Evaluating the Impact of Computer-assisted X-ray Diagnosis and Other Triage Tools to Optimise Xpert Orientated Community-based Active Case Finding for TB and COVID-19

Tuberculosis (TB) is now the commonest cause of death in many African countries. Globally, ~35% (almost 1 in 3) of TB cases are 'missed' (remain undiagnosed or undetected). In sub-Saharan Africa, 40-50% of the TB case burden remains undiagnosed within the community. These 'missed' TB cases (at primary care level) serve as a reservoir, which severely undermines TB control. With rapid advances in the development of TB screening tests, the investigators aim to determine the pragmatic utility of computer-assisted x-ray diagnosis (CAD). Recent data suggest that CAD performs on par with experienced radiologists to identify potential TB cases, hereby reducing the frequency at which Xpert tests are requested and helps to focus limited resources on the relevant cases. In addition, the investigators aim to test nascent screening technologies for TB diagnosis such as evaluating urine-based TB screening biosignatures. The COVID-19 pandemic has ravaged African peri-urban communities where TB is also common. With the pressing need to improve screening and diagnosis of COVID-19, the investigators plan to explore the potential for urine- and blood-based COVID-19 screening assays. Symptoms of COVID-19 and TB overlap, and limited affordability, as well as the stigma associated with both diseases, severely limits testing. Data are now urgently needed about the feasibility of co-screening and testing for TB and COVID-19. The utility of such an approach, if any, has not been studied in African communities.

Study Overview

• Status: Recruiting
• Conditions: HIV Infections; COVID-19; Tuberculosis
• Intervention / Treatment: Diagnostic test: CAD; Diagnostic test: Xpert

Detailed Description

Tuberculosis (TB) is now the commonest cause of death in many African countries. Several factors drive this; however, transmission is the mechanism by which these risk factors translate into active TB. Globally, ~35% (almost 1 in 3) of TB cases are 'missed' (remain undiagnosed or undetected). In sub-Saharan Africa, 40-50% of the TB case burden remains undiagnosed within the community and ~30% of such cases are microscopically smear-positive. These 'missed' TB cases (at primary care level) serve as a reservoir, which severely undermines TB control. Thus, primary care and community-based case finding should be a critical component for TB control.

Detecting cases in the community, however, has been restricted by the lack of sensitive and user-friendly Point-of-Care (POC) diagnostic tools. To address this unmet need, in 2013 the investigators planned a programme of activities (sequential interlinked studies) with the overarching aim of optimising a model for Xpert-related community-based active case finding (ACF) for TB (XACT). By 2017, through the EDCTP-funded XACT-I study, the investigators solved the impasse of rapid POC diagnosis by showing that molecular Xpert-based community-based screening was effective in identifying missing TB cases in the peri-urban 'slums' of Cape Town and Harare using a mini-truck with a generator. However, such an approach was neither broadly affordable nor scalable. The investigators therefore derived a scalable model using portable battery-operated Xpert Edge installed within a low-cost (< US$) 15 000 Nissan panel van manned by two health care workers (thus making the ACF model affordable and scalable). This completed study, XACT-II, screened over 5 000 participants in the community. The model worked well and was more effective than smear microscopy. Based on these successes, and to translate the XACT concept into policy, the Wellcome Trust and UK MRC has funded the XACT-III study. Currently commenced, XACT-III was initiated as a multi-country demonstration project in four sub-Saharan African countries.

More recently, there have been rapid advances in the development of triage testing for TB, which refers to screening tests that are generally applied in a community-based setting (either at individual community or primary care clinic level). These tests have very high sensitivity (>95%) but modest specificity (>70%) as defined by TB-specific target product profiles. A forerunner TB-orientated triage test is computer-assisted x-ray diagnosis (CAD). This entails using artificial intelligence-enabled software to read a digital x-ray and produce a probability of TB within seconds. Recent data suggest that CAD performs on par with experienced radiologists to identify potential TB cases, hereby reducing the frequency at which Xpert tests are requested and helps to focus limited resources on the relevant cases. Although these data appear promising, the feasibility of this strategy in a pragmatic field setting has not been extensively tested. There are several other unanswered questions. Is the strategy of CAD combined with Xpert cost-effective and can it reduce Xpert usage without missing an unacceptable number of TB cases? The investigators will therefore determine the utility of CAD as a triage tool to further optimise the XACT model.

