A modelling framework to support the selection and implementation of new tuberculosis diagnostic tools [State of the art series. Operational research. Number 8 in the series]

Strengthening health systems to improve the value of tuberculosis diagnostics in South Africa: A cost and cost-effectiveness analysis

BACKGROUND

In South Africa, replacing smear microscopy with Xpert-MTB/RIF (Xpert) for tuberculosis diagnosis did not reduce mortality and was cost-neutral. The unchanged mortality has been attributed to suboptimal Xpert implementation. We developed a mathematical model to explore how complementary investments may improve cost-effectiveness of the tuberculosis diagnostic algorithm.

METHODS

Complementary investments in the tuberculosis diagnostic pathway were compared to the status quo. Investment scenarios following an initial Xpert test included actions to reduce pre-treatment loss-to-follow-up; supporting same-day clinical diagnosis of tuberculosis after a negative result; and improving access to further tuberculosis diagnostic tests following a negative result. We estimated costs, deaths and disability-adjusted-life-years (DALYs) averted from provider and societal perspectives. Sensitivity analyses explored the mediating influence of behavioural, disease- and organisational characteristics on investment effectiveness.

FINDINGS

Among a cohort of symptomatic patients tested for tuberculosis, with an estimated active tuberculosis prevalence of 13%, reducing pre-treatment loss-to-follow-up from ~20% to ~0% led to a 4% (uncertainty interval [UI] 3; 4%) reduction in mortality compared to the Xpert scenario. Improving access to further tuberculosis diagnostic tests from ~4% to 90% among those with an initial negative Xpert result reduced overall mortality by 28% (UI 27; 28) at $39.70/ DALY averted. Effectiveness of investment scenarios to improve access to further diagnostic tests was dependent on a high return rate for follow-up visits.

INTERPRETATION

Investing in direct and indirect costs to support the TB diagnostic pathway is potentially highly cost-effective.

Building resource constraints and feasibility considerations in mathematical models for infectious disease: A systematic literature review

•

Mathematical model capabilities to explore complex systems now enable priority-setting to consider local resource constraints.

•

Common objectives of model-based analyses incorporating constraints are to assess real-world feasibility or allocate resources efficiently.

•

Constraints may be incorporated via (i) model-based estimation; (ii) linkage of mathematical and health system models; or (iii) optimisation.

•

Models can then project constrained intervention effects and costs and resource requirement s for delivering interventions at full scale.

•

'Health system constraints' should be systematically defined for routine operationalisation in model-based priority-setting.

Priority setting for infectious disease control is increasingly concerned with physical input constraints and other real-world restrictions on implementation and on the decision process. These health system constraints determine the ‘feasibility’ of interventions and hence impact. However, considering them within mathematical models places additional demands on model structure and relies on data availability. This review aims to provide an overview of published methods for considering constraints in mathematical models of infectious disease.

We systematically searched the literature to identify studies employing dynamic transmission models to assess interventions in any infectious disease and geographical area that included non-financial constraints to implementation. Information was extracted on the types of constraints considered and how these were identified and characterised, as well as on the model structures and techniques for incorporating the constraints.

A total of 36 studies were retained for analysis. While most dynamic transmission models identified were deterministic compartmental models, stochastic models and agent-based simulations were also successfully used for assessing the effects of non-financial constraints on priority setting. Studies aimed to assess reductions in intervention coverage (and programme costs) as a result of constraints preventing successful roll-out and scale-up, and/or to calculate costs and resources needed to relax these constraints and achieve desired coverage levels. We identified three approaches for incorporating constraints within the analyses: (i) estimation within the disease transmission model; (ii) linking disease transmission and health system models; (iii) optimising under constraints (other than the budget).

The review highlighted the viability of expanding model-based priority setting to consider health system constraints. We show strengths and limitations in current approaches to identify and quantify locally-relevant constraints, ranging from simple assumptions to structured elicitation and operational models. Overall, there is a clear need for transparency in the way feasibility is defined as a decision criteria for its systematic operationalisation within models.

Bringing a health systems modelling approach to complex evaluations: multicountry applications in HIV, TB and malaria

INTRODUCTION

Understanding how to deliver interventions more effectively is a growing emphasis in Global Health. Simultaneously, health system strengthening is a key component to improving delivery. As a result, it is challenging to evaluate programme implementation while reflecting real-world complexity. We present our experience in using a health systems modelling approach as part of a mixed-methods evaluation and describe applications of these models.

METHODS

We developed a framework for how health systems translate financial inputs into health outcomes, with in-country and international experts. We collated available data to measure framework indicators and developed models for malaria in Democratic Republic of the Congo (DRC), and tuberculosis in Guatemala and Senegal using Bayesian structural equation modelling. We conducted several postmodelling analyses: measuring efficiency, assessing bottlenecks, understanding mediation, analysing the cascade of care and measuring subnational effectiveness.

