INT J TUBERC LUNG DIS 14(12): 15181524

2010 The Union

STATE OF THE ART

Beyond accuracy: creating a comprehensive evidence base

for tuberculosis diagnostic tools
G. Mann,* S. B. Squire,* K. Bissell, P. Eliseev, E. Du Toit, A. Hesseling, M. Nicol, A. Detjen,
A. Kritski#
* Liverpool School of Tropical Medicine, Liverpool, UK; International Union Against Tuberculosis and Lung Disease,
Paris, France; Arkhangelsk Regional Anti-tuberculosis Dispensary, Arkhangelsk, Russian Federation; Desmond Tutu
TB Centre, Cape Town, National Health Laboratory Services, Johannesburg, South Africa; # Rede TB, Braslia, Brazil
SUMMARY

The need for a strong and comprehensive evidence base

to support decision making with regard to the implementation of new and improved diagnostic tools and
approaches has been highlighted by a number of stakeholders; these include members of the New Diagnostics
Working Group (NDWG) and the Subgroup for Introducing New Approaches and Tools of the Stop TB Partnership. To compile such evidence in a systematic manner, we have developed an impact assessment framework
(IAF) which links evidence on inputs to outcomes.
The IAF comprises five interconnected layers: effectiveness analysis, equity analysis, health systems analysis,
scale-up analysis and policy analysis. It can be used by
new diagnostics developers and other interested research
teams to collect as much policy-relevant data as possible

prior to, during and after the demonstration phase of

tool development. The evidence collated may be used by
international and national policy makers to support adoption, implementation and scale-up decisions. The TREAT
TB (Technology, Research, Education and Technical Assistance for TB) initiative uses the IAF in its operational
research and field evaluations of new tools and approaches for TB diagnosis. It has also been incorporated
into the NDWGs recent publication: Pathways to better diagnostics for tuberculosis: a blueprint for the development of TB diagnostics. This article describes the IAF
and the process of improving it and suggests next steps
in overcoming the challenges in its implementation.
K E Y W O R D S : impact; evidence; tuberculosis; policy;
diagnostics

Every TB patient must have access to an effective

diagnosis, treatment and cure.
The Global Plan to Stop TB 200620151

not more cost-effective than those already available.

The World Health Organization (WHO) plays a key
role in approving and developing guidelines for the
use of new tools. The policy making process, described
in a recent WHO statement,4 comprises:

THE ABOVE PRINCIPLE is key to attaining the vision of the Stop TB Partnership of seeing a tuberculosis (TB) free world by 2050.1 Access to an effective
diagnosis has long been a concern. Smear microscopy is not sufficiently sensitive to detect tuberculosis
disease in many cases, and particularly not in children and those who are co-infected with the human
immunodeficiency virus. Multi- and extensively drugresistant TB present new diagnostic and treatment
challenges. Thankfully, new diagnostic approaches
using existing tools have been recommended and new
tools are in the pipeline.2
Any new approach or tool must be evaluated before being adopted by national tuberculosis programmes (NTPs); huge sums of money are already
spent on TB diagnosticsestimated at more than
$1 billion per year globally3and resource-poor countries cannot afford to invest in interventions that are

1 identifying the need for a policy change (e.g., the

emergence of a new technology);
2 reviewing the evidence (e.g., through commissioning systematic reviews);
3 convening an expert panel to review evidence using the GRADE approach (Grading of Recommendations Assessment, Development and Evaluation, see BMJ 200859);
4 assessing draft policies and guidelines (through the
Strategic and Technical Advisory Group for TB,
STAG-TB); and
5 formulating and disseminating new policies and
guidelines.
Recent papers by Pai et al. have noted that systematic reviews, and hence the evidence reviewed in the
above policy development process, have concentrated

Correspondence to: Gillian Hazel Mann, Liverpool School of Tropical MedicineCRESTHA, Pembroke Place Liverpool
L3 5QA, UK. Tel: (+44) 151 705 3139. Fax: (+44) 151 705 3743. e-mail: [email protected]
[A version in French of this article is available from the Editorial Office in Paris and from the Union website www.theunion.org]

