The Socialist Response to the Hayes Report: Chapter 2
CENTRALISATION OF SERVICES - IS IT JUSTIFIED?
The centralisation of hospital services is an international phenomenon. "Hospitals in all systems have to deal with rising expectations and, more often that not, a need to contain the costs of health care. Outside the developing countries the generic response to this has been to reduce hospital stays and to improve the efficiency of the system, a strategy which seems to be a least partly successful. The experience of the health systems in the United States and the United Kingdom shows that cost pressures and changes in health care delivery mean that this strategy will lead to hospital mergers and closures in the longer term" (BMJ 1999;319:845-8).
The NHS already has fewer acute beds per head of the population, and patients have a shorter average length of stay in hospital, than in 20 other OECD countries (1995 figures -see Table 2).
No of acute beds per Average length of stay in
1000 population (1995) days (1995)
Australia 4.3 6.7
Austria 6.6 7.9
Belgium 4.8 7.8
France 4.6 5.9
Germany 6.9 11.4
Ireland 3.4 6.7
Italy 5.3 8.8
UK 2.2 4.8
USA 3.3 6.5
Source BMJ 1999;319:845-848
Between 1978 and 1990-91 506 small hospitals closed in England. The majority of these had less than 250 beds. With the introduction of Trust hospitals in 1991 the Department of Health stopped collecting data on the total number of hospitals, though bed numbers are still collated, so we do not know how many more hospitals have closed in the last decade. Overall two-fifths of the total bed stock closed between 1982 and 1994-95. A quarter of acute beds closed (Table 3). Since 1995 the rate of bed closure has slowed and in 2001 numbers actually rose by a few hundred in the aftermath of the then Health Minister David Blunkett’s admission that be closures had gone too far. Overall, however, a further 13,000 beds have closed since New Labour came to power.
Table 3 Changes in numbers of NHS beds in England 1982 to 1994-95
Year All Acute Geriatrics Mental Mental Maternity
Specialties Care Illness HandicapCare
1982 348104 143535 55646 83831 46983 18108
1994-95 211812 108008 36795 41827 13211 11971
% change -39 -25 -34 -50 -72 -34
Source: BMJ 1999;319:911-914
The total number of beds per 1000 population in England has fallen from 7.4 to 3.9 over the period in Table 3. The number of acute beds has fallen from 3.1 to 2.2 per 1000. At the same time as the number of beds has been falling the amount of work has been increasing sharply, though there is some controversy as to what extent the NHS workload has actually increased.
The official figures demonstrate a two-thirds increase in the total number of patients treated in the NHS in England between 1982 and 1995. The throughput rate (cases per bed) increased by 81% for acute beds between 1982 and 1995. These figures are not entirely accurate, however.
The introduction of the values of the market put the onus on NHS managers to prove that their hospitals were increasing "throughput". Consequently procedures that had previously not been included in returns to the NHS Executive were included and one stay in hospital became several "episodes" of care if a patient was moved between different wards or between different consultants.
The increased throughput figures are probably thus a combination of three things: a real, and unquantifiable, increase in activity; the counting of what was previously not counted - a false increase; and double-counting and various other slights of hand - straightforward cooking the books. Despite these caveats it is important to note that there has been a real increase in NHS capacity, because of new technology and new approaches, despite the decrease in bed numbers. It does not necessarily follow that some beds could have been safely closed. Instead, the same number of beds should have been maintained, thus allowing greater flexibility in the system and the ability to treat previously neglected illnesses.
The average length of stay in hospital has been declining since the birth of the NHS. Between 1982 and 1994 the number of acute beds fell by 2.6% per year whilst the average length of stay fell by 3.1% per year. This is largely a good thing - no one wants to remain in hospital unnecessarily - but everyone has a relative or friend who did not feel well enough to be discharged but were nevertheless asked to leave. The average length of stay cannot fall forever (eventually it would reach zero) and the argument that it is already too short is supported by evidence of increasing rates of re-admission after discharges (when patients relapse and need to return to hospital).
Outpatient attendances in the NHS increased from 35.6 million in 1992 to 41.6 million in 1997-98, an increase of 16.8%. Some of this increase at least, will be accounted for by the "more efficient" counting of cases already outlined. Total casualty (accident and emergency) department attendances have not increased, fluctuating around a mean of 13.6 million over the last 20 years. Despite this a report in October 2001 described casualty services as being in crisis across the NHS and in the same month Belfast City Hospital nurses threatened to walk out when their department ground to a halt as dozens of patients were lying on trolleys because no beds were available.
