Editorials

As for the alternatives, just like the publicity

campaign Everybody remembers a good teacher, so
every manager knows their local clinician innovators.
There is no single blueprint for these people, but typically, they combine vision, energy, and the ability to
drive developments that seem threatening while
respecting the professional interests of colleagues
enough to avoid rebellion. The extent to which these
natural leaders are involved in modernisation plans
will be a key determinant of the extent to which
sustained visible improvements can be achieved.
Alternative policies might aim to grow capacity by
growing clinical leaders through imaginative training
schemes combining clinical, managerial, and public
health training. They might encourage successful
innovators to lead neighbouring clinical teams that are
struggling with the problems highlighted above. They

might even give proven clinical leaders total clinical and

managerial control over their department, nested within
the NHS, and see what that could achieve.
Rebecca Rosen fellow in primary care
Kings Fund, London W1G 0AN

2
3
4

5
6

Department of Health. A new role for overseas and independent healthcare

providers in England. Press release 2002/0283. 25 June, 2002.
http://tap.ukwebhost.eds.com/doh/intpress.nsf/page/20020283?OpenDocument (accessed 2 Aug 2002).
Miles A, Baldwin T, Charter D. Heart surgeons stall Blairs drive to slash
waiting lists. Times 2002 June 13:1.
Rogers L, Carr-Brown J. Surgeons in rebellion over foreign doctors in the
NHS. Sunday Times 2002 July 21:2.
Department of Health. Hospital Activity Statistics. Outpatient attendances, 1998/99-2002. www.doh.gov.uk/hospitalactivity/ (accessed 2 Aug
2002).
Royal College of Ophthalmologists. College News. Q Bull R Coll Ophthalmol Spring 2002.
Performance and Innovation Unit. Better policy delivery and design.
London: Performance and Innovation Unit, March 2001.

Stroke units
Every patient with a stroke should be treated in a stroke unit

BMJ 2002;325:2912

BMJ VOLUME 325

troke is a worldwide problem with high

incidence, mortality, disability rates, and costs.1 2
Stroke units are known to improve outcome by
concentrating patients in a unit with appropriate
expertise.3 The United Kingdom has some of the worst
outcomes for stroke.2 With a national service
framework laying out milestones for the NHS to
deliver a better service by April 2004,4 the publication
of the third national sentinel stroke audit for 2001-2 by
the Royal College of Physicians is welcome, as it gives
an overview of progress to date.5
Unfortunately, it makes for desolate reading. Two
hundred and thirty five hospitals participated in this
audit of the organisation and processes of stroke care.
The outstanding feature of the report is the failure of
the NHS to provide adequate numbers of beds in
stroke units. Although the mean number of patients
with stroke in hospitals on a specific day was 30, the
mean number of acute stroke beds was four and the
median number zero. For rehabilitation beds for stroke
the mean number was 15 and the median 14. From this
lack of capacity to admit patients flows most of the failure to achieve appropriate standards of clinical care,
documented in the rest of the report, including the
relatively high mortality.5
Only one third of patients were reported to spend
any time in a stroke unit, with only a quarter spending
most of their stay in a unit, despite clear evidence from
a systematic review of randomised trials that management by a stroke unit saves lives, increases the number
of independent survivors, and reduces institutionalisation.3 The national clinical guidelines for stroke are
unequivocal that the evidence in support of this is
overwhelming and achieving this . . . should be the
highest priority of . . . clinicians and managers.6 The
national service framework explicitly recommends that
All patients who may have had a stroke . . . should be
treated by specialist stroke teams within designated
units.4 This effectively means treatment in a stroke unit
for every person with a stroke. Provision of care in a
10 AUGUST 2002