Beyond accuracy: a comprehensive evidence base

mainly on test accuracy.10,11 While such data are necessary, they are not sufficient to assess the contribution
new diagnostics can make to the universal access requirements outlined in the Global Plan to Stop TB. A
number of stakeholders, the Subgroup for Introducing New Approaches and Tools (INAT) and the New
Diagnostics Working Group (NDWG) of the Stop TB
Partnership, among others,1113 have called for a strong
and comprehensive evidence base to support decision
making with regard to implementation of new and
improved diagnostic tools.
The NDWG has published Pathways to better diagnostics for tuberculosis: a blueprint for the development of TB diagnostics,2 which outlines the required
phases from needs assessment through test development to assessment of epidemiological impact, and
all stages in between (Figure). The Stop TB Partnerships Retooling Taskforce, a precursor to INAT, stipulated the need for evidence that captures not only
the benefits of new tools but also the risks and health
systems implications associated with them.14 This
range of evidence is encapsulated in the Organisation
for Economic Cooperation and Development definition of impact subscribed to by multi- and bilateral
donors who have signed the Paris Declaration on Aid
Effectiveness (2005), which states that impact consists of:
[The] positive and negative long-term effects on
identifiable population groups produced by a development intervention, directly or indirectly, intended or unintended. These effects can be economic, socio-cultural, institutional, environmental,
technological or of other types.15

Thus, measuring the impact of a diagnostic tool or

approach involves assessing its positive and negative
effects on different stakeholders (patients, health systems, laboratories, etc).
This entails summarising evidence not only about
the tests accuracy, but also its effectiveness in field

1519

conditions in terms of diagnosing patients with various TB presentations, especially for the most infectious, and ensuring that they start appropriate treatment, its affordability, ease of implementation and
potential for scale-up (for the health system) and accessibility (especially to poor and vulnerable TB suspects). Articulating and communicating this overall
impact succinctly and with sufficient evidence is essential. It is required for national policy makers to make
rational decisions about which new diagnostic tests
and approaches to adopt, when and how to implement them, how to manage and finance them and how
to ensure sustainable access and appropriate use.14
To compile such evidence in a systematic manner,
we have developed an impact assessment framework
(IAF) that links evidence on inputs to outcomes. This
framework has been included in the NWDGs blueprint and has been adopted by the TREAT TB (Technology, Research, Education and Technical Assistance
for TB) initiative for use in its operational research
and field evaluations of new tools and approaches
that are at late stages of development or have recently
achieved international approval for use in TB diagnosis and treatment.
The present study describes the IAF, provides
examples of how it can be used and suggests
means of overcoming the remaining challenges in its
implementation.

THE IMPACT ASSESSMENT FRAMEWORK

The IAF has been developed by a multidisciplinary
team at the Liverpool School of Tropical Medicine
and collaborators, including clinicians, laboratory specialists, health economists, social scientists and health
systems analysts. It is based on a range of prior research activities in different countries that supported
different elements of the evidence base.1625 These elements have been combined to provide an overarching

Figure Pathway for the development of tuberculosis diagnostics, from needs assessment to delivery [reproduced from reference 2
with permission from the World Health Organization].

1520

The International Journal of Tuberculosis and Lung Disease

framework (the IAF) to indicate how sufficient information for policy decisions could be collected in a
systematic manner for all new diagnostic tools and
approaches. The sufficiency of information has been
considered in line with the international targets of the
Global Plan to Stop TB and the Millennium Development Goals (MDGs).26 The IAF, with references relating to different types of evidence, is shown in Table 1.
The IAF comprises five interconnected layers:
1
2
3
4
5

Layer 1: Effectiveness analysis

Layer 2: Equity analysis
Layer 3: Health systems analysis
Layer 4: Scale-up analysis
Layer 5: Policy analysis

Table 1

The impact assessment framework

Layer of
assessment

Kinds of question(s)
being addressed

References
to studies
addressing
these
questions

Layer 1
How well does the new tool work in
Effectiveness
terms of accuracy?
analysis
How many additional cases will be
identied who would otherwise not
have been identied?
How many additional cases will actually
start (and complete) treatment as a
result of using the new tool?

