Memorandum by the Intensive Care Society
1. A recent paper from the Intensive Care
Society1 used a model provided by the US Center for Disease Control
and Prevention (CDC)2 to estimate the impact of an influenza pandemic
on critical care services in England.
2. Using plausible inputs for this model
(a 25 per cent attack rate and 8 week pandemic duration) we estimated
bed occupancy for Level 3 critical care beds that exceeded 200
per cent from the pandemic alone. This scenario does not account
for prevailing high occupancy rates in UK ICUs in the absence
of a pandemic (typically 85-90 per cent).
3. Current ICU resources would be overwhelmed
if the assumed 15 per cent ICU admission rate and 7.5 per cent
ventilation rate for hospitalised patients were to materialise.
These are not outlandish figures - there is a clear problem
4. However, the exact magnitude of this
problem is difficult to estimate, given the paucity of data for
ICU admission rates with influenza. Some data are available for
the surrogate of community-acquired pneumonia,3-7 but there are
wide variations in both ICU admission rates (3-20 per cent of
hospitalised patients) and mechanical ventilation (37-88 per cent
of ICU admissions). The available evidence suggests that inter-institutional
variations in provision of ventilatory support are less than those
for ICU admission.5 Since the need for artificial ventilation
is a hard outcome in terms of health economics and resource requirement,
this may be a better end point to focus on.
5. Best North American estimates for community
acquired pneumonia suggest ICU admission rates of ®10 per
cent overall, with about half of these requiring ventilatory support.5
6. However, these figures need to be viewed
with some caution in the light of the high incidence of primary
viral pneumonia and reported mortality of >50 per cent for
recent human cases of H5N1infection in the Far East.8 The current
strain of the virus would seem to be extremely virulent, suggesting
a higher rate of admission to intensive care and need for mechanical
ventilation. In addition, the need for ICU facilities may be significantly
modified by the propensity of the virus to cause exaggerated cytokine
responses. Indeed, in past episodes, the morbidity and mortality
have been related to these aspects of viral pathogeneicity.10
7. In the event of human-to-human transmission
with a virus of this virulence and high infectivity, intensive
care colleagues in Hong Kong (population 7 million) are bracing
themselves for 2 million infected subjects with a 10 per cent
ICU requirement (200,000 patients). Many hospitals in Hong Kong
have expanded ICU isolation facilities. When the Hong Kong Infectious
Diseases Centre is finished next year it will have 12 purpose
built ICU Isolation Rooms with negative pressure ventilation (Buckley
T, Personal Communication). The whole centre, which consists of
108 beds, will have negative pressure ventilation. The total number
of ICU beds with negative ventilation in Hong Kong is reported
to be in the hundreds (this is not fully substantiated).
8. It has been reported that no patient
who has been infected by the avian H5N1 strain in Vietnam, and
has required ventilatory support, has survived. There may be questions
regarding the promptness and quality of critical care delivered
to these patients. However, if these issues are clarified and
the report is confirmed, then we need to think carefully about
when and where we deploy scarce critical care resources. Indeed,
informal reports from the Far East and Australia suggest that
they have at least considered conceding that it would be impossible
to provide conventional intensive care in this context. At the
very least, we may need to develop robust triage systems.
9. Unlike ICUs in Australasia and Canada
(which experienced the SARS outbreak), UK intensive care units
have little experience of dealing with, or even preparing for,
an infectious disease outbreak of this magnitude. There needs
to be close attention paid to infection control education and
enforcement - including and up to the designation of enforcement
officers who ensure that precautions are adhered to.
10. The availability of staff for ICU expansion,
or even running current levels of capacity, may be severely compromised
by staff sickness (25 per cent) and fear of coming to work. Indeed,
a recent publication suggests that less than 50 per cent of nursing
staff would be able to or willing to report for work in such circumstances,9
because of staff sickness, sickness in family members, or the
fear of acquiring and transmitting the disease. It is important
to recognise this latter factor, and make attempts to assure the
safety of staff, as far as is possible, by providing protective
equipment, antivirals, and the best immunisation available at
the time. It may be important to find hospital based accommodation
for key staff who are exposed to high risk of infection, so that
they can minimise the risk of taking infection home to their families.
11. These problems of inadequate critical
care would be worsened by variable resource availability, which
may be appropriate for current casemix, but not necessarily for
12. We also need to consider the impact
of disease specific epidemiology - clustering in urban areas or
near regions where disease entry into the country is likely.