bmj.com

stroke unit is not only the practice of evidence based

medicine but a matter of clinical governance.
Milestones in the national service framework specify
that all general hospitals introduce this model of care
from 2004, integrated with a stroke prevention service,
and specialist community rehabilitation teams.
In our local experience of introducing a stroke unit
changes have happened in line with the evidence base.
Overall mortality reduced by a third, and the 30 day
case fatality rate fell from 27%7 to 15%.8 Other gains
have been a low rate of institutionalisation, a high rate
of discharge from the acute hospital site, and a rolling
educational programme for staff.8 Senior management
support, and effective multidisciplinary working, with
designated protected time to manage and plan are
necessary. Key to the success of a unit is influencing
bed managers to give the stroke unit beds the same
kind of priority as the coronary care unit, and persuading management to increase capacity to match need,
with flexible bed reconfiguration within the pool of
beds. The number of beds actually used for people with
strokes should meet demand.
Hospitals without adequate capacity in stroke units
may offer a peripatetic service to patients who are not
admitted to the stroke unit, whereby the unit staff regularly advises on general wards. Such services sound
sensible but contradict the evidence, which shows that
in comparison with management in a stroke unit mortality (30% v 14%), dependency, and institutionalisation
are higher.9 Moreover there are differences in the processes of care that provide plausible explanations for the
difference in outcome: measures taken to prevent aspiration pneumonia, early involvement of therapists, and
regular multidisciplinary team meetings.10 It is such
care that the sentinel audit identifies as deficient, especially in general wards.
Another option for hospitals with inadequate
capacity is to select patients most likely to benefit
usually those in the middle band of severity. Evidence
from randomised controlled trials shows that even the
291

Editorials
most disabled patients gain from management on a
stroke unit, with less time spent in hospital and more
patients going home.11 For patients with a mild stroke,
stroke units have a role in secondary prevention, an
area where the evidence continually changes and
needs to be interpreted by a specialist. The sentinel
audit shows suboptimal management of, for example,
hypertension and smoking.5 Current evidence shows
that there is no substitute for management of all
strokes in a stroke unit. The patients are already in the
hospital, so what is needed is simply reconfiguration of
beds and training of staff to deliver the care.
What about the community? The sentinel audit
shows that only 31% of trusts have specialist
community stroke teams.5 Evidence from randomised
controlled trials shows that a focused community rehabilitation team with adequate resources, linking with a
state of the art stroke unit, reduces length of stay and, in
moderately and severely disabled patients, reduces disability and institutional care, compared with management by a stroke unit alone.12 To help deliver this
integrated service, as mandated by the national service
framework,4 requires more than the median two
sessions a week by a stroke consultant reported by the
sentinel audit. The government has released more cash
to the NHS in the recent spending review but wishes to
tie it to initiatives improving patient outcomes.
Investment in integrated stroke services will do just
that. It is up to local multidisciplinary teams, managers,
primary care trusts, and patient groups to press for
treatment in a stroke unit for every person with stroke.

Sheldon Stone senior lecturer and stroke physician

Academic Department of Geriatric Medicine, Royal Free Campus,
Royal Free and University College Medical School, London
WC1E 6BT

SS acknowledges the helpful discussions with Isa Mason (stroke

unit coordinator), Heather Appleby (occupational therapist),
and Cherry Kilbride (research physiotherapist).