16

Layer 2
Equity
analysis

27

Who benets from the new tool

(ambulant vs. hospitalised, poor/less
poor, men/women, adults/children)?
Why do these benets accrue (level
health system in which new
diagnostic is deployed, change in
time to issue of results, change in
patient costs)?

20

21

22

Layer 3
What are the human resource impliHealth system cations of introducing the new tool
analysis
(training, number and cadre of staff)?
What are the infrastructure implications
(equipment, laboratory layout, safety
installations)?
What are the procurement implications
(reagents, consumables,
documentation)?
What are the implications for quality
assurance (internal and external)?

19

Layer 4
Scale-up
analysis

What are the projected impacts of

going to scale with the new tool?
1 Cost savings to patients in relation
to income
2 Cost savings to health providers/
the health system
3 Effects on transmission of
improved infection control as a
result of the new tool

18

Layer 5
Policy
analysis

What other similar technologies are

available or likely to become
available?
How do similar existing or emerging
technologies compare in their
projected performance within each
of the layers above?

29

23

28

17

25

Layer 1: Effectiveness analysis

This layer requires evidence about the accuracy (sensitivity and specificity) of new tools and approaches,
but also flags the need to go further than this and
build evidence on effectiveness. Data on sensitivity
and specificity are universally provided by developers
of new diagnostics, and their positive and negative
predictive values have been suggested by GRADE as
proxies for patient-important outcomes in the assessments of new tools. However, estimations of the
number of patients who might start and complete appropriate treatment are typically calculated by extrapolating these parameters, rather than relying on
evidence from field trials to provide estimates of actual numbers. All too often, diagnostic evaluations
assess new tests solely in terms of their diagnostic potential (accuracy), which may not always translate
into appropriate clinical or public health management
decisions for patients within the context of health
services (effectiveness).
Layer 2: Equity analysis
This layer examines who benefits from the new intervention. The Global Plan to Stop TB highlights the
need to prioritise the needs of the poor and vulnerable, recognising that the greatest burden of TB is
found among poor people, who also face the greatest
barriers in access to care.22 Typically, however, the
systematic measurement of equity in health and
health interventions is either absent or sporadic. Although the first MDG is expressed in terms of an equitable outcome, the health and other goals that are
intended to contribute to this make no specific reference to equity or distributional issues.30
Layer 3: Health systems analysis
This layer examines the health systems requirements
of a new intervention, for example human resources,
infrastructure, operating procedures, quality assurance, procurement and maintenance.
These data are sometimes collected during the
demonstration studies (Figure)studies in optimised
operational settingsof new diagnostics, but not in
all cases. Even where they are collected, the improvement to operations necessarily provided through the
demonstration study may mask issues that become
apparent in implementation (real world) studies. This
layer provides crucial data for assessing the feasibility of implementation and for identifying where key
constraints, or bottlenecks, in the system may occur.
Layer 4: Scale-up analysis
This layer projects and models the full economic
costs as well as the clinical and epidemiological effects of going from demonstration or implementation studies to full scale (national or regional) with a
new intervention. Health system, patient and societal

Beyond accuracy: a comprehensive evidence base

perspectives are all important here. Modelling techniques can provide information concerning the epidemiological benefits of scaling up and, when combined
with patient costs from Layer 2, total additional costs
or savings for patients. At the same time mathematical systems analysis techniques can outline the potential constraints to and resources required for scale-up.
When combined with cost analyses from Layer 3, these
can give an indication of total resources required as
well as identify and quantify likely resource gaps.
Layer 5: Policy analysis
This layer critically appraises the new intervention
studied in Layers 14 against other interventions that
are available or may become available for uptake in
the short to medium term. An important part of this
layer is a scoping of the risk that a given new diagnostic test may be supplanted by newer technology
within a short period of time. It requires a rapid assessment of data on other pipeline diagnostics from
the previous four layers and a review of whether
changes made for one diagnostic may provide a better platform for the next technology or, alternatively,
whether the new technology is disruptive,31 or market transformational,32 both terms used to describe a
technology that could radically alter the way in which
TB diagnosis is achieved.

