Guidance for Studies Evaluating the Accuracy of Tuberculosis Triage Tests

Evaluation of C-Reactive Protein and Computer-Aided Analysis of Chest X-rays as Tuberculosis Triage Tests at Health Facilities in Lesotho and South Africa

BACKGROUND

To improve tuberculosis case-finding, rapid, non-sputum triage tests need to be developed according to the World Health Organization target product profile (TPP) (>90% sensitivity, >70% specificity). We prospectively evaluated and compared artificial intelligence-based, computer-aided detection software, CAD4TBv7, and C-reactive protein assay (CRP) as triage tests at health facilities in Lesotho and South Africa.

METHODS

Adults (≥18 years) presenting with ≥1 of the 4 cardinal tuberculosis symptoms were consecutively recruited between February 2021 and April 2022. After informed consent, each participant underwent a digital chest X-ray for CAD4TBv7 and a CRP test. Participants provided 1 sputum sample for Xpert MTB/RIF Ultra and Xpert MTB/RIF and 1 for liquid culture. Additionally, an expert radiologist read the chest X-rays via teleradiology. For primary analysis, a composite microbiological reference standard (ie, positive culture or Xpert Ultra) was used.

RESULTS

We enrolled 1392 participants, 48% were people with HIV and 24% had previously tuberculosis. The receiver operating characteristic curve for CAD4TBv7 and CRP showed an area under the curve of .87 (95% CI: .84-.91) and .80 (95% CI: .76-.84), respectively. At thresholds corresponding to 90% sensitivity, specificity was 68.2% (95% CI: 65.4-71.0%) and 38.2% (95% CI: 35.3-41.1%) for CAD4TBv7 and CRP, respectively. CAD4TBv7 detected tuberculosis as well as an expert radiologist. CAD4TBv7 almost met the TPP criteria for tuberculosis triage.

CONCLUSIONS

CAD4TBv7 is accurate as a triage test for patients with tuberculosis symptoms from areas with a high tuberculosis and HIV burden. The role of CRP in tuberculosis triage requires further research.

CLINICAL TRIALS REGISTRATION

Clinicaltrials.gov identifier: NCT04666311.

Diagnostic accuracy of TB screening tests in a prospective multinational cohort: Chest-X-ray with computer-aided detection, Xpert TB host response, and C-reactive protein

BACKGROUND

Accessible, accurate screening tests are necessary to advance tuberculosis (TB) case finding and early detection in high-burden countries.

METHODS

We prospectively screened adults with ≥2 weeks of cough at primary health centers in the Philippines, Vietnam, South Africa, Uganda, and India. Participants received chest-X-ray, Cepheid Xpert TB Host Response (Xpert HR) testing, and point-of-care C-reactive protein (CRP) testing (Boditech). Chest-X-ray images were processed using CAD4TB v7, a computer-aided detection algorithm. We assessed diagnostic accuracy against a microbiologic reference standard (sputum Xpert Ultra, culture). Optimal cut-points were chosen to maximize specificity at 90% sensitivity. Two-test screening algorithms were considered, using 1) sequential negative serial screening (positive defined as positive on either test) and 2) sequential positive serial screening (positive defined as positive on both tests).

RESULTS

Between July 2021 and August 2022, 1,392 participants with presumptive TB had valid index tests and reference standard results, and 303 (22%) had confirmed TB. In head-to-head comparisons, CAD4TB v7 showed the highest specificity at 90% sensitivity (70.3% vs. 65.1% for Xpert HR, difference 95% CI 1.6 to 8.9; 49.7% for CRP, difference 95% CI 17.0 to 24.3). Three two-test screening algorithms met WHO target product profile (TPP) minimum accuracy thresholds and had higher accuracy than any test alone. At 90% sensitivity, the specificity was 79.6% for Xpert HR-CAD4TB [sequential negative], 75.9% for CRP-CAD4TB [sequential negative], and 73.7% for Xpert HR-CAD4TB [sequential positive].

CONCLUSIONS

CAD4TB achieves TPP targets and outperforms Xpert HR and CRP. Combining screening tests further increased accuracy.

REGISTRATION

NCT04923958.