RESULTS

The DRC model indicated a strong relationship between shipment of commodities and utilisation thereof. In Guatemala, the strongest model coefficients were more evenly distributed. Results in Senegal varied most, but pathways related to community care had the strongest relationships. In DRC, we used model results to estimate the end-to-end cost of delivering commodities. In Guatemala, we used model results to identify potential bottlenecks and understand mediation. In Senegal, we used model results to identify potential weak links in the cascade of care, and explore subnationally.

CONCLUSION

This study demonstrates a complementary modelling approach to traditional evaluation methods. Although these models have limitations, they can be applied in a variety of ways to gain greater insight into implementation and functioning of health service delivery.

Predicting the impact of patient and private provider behavior on diagnostic delay for pulmonary tuberculosis patients in India: A simulation modeling study

BACKGROUND

Tuberculosis (TB) incidence in India continues to be high due, in large part, to long delays experienced by patients before successful diagnosis and treatment initiation, especially in the private sector. This diagnostic delay is driven by patients' inclination to switch between different types of providers and providers' inclination to delay ordering of accurate diagnostic tests relevant to TB. Our objective is to quantify the impact of changes in these behavioral characteristics of providers and patients on diagnostic delay experienced by pulmonary TB patients.

METHODS AND FINDINGS

We developed a discrete event simulation model of patients' diagnostic pathways that captures key behavioral characteristics of providers (time to order a test) and patients (time to switch to another provider). We used an expectation-maximization algorithm to estimate the parameters underlying these behavioral characteristics, with quantitative data encoded from detailed interviews of 76 and 64 pulmonary TB patients in the 2 Indian cities of Mumbai and Patna, respectively, which were conducted between April and August 2014. We employed the estimated model to simulate different counterfactual scenarios of diagnostic pathways under altered behavioral characteristics of providers and patients to predict their potential impact on the diagnostic delay. Private healthcare providers including chemists were the first point of contact for the majority of TB patients in Mumbai (70%) and Patna (94%). In Mumbai, 45% of TB patients first approached less-than-fully-qualified providers (LTFQs), who take 28.71 days on average for diagnosis. About 61% of these patients switched to other providers without a diagnosis. Our model estimates that immediate testing for TB by LTFQs at the first visit (at the current level of diagnostic accuracy) could reduce the average diagnostic delay from 35.53 days (95% CI: 34.60, 36.46) to 18.72 days (95% CI: 18.01, 19.43). In Patna, 61% of TB patients first approached fully qualified providers (FQs), who take 9.74 days on average for diagnosis. Similarly, immediate testing by FQs at the first visit (at the current level of diagnostic accuracy) could reduce the average diagnostic delay from 23.39 days (95% CI: 22.77, 24.02) to 11.16 days (95% CI: 10.52, 11.81). Improving the diagnostic accuracy of providers per se, without reducing the time to testing, was not predicted to lead to any reduction in diagnostic delay. Our study was limited because of its restricted geographic scope, small sample size, and possible recall bias, which are typically associated with studies of patient pathways using patient interviews.

CONCLUSIONS

In this study, we found that encouraging private providers to order definitive TB diagnostic tests earlier during patient consultation may have substantial impact on reducing diagnostic delay in these urban Indian settings. These results should be combined with disease transmission models to predict the impact of changes in provider behavior on TB incidence.

Estimating the Impact of Tuberculosis Case Detection in Constrained Health Systems: An Example of Case-Finding in South Africa

Mathematical models are increasingly being used to compare strategies for tuberculosis (TB) control and inform policy decisions. Models often do not consider financial and other constraints on implementation and may overestimate the impact that can be achieved. We developed a pragmatic approach for incorporating resource constraints into mathematical models of TB. Using a TB transmission model calibrated for South Africa, we estimated the epidemiologic impact and resource requirements (financial, human resource (HR), and diagnostic) of 9 case-finding interventions. We compared the model-estimated resources with scenarios of future resource availability and estimated the impact of interventions under these constraints. Without constraints, symptom screening in public health clinics and among persons receiving care for human immunodeficiency virus infection was predicted to lead to larger reductions in TB incidence (9.5% (2.5th–97.5th percentile range (PR), 8.6–12.2) and 14.5% (2.5th–97.5th PR, 12.2–16.3), respectively) than improved adherence to diagnostic guidelines (2.7%; 2.5th–97.5th PR, 1.6–4.1). However, symptom screening required large increases in resources, exceeding future HR capacity. Even under our most optimistic HR scenario, the reduction in TB incidence from clinic symptom screening was 0.2%–0.9%—less than that of improved adherence to diagnostic guidelines. Ignoring resource constraints may result in incorrect conclusions about an intervention’s impact and may lead to suboptimal policy decisions. Models used for decision-making should consider resource constraints.

Empirical estimation of resource constraints for use in model-based economic evaluation: an example of TB services in South Africa

BACKGROUND

Evidence on the relative costs and effects of interventions that do not consider 'real-world' constraints on implementation may be misleading. However, in many low- and middle-income countries, time and data scarcity mean that incorporating health system constraints in priority setting can be challenging.