USING THE IAF

The IAF can be used by diagnostics research teams
during the demonstration and evidence for scale-up,
delivery and access phases of development shown in
the Figure. The latter may take the form of field evaluation, or implementation, studies in non-optimised
settings, or of other operational research activities.
The framework can also be used by international policy makers during the policy development process to
systematically assess a broader range of evidence, and
by national policy makers to support adoption, implementation and scale-up decisions.
The IAF has already been used for the development
of protocols for a multi-country research programme
to study the implementation of line-probe assays
(LPAs), which were recommended by WHO STAGTB in 2008.33 Representatives from three countries
(Russia, Brazil and South Africa), all clinicians or laboratory specialists, with other members of the TREAT
TB core group, discussed the priority research questions they would like to answer with regard to the
use of LPAs, and mapped these questions to each
layer of the IAF. All the questions raised mapped to
one layer of the framework, and all layers were addressed; the resulting framework is shown in Table 2.
Each of these teams now has a different protocol for
collecting the evidence, due to the stage at which their
NTPs are with regard to rolling out LPA. Nevertheless, each will provide data against the same set of

1521

Table 2 Use of the IAF for designing LPA eld studies

Layer of
assessment

Kinds of questions being addressed: questions and

issues raised by multi-country research teams

Layer 1
Effectiveness
analysis

How many additional cases will be identied who

would otherwise not have been identied?
How many additional cases will actually start
treatment/achieve cure/avoid death as a result of
using LPAs?
What will be the effect on tuberculosis
transmission?
How will LPA affect the timeliness in results
inuencing a clinical or treatment decision?
Layer 2
Who is beneting from LPA implementation
Equity
and why?
analysis
Is the test sufciently accurate for all patients?
What are the risks to patients/others?
What costs will patients face?
How acceptable is the test to patients?
Are there inequalities in access to LPA?
Layer 3
What is the effectiveness and/or efciency from a
Health system
health system perspective?
analysis
What effect will LPA have on how cases are
managed in the health system?
What quality assurance mechanisms need to be
in place?
What information systems need to be in place?
What are the human resource requirements in the
health system?
What are the laboratory issues (including
infrastructure, e.g., utilities, space; personnel,
e.g., numbers and skills; monitoring system for
laboratory)?
How will the challenge of mixed infections be
addressed?
What are the safety issues?
How will the results be interpreted and
standardised?
Layer 4
What are the obstacles to the rollout?
Scale-up
What are the human resource and training
analysis
requirements for full national scale-up?
Layer 5
How does LPA compare with conventional old
Policy
methods vs. other new methods that may be
analysis
available in the short to medium term?
How does LPA interface with other existing and
new diagnostics that will be recommended and
implemented in the future (e.g., GeneXpert)?34
Should routine drug susceptibility testing be
completely dropped and replaced by LPA?
IAF = impact assessment framework; LPA = line-probe assay.

outcome indicators, facilitating comparisons across

different epidemiological settings.
The central methodology that we advocate to feed
robust evidence into Layers 13 is the prospective
randomised controlled trial (RCT). This design permits comparison between the existing technology
and approach (control) and the new (intervention),
as follows:
For Layer 1, a comparison of effects on 1) numbers
of patients achieving important outcomes (including
diagnosis, start of treatment and treatment completion), and 2) time to achieving these outcomes.
For Layer 2, a comparison of effects on different
patient sub-groups (e.g., poor vs. less poor, adults vs.
children). Equity may be assessed based on outcome
indicators among different groups, in terms of morbidity or mortality measures, or process indicators such