Effect of tuberculosis-specific antigen stimulation on the diagnostic accuracy of interferon-γ inducible protein-10 in distinguishing active and latent tuberculosis infection: a meta-analysis

BACKGROUND

Identifying active tuberculosis (ATB) from latent tuberculosis infection (LTBI) persists as a challenge, and interferon-γ inducible protein-10 (IP-10) has been employed as the solution. To further improve its diagnostic performance, the sample can be stimulated with TB specific antigen (TBAg).

AIM

To perform meta-analysis on diagnostic accuracy of unstimulated and TBAg-stimulated IP-10 in differentiating ATB from LTBI.

METHODS

Systematic search was performed on five major scientific databases as of 29 November 2023. Observational studies reporting diagnostic values of unstimulated or TBAg-stimulated IP-10 in identifying ATB from LTBI were included. Meta-analysis was carried out using two-level mixed-effect logistic regression model.

RESULTS

Twenty-five studies recruiting 2301 patients (1137 ATB versus 1164 LTBI) were included in the quantitative analysis. The pooled sensitivity and specifity of IP-10 were 72% (95%CI: 0.59-0.82) and 78% (95%CI: 0.63-0.88), respectively. As for TBAg-stimulated IP-10, the sensitivity and specifity were 82% (95%CI: 0.76-0.87) and 85% (95%CI: 0.73-0.92), respectively. The senstivity was reduced signiticantly (p < 0.01) when the patients with human immunodeficiency virus infection were included, except after the TBAg stimulation.

CONCLUSION

Stimulating IP-10 with TBAg could improve the diagnostic accuracy in differentiating ATB from LTBI.

Head-to-head comparison of diagnostic accuracy of TB screening tests: Chest-X-ray, Xpert TB host response, and C-reactive protein

Background

Accessible, accurate screening tests are necessary to advance tuberculosis (TB) case finding and early detection in high-burden countries. We compared the diagnostic accuracy of available TB triage tests.

Methods

We prospectively screened consecutive adults with ≥2 weeks of cough presenting to primary health centers in the Philippines, Vietnam, South Africa, Uganda, and India. All participants received the index tests: chest-X-ray (CXR), venous or capillary Cepheid Xpert TB Host Response (HR) testing, and point-of-care C-reactive protein (CRP) testing (Boditech iChroma II). CXR images were processed using computer-aided detection (CAD) algorithms. We assessed diagnostic accuracy against a microbiologic reference standard (sputum Xpert Ultra, culture). Optimal cut-points were chosen to achieve sensitivity ≥90% and maximize specificity. Two-test screening algorithms were considered, using two approaches: 1) sequential negative serial screening in which the second screening test is conducted only if the first is negative and positive is defined as positive on either test and 2) sequential positive serial screening, in which the second screening test is conducted only if the first is positive and positive is defined as positive on both tests.

Results

Between July 2021 and August 2022, 1,392 participants with presumptive TB had valid results on index tests and the reference standard, and 303 (22%) had confirmed TB. In head-to-head comparisons, CAD4TB v7 showed the highest specificity when using a cut-point that achieves 90% sensitivity (70.3% vs. 65.1% for Xpert HR, difference 95% CI 1.6 to 8.9; 49.7% for CRP, difference 95% CI 17.0 to 24.3). Among the possible two-test screening algorithms, three met WHO target product profile (TPP) minimum accuracy thresholds and had higher accuracy than any test alone. At 90% sensitivity, the specificity was 79.6% for Xpert HR-CAD4TB [sequential negative], 75.9% for CRP-CAD4TB [sequential negative], and 73.7% for Xpert HR-CAD4TB [sequential positive].

Conclusions

CAD4TB achieves TPP targets and outperforms Xpert HR and CRP. Combining screening tests further increased accuracy. Cost and feasibility of two-test screening algorithms should be explored.

Registration

NCT04923958.

Magnitude and factors associated with delay in treatment-seeking among new pulmonary tuberculosis patients in public health facilities in Habro district, eastern Ethiopia

Background

In developing nations, almost two-thirds of people with active tuberculosis (TB) remain undetected and untreated. Delays in seeking treatment increase the severity of the illness, the likelihood of mortality, and the risk of the infection spreading to others in the community. Thus, this study aimed to assess the magnitude of delay in treatment-seeking and its associated factors among new adult pulmonary tuberculosis patients attending public health facilities in Habro District, West Harerge Zone, Oromia Region, Eastern Ethiopia.