1. The ICU community needs to be involved
in contingency planning (the ICS has specific individuals with
responsibilities for these issues).
2. We need better modelling with hard UK
data to address the magnitude of the problem and the extent of
available solutions. Dr Menon and the ICS have begun to address
modelling with UK data as part of a dialogue with the DoH (collaboration
with Daniel Wood and Peter Grove, with potential involvement of
David Harrison of the Intensive Care National Audit and Research
3. There are potential means of expanding
ICU capacity temporarily, which have been explored in published
contingency plans.11 In the UK setting, this may be initially
achieved by converting Level 2 to Level 3 beds, and creating temporary
critical care resource in operating theatres and recovery areas.
4. Individuals who work in these areas (recovery
nurses and operating department assistants) have the requisite
clinical and technical skills to support critical care staff.
It is essential that we accelerate rotation of non-ICU staff through
ICU areas to build up a larger cohort of staff that could be called
on to help in the context of a pandemic.
5. There is likely to be a fine balance
between useful ICU expansion (with preserved basic care standards)
and over-expansion (to an extent that the care systems, clinical
skills, and discipline inherent in critical care are lost). There
is no point in admitting patients to expanded intensive
care areas" if they do not get a higher standard of care.
Discussions with colleagues in Hong Kong with experience of the
SARS epidemic suggest that a 50 per cent expansion may be tolerated,
but this will vary from hospital to hospital.
6. We should consider, on a hospital-by-hospital
basis, the development of critical care services in isolation/infectious
disease wards. If this is thought appropriate we need to organise
the services that will be required in these areas - piped oxygen,
suction and air, and protected power supply.
7. Non-NHS (eg private sector) critical
care resources should be logged and quantified. The NHS should
develop agreements that will allow us to use these resources in
the event of a pandemic. Consideration should be given to using
these facilities, where adequate and appropriate, as centres for
clean" (ie non pandemic related) essential clinical
8. We need to have clear central triggers
to move from standard to crisis mode - ideally as a staged escalation.
This will avoid any confusion about when current priorities (eg
waiting lists etc.) can be set aside to deal with the pandemic.
9. We need to create and maintain libraries
of critical care equipment in hospitals. Equipment shortages may
also be addressed (at least partly) by dialogue with industry
and imaginative resource expansion schemes. For example, it may
be possible to induce vendors of clinical monitoring, ventilation,
and infusion devices to increase their stock levels. This would
provide equipment that was available for rapid deployment. Increases
in stock levels may require the payment of a relatively small
10. Such ICU expansion is not compatible
with continuing adherence to conventional clinical standards and
targets - public education is essential to manage expectations.
11. There need to be clear plans at a hospital
level, including rapid access to protective equipment for staff
(perhaps chemoprophylaxis for carefully defined subgroups of key
hospital workers). We also need to stockpile antibiotics (especially
antistaphylococcal antibiotics) for the treatment of secondary
12. Regardless of preparations, resources
may be overwhelmed in many hospitals, and we need to consider
two options (both may be required, and should certainly be explored):
a. transfers will occur, and we need to make
provision for this;
b. we may need to bring skilled staff to
hotspots, rather than transfer patients.
13. We need to make plans for the recovery
and aftermath - catching up with waiting lists etc.
14. It may be prudent to consider whether
current indemnities will provide adequate cover, both in terms
of individual life insurance (if clinical staff voluntarily go
into areas with high infection risk) or organisational responsibility
(where the NHS utilises staff working at the margins of their
While an influenza pandemic may be clinically
devastating, it may also offer unique research opportunities.