1522

The International Journal of Tuberculosis and Lung Disease

as health service use.30 Analysis of socio-economic status may use asset-based measures to define different
socio-economic groups.35 Demographic and health
surveys and more recent TB prevalence surveys are
increasingly using these methods.36,37
For Layer 3, a comparison of the health system inputs is required. Data for this may be gained through
economic analyses of standard vs. new diagnostic interventions, focusing on the health system and not
just the tool, and through interviews with health systems personnel.
Data for these comparisons can be obtained across
all study participants in both intervention and control arms, or through nested sub-studies on more limited numbers. For example, in-depth qualitative and
quantitative research on patient costs incurred during a diagnostic process (either control or intervention) is time consuming, and data are thus only collected for a subgroup of study participants. Data
from Layers 13 can then be fed into the modelling
and other methodologies required in Layers 4 and 5.
We recognise that the type of randomised trial
employed will depend on the stage of diagnostic development to which the IAF is being applied. During demonstration studies (which may be conducted
prior to STAG-TB approval), an explanatory RCT
with well-controlled study conditions and data collection instruments is appropriate. During subsequent
implementation or operational research, a pragmatic
RCT (PRCT) approach using existing health system
data will be more suitable (for a fuller description of
the difference between explanatory and pragmatic
RCTs see Zwarenstein et al.38) There are concerns
that RCT designs deny some patients (those in the control arm) the assumed benefits of a new technology
especially in the implementation research of STAGapproved technologies. Such ethical concerns need
to be addressed, for example by ensuring that the
PRCT includes a scale-up plan, such as through a
step wedged approach in which all sites access the
technology, but in a phased manner, to allow for
comparisons between those with and those without
the technology.

NEXT STEPS AND OVERCOMING

CHALLENGES TO USING THE IMPACT
ASSESSMENT FRAMEWORK
The framework will continue to be revised as experience in using it for research design and implementation continues. It will have value for other diagnostics
tools and also for drugs and vaccines. The research
methodologies for addressing each of the different
layers are under constant development. As the multidisciplinary research teams needed to implement these
methodologies are currently uncommon in many
countries, capacity building involving training, men-

toring and partnership between service delivery programmes, academic organisations and patient organisations will be required. The increased focus on
patient-centred outcomes in particular will provide
opportunities for patient representatives and organisations to become more engaged in the research process. If patient groups are empowered to collect and
analyse relevant dataparticularly in Layer 2it
will give them a greater voice in policy decisionmaking at the national and international levels.
When it was originally developed, the IAF was envisaged as being applicable to single new tests. However, as we move further into implementation research it is clear that it will also need to be applied to
packages of tests, or combinations of existing and
novel tests, along with all the additional inputs required to introduce such packages and combinations
in different algorithms; this challenge is currently being addressed under the TREAT TB initiative.
The questions in the different layers of the IAF do
not all necessarily carry equal weight in any given circumstance. For example, a new test for detecting
drug resistance that is best suited for deployment in a
central reference laboratory may be more important
in monitoring drug resistance patterns than in directly
improving patient access. The questions about which
patient group or type of patient benefits (Layer 2) may
then assume lesser importance, whereas these may be
key research questions in a diagnostic approach or
test that is aimed at point of care.
We also recognise that while the IAF provides a
body of evidence for policy makers, evidence alone is
often not the only driver of policy change; process,
context and sometimes subjective factors, for example expert or political opinion, can also play a substantial role. These factors need clearer and more systematic documentation and analysis in the process
of implementation research. This is the subject of a
forthcoming study by Bissell et al.
There are concerns that accumulating a comprehensive evidence base such as the one we advocate
here will take too long and be too costly; rather than
promoting the rational uptake of new technologies, it
will instead impede the introduction of much needed
innovations. Such concerns are valid, but they must
be balanced against the dangers of the premature
introduction of tools into unprepared and underresourced health systems, often as a result of lobbying or forceful marketing. To counter both sides of
this argument, research and implementation partners
need to come together and collaborate on an unprecedented scale, and with a renewed sense of urgency.
By directly addressing the concerns of policy makers
through the research process, the adoption and implementation of new tools should be achieved more
rapidly, sustainably and with beneficial effects for affected populations.

Beyond accuracy: a comprehensive evidence base

Acknowledgements
The authors thank the New Diagnostics Working Group (NDWG)
for funding the initial development of this framework and the
United States Agency for International Development for funding
the TREAT TB (Technology, Research, Education and Technical
Assistance for TB) initiative, which has enabled its evolution. They
also thank all those people who have contributed to the process of
developing the impact assessment framework outlined in this article; these include R Thomson, R Beddell, M Van Lettow, A Harries,
R Dacombe, members of the NDWG and members of the TREAT
TB management team.