The COVID-19 pandemic, due to SARS-CoV-2, has ravaged African peri-urban communities where TB is also common. Symptoms of COVID-19 and TB overlap, and limited affordability, as well as the stigma associated with both diseases, severely limits testing. Data are now urgently needed about the feasibility of co-screening and testing for TB and COVID-19. The utility of such an approach, if any, has not been studied in African communities. As Xpert POC TB testing and x-rays for CAD will be performed in the proposed study, it affords a unique and easy opportunity to seamlessly screen for both diseases when appropriate.

Other nascent screening technologies are rapidly emerging for TB and COVID-19, including urine- and blood-based triage tests. XACT-19 provides a unique opportunity to collect the relevant samples and test new technologies in a pragmatic community-based setting.

In summary, the XACT-19 study results will have substantial implications for public health policy and practice and will likely define a new standard for community-based ACF for TB, and potentially COVID-19 in tandem.

• Study Type: Interventional
• Enrollment (Anticipated): 26200
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Alex J Scott, MBChB; Phone Number: 00274066669; Email: alex.scott@uct.ac.za
• Study Contact Backup: Name: Aliasgar Esmail, MD; Phone Number: 00274066119; Email: a.esmail@uct.ac.za

Study Locations

• South Africa
  • Western Cape
    • Cape Town, Western Cape, South Africa
      • Recruiting
      • University of Cape Town
      • Contact: Keertan Dheda, MD/PhD; Email: keertan.dheda@uct.ac.za
      • Contact: Ali Esmail, MD; Email: a.esmail@uct.ac.za
• Zambia
  • Lusaka, Zambia
    • Not yet recruiting
    • Helen Ayles
    • Contact: Helen Ayles, MBChB
• Zimbabwe
  • Harare, Zimbabwe
    • Not yet recruiting
    • Junior Mutsvangwa
    • Contact: Junior Mutsvangwa, MBChB

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 18 years and older (Adult, Older Adult)
• Accepts Healthy Volunteers: No
• Genders Eligible for Study: All

Description

Inclusion Criteria:

• Participants willing to complete community-based symptom screening, finger-prick and venepuncture blood sampling, urine testing, and/or undergo TB and/or COVID-19 diagnostic testing.
• Provision of informed consent.
• Participant 18 years or above.
• HIV-positive or negative participants will be included.

Exclusion Criteria:

• Inability to provide informed consent (e.g., mentally impaired).
• Participants who have completed TB treatment in the last two months, or who have self-presented to their local TB clinic and are currently being worked up for suspected TB.
• Participants already diagnosed with active TB on treatment.
• Participants unable to commit to at least a two-month follow-up.
• Female participants who are pregnant or who refuse a urine pregnancy test.
• Participants in the community who cannot access healthcare due to severe ill health or lack of access to the local clinic.

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Screening
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: CAD + POC Xpert; CAD followed by Xpert in CAD-positive participants (performed at POC) employing a low-cost panel van that is staffed by three health care workers. CAD-negative participants will be followed up, while CAD-positive participants will be offered POC Xpert. Xpert-positive participants will be referred for TB treatment initiation, while Xpert-negative (but CAD-positive) participants will undergo a clinical review. Thus, the active case finding (ACF) interventional package is one of CAD + POC Xpert (only in CAD positive participants).	Diagnostic test: CAD; It is an artificial intelligence (AI) system for detection of TB on CXR images. The system input is a frontal CXR, and the outputs are 1) a heatmap indicating suspicious regions on the image; and 2) a score (0-100) which implies the likelihood that the x-ray image shows TB.; Other Names: CAD4TB and/or other AI/CAD software; Diagnostic test: Xpert; A novel diagnostic for active case finding (GeneXpert MTB/RIF) for TB on sputum collected and performed at POC in a mobile van.; Other Names: GeneXpert System
Active Comparator: POC Xpert only; Participants who are Xpert-positive will be referred for TB treatment initiation while Xpert-negative participants will be followed up. Thus, the active case finding (ACF) standard of care package is POC Xpert.	Diagnostic test: Xpert; A novel diagnostic for active case finding (GeneXpert MTB/RIF) for TB on sputum collected and performed at POC in a mobile van.; Other Names: GeneXpert System

