Review 1: "Sputum and Tongue Swab Molecular Testing for the In-Home Diagnosis of Tuberculosis in Unselected Household Contacts: A Cost and Cost-Effectiveness Analysis"

Peter Auguste

doi:doi:10.1162/2e3983f5.942a01b3

Sputum and tongue swab molecular testing for the in-home diagnosis of tuberculosis in unselected household contacts: a cost and cost-effectiveness analysis

by Charl Bezuidenhout, Lawrence Long, Brooke Nichols, Gesine Meyer-Rath, Matthew P Fox, Grant Theron, Bernard Fourie, Sharon Olifant, Adam Penn-Nicholson, Morten Ruhwald, and Andrew Medina-Marino

Show Description
Published on Oct 18, 2024
www.medrxiv.org

Description

ABSTRACT Background Delayed and missed diagnosis are a persistent barrier to tuberculosis control, partly driven by limitations associated with sputum collection and an unmet need for decentralized testing. Household contact investigation with point-of-care testing of non-invasive specimens like tongue swabs are hitherto undescribed and may be a cost-effective solution to enable community-based active case finding.Methods In-home, molecular point-of-care testing was conducted using sputum and tongue specimens collected from all household contacts of confirmed tuberculosis cases. A health economic assessment was executed to estimate and compare the cost and cost-effectiveness of different in-home, point-of-care testing strategies. Incremental cost effectiveness ratios of strategies utilizing different combination testing algorithms using sputum and/or tongue swab specimens were compared.Findings The total implementation cost of delivering the standard of care for a 2-year period was $84 962. Strategies integrating in-home point-of-care testing ranged between $87 844 - $93 969. The cost-per-test for in-home, POC testing of sputum was the highest at $20·08 per test. Two strategies, Point-of-Care Sputum Testing and Point-of-Care Combined Sputum and Individual Tongue Swab Testing were the most cost-effective with ICERs of $543·74 and $547·29 respectively, both below a $2,760 willingness-to-pay threshold.Interpretation An in-home, point-of-care molecular testing strategy utilizing combination testing of tongue swabs and sputum specimens would incur an additional 10.6% program cost, compared to SOC, over a 2-year period. The increased sample yield from tongue swabs combined with immediate result notification following, in-home POC testing would increase the number of new TB cases detected and linked to care by more than 800%.Research in context Evidence before this study We searched PubMed for original research published between January 1, 1950 and June 30, 2024 that evaluated the cost-effectiveness of in-home POC molecular testing, as part of HCI strategies for tuberculosis. PubMed search terms used included [“household contact investigation” OR “household contact tracing”] AND “tuberculosis” AND “cost-effectiveness”. The search revealed 8 studies, of which one was removed as HCIs were leveraged for the provision of short course preventative therapy and not tuberculosis testing. None of the studies were conducted in South Africa. All seven remaining studies relied on a hub-and-spoke model of sputum collection and transportation with sputum tested at a centralized laboratory facility. Although active case finding strategies like HCIs are endorsed by the WHO to improve early case detection and treatment initiation, limited research has been done to assess its cost-effectiveness in low- and middle-income countries.Added value of this study To our knowledge, this is the first example of in-home molecular point-of-care (POC) testing as part of HCI. The use of primary data to estimate and compare the incremental cost effectiveness of different combination, in-home testing strategies utilizing alternative sample types equips policy makers with a selection of strategy options to choose from. The tradeoff between sample types with high collection yield and those with increased accuracy becomes evident in the economic analysis, highlighting the need to consider both yield and accuracy in effective clinical decision making and use-case development. The success of in-home, POC tongue swab testing of all contacts, irrespective of symptom presentation shows great promise for universal testing programs.Implications of all available evidence Results from our economic modeling provide evidence in support for the integration of in-home, POC tuberculosis (TB) testing during HCI. The use of less invasive tongue swab samples to increase sample yield in the absence of sputum expectoration highlights the value of combination testing strategies. Immediate result notification resulting from rapid, in-home POC testing shows great promise for increasing early case detection and improving treatment uptake. In-home, POC testing strategies, when incorporated into HCI could curb ongoing community transmission and reduce the overall burden of TB. Considerations for adopting novel POC testing strategies in future active case finding programs like HCI should strongly be considered.Summary We evaluated the cost-effectiveness of in-home, point-of-care TB testing of household contacts. The findings indicate that combined testing strategies using tongue swab and sputum specimens could significantly increase TB case detection, with modest additional program costs.

RR\ID Evidence Scale rating by reviewer:

Potentially informative. The main claims made are not strongly justified by the methods and data, but may yield some insight. The results and conclusions of the study may resemble those from the hypothetical ideal study, but there is substantial room for doubt. Decision-makers should consider this evidence only with a thorough understanding of its weaknesses, alongside other evidence and theory. Decision-makers should not consider this actionable, unless the weaknesses are clearly understood and there is other theory and evidence to further support it.

***************************************

Review: The authors undertook a cost-effectiveness analysis that compared in-home versus point-of-care testing for TB in household contact. They found that the strategy of POC sputum testing was more costly and more effective and generated an ICER of approximately US$540 per new case detected and linked to treatment.

Abstract

The abstract contains information about the conduct of the economic analysis. One minor comment is that it is customary to include in the methods information about the model structure used as well as programme to undertake the economic analysis. Additionally, provide information about the outcomes and measures of effectiveness.

Introduction

The authors provided information about the global burden of TB, but in the methods the focus is South Africa. Would it be helpful to include the burden of TB in South Africa as well?
Would the authors mind spelling out HCI? It is unclear what the acronym represents.

Methods

What was the rationale for using a decision tree structure? Was the conceptual model developed with input from clinical experts. If so, this should be declared.
Authors included a combination strategy comprising sputum and individual tongue swab. Sensitivity and specificity have been reported for each individual strategy. However, it is unclear what assumptions have been made when combining each test. What constitutes a true positive TB result? For example, both individual tests positive? It isn’t clear based on the inputs for the combination strategy how they have been used in the illustrative structure.
Authors converted South African Rand to US dollars using the 2023 World Bank conversion rate. I think it is more appropriate to convert using the purchasing power parity (PPP).
Point estimates along with their low and high values are reported. Please clarify in the table of inputs if these are ranges or confidence intervals.
It was not clear on the evidence synthesis methods used to derive test sensitivity and specificity, especially where information had been obtained from more than one source.

Results

Authors reported results in the form of an ICER expressed as the cost per new case detected and linked to treatment. Results for standard of care appeared to be the cheapest and least effective. The strategy of POC sputum testing was more costly and more effective and generated an ICER of approximately US$540 per new case detected and linked to treatment. Authors reported that the combined strategy of sputum and individual tongue swabbing testing was the most expensive and most effective treatment generating an ICER of approximately US$550 when compared to POC sputum testing. Given the concerns raised with regards to the combination strategy (and without access to the model), it is unclear if the results accurately reflect this strategy.
In the supplementary material, the authors presented a cost-effectiveness acceptability curve, which indicates the probability of an intervention being cost-effective across a range of willingness-to-pay thresholds. However, the authors have not reported any scatterplots or reported these results in the main paper. Additionally, it is common to report the distributions and their values for parameters included in the probabilistic sensitivity analysis (PSA). For example, what distribution was used for the sensitivity and specificity. Furthermore, how have the authors maintained the correlation between sensitivity and specificity in the PSA?

Review 1: "Sputum and Tongue Swab Molecular Testing for the In-Home Diagnosis of Tuberculosis in Unselected Household Contacts: A Cost and Cost-Effectiveness Analysis"

Connections