References
1 Stop TB Partnership, World Health Organization. The global
plan to stop TB 20062015. Geneva, Switzerland: WHO, 2006.
2 World Health Organization Stop TB Partnership New Diagnostics Working Group. Pathways to better diagnostics for tuberculosis: a blueprint for the development of TB diagnostics.
Geneva, Switzerland: WHO, 2009. http://whqlibdoc.who.int/
publications/2009/9789241598811_eng.pdf Accessed September 2010.
3 World Health Organization Special Programme for Research
and Training in Tropical Diseases/Foundation for Innovative
New Diagnostics. Diagnostics for tuberculosis: global demand
and market potential. Geneva, Switzerland: WHO, 2006.
4 World Health Organization. Moving research findings into new
WHO policies. Geneva, Switzerland: WHO, 2008. http://www.
who.int/tb/dots/laboratory/policy/en/index4.html Accessed September 2010.
5 Guyatt G H, Oxman A D, Vist G, et al. Rating quality of evidence and strength of recommendations GRADE: an emerging
consensus on rating quality of evidence and strength of recommendations. BMJ 2008; 336: 924926.
6 Guyatt G H, Oxman A D, Kunz R, et al. Rating quality of evidence and strength of recommendations: what is quality of
evidence and why is it important to clinicians? BMJ 2008; 336:
995998.
7 Schunemann H J, Oxman A D, Brozek J, et al. Grading quality
of evidence and strength of recommendations for diagnostic
tests and strategies. BMJ 2008; 336: 11061110.
8 Guyatt G H, Oxman A D, Kunz R, et al. Rating quality of evidence and strength of recommendations: incorporating considerations of resources use into grading recommendations. BMJ
2008; 336: 11701173.
9 Guyatt G H, Oxman A D, Kunz R, et al. Rating quality of evidence and strength of recommendations: going from evidence
to recommendations. BMJ 2008; 336: 10491051.
10 Pai M, Ramsay A, OBrien R. Evidence-based tuberculosis
diagnosis. PLoS Med 2008; 5: 10431049.
11 Pai M, Minion J, Steingart K, Ramsay A. New and improved
tuberculosis diagnostics: evidence, policy, practice, and impact.
Curr Opin Pulm Med 2010; 16: 271284.
12 Pai M, Ramsay A, OBrien R J. Comprehensive new resource
for evidence-based TB diagnosis. Expert Rev Mol Diagn 2009;
9: 637639.
13 Marais B J, Raviglione M C, Donald P R, et al. Scale-up of
services and research priorities for diagnosis, management, and
control of tuberculosis: a call to action. Lancet 2010; 375:
21792191.
14 Stop TB Partnership, World Health Organization Retooling
Task Force. New technologies for tuberculosis control: a framework for their adoption, introduction and implementation.
WHO/HTM/STB/2007.40. Geneva, Switzerland: WHO, 2007.
15 Organisation for Economic Cooperation and Development
Development Assistance Committee (OECD-DAC). Glossary of
key terms in evaluation and results-based management proposed harmonized terminology. 2002. Paris, France: OECD-