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Time to detection of microbiologically proven TB	The microbiological reference standard for TB will be culture and/or Xpert positivity. Thus, the overall time to detection (using a proportional hazards model) and the proportion of TB cases detected at a specific time-point (e.g., 14-, 30- and 60-days) with and without culture (Xpert alone) will be reported.	Through study completion, up to 48 months

Secondary Outcome Measures

Outcome Measure	Time Frame
Feasibility of CAD + POC Xpert performed by minimally trained healthcare workers	Through study completion, up to 48 months
Number of infectious TB cases detected (defined by cough aerosol sampling system [CASS] and/or smear and/or cavitatory disease positive)	Through study completion, up to 48 months
Time-specific proportion of participants initiated on TB treatment up to 60 days post-sample donation in each arm (7-, 14-, 30- and 60-days)	Through study completion, up to 48 months
Time to TB treatment initiation (both the median time to treatment in each group and time to event [treatment] analyses will be conducted)	Through study completion, up to 48 months
Yield of culture positive TB in household contacts of index participants	Through study completion, up to 48 months
NPV and false negative rate (TB cases missed per 1 000 persons screened) of CAD and other screening tests for TB	Through study completion, up to 48 months
Reduction in number of sputum induction procedures and/or Xpert tests performed	Through study completion, up to 48 months
Global and country-specific cost-effectiveness analysis for each strategy	Through study completion, up to 48 months
Transmission and disease burden impact using modelling based on exposure scores, imaging, and CASS	Through study completion, up to 48 months
Rates or prevalence of microbiological versus probable (clinical TB)	Through study completion, up to 48 months
Proportion of culture-positive TB cases completing three- and six-months of TB treatment in each study arm	Through study completion, up to 48 months
Middleware/dashboard design requirements and deployment models for each strategy	Through study completion, up to 48 months
Feasibility and yield of POC Xpert (Xpress cartridge) for COVID-19 detection	Through study completion, up to 48 months
Feasibility and performance of CAD4COVID for PCR-positive COVID-19 detection	Through study completion, up to 48 months
Feasibility of a novel mass screening strategy for COVID-19 that uses pooling of specimen from a group of COVID-19 suspects	Through study completion, up to 48 months

Other Outcome Measures

Outcome Measure	Time Frame
Economic outcome: Cost effectiveness of CAD + POC Xpert (cost per TB case diagnosed and/or averted, and cost per death and disability-adjusted life year [DALY] averted)	Through study completion, up to 48 months
Economic outcome: Direct comparison of the cost effectiveness of ACF compared to passive case finding (the current public health practice)	Through study completion, up to 48 months
Economic outcome: Cost effectiveness considering drug resistant TB (DR-TB) and HIV prevention	Through study completion, up to 48 months

Collaborators and Investigators

• Sponsor: University of Cape Town
• Collaborators: Radboud University Medical Center; University of Stellenbosch; Foundation for Innovative New Diagnostics, Switzerland; European and Developing Countries Clinical Trials Partnership (EDCTP); Ospedale San Raffaele; Biomedical Research and Training Institute; Zambart
• Investigators: Principal Investigator: Keertan Dheda, PhD, University of Cape Town

Study record dates

Study Major Dates

• Study Start (Actual): February 23, 2022
• Primary Completion (Anticipated): January 1, 2025
• Study Completion (Anticipated): March 1, 2025

Study Registration Dates

• First Submitted: February 1, 2022
• First Submitted That Met QC Criteria: February 1, 2022
• First Posted (Actual): February 2, 2022

Study Record Updates

• Last Update Posted (Actual): April 5, 2022
• Last Update Submitted That Met QC Criteria: March 25, 2022
• Last Verified: March 1, 2022

More Information

Terms related to this study

• Keywords: COVID-19; Screening; Tuberculosis; Active case finding; Computer assisted diagnosis
• Additional Relevant MeSH Terms: Coronavirus Infections; Coronaviridae Infections; Nidovirales Infections; RNA Virus Infections; Virus Diseases; Infections; Respiratory Tract Infections; Respiratory Tract Diseases; Pneumonia, Viral; Pneumonia; Lung Diseases; Bacterial Infections; Bacterial Infections and Mycoses; Gram-Positive Bacterial Infections; Actinomycetales Infections; Mycobacterium Infections; COVID-19; Tuberculosis
• Other Study ID Numbers: XACT-19

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No