Additional information from Mrs Gini Williams
10. Can you describe a whole systems approach
in relation to the control of tuberculosis; are there any general
lessons that can be learned from this for surveillance of other
This paper aims to describe how a whole systems
approach might be beneficial in the fight against tuberculosis
(TB) as well as making suggestions as to how this could be achieved.
Changing the focus from the disease to the patient
In TB control there has been a tendency to focus
on the immediate biomedical aspects of the disease and its management.
Doctors, nurses, microbiologists and consultants of communicable
disease control (CCDCs) are seen as the key individuals involved
with other medical departments (eg genito-urinary medicine and
paediatrics) and external organisations (eg housing and social
services) being regarded to a greater or lesser extent as peripheral.
This is hardly surprising due to the public
health imperative to control the disease, which can be most effectively
achieved by finding and treating active cases. Since we have had
a cure for TB, control efforts have had a predominantly biomedical
focus, the assumption being that once diagnosed with TB every
patient will take the treatment prescribed.
This system works well for people who can easily
access health care but, with the rising levels of TB, it is becoming
apparent that a significant number of patients have difficulties
accessing services for diagnosis and/or treatment. If we fail
to address the broader problems that many TB patients face, such
as homelessness, loss of work, drug or alcohol addiction and so
on, we will not be successful in controlling the disease.
Addressing fragmentation of services
Issues relating to fragmentation of services
and inconsistent approaches, which have been developing in different
areas (see previous evidence) need to be addressed. Services need
to be equitable as well as locally appropriate. A whole systems
approach would bring together the relevant organisations to make
sure that effective and sustainable approaches were developed
according to each local situation.
A "system", in the context of whole
systems working, can be seen as ". . . something that assembles
itself around a shared meaning or purpose" or more generally
as ". . . `a perceived whole whose essential properties arise
from the relationships between its parts'" (Pratt et al,
1999:xiv). Whole systems working tends to be adaptable as
it concentrates on connections, relationships and meaning in contrast
to a mechanical approach which focuses on design and control mechanisms
(Pratt et al, 1999). Adaptability is essential in an environment,
which is constantly changing from the organisation of public services
to the make up of the client group.
Health policy and commissioning
There has been a growing emphasis on partnership
working and collaboration in health care policy over the last
few years and a number of local planning processes have been developed
in order to achieve it: for example joint investment plans, locality
health and social care partnership boards, health improvement
and modernisation programmes and so on. Primary care trusts are
now responsible for commissioning services for TB care and control
and have themselves been developing local mechanisms for partnership
In summary, a whole systems approach would have
a great deal to offer TB control efforts for a number of reasons:
1. It would enable costs to be shared by
a number of organisations and services.
2. Public services are required to work in
partnership with lay people and other organisations.
3. There is a growing recognition that many
people with TB need more than the offer of medication in order
to make a full recovery.
4. TB is increasing and will continue to
do so if we continue to approach it as we have done so far.
5. TB control requires a sustained long term
effort as has been demonstrated in New York and other parts of
6. The system to control TB must be adaptable
both in terms of what it does and the way it works to respond
to ongoing changes appropriately.
TB WOULD LOOK
In order to demonstrate what a whole systems
approach to TB control would look like it is necessary to first
appreciate what the system is trying to achieve. Over the years
in working in different parts of the system I have been developing
what I see as a patient-centred model for TB control (figure 1).
The model aims to demonstrate how the essential elements of TB
control work together to maintain the necessary standardised approach
to treatment and monitoring, while offering individualised patient-centred
Case finding and patient holding are identified
as the two key elements of TB control and management and are presented
in a cyclical format. This is due to the fact that (a) cases are
constantly being found, prompting further screening; and (b) patients'
needs may change while they are on treatment and so regular evaluation
is required. If there is weakness at any point there will be a
risk of losing potential or actual cases. It is therefore essential
to ensure that each element is as strong as the other. Emphasis
on any one of the elements and a lack of attention to another
will lead to a failure in the system.
TO TB CONTROL
Rather than attempt to describe the entire system,
I will give examples of how the principles of whole systems working
can be applied at each stage highlighted in the model (figure
1). On the whole it is about taking a fresh approach to finding
solutions. Thinking carefully about who is or could be involved
at each stage will enable us to see who is involved in the system
as a whole.
It must be noted that this model is itself within
a much wider framework of efforts to fight poverty and reduce
inequality. It is clear that the majority of patients are from
the poorest parts of society both in the UK and worldwide. Efforts
to control TB on the ground will inevitably be hampered by poverty
and deprivation so policies to address these in the UK as well
as in the rest of the world will also help to reduce the overall
incidence of poverty-related diseases such as TB, AIDS and malaria.