Methods

A health facility-based cross-sectional study design was conducted among 420 randomly selected patients with pulmonary tuberculosis who visited public health facilities in Habro District from September 5 to October 5, 2022. Binary logistic regression analysis was used to determine the relationship between the dependent and independent variables, and a 95% confidence interval was used to select significant variables.

Result

Twenty-one days after the start of their illness, 62.38% (95% CI: 57.4%, 66.6%) of the patients sought an initial consultation. Being female (AOR = 2.14, 95% CI: 1.26, 3.65), having poor knowledge about TB (AOR = 3.10, 95% CI: 1.77, 5.43), having no contact history with TB patients (AOR = 3.52, 95% CI: 1.29, 9.58), having clinically diagnosed pulmonary tuberculosis (AOR = 2.16, 95% CI: 1.26, 3.67), and living a long distance from the nearest health facility (AOR = 2.87, 95% CI: 1.31, 6.23) were important predictors contributing to TB patient treatment delay.

Conclusion

In the current study, more than three-fifths of TB patients delayed seeking treatment. Thus, awareness of tuberculosis should be created, especially targeting females and communities found at a distance from the health facility. Additionally, health professionals should emphasise the importance of getting medical attention early and knowing how to suspect and identify tuberculosis symptoms.

Accuracy of digital chest x-ray analysis with artificial intelligence software as a triage and screening tool in hospitalized patients being evaluated for tuberculosis in Lima, Peru

Tuberculosis (TB) transmission in healthcare facilities is common in high-incidence countries. Yet, the optimal approach for identifying inpatients who may have TB is unclear. We evaluated the diagnostic accuracy of qXR (Qure.ai, India) computer-aided detection (CAD) software versions 3.0 and 4.0 (v3 and v4) as a triage and screening tool within the FAST (Find cases Actively, Separate safely, and Treat effectively) transmission control strategy. We prospectively enrolled two cohorts of patients admitted to a tertiary hospital in Lima, Peru: one group had cough or TB risk factors (triage) and the other did not report cough or TB risk factors (screening). We evaluated the sensitivity and specificity of qXR for the diagnosis of pulmonary TB using culture and Xpert as primary and secondary reference standards, including stratified analyses based on risk factors. In the triage cohort (n = 387), qXR v4 sensitivity was 0.91 (59/65, 95% CI 0.81-0.97) and specificity was 0.32 (103/322, 95% CI 0.27-0.37) using culture as reference standard. There was no difference in the area under the receiver-operating-characteristic curve (AUC) between qXR v3 and qXR v4 with either a culture or Xpert reference standard. In the screening cohort (n = 191), only one patient had a positive Xpert result, but specificity in this cohort was high (>90%). A high prevalence of radiographic lung abnormalities, most notably opacities (81%), consolidation (62%), or nodules (58%), was detected by qXR on digital CXR images from the triage cohort. qXR had high sensitivity but low specificity as a triage in hospitalized patients with cough or TB risk factors. Screening patients without cough or risk factors in this setting had a low diagnostic yield. These findings further support the need for population and setting-specific thresholds for CAD programs.

Computer-aided detection thresholds for digital chest radiography interpretation in tuberculosis diagnostic algorithms

Objectives

Use of computer-aided detection (CAD) software is recommended to improve tuberculosis screening and triage, but threshold determination is challenging if reference testing has not been performed in all individuals. We aimed to determine such thresholds through secondary analysis of the 2019 Lesotho national tuberculosis prevalence survey.

Methods

Symptom screening and chest radiographs were performed in participants aged ≥15 years; those symptomatic or with abnormal chest radiographs provided samples for Xpert MTB/RIF and culture testing. Chest radiographs were processed using CAD4TB version 7. We used six methodological approaches to deal with participants who did not have bacteriological test results to estimate pulmonary tuberculosis prevalence and assess diagnostic accuracy.