There is evidence that whilst we are not significantly more likely to attend an outpatient department or a casualty department than we were in 1982, we are more likely to be admitted to hospital. Just why this is, is not clear. The introduction of new technology and new interventions must play some role - there are now treatments available for certain conditions were before there were none.
Overall those figures do not bear out the argument that the NHS is in difficulty largely because of an overwhelming tide of spurious demand. Could it just be that any increase in demand has been entirely manageable, any difficulties are a result of cutbacks?
Hayes utilises the fact that Northern Ireland has more acute beds per head of the population than England to argue that we have "too many" beds. Given that almost everyone, including the New Labour government, now accepts that bed closures in England have gone too far it is more logical to argue that we have too few beds, and that England is in an even worse situation. When David Blunkett accepted that bed closures had to stop and that bed numbers should in fact increase slightly (by 3000) he could do no other, given the weight of evidence. His pledges mean little in the context of the implementation of the Private Finance Initiative, but it is a useful weapon for hospital campaigners to know that the case for no more bed closures has been conceded.
Is bigger better?
There seems to be an unstoppable tide in the favour of the closure of smaller hospitals and the merger of two or more hospitals into super hospitals. The rationale for such action is the belief that bigger hospitals are more cost effective and safer. It is argued that bigger hospitals reduce average costs through the operation of economies of scale and that outcomes improve because of increasing average volumes of activity per clinician (that is doctors specialise in a very small area and get better through more practice).
John Posnett, director of the York Health Economics Consortium, has argued in a recent article that "this logic is not support by the evidence" (BMJ 1999;319:1063-1065). There is evidence that the most cost effective hospitals have between 200 and 400 beds. Those with fewer than 200 are more expensive but so too are those with 400-600 beds. There are still a large number of hospitals in England with fewer than 200 beds (see Table 4) and of course the closure of hundreds of smaller hospitals has not cured the ills of the English NHS.
Table 4 Distribution of acute hospitals in England by size.
No of beds No of hospitals Share of total Share of total
hospitals (%) beds (%)
<200 149 36 10
200-400 106 26 23
>400 154 38 68
The sparse evidence that there is suggests that two hospitals of 400 beds are more efficient (in financial terms) than one single site hospital of 800 beds. Generally when small hospitals are replaced by larger ones total management costs either increase or remain unchanged - which depends on the size of the new organisation. There is no basis for the argument that replacing multiple small sites with super hospitals reduces overall management costs.
Sometimes new hospitals are cheaper than the ones they replace, in terms of total costs. This is simple because capacity - the number of beds - has been cut. And of course new hospitals built under the PFI will eat up any savings which may result from reducing the number of sites in any case (this is explained further in Chapter Three).
The Hayes Report does accept that "the evidence on whether concentrating services in a reduced number of specialist centres results in improved outcomes is not clear cut".
According to Posnett "the literature shows quite conclusively that there can be no general presumption that larger units produce better outcomes for patients. The evidence of a positive relation between volume and outcome for a small number of defined procedures is reliable, but these effects operate at comparatively low levels of activity, certainly not large enough to justify notable concentration". What Posnett means by this latter remark is that hospitals do not need to be that large to gain the benefits that accrue from increased practice at a particular procedure for doctors.
In Posnett's view most of the published evidence that demonstrates that outcomes in bigger hospitals are better is unreliable. In the case of intensive care units, the supposed superiority of larger units disappears when the severity of patients' conditions when they enter the unit is taken into consideration. Smaller units admit more severely ill patients and this is why their outcomes are worse.
Why some doctors or hospitals have better outcomes than others is not well understood. Greater levels of activity may not be the key. The availability of support services (such as imaging and intensive care), good ongoing training for doctors and high quality co-operation between doctors and between other members of staff may be equally, or more, important.
Posnett summaries his arguments in this way: "On the basis of available research evidence, bigger is not better: at present there is no reason to believe that further concentration in the provision of hospitals will lead to any automatic gains in efficiency or patient outcomes. Maybe the research base is inadequate, but the onus is on those who advocate the benefits of concentration to prove their case. In the future as general practitioners assume an increasingly influential role in planning the provision of health services, the perceived benefits of accessible local services may begin to turn the tide of professional opinion" (BMJ 1999;319:1063-1065).