Notification of all cases of TB is a legal requirement.
It is vital for initiating contact tracing and for gathering information
for epidemiological purposes and service provision. Notification
of infectious diseases is statutorily the doctor's responsibility,
however, the fact that the TB nurse aims to provide care to all
TB patients from the time they are diagnosed, gives them the ideal
opportunity to organise notification of TB. There is evidence
to suggest that this improves the rate of notification (Pym et
al, 1995) especially where cases are being diagnosed in departments
other than respiratory medicine eg genito-urinary medicine, renal,
In some, more often, lower incidence areas,
TB nurses or health visitors rely on the notification process
to find out about local cases, before they become involved in
the care and initiate contact tracing. As well as producing in
accurate epidemiology, under-notification can lead to patients
being denied access to care and contact screening not taking place.
Contacts of known cases of TB are at the highest risk of developing
TB and additional cases can be missed if screening does not take
At the time of notification, additional information
is collected in the form of enhanced surveillance, which provides
valuable data on trends associated with TB. The data are largely
demographic and disease-related and only clinical outcomes are
assessed. Routine collection of socio-economic data would assist
in both planning services, which can respond appropriately to
a wider range of patients' needs, and implementing more carefully
targeted screening programmes.
Key Points: Appropriate systems are
needed to collect relevant information, with clearly identified
responsibilities for all those involved in diagnosing and treating
TB (inside and outside the NHS).
The screening of high-risk groups is an important
way of finding previously unknown cases of TB in the community.
These cases can then be treated promptly and further contact tracing
can be initiated. Screening is organised according to a person's
risk of exposure to TB.
Contacts of known infectious cases are the most
likely to have been infected and/or to have developed active disease.
Their level of risk will vary according to the amount of exposure
and their general state of health. It has previously been shown
that 10 per cent of tuberculosis cases have been identified through
contact tracing (Kumar et al, 1992, Ormerod 1993). Contact
tracing of children is also particularly important as they are
vulnerable to more serious forms of TB, such as TB meningitis
and miliary TB. Paediatricians may need to be consulted to read
paediatric chest x-rays and manage treatment.
In sputum smear positive index cases, the aim
is to investigate household, occupational and social contacts.
In smear negative cases and children with non-pulmonary TB, only
household contacts may be investigated. Almost without exception,
contact tracing will be the responsibility of the local TB services.
Good collaboration between services is needed when contacts live
over a large geographical area. If a different approach to contact
tracing is practiced in different areas, there may be duplication
or omission and cases may be missed. Apart from this, mixed messages
may be given to people attending different services for their
Screening of other high risk groups
While contact tracing identifies specific individuals
who are known to have been exposed to TB, the screening of other
people considered to be at risk, relies on the identification
of groups or communities where TB is known to be more common.
Their risk can be exacerbated by poor general health and lack
of access to health care facilities. This relies on routinely
collected data to identify high risk groups and target screening
effectively. Having identified who needs screening, it is then
essential to work out how best to organise it.
In the past, the recommendation to screen all
new entrants to the UK from countries of high incidence has been
very difficult to implement due to the fact that information provided
to the chest clinics is often out of date by the time an invitation
is sent. Uptake is often very low and time is wasted looking for
people who have already moved on. As a result there is now a great
variety between services as to who is offered screening and how
it is done. Different methods are being evaluated, for instance,
in Hackney, there is a study looking into screening on registration
with a GP practice and in Newham, there has been a screening project
running in collaboration with the asylum team. It is likely that
a number of different approaches will be needed in all areas in
order to be effective.
The planned acquisition of a mobile digital
x-ray unit, such as those commonly used in Holland, will provide
an opportunity to carry out targeted mass screening in a relatively
quick, cheap and acceptable manner. It will be available to implement
screening in prisons, homeless hostels, and so on all over London,
if not further afield. It will also be available for one-off mass
contact screening in a potential outbreak situation.
In order for any screening activities to function
properly, they have to be fully integrated with existing services.
That is, there has to be a process for planning where and when
screening will take place with the collaboration of local TB services
as well as the institutions involved. Any screening requires collaboration
between numerous partners. Screening initiatives require detailed
preparation and publicity beforehand, and continued commitment
to following up suspected and confirmed cases identified during
the screening. Everyone involved needs to understand their responsibility
before, during and after screening has taken place.
Mass tuberculin skin testing continues to be
carried out when children are offered their routine BCG in school.