Results

Among 17 070 participants, 5214 (31%) had their tuberculosis status determined; 142 had tuberculosis. Prevalence estimates varied between methodological approaches (0.83-2.72%). Using multiple imputation to estimate tuberculosis status for those eligible but not tested, and assuming those not eligible for testing were negative, a CAD4TBv7 threshold of 13 had a sensitivity of 89.7% (95% CI 84.6-94.8) and a specificity of 74.2% (73.6-74.9), close to World Health Organization (WHO) target product profile criteria. Assuming all those not tested were negative produced similar results.

Conclusions

This is the first study to evaluate CAD4TB in a community screening context employing a range of approaches to account for unknown tuberculosis status. The assumption that those not tested are negative - regardless of testing eligibility status - was robust. As threshold determination must be context specific, our analytically straightforward approach should be adopted to leverage prevalence surveys for CAD threshold determination in other settings with a comparable proportion of eligible but not tested participants.

Accuracy of digital chest x-ray analysis with artificial intelligence software as a triage and screening tool in hospitalized patients being evaluated for tuberculosis in Lima, Peru

Introduction

Tuberculosis (TB) transmission in healthcare facilities is common in high-incidence countries. Yet, the optimal approach for identifying inpatients who may have TB is unclear. We evaluated the diagnostic accuracy of qXR (Qure.ai, India) computer-aided detection (CAD) software versions 3.0 and 4.0 (v3 and v4) as a triage and screening tool within the FAST (Find cases Actively, Separate safely, and Treat effectively) transmission control strategy.

Methods

We prospectively enrolled two cohorts of patients admitted to a tertiary hospital in Lima, Peru: one group had cough or TB risk factors (triage) and the other did not report cough or TB risk factors (screening). We evaluated the sensitivity and specificity of qXR for the diagnosis of pulmonary TB using culture and Xpert as primary and secondary reference standards, including stratified analyses based on risk factors.

Results

In the triage cohort (n=387), qXR v4 sensitivity was 0.91 (59/65, 95% CI 0.81-0.97) and specificity was 0.32 (103/322, 95% CI 0.27-0.37) using culture as reference standard. There was no difference in the area under the receiver-operating-characteristic curve (AUC) between qXR v3 and qXR v4 with either a culture or Xpert reference standard. In the screening cohort (n=191), only one patient had a positive Xpert result, but specificity in this cohort was high (>90%). A high prevalence of radiographic lung abnormalities, most notably opacities (81%), consolidation (62%), or nodules (58%), was detected by qXR on digital CXR images from the triage cohort.

Conclusions

qXR had high sensitivity but low specificity as a triage in hospitalized patients with cough or TB risk factors. Screening patients without cough or risk factors in this setting had a low diagnostic yield. These findings further support the need for population and setting-specific thresholds for CAD programs.

Volatile Organic Compound Identification-Based Tuberculosis Screening among TB Suspects: A Diagnostic Accuracy Study

Tuberculosis (TB) affects a third of the global population, and a large population of infected individuals still remain undiagnosed-making the visible burden only the tip of the iceberg. The detection of tuberculosis in close-proximity patients is one of the key priorities for attaining the Sustainable Development Goals (SDG) of TB elimination by 2030. With the current battery of screening tests failing to cover this need, the authors of this paper examined a simple and inexpensive point-of-care breath analyzer (TSI-3000(I)), which is based on detecting the volatile organic compounds that are emitted from infected cells and released in exhaled breath as a screening tool for the detection of TB. A single-center pilot study for assessing the diagnostic accuracy of the point-of-care Tuberculosis Breath Analyzer was conducted, and it was compared against the WHO-recommended TrueNat assay, which is a rapid molecular test and was also treated as the reference standard in this study. Of the 334 enrolled participants with TB signs/symptoms, 42.51% were TrueNat positive for Mycobacterium tuberculosis. The sensitivity of the Tuberculosis Breath Analyzer was found to be 95.7%, with a specificity of 91.3% and a ROC area of 0.935. The test kit showed considerable/significant high sensitivity and specificity as reliability indicators. The performance of the Tuberculosis Breath Analyzer tested was found to be comparable in efficiency to that of the TrueNat assay. A large cohort-based multicentric study is feasibly required to further validate and extrapolate the results of the pilot study.