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

1523

DAC, 2002. http://www.undg.org/index.cfm?P=224 Accessed

September 2010.
Wilkinson D, Newman W, Reid A, Squire S B, Sturm A W, Gilks
C F. Trial-of-antibiotic algorithm for the diagnosis of tuberculosis in a district hospital in a developing country with high
HIV prevalence. Int J Tuberc Lung Dis 2000; 4: 513518.
Mundy C J F, Harries A D, Banerjee A, Salaniponi F M L, Gilks
C F, Squire S B. Quality assessment of sputum transportation,
smear preparation and AFB microscopy in a rural district in
Malawi. Int J Tuberc Lung Dis 2002; 6: 4754.
Mundy C J F, Bates I, Nkhoma W, et al. The operation, quality
and costs of a district hospital laboratory service in Malawi.
Trans R Soc Trop Med Hyg 2003; 97: 403408.
Squire S B, Belaye A K, Kashoti A, et al. Lost smear-positive
pulmonary tuberculosis cases: where are they and why did we
lose them? Int J Tuberc Lung Dis 2005; 9: 2531.
Ramsay A, Squire S B, Siddiqi K, Cunningham J, Perkins M.
The bleach microscopy method and case detection for TB control. Int J Tuberc Lung Dis 2006; 10: 256258.
Cambanis A, Yassin M A, Ramsay A, Squire S B, Arbide I, Cuevas L E. A one-day method for the diagnosis of pulmonary tuberculosis in rural Ethiopia. Int J Tuberc Lung Dis 2006; 10:
230232.
Kemp J R, Mann G H, Nhlema Simwaka B, Salaniponi F M L,
Squire S B. Can Malawis poor afford free TB services? Patient and household costs associated with a TB diagnosis in
Lilongwe. Bull World Health Organ 2006; 85: 580585.
Liu X, Thomson R, Gong Y, et al. How affordable are TB diagnosis and treatment in rural China? An analysis from community and TB patient perspectives. Trop Med Int Health 2007;
12: 14641471.
Theobald S, Taegtmeyer M, Squire S B, et al. Towards building
equitable health systems in sub-Saharan Africa: what can we
learn from case studies on operational research? Health Res
Pol Syst 2009; 7: 26.
Ramsay A, Cuevas L E, Mundy C, et al. New policies, new
technologies: modelling the potential for improved smear microscopy services in Malawi. PLoS ONE 2009; 4: e7760.
World Health Organization, The Stop TB Partnership. The
Stop TB strategy: building on and enhancing DOTS to meet
the TB-related Millennium Development Goals. Geneva, Switzerland: WHO, 2006.
Yagui M, Perales M T, Asencios L, et al. Timely diagnosis of
MDR-TB under program conditions: is rapid drug susceptibility testing sufficient? Int J Tuberc Lung Dis 2006; 10: 838843.
Ramsay A, Bonnet M, Gagnidze L, Githui W, Varaine F, Guerin
P J. Sputum sex and scanty smears: new case definition may reduce sex disparities in smear-positive tuberculosis. Int J Tuberc
Lung Dis 2009; 13: 613619.
Harries A D, Michongwe J, Nyirenda T E, et al. Using a bus
service for transporting sputum specimens to the Central Reference Laboratory: effect on the routine TB culture service in
Malawi. Int J Tuberc Lung Dis 2004; 8: 204210.
Gwatkin D R. Health inequalities and the health of the poor:
what do we know? What can we do? Bull World Health Organ
2000; 78: 318.
Christensen C M. The innovators dilemma: when new technologies cause great firms to fail. Boston, MA, USA: Harvard
Business School Press, 1997.
Blumstein C, Goldstone S, Lutzenhiser L. A theory-based approach to market transformation. Energy Policy 2000; 28: 137
144.
World Health Organization. New WHO policies. Geneva, Switzerland: WHO, 2010. http://www.who.int/tb/dots/laboratory/
policy/en/index.html Accessed September 2010.
Blakemore R, Story E, Helb D, et al. Evaluation of the analytical performance of the Xpert MTB/RIF Assay. J Clin Microbiol
2010; 48: 24952501.
Wagstaff A. Socio-economic inequalities in child mortality:

1524

The International Journal of Tuberculosis and Lung Disease

comparisons across developing countries. Bull World Health

Organ 2000; 78: 1929.
36 Lnnroth K, Holtz T H, Cobelens F, et al. Inclusion of information on risk factors, socio-economic status and health seeking
in a tuberculosis prevalence survey. Int J Tuberc Lung Dis
2009; 13: 171176.

37 Muniyandi M, Rajeswari R. Socio-economic inequalities of tuberculosis in India. Expert Opin Pharmacother 2008; 9: 1623
1628.
38 Zwarenstein M, Treweek S, Gagnier J J, et al. Improving the
reporting of pragmatic trials: an extension of the CONSORT
statement. BMJ 2008; 337: a2390.