There is a strong argument for identifying research funding, developing
research protocols, and applying for regulatory authorisation
to conduct such studies now, so that we are prepared if and when
a pandemic strikes. This process needs to be urgently undertaken,
and should involve (but not be limited to) epidemiologists, public
health physicians, virologists, immunologists, general physicians
We should plan to get regulatory approval from
MREC that covers issues relating to the Mental Capacity Act (the
sickest patients will be unable to provide consent), EU Clinical
Trials Directive (to assess antiviral drug and vaccine efficacy
and safety), the Human Tissue Act (research will almost inevitably
involve blood samples), and the Data Protection Act (to allow
clinical data collection and processing). I would suggest the
[a] We define a minimum clinical dataset
from all patients that would be used for epidemiological analysis,
and could provide data for use across many studies. This would
be best designed as a series of nested data collection instruments,
with increasing complexity of data collection as the severity
of disease increased. NHS Direct are keen to provide the base
for this, and I would see them collecting basic data on demographics,
comorbidities and symptomatology. A proportion of patients would
present to GPs, either directly or via referral by NHS Direct,
and could have additional data collected on these topics, as well
as information about vaccination and antiviral drug use. A smaller
number would present to hospital, where detailed clinical and
laboratory data could be collected. Finally, we expect a small
proportion of hospitalised patients to come into ICUs, where additional
data will be available. It is essential that we have good quality
outcome data for the hospital and ICU segments of the populations,
including length of hospital stay, treatment and mortality.
[b] It would be ideal if we could achieve
such data collections within the framework of existing administrative
datasets. As far as the ICU segment was concerned, the perfect
framework would be the ICNARC Casemix Program, which provides
high quality audit for the vast majority of ICUs in England and
Wales (including data on disease severity and organ failure quantitation
and 28 day mortality). The additional resource required to obtain
data of specific relevance to the pandemic have not been quantified,
but are likely to be small. We should explore the use and expansion
of similar administrative datasets in less acute settings.
[c] We should identify a matrix of GP surgeries,
hospitals and critical care units wishing to participate as stakeholders
in a research collaboration that pursued a range of pre-specified
studies. A collaboration covering 2 million individuals (®2
SHAs), would result in an estimated 100,000 GP consultations,
3,000 hospitalisations, 1000 ICU admissions, and 2,000 deaths
(based on HPA estimates).
[c] In the subset of the NHS identified in
[b] above, patients could be involved in more detailed studies.
In these patients we should bank a sample of blood for DNA extraction,
and store acute and convalescent blood samples to assess host
responses, including the development of immunity. Possible synergies
with the UK Biobank initiative that should be explored.
Two potential research areas are described below
as examples. I have listed studies at two ends of the research
spectrum to provide a view of the range of opportunities available:
1. Clinical trials of antiviral therapy
and vaccination in the early phases of the epidemic
When the pandemic first strikes we will
be using neuraminidase inhibitors and/or novel vaccines with an
extremely small (or possibly nonexistent) evidence base. Rapid
processing of results from early in the epidemic may allow more
optimal use of a range of therapeutic options in subsequent pandemic
waves, or even later in the course of the first wave of the pandemic
(especially if it is prolonged). Similar considerations apply
to the use of protective measures. Assessing the efficacy of barrier
protection and antiviral chemoprophylaxis in hospital staff at
high risk of infection might provide important early pointers
for the use of these interventions in the general population.
We need to identify the key interventions that we wish to evaluate
now, but provide enough flexibility in funding and study organisation
to respond to unexpected challenges and opportunities.
2. Genotype and outcome from major illness There
has been much interest in how genetic regulation of the innate
immune response affects survival from critical illness. At one
extreme, an inadequate proinflammatory cytokine response may result
in high mortality from infection. At the other extreme, an excessive
proinflammatory host response may, in itself, predispose to multiple
organ failure and death.
The ability to define critical components
and optimal levels of the immune response that enable survival
could allow risk stratification and selection of patients for
intensive monitoring and specific anti-inflammatory therapy. However,
clear understanding in this area has been limited by small patient
numbers in most studies, and by the heterogeneity of critical
illness (in terms of aetiology, disease severity, therapy, and
An influenza pandemic presents a unique
opportunity to address these questions in situations where a large
number of individuals (potentially up to 25 per cent of the UK
population) have a stereotyped biological insult, have similar
therapy, and clear outcome measure (mortality). The data that
result from a genetic study which involved only a fraction of
individuals in this setting would be immensely valuable, not only
in the context of influenza, but also for other infections and
for non-infectious critical illness.
1. Complete the ICS/ICNARC/DH calculation
of ICU admission rates for seasonal flu like illness.
2. Individual hospitals to run Flusurge
2.0 for their practice, with a range of inputs decided after the
ICS/ICNARC/DH simulation exercise.
3. The DH should talk to the independent
sector nationally, and encourage individual hospitals to do the
same on a local basis. It is important to identify the ICU facilities
that are available in these hospitals, and agree on ways in which
these can be used in the event of a pandemic.