Rapid and Accurate Diagnosis of Pediatric Tuberculosis Disease: A Diagnostic Accuracy Study for Pediatric Tuberculosis

INTRODUCTION

An estimated 1.2 million children develop tuberculosis (TB) every year with 240,000 dying because of missed diagnosis. Existing tools suffer from lack of accuracy and are often unavailable. Here, we describe the scientific and clinical methodology applied in RaPaed-TB, a diagnostic accuracy study.

METHODS

This prospective diagnostic accuracy study evaluating several candidate tests for TB was set out to recruit 1000 children <15 years with presumptive TB in 5 countries (Malawi, Mozambique, South Africa, Tanzania, India). Assessments at baseline included documentation of TB signs and symptoms, TB history, radiography, tuberculin skin test, HIV testing and spirometry. Respiratory samples for reference standard testing (culture, Xpert Ultra) included sputum (induced/spontaneous) or gastric aspirate, and nasopharyngeal aspirate (if <5 years). For novel tests, blood, urine and stool were collected. All participants were followed up at months 1 and 3, and month 6 if on TB treatment or unwell. The primary endpoint followed NIH-consensus statements on categorization of TB disease status for each participant. The study was approved by the sponsor's and all relevant local ethics committees.

DISCUSSION

As a diagnostic accuracy study for a disease with an imperfect reference standard, Rapid and Accurate Diagnosis of Pediatric Tuberculosis Disease (RaPaed-TB) was designed following a rigorous and complex methodology. This allows for the determination of diagnostic accuracy of novel assays and combination of testing strategies for optimal care for children, including high-risk groups (ie, very young, malnourished, children living with HIV). Being one of the largest of its kind, RaPaed-TB will inform the development of improved diagnostic approaches to increase case detection in pediatric TB.

Utility of Xpert MTB/RIF Ultra and digital chest radiography for the diagnosis and treatment of TB in people living with HIV: a randomised controlled trial (XACT-TB)

BACKGROUND

TB is a leading cause of morbidity among HIV positive individuals. Accurate algorithms are needed to achieve early TB diagnosis and treatment. We investigated the use of Xpert MTB/RIF Ultra in combination with chest radiography for TB diagnosis in ambulatory HIV positive individuals.

METHODS

This was a randomised controlled trial with a 2-by-2 factorial design. Outpatient HIV clinic attendees with cough were randomised to four arms: Arm 1-Standard Xpert/no chest radiography (CXR); Arm 2-Standard Xpert/CXR; Arm 3-Xpert Ultra/no CXR; and Arm 4-Xpert Ultra/CXR. Participants were followed up at days 28 and 56 to assess for TB treatment initiation.

RESULTS

We randomised 640 participants. Bacteriologically confirmed TB treatment initiation at day 28 were: Arm 1 (8.4% [14/162]), Arm 2 (6.9% [11/159]), Arm 3 (8.2% [13/159]) and Arm 4 (5.6% [9/160]) and between Xpert Ultra group (Arms 3 and 4) (6.9% [22/319]) vs Standard Xpert group (Arms 1 and 2) (7.8% [25/321]), risk ratio 0.89 (95% CI 0.51 to 1.54). By day 56, there were also similar all-TB treatment initiations in the x-ray group (Arms 2 and 4) (16.0% [51/319]) compared with the no x-ray group (Arms 1 and 3) (13.1% [42/321]), risk ratio 1.22 (95% CI 0.84 to 1.78); however, the contribution of clinically diagnosed treatment initiations were higher in x-ray groups (50.9% vs 19.0%).

CONCLUSIONS

Xpert Ultra performed similarly to Xpert MTB/RIF. X-rays are useful for TB screening but further research should investigate how to mitigate false-positive treatment initiations.

Clinical utility of C-reactive protein-based triage for presumptive pulmonary tuberculosis in South African adults

BACKGROUND

Identification of an accurate, low-cost triage test for pulmonary TB among people presenting to healthcare facilities is an urgent global research priority. We assessed the diagnostic accuracy and clinical utility of C-reactive protein (CRP) for TB triage among symptomatic adult outpatients, irrespective of HIV status.