Thorvaldsen P, Kuulasmaa K, Rajakangas AM, Rastenyte D, Sarti C,

Wilhelmsen L. Stroke trends in the WHO MONICA Project. Stroke
1997;28:500-6.
2 Wolfe CDA, Tilling K, Bech R, Rudd AG. Variations in case fatality and
dependency from stroke in western and central Europe. Stroke
1999;30:350-6.
3 Stroke Unit Trialists' Collaboration. Organised inpatient (stroke unit)
care for stroke. Cochrane Database Syst Rev 2000;(2):CD000197.
4 Department of Health. National service framework for older people.
Standard 5: Stroke. London: Stationery Office, March 2001.
5 Intercollegiate Stroke Working Party. Summary report on the National Sentinel Stroke Audit 2001/2. Clinical Effectiveness and Evaluation Unit, Royal
College of Physicians of London. www.rcplondon.ac.uk/pubs/
strokeaudit01-02.pdf (accessed 2 Aug 2002).
6 Intercollegiate Stroke Working Party, Royal College Physicians. National
clinical guidelines for stroke. www.rcplondon.ac.uk/pubs/books/stroke
(accessed 29 Jul 2002).
7 Stone SP, Whincup P. Standards for the hospital management of stroke
patients. J R Coll Physicians 1994;28:52-8.
8 Intercollegiate Stroke Working Party. National Sentinel Stroke Audit 2001/2
Trust report. Clinical Effectiveness and Evaluation Unit. Royal College of
Physicians 2002.
9 Kalra L, Evans A, Perez I, Knapp M, Donaldson N, Swift CG. Alternative
strategies for stroke care: a prospective randomised controlled trial of
stroke unit, stroke team and domiciliary management of stroke. Lancet
2000;356:894-9.
10 Evans A, Perez I, Harraf F, Melbourn A, Steadman J, Donaldson N, et al.
Can differences in management processes explain different outcomes
between stroke unit and stroke-team care? Lancet 2001;358:1586-92.
11 Kalra L, Eade J. Role of stroke rehabilitation units in managing severe
disability after stroke. Stroke 1995:26:2031-4.
12 Indredavik B, Fjaertoft H, Ekeberg G, Loge AD, Morch B. Benefit of
an extended stroke unit service with early supported discharge:
A randomized, controlled trial. Stroke 2000;31:2989-94.

Polysaccharide pneumococcal vaccines

Existing guidance is at variance with the evidence

nfection with Streptococcus pneumoniae (pneumococcus) is a leading cause of illness in young

children and of illness and death in elderly people
and people with immune deficiencies and chronic
illness. Pneumococcus causes a spectrum of disease:
infections of the upper respiratory tract, otitis media,
invasive infections such as bacteraemia and meningitis,
and infections of the lower respiratory tract such as
pneumonia.w1
Polyvalent pneumococcal polysaccharide vaccines
containing more than one capsular antigen of
pneumococcus have been around since the 1940s. The
current 23-valent pneumococcal polysaccharide vaccine was introduced in the United States in 1983.w2 Its
antigenic composition reflects the incidence of the 23
serotypes in the causation of invasive disease, as well as
coverage of the most prevalent serotypes.w2 As each of
the many pneumococcal serotypes has a different epidemiology and probably invasiveness this combined
vaccine has a very broad spectrum.w3
Pneumococcal polysaccharide vaccine is recommended in the United States for use in all people aged
65 years or more and in people at risk aged 2 years or
more.w1 In several other developed countries it is
recommended for similar indications.w2 The evidence
to support such indications for its use is, however, far
292

from clear cut.1 w4 The decision to introduce a vaccine

into an immunisation programme should be based on
an assessment of the importance of the target health
problem, as well as the vaccines impact on the
problem, its safety record, economic profile, and
acceptability to users.w5
The epidemiology of pneumococcal disease seems
to vary by serotype, population, age, and setting.w3 The
incidence of confirmed pneumococcal disease in the
control arms of trials of pneumococcal polysaccharide
vaccines carried out in developed countries varied
between 1.3% and 19% for pneumococcal pneumonias
in people at high risk and between 0.31% and 2.4% in
people aged 65 years or more.1 In England and Wales
pneumococcal bacteraemia is rare (7/100 000 in the
general population),w6 whereas up to 40% of hospital
admissions for community acquired pneumonia are due
to pneumococcus.w7 In developing countries the
proportions may be higher, but difficulties of surveillance and serological diagnosis hinder quantification.w2
Evidence of the effectiveness of pneumococcal
polysaccharide vaccines comes from several trials
carried out in the past decades, which have been the
subject of seven systematic reviews published between
1994 and 2002.17 The methodological quality of the
reviews is variable. No review stratified its meta-analysis
BMJ VOLUME 325

Additional
references appear
on bmj.com
BMJ 2002;325:2923

10 AUGUST 2002

bmj.com