Children who have a strong tuberculin reaction are referred to
the local chest clinic. They may have been previously vaccinated
or alternatively they may have been exposed to TB and require
preventive therapy. They are sent for a chest x-ray and depending
on the result and the presence or absence of symptoms, they are
either started on full treatment or are given preventive therapy.
They should then be followed-up regularly by the TB nurses in
collaboration, if necessary with the school nurse.
The role of occupational health departments,
particularly in higher education and organisations recruiting
from overseas is becoming ever more important to TB control. There
should be good links between occupational health departments and
TB services in order to ensure that adequate screening is performed
and appropriate action is taken when TB is suspected. This is
particularly true of health care providers from hospitals to community
trusts. This is required to protect both staff and patients and
Key Points: Contact screening procedures
should be consistent with good communication between centres,
to enable the appropriate management of contacts, if necessary,
across a large geographical area.
Any type of mass screening must be
properly planned (in terms of publicity, process and follow up)
and integrated with the local TB service.
Passive case finding relies on people's general
level of awareness of TB symptoms, having somewhere to go for
the necessary help and the appropriate action being taken. Good
communication with GPs and organisations working with groups such
as the homeless and refugees is essential in order for potential
TB cases to be recognised, referred and investigated quickly.
The earlier a case of TB is found the less likely it is to be
infectious and the more successful (in terms of lasting damage
to the lung and other body tissues) it is to treat.
The highest risk groups often have the poorest
access to both passive and active screening so a network of linked
organisations is required to give appropriate guidance to people
needing help. Good access to primary care which can respond appropriately
to patients presenting with TB symptoms is essential to effective
TB control. A number of chest clinics provide the facility for
the immediate referral of suspect cases picked-up in screening
projects, by GPs, by themselves and so on. This can significantly
cut down on delays in diagnosis thereby reducing the length of
time a patient may be infectious.
Although most attention is paid to groups with
the highest risk of contracting TB, it is vital to guard against
delayed diagnoses in lower risk groups. The latest outbreaks have
occurred in these groups ie where TB has been least expected.
There needs to be a minimum level of awareness in primary care
and local network of community, voluntary and statutory organisations
in order to avoid this.
In order that the service is used appropriately,
it is necessary for TB service staff to act as a resource, providing
advice and information, on the telephone as well as delivering
talks to staff, students, refugee groups, the homeless and organisations
working with people who may be at higher risk of suffering from
TB. In this way more people can be alerted more promptly to potential
cases and can react appropriately. It is also a way of making
the service more accessible to people who would otherwise be suspicious
and to feel apprehensive about attending for screening.
Key Point: Primary care staff and
other key local partners must be regularly updated on TB and the
service available locally.
Prompt diagnosis is essential to controlling
TB in the community. Achieving it, relies on an appropriate index
of suspicion, knowledge of appropriate tests and local referral
procedures as well as good diagnostic services. It is recommended
that physicians and nurses with appropriate training and experience
manage TB and this can only be achieved if there is good networking
across different specialities. As with many other diseases, there
is a need for specialists to manage the particular challenges
associated with TB but people working in related specialities
play a vital role in recognising and diagnosing this disease.
Local diagnostic services including microbiology,
radiology and histology are all essential to the prompt diagnosis
of TB. There needs to be good communication between these and
the TB services with provision for the rapid feedback of results.
This is essential for a number of reasons. Firstly, it allows
for patients to start treatment as soon as possible rendering
them non-infectious and reducing the opportunity for further transmission.
Secondly, it informs decisions regarding infection control ie
whether a patient should remain isolated or not. If not, it is
important that the information is fed back as soon as possible
to the ward in order to "release" the patient and free
up the side room. Thirdly, patients, who are under investigation
for TB can be very anxious and it is important that they receive
rapid feedback regarding their results. External reference laboratories
are also essential for the identification of drug resistant cases.
Diagnosis can become complicated if it involves
different institutions, for instance, when a case is suspected
within the prison service. Not only does there have to be equally
good communication with the local diagnostic services, the mobility
of prisoners both in and out of community and between different
prisons requires careful management. It is essential to have a
clear process for tracking potential and actual cases through
the prison service, with good links to local NHS TB services wherever
the prisoner is based. There is an opportunity to develop a more
robust process than currently exists, when the prison health services
become allied with the NHS in April.
Apart from being managed between more than institution
patients may also have a dual diagnosis which requiring input
from more than one specialist service. TB patients may suffer
from a range of other conditions including drug and alcohol problems,
HIV, diabetes, renal problems, psychiatric problems etc. One condition
cannot be successfully managed without addressing the other and
good collaboration between services is essential in order that
any intervention will accommodate rather than compromise the other.