METHODS

We prospectively enrolled adults reporting at least one (for people with HIV) or two (for people without HIV) symptoms of cough, fever, night sweats, or weight loss at two TB clinics in Cape Town, South Africa. Participants provided sputum for culture and Xpert MTB/RIF Ultra. We evaluated the diagnostic accuracy of CRP (measured using a laboratory-based assay) against a TB-culture reference standard as the area under the receiver operating characteristic curve (AUROC), and sensitivity and specificity at pre-specified thresholds. We assessed clinical utility using decision curve analysis and benchmarked against WHO recommendations.

RESULTS

Of 932 included individuals, 255 (27%) had culture-confirmed pulmonary TB and 389 (42%) were living with HIV. CRP demonstrated an AUROC of 0·80 (95% confidence interval 0·77-0·83), with sensitivity 93% (89-95%) and specificity 54% (50-58%) using a primary cut-off of ≥10 mg/L. Performance was similar among people with HIV to those without. In decision curve analysis, CRP-based triage offered greater clinical utility than confirmatory testing for all up to a number willing to test threshold of 20 confirmatory tests per true positive pulmonary TB case diagnosed (threshold probability 5%). If it is possible to perform more confirmatory tests than this, a 'confirmatory test for all' strategy performed better.

CONCLUSIONS

CRP achieved the WHO-defined sensitivity, but not specificity, targets for a triage test for pulmonary TB and showed evidence of clinical utility among symptomatic outpatients, irrespective of HIV status.

FUNDING

South African Medical Research Council, EDCTP2, Royal Society Newton Advanced Fellowship, Wellcome Trust, National Institute of Health Research, Royal College of Physicians.

Patient Preferences for Strategies to Improve Tuberculosis Diagnostic Services in Zambia

IMPORTANCE

Delayed engagement in tuberculosis (TB) services is associated with ongoing transmission and poor clinical outcomes.

OBJECTIVE

To assess whether patients with TB have differential preferences for strategies to improve the public health reach of TB diagnostic services.

DESIGN, SETTING, AND PARTICIPANTS

A cross-sectional study was undertaken in which a discrete choice experiment (DCE) was administered between September 18, 2019, and January 17, 2020, to 401 adults (>18 years of age) with microbiologically confirmed TB in Lusaka, Zambia. The DCE had 7 attributes with 2 to 3 levels per attribute related to TB service enhancements. Latent class analysis was used to identify segments of participants with unique preferences. Multiscenario simulations were used to estimate shares of preferences for different TB service improvement strategies.

MAIN OUTCOMES AND MEASURES

The main outcomes were patient preference archetypes and estimated shares of preferences for different strategies to improve TB diagnostic services. Collected data were analyzed between January 3, 2022, to July 2, 2022.

RESULTS

Among 326 adults with TB (median [IQR] age, 34 [27-42] years; 217 [66.8%] male; 158 [48.8%] HIV positive), 3 groups with distinct preferences for TB service improvements were identified. Group 1 (192 participants [58.9%]) preferred a facility that offered same-day TB test results, shorter wait times, and financial incentives for testing. Group 2 (83 participants [25.4%]) preferred a facility that provided same-day TB results, had greater privacy, and was closer to home. Group 3 (51 participants [15.6%]) had no strong preferences for service improvements and had negative preferences for receiving telephone-based TB test results. Groups 1 and 2 were more likely to report at least a 4-week delay in seeking health care for their current TB episode compared with group 3 (29 [51.3%] in group 1, 95 [35.8%] in group 2, and 10 [19.6%] in group 3; P < .001). Strategies to improve TB diagnostic services most preferred by all participants were same-day TB test results alone (shares of preference, 69.9%) and combined with a small financial testing incentive (shares of preference, 79.3%), shortened wait times (shares of preference, 76.1%), or greater privacy (shares of preference, 75.0%). However, the most preferred service improvement strategies differed substantially by group.

CONCLUSIONS AND RELEVANCE

In this study, patients with TB had heterogenous preferences for TB diagnostic service improvements associated with differential health care-seeking behavior. Tailored strategies that incorporate features most valued by persons with undiagnosed TB, including same-day results, financial incentives, and greater privacy, may optimize reach by overcoming key barriers to timely TB care engagement.

