/bmj.l1 6. From the London Study - ‘Impact of centralising acute stroke services in English metropolitan areas on mortality and length of hospital stay: difference-in-differences analysis’ http://dx.doi.org/10.1136/bmj.g4757, “..the results ... might be less relevant to services in rural settings. The greater travel times in rural areas make centralisation challenging and might necessitate other solutions..” p10. Section B: Distance Does Matter - https://bit.ly/2QcXnd1 Section E: BMA ‘Specialty Fill Rates by Region’ 2013-16: https://www.bma.org.uk/collective-voice/policy-and-research/education-training-and-workforce/state-of- medical-recruitment The Excel sheet can be found and downloaded under ‘Variations between Specialties’ heading. Section L: 7. SECamb covers the southeast, serving 4.5m people. The population covered by the stroke review (Kent and borders) is 2.2m people, therefore this is 48% of the population that SECamb cover. This means Secamb's running costs that are attributable to Kent and borders will be £101,358,240 pa. 8. Private ambulances: https://www.theguardian.com/society/2018/oct/21/nhs-private-ambulances-trusts- wasted-235-million-pounds Second Hand Ambulances: https://www.bbc.co.uk/news/uk-england-45571521
Links to Studies:
MANCHESTER & LONDON HASU STUDY: STROKE "Impact of centralising acute stroke services in English metropolitan areas on mortality and length of hospital stay: difference-in-differences analysis" https://www.bmj.com/content/349/bmj.g4757 NORTHUMBRIA CENTRALISATION: STROKE "The impact of acute stroke service centralisation: a time series evaluation" http://futurehospital.rcpjournal.org/content/5/3/181.full 5 A&Es: A&E CENTRALISATION STUDY "Closing five Emergency Departments in England between 2009 and 2011: the closed controlled interrupted time- series analysis" https://www.journalslibrary.nihr.ac.uk/hsdr/hsdr06270/#/abstract ‘Impact of centralising acute stroke services in English metropolitan areas on mortality and length of hospital stay: difference-in-differences analysis’ Page 8, Strengths and weaknesses The main strengths of our study are that we used a large national dataset containing detailed information on outcomes and patient characteristics, and the robust quasi-experimental framework; these allowed us to control for trends in the rest of England and other factors that could affect outcomes during the same period. There are, however, several weaknesses. Firstly, the hospital episode statistics database does not include information on severity of stroke, which is an important predictor of mortality.31 Data collected after the reconfigurations from the Stroke Improvement National Audit Programme32 show that indicators of stroke severity, such as worst level of consciousness in first 24 hours after stroke and neurological deficits on admission, varied between Greater Manchester, London, and the rest of England (see table A, appendix), but there was no discernible trend over all the indicators. Despite this, and even though our outcomes are risk adjusted for several patient level factors and we accounted flexibly for differences between hospitals and trends over time, we cannot rule out the possibility that the differences in outcomes could be caused by variations in severity of stroke between Greater Manchester and London and the rest of England. Secondly, we were unable to assess the impact of the reconfigurations on other outcomes, such as quality of life, disability, or neurological and functional impairment, as these measures were not collected in the hospital episode statistics database. We could, however, investigate the impact of the reconfigurations on mortality and length of hospital stay, outcomes that have not been previously reported. Thirdly, the hospital episode statistics database includes only patients admitted to hospital. It does not include any information about patients who died before they reached the hospital, nor does it include information on the time of stroke; hence our analyses of mortality were based on time from admission. If patients with stroke in London were more likely to die before reaching the hospital because of longer travel distances to hyperacute stroke units then the effects of the reconfigurations on mortality would be overestimated. Evidence suggests this is unlikely because ambulance journey times for patients with stroke did not increase appreciably after the reconfiguration in London, with mean times from scene to hospital of 14 minutes from January 2005 to March 20083 and 16 minutes from April 2011 to March 2012.33 Also, severity of stroke in London after reconfiguration was similar to the severity in the rest of England (table A, appendix); if more patients with severe stroke died in London before reaching the hospital the level of severity in the audit data for London would be lower than elsewhere. Fourthly, length of hospital stay was measured as the difference between date of admission and date of discharge. We assumed that when patients were discharged from one hospital and readmitted to another hospital on the same day this was a transfer related to the original stroke, capturing the movement between components of the stroke care pathway (for example, between a hyperacute stroke unit and a stroke unit in London). Conversely, we assumed that when a subsequent admission occurred one or more days later after discharge this was a recurrent stroke (the risk of recurrence of stroke in the first month after discharge is 1.1-15%).34 Finally, there was a higher than expected number of patients with stroke per month in London during the period after reconfiguration. One possible reason is that after the reconfiguration London units treated more patients from surrounding areas, who might previously have gone to their local emergency department. This could bias the results in favour of the London reconfiguration if the additional admissions were for less severe strokes, but there are no discernible differences in severity between London, Greater Manchester, and the rest of England (table A, appendix).