METHODS

We developed a 'proof of concept' method to empirically estimate health system constraints for inclusion in model-based economic evaluations, using intensified case-finding strategies (ICF) for tuberculosis (TB) in South Africa as an example. As part of a strategic planning process, we quantified the resources (fiscal and human) needed to scale up different ICF strategies (cough triage and WHO symptom screening). We identified and characterised three constraints through discussions with local stakeholders: (1) financial constraint: potential maximum increase in public TB financing available for new TB interventions; (2) human resource constraint: maximum current and future capacity among public sector nurses that could be dedicated to TB services; and (3) diagnostic supplies constraint: maximum ratio of Xpert MTB/RIF tests to TB notifications. We assessed the impact of these constraints on the costs of different ICF strategies.

RESULTS

It would not be possible to reach the target coverage of ICF (as defined by policy makers) without addressing financial, human resource and diagnostic supplies constraints. The costs of addressing human resource constraints is substantial, increasing total TB programme costs during the period 2016-2035 by between 7% and 37% compared to assuming the expansion of ICF is unconstrained, depending on the ICF strategy chosen.

CONCLUSIONS

Failure to include the costs of relaxing constraints may provide misleading estimates of costs, and therefore cost-effectiveness. In turn, these could impact the local relevance and credibility of analyses, thereby increasing the risk of sub-optimal investments.

Dynamic modeling approaches to characterize the functioning of health systems: A systematic review of the literature

Universal Health Coverage (UHC) is one of the targets for the United Nations Sustainable Development Goal 3. The impetus for UHC has led to an increased demand for time-sensitive tools to enhance our knowledge of how health systems function and to evaluate impact of system interventions. We define the field of "health system modeling" (HSM) as an area of research where dynamic mathematical models can be designed in order to describe, predict, and quantitatively capture the functioning of health systems. HSM can be used to explore the dynamic relationships among different system components, including organizational design, financing and other resources (such as investments in resources and supply chain management systems) - what we call "inputs" - on access, coverage, and quality of care - what we call "outputs", toward improved health system "outcomes", namely increased levels and fairer distributions of population health and financial risk protection. We undertook a systematic review to identify the existing approaches used in HSM. We identified "systems thinking" - a conceptual and qualitative description of the critical interactions within a health system - as an important underlying precursor to HSM, and collated a critical collection of such articles. We then reviewed and categorized articles from two schools of thoughts: "system dynamics" (SD)" and "susceptible-infected-recovered-plus" (SIR+). SD emphasizes the notion of accumulations of stocks in the system, inflows and outflows, and causal feedback structure to predict intended and unintended consequences of policy interventions. The SIR + models link a typical disease transmission model with another that captures certain aspects of the system that impact the outcomes of the main model. These existing methods provide critical insights in informing the design of HSM, and provide a departure point to extend this research agenda. We highlight the opportunity to advance modeling methods to further understand the dynamics between health system inputs and outputs.

Diagnostic accuracy of three technologies for the diagnosis of multi-drug resistant tuberculosis

Introducción. La tuberculosis multirresistente (TB-MDR) y la extremadamente resistente (TB-XDR) constituyen un problema de salud pública a nivel mundial. Su detección oportuna permitiría reducir la carga de la enfermedad y su impacto económico en los sistemas de salud.Objetivo. Revisar sistemáticamente la información relacionada con la precisión diagnóstica de tres pruebas moleculares para detectar la tuberculosis multirresistente y la extremadamente resistente.Materiales y métodos. Se hizo una revisión sistemática de la literatura, según los lineamientos de Cochrane, de los estudios en población inmunocompetente relacionados con la precisión diagnóstica de tres pruebas moleculares para detectar la tuberculosis multirresistente y la extremadamente resistente. La búsqueda de los estudios publicados a partir del 2007 se hizo en Medline y Embase. La precisión diagnóstica de las pruebas se estableció con base en los valores máximos y mínimos de sensibilidad y especificidad, y en los valores predictivos positivos y negativos.Resultados. Se detectaron ocho estudios sobre la precisión diagnóstica de la prueba GeneXpert MTB/RIF®, 12 sobre la de GenoType MTBDRplus® y 13 sobre la de GenoType MTBDRsl®. La especificidad de GeneXpert MTB/RIF® osciló entre 91 y 100 % y su sensibilidad, entre 33,3 y 100 %. La sensibilidad de GenoType MTBDRplus® varió entre 82 y 100 %, en tanto que la sensibilidad y la especificidad de GenoType® MTBDRsl fluctuaron entre 56 y 100 % y 21 y 100 %, respectivamente.Conclusión. Según los estudios consultados, los tres métodos de diagnóstico evaluados presentaban una adecuada eficacia diagnóstica para detectar la tuberculosis multirresistente y la extremadamente resistente.

Modular programming for tuberculosis control, the "AuTuMN" platform

BACKGROUND

Tuberculosis (TB) is now the world's leading infectious killer and major programmatic advances will be needed if we are to meet the ambitious new End TB Targets. Although mathematical models are powerful tools for TB control, such models must be flexible enough to capture the complexity and heterogeneity of the global TB epidemic. This includes simulating a disease that affects age groups and other risk groups differently, has varying levels of infectiousness depending upon the organ involved and varying outcomes from treatment depending on the drug resistance pattern of the infecting strain.