RSUM

La ncessit dune base de preuves solide et complte

pour servir la prise de dcisions en ce qui concerne la
mise en uvre doutils de diagnostic et dapproches nouvelles et amliores a t souligne par un certain nombre de responsables ; parmi ceux-ci, des membres du
groupe de travail sur les nouveaux outils diagnostiques
New Diagnostics Working Group (NDWG) et du sousgroupe pour lintroduction dapproches et doutils nouveaux (Subgroup for Introducing New Approaches and
Tools) du Partenariat Stop TB. Afin de rassembler ces
vidences de manire systmatique, nous avons labor
un rseau dvaluation dimpacts (IAF) qui fait le lien
entre apports et rsultats finaux.
LIAF comporte cinq couches interconnectes : analyse defficience, analyse dquit, analyse des systmes de
sant, analyse de lextension et analyse de la politique. Il
peut tre utilis par ceux qui laborent de nouvelles
techniques de diagnostic et par dautres quipes de re-

cherche intresses rassembler autant de donnes possibles en rapport avec la politique suivre avant, pendant et aprs la phase de dmonstration de llaboration
de loutil. Les vidences rassembles peuvent tre utilises par les dcideurs politiques internationaux et nationaux pour soutenir des dcisions dadoption, de mise
en uvre et dextension. Linitiative TREAT TB (Technologie, Recherche, Education et Assistance Technique)
utilise lIAF dans sa recherche oprationnelle et dans ses
valuations sur terrain des nouveaux outils et des nouvelles approches du diagnostic de la tuberculose ; lIAF a
t incorpor dans la publication rcente du NDWG :
Pathways to better diagnostics for tuberculosis: a blueprint for the development of TB diagnostics . Cet article dcrit lIAF et les processus employs pour son amlioration et suggre les tapes ultrieures pour surmonter
les dfis que comporte sa mise en uvre.

RESUMEN

Varios interesados directos, entre ellos el Grupo de Trabajo sobre Nuevos Diagnsticos (NDWG) y el subgrupo
de introduccin de nuevos enfoques e instrumentos de
la Alianza Alto a la Tuberculosis, han destacado la necesidad de contar con una base de datos cientficos slida
y exhaustiva, a fin de respaldar la toma de decisiones
relacionadas con la introduccin de nuevos enfoques e
instrumentos perfeccionados de diagnstico. Con el
objeto de recoger estos datos de manera sistemtica, se
ha diseado un marco de evaluacin del impacto (IAF),
que vincula los datos aportados con los resultados
obtenidos.
El marco de evaluacin del impacto comporta cinco
estratos interconectados: el anlisis de eficacia, el anlisis de equidad, el anlisis de los sistemas de salud, el
anlisis de la ampliacin de escala y el anlisis de las
polticas. Este marco pueden usarlo los creadores de
nuevos mtodos diagnsticos y otros grupos cientficos
interesados, durante el desarrollo de un nuevo instrumento con el fin de recoger la mxima cantidad de datos

pertinentes a las polticas, antes de la fase de demostracin, durante la misma o despus de ella. Los datos
cientficos recogidos pueden ser tiles a los encargados
de definir las polticas a escala nacional o internacional,
a fin de respaldar las decisiones de adopcin, ejecucin
o ampliacin de escala. En la iniciativa TREAT TB (Tecnologa, Investigacin, Educacin y Asistencia Tcnica
para la Tuberculosis) se aplica el IAF en las evaluaciones
de la investigacin operativa y de terreno de los nuevos
instrumentos y estrategias utilizados en el diagnstico
de la tuberculosis. Tambin se ha incorporado en la
publicacin reciente del NDWG: Pathways to better diagnostics for tuberculosis: a blueprint for the development of TB diagnostics (Estrategias encaminadas a mejorar los mtodos diagnsticos de la tuberculosis: un
plan de accin para el desarrollo de medios diagnsticos). En el presente artculo se describe el IAF y los mecanismos que permiten mejorarlo y se sugieren nuevos
pasos que contribuyan a superar las dificultades que
plantea su ejecucin.