The NHS Centre for Reviews and Dissemination has also reviewed the evidence and concludes: "there is no compelling reason to believe that further concentration of hospital services will result in improved efficiency or lead to automatic improvements in the quality of outcomes. In assessing the potential effects of increased concentration on access and utilisation the implications for disadvantaged groups in particular should not be overlooked" (Report 8,1997).
The medical hierarchy appears to be of one voice on the future of small hospitals. A joint working party of the British Medical Association or BMA (the main doctors’ trade union), the Royal College of Physicians of London and the Royal College of Surgeons of England argues that "comprehensive medical and surgical care of the highest quality requires the concentration of resources and skills into larger organisational units" (Provision of Acute General Hospital Services. London RCS, 1998).
The Royal College of Surgeons would really like to see super hospitals servicing populations of half a million or more. This would allow the "dream set up" of 15 consultant surgeons, 15 consultant orthopaedic surgeons, 30 anaesthetists, 24 hour a day operating, an intensive care unit and 24 hour pathology and X-ray services. This would mean only three acute hospitals in the North. The Royal College of Physicians are in favour of eight to ten acute hospitals for a population of Northern Ireland’s size.
The medical Royal Colleges are professional bodies and not trade unions. They provide advice to the government on the required numbers of doctors nationally and on the training needs of doctors. All physicians must belong to the Royal College of Physicians and all surgeons to the Royal College of Surgeons (passing the required exams and receiving adequate training are the conditions of membership). The Colleges undoubtedly play a useful role and help to maintain high standards, though they are somewhat archaic in their titles and procedures and tend to be dominated by senior doctors in London and the Home Counties. As an extension of their role they have a lot to say about the optional configuration of hospital services. The main thrust of their arguments is that large hospitals allow for improved training, necessary sub specialisation and thus improve outcomes.
They surmise that if small hospitals (servicing populations less than 150,000) are to survive then they need to be comparatively overstaffed. In essence, accepting the current restraints of the system, they issue edicts in the full knowledge that hospital closures will follow. Public opinion is ignored and little consideration is given to the special needs of rural areas.
The guidance of the Royal Colleges, supposedly given on training issues, can be very damaging for local services. To take one example, an accident and emergency department will lose its training recognition (the right to train junior doctors) if one of four "essential" services is removed from the hospital (the four are general medicine, general surgery, trauma and orthopaedics) and will thus have to close. A knock-on effect leads to the closure of one department after another.
In this way the Royal Colleges were instrumental in the closure of the South Tyrone Hospital, withdrawing training recognition from several departments. The Colleges argue that closure is not their intention, or their responsibility. In one sense this is true but it is evasive and disingenuous to try to avoid any responsibility. The eventual outcome of the withdrawal of training recognition is entirely predictable.
Of course the government, ultimately responsible for closures, accept Royal College decisions with barely disguised glee. The closure of the South Tyrone suited them down to the ground and they too were able to absolve themselves of responsibility. Indeed it appears that no one was responsible!
Not all doctors support the views of the BMA and the Royal Colleges. These bodies take a conservative stance, focus on the needs of doctors in isolation from other staff and patients, and are London-orientated. They do not focus on the needs of local people in rural areas. They do not consider the possibility that good quality training is available in small hospitals. They do not use their strength to bolster small hospitals.
Whilst the medical hierarchy are in favour of greater centralisation there are dissenting voices. When the Joint Consultants Committee published a report in 1999 on the future of hospitals (Organisation of Acute General Hospital Services) its conclusions were attacked by Mr James Glancy, a consultant cardiologist and physician at County Hospital, Hereford. In his view "the conclusions of this report represent a scandalous misrepresentation of what little data exists on comparison in outcomes between small and large hospitals " (Hospital Doctor, 14th October 1999). He added: "Yet again doctors' leaders have shown how hopelessly out of touch they are with the grass roots of the profession and patients".
Dr Susan Coe in Perth challenged the idea that smaller hospitals are not safe. "I feel the need to challenge this idea that technology equals excellence. There are many good doctors who choose not to live in big cities and work in university hospitals. They know that sometimes they will have to refer patients on to a more specialised practitioner. They also know that their teaching hospital colleagues do not always get it right" (Hospital Doctor, 7th October 1999).
Dr D Forbes, also writing from Perth, questioned the role of "staff from large hospitals who wield political and academic power". In his view “clinical networks” can be developed to allow for specialisation and the closure of units is not required. Other letters in Hospital Doctor (30th September 1999) argued that smaller hospitals are in fact safer than large hospitals as one is less likely to contract serious, untreatable infections such as MRSA is in a small hospital.