RESULTS

We adopted sound basic principles of software engineering to develop a modular software platform for simulation of TB control interventions ("AuTuMN"). These included object-oriented programming, logical linkage between modules and consistency of code syntax and variable naming. The underlying transmission dynamic model incorporates optional stratification by age, risk group, strain and organ involvement, while our approach to simulating time-variant programmatic parameters better captures the historical progression of the epidemic. An economic model is overlaid upon this epidemiological model which facilitates comparison between new and existing technologies. A "Model runner" module allows for predictions of future disease burden trajectories under alternative scenario situations, as well as uncertainty, automatic calibration, cost-effectiveness and optimisation. The model has now been used to guide TB control strategies across a range of settings and countries, with our modular approach enabling repeated application of the tool without the need for extensive modification for each application.

CONCLUSIONS

The modular construction of the platform minimises errors, enhances readability and collaboration between multiple programmers and enables rapid adaptation to answer questions in a broad range of contexts without the need for extensive re-programming. Such features are particularly important in simulating an epidemic as complex and diverse as TB.

Optimally capturing latency dynamics in models of tuberculosis transmission

Although different structures are used in modern tuberculosis (TB) models to simulate TB latency, it remains unclear whether they are all capable of reproducing the particular activation dynamics empirically observed. We aimed to determine which of these structures replicate the dynamics of progression accurately. We reviewed 88 TB-modelling articles and classified them according to the latency structure employed. We then fitted these different models to the activation dynamics observed from 1352 infected contacts diagnosed in Victoria (Australia) and Amsterdam (Netherlands) to obtain parameter estimates. Six different model structures were identified, of which only those incorporating two latency compartments were capable of reproducing the activation dynamics empirically observed. We found important differences in parameter estimates by age. We also observed marked differences between our estimates and the parameter values used in many previous models. In particular, when two successive latency phases are considered, the first period should have a duration that is much shorter than that used in previous studies. In conclusion, structures incorporating two latency compartments and age-stratification should be employed to accurately replicate the dynamics of TB latency. We provide a catalogue of parameter values and an approach to parameter estimation from empiric data for calibration of future TB-models.

Modeling the patient and health system impacts of alternative xpert® MTB/RIF algorithms for the diagnosis of pulmonary tuberculosis in Addis Ababa, Ethiopia

Background

To reduce global tuberculosis (TB) burden, the active disease must be diagnosed quickly and accurately and patients should be treated and cured. In Ethiopia, TB diagnosis mainly relies on spot-morning-spot (SMS) sputum sample smear analysis using Ziehl-Neelsen staining techniques (ZN). Since 2014 targeted use of xpert has been implemented. New diagnostic techniques have higher sensitivity and are likely to detect more cases if routinely implemented. The objective of our study was to project the effects of alternative diagnostic algorithms on the patient, health system, and costs, and identify cost-effective algorithms that increase TB case detection in Addis Ababa, Ethiopia.

Methods

An observational quantitative modeling framework was applied using the Virtual Implementation approach. The model was designed to represent the operational and epidemiological context of Addis Ababa, the capital city of Ethiopia. We compared eight diagnostic algorithm with ZN microscopy, light emitting diode (LED) fluorescence microscopy and Xpert MTB/RIF. Interventions with an annualized cost per averted disability adjusted life year (DALY) of less than the Gross Domestic Product (GDP) per capita are considered cost-effective interventions.

Results

With a cost lower than the average per-capita GDP (US$690 for Ethiopia) for each averted disability adjusted life year (DALY), three of the modeled algorithms are cost-effective. Implementing them would have important patient, health system, and population-level effects in the context of Addis Ababa

❖ The full roll-out of Xpert MTB/RIF as the primary test for all presumptive TB cases would avert 91170 DALYs (95% credible interval [CrI] 54888 – 127448) with an additional health system cost of US$ 11.6 million over the next 10 years. The incremental cost-effectiveness ratio (ICER) is $370 per DALY averted.

❖ Same day LED fluorescence microscopy for all presumptive TB cases combined with Xpert MTB/RIF targeted to HIV-positive and High multidrug resistant (MDR) risk groups would avert 73600 DALYs( 95% CrI 48373 - 99214) with an additional cost of US$5.1 million over the next 10 years. The ICER is $169per DALY averted.

❖ Same-day LED fluorescence microscopy for all presumptive TB cases (and no Xpert MTB/RIF) would avert 43580 DALYs with a reduction cost of US$ 0.2 million over the next 10years. The ICER is $13 per DALY averted.

Conclusions

The full roll-out of Xpert MTB/RIF is predicted to be the best option to substantially reduce the TB burden in Addis Ababa and is considered cost effective. However, the investment cost to implement this is far beyond the budget of the national TB control program. Targeted use of Xpert MTB/RIF for HIV positive and high MDR risk groups with same-day LED fluorescence microscopy for all other presumptive TB cases is an affordable alternative.

Electronic supplementary material

The online version of this article (doi:10.1186/s12879-017-2417-6) contains supplementary material, which is available to authorized users.