Are mRNA based transcriptomic signatures ready for diagnosing tuberculosis in the clinic? - A review of evidence and the technological landscape

Funding

Diagnostic Accuracy of the Cepheid 3-gene Host Response Fingerstick Blood Test in a Prospective, Multi-site Study: Interim Results

BACKGROUND

The development of a fast and accurate, non-sputum-based point-of-care triage test for tuberculosis (TB) would have a major impact on combating the TB burden worldwide. A new fingerstick blood test has been developed by Cepheid (the Xpert MTB Host Response [MTB-HR] prototype), which generates a "TB score" based on messenger RNA (mRNA) expression of 3 genes. Here we describe the first prospective findings of the MTB-HR prototype.

METHODS

Fingerstick blood from adults presenting with symptoms compatible with TB in South Africa, The Gambia, Uganda, and Vietnam was analyzed using the Cepheid GeneXpert MTB-HR prototype. Accuracy of the Xpert MTB-HR cartridge was determined in relation to GeneXpert Ultra results and a composite microbiological score (GeneXpert Ultra and liquid culture) with patients classified as having TB or other respiratory diseases (ORD).

RESULTS

When data from all sites (n = 75 TB, 120 ORD) were analyzed, the TB score discriminated between TB and ORD with an area under the curve (AUC) of 0.94 (95% confidence interval [CI], .91-.97), sensitivity of 87% (95% CI, 77-93%) and specificity of 94% (88-97%). When sensitivity was set at 90% for a triage test, specificity was 86% (95% CI, 75-97%). These results were not influenced by human immunodeficiency virus (HIV) status or geographical location. When evaluated against a composite microbiological score (n = 80 TB, 111 ORD), the TB score was able to discriminate between TB and ORD with an AUC of 0.88 (95% CI, .83-.94), 80% sensitivity (95% CI, 76-85%) and 94% specificity (95% CI, 91-96%).

CONCLUSIONS

Our interim data indicate the Cepheid MTB-HR cartridge reaches the minimal target product profile for a point of care triage test for TB using fingerstick blood, regardless of geographic area or HIV infection status.

Addressing TB-related mortality in adults living with HIV: a review of the challenges and potential solutions

Tuberculosis (TB) is the leading cause of death in people living with HIV (PLHIV) globally, causing 208,000 deaths in PLHIV in 2019. PLHIV have an 18-fold higher risk of TB, and HIV/TB mortality is highest in inpatient facilities, compared with primary care and community settings. Here we discuss challenges and potential mitigating solutions to address TB-related mortality in adults with HIV. Key factors that affect healthcare engagement are stigma, knowledge, and socioeconomic constraints, which are compounded in people with HIV/TB co-infection. Innovative approaches to improve healthcare engagement include optimizing HIV/TB care integration and interventions to reduce stigma. While early diagnosis of both HIV and TB can reduce mortality, barriers to early diagnosis of TB in PLHIV include difficulty producing sputum specimens, lower sensitivity of TB diagnostic tests in PLHIV, and higher rates of extra pulmonary TB. There is an urgent need to develop higher sensitivity biomarker-based tests that can be used for point-of-care diagnosis. Nonetheless, the implementation and scale-up of existing tests including molecular World Health Organization (WHO)-recommended diagnostic tests and urine lipoarabinomannan (LAM) should be optimized along with expanded TB screening with tools such as C-reactive protein and digital chest radiography. Decreased survival of PLHIV with TB disease is more likely with late HIV diagnosis and delayed start of antiretroviral (ART) treatment. The WHO now recommends starting ART within 2 weeks of initiating TB treatment in the majority of PLHIV, aside from those with TB meningitis. Dedicated TB treatment trials focused on PLHIV are needed, including interventions to improve TB meningitis outcomes given its high mortality, such as the use of intensified regimens using high-dose rifampin, new and repurposed drugs such as linezolid, and immunomodulatory therapy. Ultimately holistic, high-quality, person-centered care is needed for PLHIV with TB throughout the cascade of care, which should address biomedical, socioeconomic, and psychological barriers.