4. The DH should talk to medical equipment
manufacturers and define current stock levels. We should also
explore ways of rapidly making additional equipment available
(eg by increasing stock levels).
5. Published infection control measures
need to be cascaded down to ICUs, and their preparedness audited
(both in terms of knowledge and availability of equipment).
6. Critical Care Medical and Nursing leaders
to identify areas in which non-ICU staff can provide useful support
- can be within the auspices of the Critical care Contingency
Planning Group. Similarly, the Intensive Care Society should identify
core medical skills that will allow expansion of ICU medical staff.
7. Hospitals should undertake evaluation
of the ability to expand or upgrade the physical space available
for critical care. This may require a specific simulation exercise
which should be funded. Such expansion will include conversion
of Level 2 to Level 3 facilities, and utilisation of operating
theatres and recovery areas. Where geography is favourable, individual
hospitals may consider expanding infrastructure for critical care
(medical gases, suction and protected power supply) to infectious
8. Initiate hospital based cataloguing of
spare ventilatory and monitoring equipment. When operating theatre
and recovery ward equipment are taken into account, every hospital
ought to be able to expand ICU capacity by 50 per cent if elective
surgery is stopped. Consider the case for purchase of small amounts
of additional equipment.
9. Identify the theatre and recovery staff
theatre would be freed by above changes. These staff would be
the first port of call for the ancillary ICU training identified
by 3 above. See if this can allow expansion of ICU nursing
numbers to allow staffing of 150 per cent of current capacity,
even if there was a 25 per cent sickness/absenteeism rate in both
cohorts of staff.
10. The DH should provide additional funds
for medical and nursing staff to undertake core critical care
medicine training, and find ways to ensure their willingness to
be called in the event of an emergency (eg payment of a retainer
to undergo initial and refresher training, and be available in
the event of an emergency; these would be the equivalent of the
Territorial Army for the hospital's critical care services).
11. There needs to be explicit statement
and restatement of the principle that staff involved in the care
of patients with influenza will receive the best available personal
protection, in terms of vaccination, antivirals and personal protective
1. Menon DK, Taylor BL, Ridley SA. Intensive
Care Society, UK. Modelling the impact of an influenza pandemic
on critical care services in England. Anaesthesia 2005; 60: 952-954.
2. Zhang X, Meltzer MI, Wortley P. FluSurge
1.0: a manual to assist state and local public health officials
and hospital administrators in estimating the impact of an influenza
pandemic on hospital surge capacity (Beta test version). Centers
for Disease Control and Prevention, U.S. Department of Health
and Human Services; 2004. http://www.cdc.gov/flu/flusurge.htm
3. Mandell LA, Marrie TJ, Grossman RF et
al. Summary of Canadian Guidelines for the initial management
of community-acquired pneumonia: An evidence-based update by the
Canadian Infectious Disease Society and the Canadian Thoracic
Society. Can Respir J 2000; 7: 371-382.
4. Marrie TJ, Lau CY, Wheeler SL et al.
A controlled trial of a Critical Pathway for treatment of community-acquired
pneumonia. JAMA 2000; 283: 749-755.
5. Angus DC, Marrie TJ, Obrosky DS et al.
Severe community-acquired pneumonia. Use of intensive care services
and evaluation of American and British Society diagnostic criteria.
Am J Respir Crit Care Med 2002; 166; 717-723.
6. Riley PD, Aronsky D, Dean NC. Validation
of the 2001 American Thoracic Society criteria for severe community-acquired
pneumonia. Chest 2004; 32: 2398-2402.
7. Halm EA, Horowitz C, Silver A, et al.
Limited impact of a multicentre intervention to improve the quality
and efficiency of pneumonia care. Chest 2004; 126: 100-107.
9. Qureshi K, Gershon RR, Sherman MF, et
al. Health care workers' ability and willingness to report to
duty during catastrophic disasters. J Urban Health 2005; 82: 378_88.
10. Osterholm MT. Preparing for the next
pandemic. N Engl J Med 2005; 352: 1839_1842.
11. Rubinson L, Nuzzo JB, Talmor DS, et
al. Augmentation of hospital critical care capacity after bioterrorist
attacks or epidemics: Recommendations of the Working Group on
Emergency Mass Critical Care. Crit Care Med 2005; 33: 2393_2403.
12. Personal communication. Frances Chinemana,
Access and Health Protection, NHS Direct.
27 October 2005