Impact and cost-effectiveness of current and future tuberculosis diagnostics: the contribution of modelling

The landscape of diagnostic testing for tuberculosis (TB) is changing rapidly, and stakeholders need urgent guidance on how to develop, deploy and optimize TB diagnostics in a way that maximizes impact and makes best use of available resources. When decisions must be made with only incomplete or preliminary data available, modelling is a useful tool for providing such guidance. Following a meeting of modelers and other key stakeholders organized by the TB Modelling and Analysis Consortium, we propose a conceptual framework for positioning models of TB diagnostics. We use that framework to describe modelling priorities in four key areas: Xpert^® MTB/RIF scale-up, target product profiles for novel assays, drug susceptibility testing to support new drug regimens, and the improvement of future TB diagnostic models. If we are to maximize the impact and cost-effectiveness of TB diagnostics, these modelling priorities should figure prominently as targets for future research.

Detection of lipoarabinomannan (LAM) in urine is an independent predictor of mortality risk in patients receiving treatment for HIV-associated tuberculosis in sub-Saharan Africa: a systematic review and meta-analysis

BACKGROUND

Simple immune capture assays that detect mycobacterial lipoarabinomannan (LAM) antigen in urine are promising new tools for the diagnosis of HIV-associated tuberculosis (HIV-TB). In addition, however, recent prospective cohort studies of patients with HIV-TB have demonstrated associations between LAM in the urine and increased mortality risk during TB treatment, indicating an additional utility of urinary LAM as a prognostic marker. We conducted a systematic review and meta-analysis to summarise the evidence concerning the strength of this relationship in adults with HIV-TB in sub-Saharan Africa, thereby quantifying the assay's prognostic value.

METHODS

We searched MEDLINE and Embase databases using comprehensive search terms for 'HIV', 'TB', 'LAM' and 'sub-Saharan Africa'. Identified studies were reviewed and selected according to predefined criteria.

RESULTS

We identified 10 studies eligible for inclusion in this systematic review, reporting on a total of 1172 HIV-TB cases. Of these, 512 patients (44 %) tested positive for urinary LAM. After a variable duration of follow-up of between 2 and 6 months, overall case fatality rates among HIV-TB cases varied between 7 % and 53 %. Pooled summary estimates generated by random-effects meta-analysis showed a two-fold increased risk of mortality for urinary LAM-positive HIV-TB cases compared to urinary LAM-negative HIV-TB cases (relative risk 2.3, 95 % confidence interval 1.6-3.1). Some heterogeneity was explained by study setting and patient population in sub-group analyses. Five studies also reported multivariable analyses of risk factors for mortality, and pooled summary estimates demonstrated over two-fold increased mortality risk (odds ratio 2.5, 95 % confidence interval 1.4-4.5) among urinary LAM-positive HIV-TB cases, even after adjustment for other risk factors for mortality, including CD4 cell count.

CONCLUSIONS

We have demonstrated that detectable LAM in urine is associated with increased risk of mortality during TB treatment, and that this relationship remains after adjusting for other risk factors for mortality. This may simply be due to a positive test for urinary LAM serving as a marker of higher mycobacterial load and greater disease dissemination and severity. Alternatively, LAM antigen may directly compromise host immune responses through its known immunomodulatory effects. Detectable LAM in the urine is an independent risk factor for mortality among patients receiving treatment for HIV-TB. Further research is warranted to elucidate the underlying mechanisms and to determine whether this vulnerable patient population may benefit from adjunctive interventions.

Incorporating Demand and Supply Constraints into Economic Evaluations in Low-Income and Middle-Income Countries

Global guidelines for new technologies are based on cost and efficacy data from a limited number of trial locations. Country-level decision makers need to consider whether cost-effectiveness analysis used to inform global guidelines are sufficient for their situation or whether to use models that adjust cost-effectiveness results taking into account setting-specific epidemiological and cost heterogeneity. However, demand and supply constraints will also impact cost-effectiveness by influencing the standard of care and the use and implementation of any new technology. These constraints may also vary substantially by setting. We present two case studies of economic evaluations of the introduction of new diagnostics for malaria and tuberculosis control. These case studies are used to analyse how the scope of economic evaluations of each technology expanded to account for and then address demand and supply constraints over time. We use these case studies to inform a conceptual framework that can be used to explore the characteristics of intervention complexity and the influence of demand and supply constraints. Finally, we describe a number of feasible steps that researchers who wish to apply our framework in cost-effectiveness analyses.