Key Point: All those involved in
diagnosing TB need to work together to avoid delays and losing
patients. Special care is needed when more than one organisation
(eg prison service, private sector etc) or department is involved.
In order to achieve a successful outcome it
is essential to find out what is important to the patient as well
as what is important to treating the disease. There has been a
tendency to focus on the latter with the result that people have
tried to control the disease by controlling the patient. Although
a standardised treatment regimen should be used, care needs to
be planned on an individual basis. Different patients will have
different concerns and will require a different level of support
in order to complete treatment as well as ideally improving any
circumstances which may have led to the patient becoming ill with
TB in the first place. A patient-centred approach requires patient
involvement and good local networks with strong links to voluntary
and statutory organisations such as housing and social services.
If on assessment it appears likely that the
patient will have difficulties taking the treatment they may be
offered additional support such as directly observed therapy.
This has to be arranged with the agreement of the patient and
if possible offering some flexibility with regard to time and
location of treatment as far as possible to suit the patient.
Staff from other organisations such as district nurses, school
nurses or hostel workers may be involved (with the support of
the local TB nurse) if they are able to offer a more convenient
service to the patient.
Key Point: Holistic individual assessment
of each patient is required in order to develop an appropriate
plan of care which will maximise the chances of a successful outcome.
Care has to be planned with the patientif
a patient is unable to attend at certain times or does not understand
certain instructions then the treatment will fail. It also has
to be realistic and planned in collaboration with any services
being offered, so that patients maintain confidence in the system.
It is no good promising something, which cannot be delivered.
When negotiating the involvement of other services
it is important that they also participate in discussions as to
how TB is managed locally at a strategic level. For example, if
local social services departments are asked to become involved
on a patient to patient basis it will be difficult for them to
see the relevance of that involvement and offer it any type of
priority within their already pressurised workload. Partnership
relies on participants being fully engaged with a clear role to
play as part of the whole system. Local authority departments
and other services need to understand the contribution they make
to TB control from regeneration to case finding to case holding.
Planning can often be compromised by inequities
in services in different areas. For instance, some services are
able to offer incentives such as food and travelcards to patients
attending regularly for directly observed treatment, and others
are not. There continues to be discrepancies regarding the cost
of treatment. Most health care professionals believe that TB treatment
should be free of prescription charges but this is not the case.
A number of clinics try to ease the cost for patients who are
eligible to pay and are invariably on low incomes. Where this
is not the case, patients may be faced with the high costs of
paying for repeat prescriptions for the combination of drugs they
need over a six month period. Cost is a significant barrier to
treatment and with an infectious disease such as TB it should
not be borne by the patient.
The availability of negative pressure isolation
rooms is also an issuenot only the number of rooms available,
but also their monitoring and maintenance. Different hospitals
have different types of rooms which may not, strictly speaking,
meet the criteria of a negative pressure isolation room. For instance,
they may not have a lobby or en suite facilities and the extractor
fan may not be filtered. Some can offer isolated patients TV and
telephones, which should be a minimum requirement, some allow
patients to smoke, while others offer none of the above to ease
the boredom and frustration of isolation.
Key Points: Services which may be
involved in the care of TB patients should also be involved in
planning at a strategic level.
Local services should meet nationally
agreed standards eg with regard to provision and type of isolation
facilities, efforts to reduce costs for patients, incentive schemes
Care should be implemented by those identified
during the planning process and ongoing communication is required
throughout the treatment period. Minimum routine follow-up involves
the patient being seen at least monthly by the TB nurse and at
significant times for the treatment (ie diagnosis, change from
initial to continuation phase and on completion) by a physician
who has specialist knowledge of TB. If any clinical problem arises
at any other time, the patient is referred to the physician.
The TB nurse sees each patient on their caseload
at least monthly to monitor treatment and progress, offer support
and advice and pick up any problems promptly. All patients should
be able to phone the TB nurse specialist directly if they have
any problems. They can then be given information or advice they
need or an earlier appointment can be made to see the physician
if necessary. Clear processes are needed for the follow up of
patients who fail to attend for routine appointments. If support
and advice are accessible and problems are dealt with promptly,
breaks in treatment can be reduced and compliance maintained.
Providing care can be complex if the patient
has a dual diagnosis or lives outside the area covered by the
treatment centre. Each patient should have a case manager who
is able to liaise with a number of colleagues across organisational
and geographical boundaries. The case manager is usually a TB
nurse. While they may not always be the most significant person
involved in a patient's care, nurses can organise the service
so that the patient gets the best support from the most appropriate
person or people at the appropriate time.
Key Point: Flexibility is required
across organisational and geographical boundaries to ensure that
problems are identified and dealt with promptly and appropriately.