The stated view of the BMA in Northern Ireland is that "hospitals servicing rural communities are not going to be staffed without considerable expansion and there's not enough resources to do that "(Dr Caroline Marriott, chair NI BMA Central Consultants and Specialists Committee, quoted in Hospital Doctor, 9th December 1999). "Putting It Right" (published by John McFall in 1999) dismissed the idea that the then 17 acute hospitals could remain open as it would require a 50% increase in the numbers of consultant surgeons and physicians to achieve this. Now the Hayes Report proposes just such an increase at the same time as it proposes the closure of six acute hospitals. What has happened between the publication of the McFall report and the Hayes Report to justify this change in approach! Surely the BMA should now be reconsidering their position - if the staff are there the rural and small hospitals can remain open.
The medical hierarchy put a low premium on access to services, perhaps believing that everyone is as mobile as those in the moneyed circles they move in. There is evidence, however, that the further away some services are the less likely they are to be utilised. This applies in particular to consultations with general practitioners, to self-referral to accident and emergency departments and to attendance at breast and cervical cancer screening services. It appears that distance is less often a problem when someone requires acute care - patients will seek help regardless of the distance faced – but there is evidence for a greater deterrent effect for the poor.
Costs are shifted from the NHS to the patient as services become more concentrated, largely through increased travel costs but sometimes also through the need to arrange overnight accommodation. According to Posnett: "This effect is unlikely to be uniform across different sections of the population and the evidence is consistent with large deterrent effects for particular groups, such as those with low personal mobility or those in particular socio-economic groups" (BMJ 1999;319:1063-1065).
And it can be argued that the closure of rural hospitals will cost lives. Dr Kieran Deeny, chairman of the Omagh and District GP Association, wrote in the Belfast Telegraph (19/6/02) “A few years ago in our Carrickmore practice we had seven cases of meningitis in a 13 month period and I have no doubt that three of four children would have died had it not been for the close proximately of the Tyrone County and South Tyrone Hospitals”.
In the mid 1990’s an earlier report on the future of health care in Belfast - the McKenna Report- recommended that the City Hospital Accident and Emergency Department should close. The Report was met by an avalanche of criticism. One of the points raised was that the closure would reduce accessibility to good health care.
The Report, however, coolly stated that “accessibility is not the problem”. It went on to say that 90% of patients drive to or are driven to hospital (to use A&E services). The authors of this report are clearly immune to the realities of working class life in Belfast. At night, and especially at times of increased tension, working class people are fearful of travelling outside their own area.
This is not an irrational fear but a very realistic one. At such times one hundred yards might as well be one hundred miles. More than half of working class households do not own a car. Ninety-five percent of the households of those in “professional” occupations have a car, compared to only 38% of the households of manual workers. If rioting has stopped the buses or if sectarian attacks make travelling with certain taxi firms hazardous, what are they to do?
The Royal Victoria Hospital and the Belfast City Hospital are not directly comparable to hospitals in any other city on these islands, precisely because Belfast is not directly comparable. To think otherwise betrays much about the cosseted middle class lives of the authors. If local communities were represented on the various review groups these common sense points would not be missed.
Accessibility is also an issue in every other area of Northern Ireland. Local communities are comfortable with their local hospitals, built up over decades. Relatives can visit with ease, patients feel at home. In England, hospitals are much further apart but it does not necessarily follow that a similar model should apply here. Hospitals ought to be large enough to be viable but not so far apart that lives are put at risk. In more rural areas, extra minutes can be vital. Every town and village cannot have its own hospital but there should be no area without ready access to a good hospital.
Where hospitals already exist, and have done so for generations, a good argument needs to be made before such a facility is closed. Local communities are perfectly entitled to be suspicious of closure plans which promise a better service a little (or a lot) further away. This does not mean that from time to time, however, such a move would be a genuine advance.
In conclusion, there is little or no evidence in favour of centralisation of our acute hospital services. It is difficult to demonstrate that the closure of smaller hospitals actually costs lives, just as it cannot easily be proven that larger hospitals are safer, but common sense tells us that, at least occasionally, the extra journey involved will lead to a fatal delay. And the delay is more likely to be fatal in poorer households without easy access to a car. People want their local hospital. They provide good quality care for the majority of patients, they provide much needed local employment and they are often a vital hub of the local community. In the absence of evidence that “bigger is better” the smaller rural hospitals should remain open.