CAHRD Consultation 2014: the 10-20 year Horizon Introduction and Overview - as circulated to Consultation participants

The overall aim of the 2014 Consultation is to bring together internal and external partners to help shape the strategic direction for CAHRD over the 10 to 20 year horizon. Our strategic thinking will be guided by our vision of a healthy future for low and middle income populations and our mission to transform health systems to improve the health of these populations. Partnership between northern and southern institutions is integral to this work and critical in the consultation process. The Consultation considers four selected areas of the current work of CAHRD: Lung Health, Maternal & Newborn Health, Neglected Tropical Diseases, and Health Systems. We aim to foster dialogue and learning between these and across contexts and disciplines. The major challenges that will need to be addressed over the next 10 to 20 years will be scoped and pathways to possible solutions proposed. The overall vision is a process of co-production of knowledge

Assessment of the patient, health system, and population effects of Xpert MTB/RIF and alternative diagnostics for tuberculosis in Tanzania: an integrated modelling approach

BACKGROUND

Several promising new diagnostic methods and algorithms for tuberculosis have been endorsed by WHO. National tuberculosis programmes now face the decision on which methods to implement and where to place them in the diagnostic algorithm.

METHODS

We used an integrated model to assess the effects of different algorithms of Xpert MTB/RIF and light-emitting diode (LED) fluorescence microscopy in Tanzania. To understand the effects of new diagnostics from the patient, health system, and population perspective, the model incorporated and linked a detailed operational component and a transmission component. The model was designed to represent the operational and epidemiological context of Tanzania and was used to compare the effects and cost-effectiveness of different diagnostic options.

FINDINGS

Among the diagnostic options considered, we identified three strategies as cost effective in Tanzania. Full scale-up of Xpert would have the greatest population-level effect with the highest incremental cost: 346 000 disability-adjusted life-years (DALYs) averted with an additional cost of US$36·9 million over 10 years. The incremental cost-effectiveness ratio (ICER) of Xpert scale-up ($169 per DALY averted, 95% credible interval [CrI] 104-265) is below the willingness-to-pay threshold ($599) for Tanzania. Same-day LED fluorescence microscopy is the next most effective strategy with an ICER of $45 (95% CrI 25-74), followed by LED fluorescence microscopy with an ICER of $29 (6-59). Compared with same-day LED fluorescence microscopy and Xpert full rollout, targeted use of Xpert in presumptive tuberculosis cases with HIV infection, either as an initial diagnostic test or as a follow-on test to microscopy, would produce DALY gains at a higher incremental cost and therefore is dominated in the context of Tanzania.

INTERPRETATION

For Tanzania, this integrated modelling approach predicts that full rollout of Xpert is a cost-effective option for tuberculosis diagnosis and has the potential to substantially reduce the national tuberculosis burden. It also estimates the substantial level of funding that will need to be mobilised to translate this into clinical practice. This approach could be adapted and replicated in other developing countries to inform rational health policy formulation.

How can mathematical models advance tuberculosis control in high HIV prevalence settings?

Existing approaches to tuberculosis (TB) control have been no more than partially successful in areas with high human immunodeficiency virus (HIV) prevalence. In the context of increasingly constrained resources, mathematical modelling can augment understanding and support policy for implementing those strategies that are most likely to bring public health and economic benefits. In this paper, we present an overview of past and recent contributions of TB modelling in this key area, and suggest a way forward through a modelling research agenda that supports a more effective response to the TB-HIV epidemic, based on expert discussions at a meeting convened by the TB Modelling and Analysis Consortium. The research agenda identified high-priority areas for future modelling efforts, including 1) the difficult diagnosis and high mortality of TB-HIV; 2) the high risk of disease progression; 3) TB health systems in high HIV prevalence settings; 4) uncertainty in the natural progression of TB-HIV; and 5) combined interventions for TB-HIV. Efficient and rapid progress towards completion of this modelling agenda will require co-ordination between the modelling community and key stakeholders, including advocates, health policy makers, donors and national or regional finance officials. A continuing dialogue will ensure that new results are effectively communicated and new policy-relevant questions are addressed swiftly.

Mathematical Modelling and Tuberculosis: Advances in Diagnostics and Novel Therapies

As novel diagnostics, therapies, and algorithms are developed to improve case finding, diagnosis, and clinical management of patients with TB, policymakers must make difficult decisions and choose among multiple new technologies while operating under heavy resource constrained settings. Mathematical modelling can provide helpful insight by describing the types of interventions likely to maximize impact on the population level and highlighting those gaps in our current knowledge that are most important for making such assessments. This review discusses the major contributions of TB transmission models in general, namely, the ability to improve our understanding of the epidemiology of TB. We focus particularly on those elements that are important to appropriately understand the role of TB diagnosis and treatment (i.e., what elements of better diagnosis or treatment are likely to have greatest population-level impact) and yet remain poorly understood at present. It is essential for modellers, decision-makers, and epidemiologists alike to recognize these outstanding gaps in knowledge and understand their potential influence on model projections that may guide critical policy choices (e.g., investment and scale-up decisions).

Modeling of novel diagnostic strategies for active tuberculosis - a systematic review: current practices and recommendations

Introduction

The field of diagnostics for active tuberculosis (TB) is rapidly developing. TB diagnostic modeling can help to inform policy makers and support complicated decisions on diagnostic strategy, with important budgetary implications. Demand for TB diagnostic modeling is likely to increase, and an evaluation of current practice is important. We aimed to systematically review all studies employing mathematical modeling to evaluate cost-effectiveness or epidemiological impact of novel diagnostic strategies for active TB.