Evaluation should occur for each individual
patient, which will in turn provide information as to how the
system is performing overall in relation to its aims. There are
a number of clinical factors such as treatment outcome, drug resistance
levels etc which should be routinely evaluated. As well as evaluating
the patient's clinical progress, there should also be ongoing
evaluation of the appropriateness of the care being offered and
a review of that care according to the patient's needs and circumstances.
As well as during routine follow up, there should be other mechanisms
for patients to give feedback, anonymously if necessary, on the
There should also be the opportunity to evaluate
more structural factors such as the effectiveness of relationships
with other organisations and the need, availability and use of
negative pressure facilities. Useful evaluation requires agreed
standards with processes for performance management and clear
lines of responsibility. Having said this, if the system is working
as a whole, responsibility is shared so that problems, such as
the failure to meet targets, can be addressed in a collective
and supportive way in order to find an appropriate solution.
Key Point: Evaluation of factors
from treatment outcome to the provision of care to liaison with
other organisations, is essential to ensure that the system can
measure and maintain its effectiveness at all levels.
How this can be applied to other diseases
From looking at the different elements involved,
it is clear to see how important it is to be working with a whole
network of services in order to achieve the necessary levels of
case finding and patient holding. The principles are likely to
be similar for other infectious diseases, especially when there
are similarities in terms of the stigma associated with the disease;
social aspects (lifestyle, environment, education, contact etc)
outweigh clinical aspects; and management and patient behaviour
are crucial not only to the patient but also to their "community".
People with infectious diseases must be made
part of the solution and treated as partners so the need for patient-centred
approaches is essential. People with non-infectious diseases can
choose their course of action without affecting others physically,
however, they may cost the health service more by suffering increased
acute episodes. In this sense, management of TB may even have
lessons for non-infectious diseases.
In terms of related infectious diseases, no
strategy for either HIV/AIDS or TB should be independent from
the other as they are so closely associated. The development of
specialist areas has led to different diseases being seen as separate
issues, which has been exacerbated by the need to compete for
limited resources. Within the old health authority structures,
communicable diseases have often fallen between public health
and infection control with little involvement of health care providers.
The Health Protection Agency can help to address this through
close links with primary care trusts and local providers such
as GPs, acute trusts, private providers, local authority organisations
etc. Provision also needs to be made for prison health care when
it comes within the NHS remit. This will be a particular challenge
due to the mobility of prisoners in and out of the community as
well as between different prisons and other institutions.
In Getting Ahead of the Curve (DOH 2002)
the Department of Health acknowledges the need to address the
TB situation on a nationwide basis. This gives us the opportunity
to look with fresh eyes at the issue of TB control rather than
assuming that we simply need more of the same. There is no point
in sorting out one issue while ignoring others as all parts of
the system are related in a number of ways. It is not a case of
either or (eg screening or negative pressure), the whole system
needs strengthening with collaborative planning, costing and implementing.
Once TB control is seen as a whole system it is possible to recognise
how many other agencies are integral rather than peripheral to
the fight against the disease.
This paper has described how diagnosis may be
made and appropriate treatment prescribed but without notification
or appropriate referral to specialist services, care and contact
tracing may not be adequate. It has also pointed out that although
treatment may be started within the TB service and all information
provided, if a patient remains homeless or is suddenly required
to move (eg as a prisoner or an asylum seeker) they may be lost.
A broad range of services involved in case finding as well as
patient holding. If we stop thinking of TB control as a list of
separate tasks and look at it as a whole system, fragmentation
and inconsistency can be addressed and people can be offered the
access to the care they need.
It is essential that the systems we have for
controlling TB and other infectious diseases remain adaptable
and able to respond promptly to inevitable ongoing policy changes.
Reconfiguration of health services change commissioning processes;
asylum law has an impact on the movement, location and, ultimately,
health of asylum seekers; prison health becoming part of the NHS:
all these things present challenges and opportunities for the
management of infectious diseases such as TB.
Department of Health (2002) Getting Ahead
of the Curve: A strategy for combating infectious diseases.
Kumar S, Innis J, Skinner C, Yield from Tuberculosis
Contact Screening in Birmingham, Thorax 1992;42: pp 179-182.
Pratt J, Gordon P, Plampling D (1999) Working
Whole Systems: Putting theory into practice in organisations.
London: King's Fund.
Pym A, Churchill D, Coker R, Gleissberg V, Reasons
for increased incidence of tuberculosis BMJ 311: 570, 1995
Ormerod LP, Tuberculosis Contact Tracing,
Blackburn 1982-90. Respir Med 1993; 87: 127-132.