Methods

Pubmed, personal libraries and reference lists were searched to identify eligible papers. We extracted data on a wide variety of model structure, parameter choices, sensitivity analyses and study conclusions, which were discussed during a meeting of content experts.

Results & Discussion

From 5619 records a total of 36 papers were included in the analysis. Sixteen papers included population impact/transmission modeling, 5 were health systems models, and 24 included estimates of cost-effectiveness. Transmission and health systems models included specific structure to explore the importance of the diagnostic pathway (n = 4), key determinants of diagnostic delay (n = 5), operational context (n = 5), and the pre-diagnostic infectious period (n = 1). The majority of models implemented sensitivity analysis, although only 18 studies described multi-way sensitivity analysis of more than 2 parameters simultaneously. Among the models used to make cost-effectiveness estimates, most frequent diagnostic assays studied included Xpert MTB/RIF (n = 7), and alternative nucleic acid amplification tests (NAATs) (n = 4). Most (n = 16) of the cost-effectiveness models compared new assays to an existing baseline and generated an incremental cost-effectiveness ratio (ICER).

Conclusion

Although models have addressed a small number of important issues, many decisions regarding implementation of TB diagnostics are being made without the full benefits of insight from mathematical models. Further models are needed that address a wider array of diagnostic and epidemiological settings, that explore the inherent uncertainty of models and that include additional epidemiological data on transmission implications of false-negative diagnosis and the pre-diagnostic period.

Operational modelling to guide implementation and scale-up of diagnostic tests within the health system: exploring opportunities for parasitic disease diagnostics based on example application for tuberculosis

Research and innovation in the diagnosis of infectious and parasitic diseases has led to the development of several promising diagnostic tools, for example in malaria there is extensive literature concerning the use of rapid diagnostic tests. This means policymakers in many low and middle income countries need to make difficult decisions about which of the recommended tools and approaches to implement and scale-up. The test characteristics (e.g. sensitivity and specificity) of the tools alone are not a sufficient basis on which to make these decisions as policymakers need to also consider the best combination of tools, whether the new tools should complement or replace existing diagnostics and who should be tested. Diagnostic strategies need dovetailing to different epidemiology and structural resource constraints (e.g. existing diagnostic pathways, human resources and laboratory capacity). We propose operational modelling to assist with these complex decisions. Projections of patient, health system and cost impacts are essential and operational modelling of the relevant elements of the health system could provide these projections and support rational decisions. We demonstrate how the technique of operational modelling applied in the developing world to support decisions on diagnostics for tuberculosis, could in a parallel way, provide useful insights to support implementation of appropriate diagnostic innovations for parasitic diseases.

Do high rates of empirical treatment undermine the potential effect of new diagnostic tests for tuberculosis in high-burden settings?

In tuberculosis-endemic settings, patients are often treated empirically, meaning that they are placed on treatment based on clinical symptoms or tests that do not provide a microbiological diagnosis (eg, chest radiography). New tests for tuberculosis, such as the Xpert MTB/RIF assay (Cepheid, Sunnyvale, CA, USA), are being implemented at substantial cost. To inform policy and rationally drive implementation, data are needed for how these tests affect morbidity, mortality, transmission, and population-level tuberculosis burden. If people diagnosed by use of new diagnostics would have received empirical treatment a few days later anyway, then the incremental benefit might be small. Will new diagnostics substantially improve outcomes and disease burden, or simply displace empirical treatment? Will the extent and accuracy of empirical treatment change with the introduction of a new test? In this Personal View, we review emerging data for how empirical treatment is frequently same-day, and might still be the predominant form of treatment in high-burden settings, even after Xpert implementation; and how Xpert might displace so-called true-positive, rather than false-positive, empirical treatment. We suggest types of studies needed to accurately assess the effect of new tuberculosis tests and the role of empirical treatment in real-world settings. Until such questions can be addressed, and empirical treatment is appropriately characterised, we postulate that the estimated population-level effect of new tests such as Xpert might be substantially overestimated.

Computational models for neglected diseases: gaps and opportunities

Neglected diseases, such as Chagas disease, African sleeping sickness, and intestinal worms, affect millions of the world's poor. They disproportionately affect marginalized populations, lack effective treatments or vaccines, or existing products are not accessible to the populations affected. Computational approaches have been used across many of these diseases for various aspects of research or development, and yet data produced by computational approaches are not integrated and widely accessible to others. Here, we identify gaps in which computational approaches have been used for some neglected diseases and not others. We also make recommendations for the broad-spectrum integration of these techniques into a neglected disease drug discovery and development workflow.

Alignment of new tuberculosis drug regimens and drug susceptibility testing: a framework for action

New tuberculosis drug regimens are creating new priorities for drug susceptibility testing (DST) and surveillance. To minimise turnaround time, rapid DST will need to be prioritised, but developers of these assays will need better data about the molecular mechanisms of resistance. Efforts are underway to link mutations with drug resistance and to develop strain collections to enable assessment of new diagnostic assays. In resource-limited settings, DST might not be appropriate for all patients with tuberculosis. Surveillance data and modelling will help country stakeholders to design appropriate DST algorithms and to decide whether to change drug regimens. Finally, development of practical DST assays is needed so that, in countries where surveillance and modelling show that DST is advisable, these assays can be used to guide clinical decisions for individual patients. If combined judiciously during both development and implementation, new tuberculosis regimens and new DST assays have enormous potential to improve patient outcomes and reduce the burden of disease.

The impact of new tuberculosis diagnostics on transmission: why context matters

OBJECTIVE

To estimate the impact of new tuberculosis diagnostics on tuberculosis transmission given the complex contextual factors that can lead to patient loss before diagnosis or treatment.

METHODS

An epidemic model of tuberculosis specifying discrete steps along the tuberculosis diagnostic pathway was constructed. The model was calibrated to the epidemiology of tuberculosis and human immunodeficiency virus (HIV) infection in the United Republic of Tanzania and was used to assess the impact of a new diagnostic tool with 70% sensitivity for smear-negative pulmonary tuberculosis. The influence of contextual factors on the projected epidemic impact of the new diagnostic tool over the decade following introduction was explored.

FINDINGS

With the use of smear microscopy, the incidence of tuberculosis will decline by an average of 3.94% per year. If the new tool is added, incidence will decline by an annual 4.25%. This represents an absolute change of 0.31 percentage points (95% confidence interval: 0.04-0.42). However, the annual decline in transmission with use of the new tool is less when existing strategies for the diagnosis of smear-negative cases have high sensitivity and when symptomatic individuals delay in seeking care. Other influential contextual factors include access to tuberculosis care, patient loss before diagnosis, initial patient default after diagnosis and treatment success rate.

CONCLUSION

When implementing and scaling up the use of a new diagnostic tool, the operational context in which diagnosis and treatment take place needs to be considered.

Modelling the impacts of new diagnostic tools for tuberculosis in developing countries to enhance policy decisions

The introduction and scale-up of new tools for the diagnosis of Tuberculosis (TB) in developing countries has the potential to make a huge difference to the lives of millions of people living in poverty. To achieve this, policy makers need the information to make the right decisions about which new tools to implement and where in the diagnostic algorithm to apply them most effectively. These decisions are difficult as the new tools are often expensive to implement and use, and the health system and patient impacts uncertain, particularly in developing countries where there is a high burden of TB. The authors demonstrate that a discrete event simulation model could play a significant part in improving and informing these decisions. The feasibility of linking the discrete event simulation to a dynamic epidemiology model is also explored in order to take account of longer term impacts on the incidence of TB. Results from two diagnostic districts in Tanzania are used to illustrate how the approach could be used to improve decisions.

Tuberculosis diagnostics and biomarkers: needs, challenges, recent advances, and opportunities

Tuberculosis is unique among the major infectious diseases in that it lacks accurate rapid point-of-care diagnostic tests. Failure to control the spread of tuberculosis is largely due to our inability to detect and treat all infectious cases of pulmonary tuberculosis in a timely fashion, allowing continued Mycobacterium tuberculosis transmission within communities. Currently recommended gold-standard diagnostic tests for tuberculosis are laboratory based, and multiple investigations may be necessary over a period of weeks or months before a diagnosis is made. Several new diagnostic tests have recently become available for detecting active tuberculosis disease, screening for latent M. tuberculosis infection, and identifying drug-resistant strains of M. tuberculosis. However, progress toward a robust point-of-care test has been limited, and novel biomarker discovery remains challenging. In the absence of effective prevention strategies, high rates of early case detection and subsequent cure are required for global tuberculosis control. Early case detection is dependent on test accuracy, accessibility, cost, and complexity, but also depends on the political will and funder investment to deliver optimal, sustainable care to those worst affected by the tuberculosis and human immunodeficiency virus epidemics. This review highlights unanswered questions, challenges, recent advances, unresolved operational and technical issues, needs, and opportunities related to tuberculosis diagnostics.

Opportunities and challenges for cost-efficient implementation of new point-of-care diagnostics for HIV and tuberculosis

Stakeholders agree that supporting high-quality diagnostics is essential if we are to continue to make strides in the fight against human immunodeficiency virus (HIV) and tuberculosis. Despite the need to strengthen existing laboratory infrastructure, which includes expanding and developing new laboratories, there are clear diagnostic needs where conventional laboratory support is insufficient. Regarding HIV, rapid point-of-care (POC) testing for initial HIV diagnosis has been successful, but several needs remain. For tuberculosis, several new diagnostic tests have recently been endorsed by the World Health Organization, but a POC test remains elusive. Human immunodeficiency virus and tuberculosis are coendemic in many high prevalence locations, making parallel diagnosis of these conditions an important consideration. Despite its clear advantages, POC testing has important limitations, and laboratory-based testing will continue to be an important component of future diagnostic networks. Ideally, a strategic deployment plan should be used to define where and how POC technologies can be most efficiently and cost effectively integrated into diagnostic algorithms and existing test networks prior to widespread scale-up. In this fashion, the global community can best harness the tremendous capacity of novel diagnostics in fighting these 2 scourges.