Automatic cough classification for tuberculosis screening in a real-world environment

Objective.The automatic discrimination between the coughing sounds produced by patients with tuberculosis (TB) and those produced by patients with other lung ailments.Approach.We present experiments based on a dataset of 1358 forced cough recordings obtained in a developing-world clinic from 16 patients with confirmed active pulmonary TB and 35 patients suffering from respiratory conditions suggestive of TB but confirmed to be TB negative. Using nested cross-validation, we have trained and evaluated five machine learning classifiers: logistic regression (LR), support vector machines, k-nearest neighbour, multilayer perceptrons and convolutional neural networks.Main Results.Although classification is possible in all cases, the best performance is achieved using LR. In combination with feature selection by sequential forward selection, our best LR system achieves an area under the ROC curve (AUC) of 0.94 using 23 features selected from a set of 78 high-resolution mel-frequency cepstral coefficients. This system achieves a sensitivity of 93% at a specificity of 95% and thus exceeds the 90% sensitivity at 70% specificity specification considered by the World Health Organisation (WHO) as a minimal requirement for a community-based TB triage test.Significance.The automatic classification of cough audio sounds, when applied to symptomatic patients requiring investigation for TB, can meet the WHO triage specifications for the identification of patients who should undergo expensive molecular downstream testing. This makes it a promising and viable means of low cost, easily deployable frontline screening for TB, which can benefit especially developing countries with a heavy TB burden.

Computer-aided X-ray screening for tuberculosis and HIV testing among adults with cough in Malawi (the PROSPECT study): A randomised trial and cost-effectiveness analysis

BACKGROUND

Suboptimal tuberculosis (TB) diagnostics and HIV contribute to the high global burden of TB. We investigated costs and yield from systematic HIV-TB screening, including computer-aided digital chest X-ray (DCXR-CAD).

METHODS AND FINDINGS

In this open, three-arm randomised trial, adults (≥18 years) with cough attending acute primary services in Malawi were randomised (1:1:1) to standard of care (SOC); oral HIV testing (HIV screening) and linkage to care; or HIV testing and linkage to care plus DCXR-CAD with sputum Xpert for high CAD4TBv5 scores (HIV-TB screening). Participants and study staff were not blinded to intervention allocation, but investigator blinding was maintained until final analysis. The primary outcome was time to TB treatment. Secondary outcomes included proportion with same-day TB treatment; prevalence of undiagnosed/untreated bacteriologically confirmed TB on day 56; and undiagnosed/untreated HIV. Analysis was done on an intention-to-treat basis. Cost-effectiveness analysis used a health-provider perspective. Between 15 November 2018 and 27 November 2019, 8,236 were screened for eligibility, with 473, 492, and 497 randomly allocated to SOC, HIV, and HIV-TB screening arms; 53 (11%), 52 (9%), and 47 (9%) were lost to follow-up, respectively. At 56 days, TB treatment had been started in 5 (1.1%) SOC, 8 (1.6%) HIV screening, and 15 (3.0%) HIV-TB screening participants. Median (IQR) time to TB treatment was 11 (6.5 to 38), 6 (1 to 22), and 1 (0 to 3) days (hazard ratio for HIV-TB versus SOC: 2.86, 1.04 to 7.87), with same-day treatment of 0/5 (0%) SOC, 1/8 (12.5%) HIV, and 6/15 (40.0%) HIV-TB screening arm TB patients (p = 0.03). At day 56, 2 SOC (0.5%), 4 HIV (1.0%), and 2 HIV-TB (0.5%) participants had undiagnosed microbiologically confirmed TB. HIV screening reduced the proportion with undiagnosed or untreated HIV from 10 (2.7%) in the SOC arm to 2 (0.5%) in the HIV screening arm (risk ratio [RR]: 0.18, 0.04 to 0.83), and 1 (0.2%) in the HIV-TB screening arm (RR: 0.09, 0.01 to 0.71). Incremental costs were US$3.58 and US$19.92 per participant screened for HIV and HIV-TB; the probability of cost-effectiveness at a US$1,200/quality-adjusted life year (QALY) threshold was 83.9% and 0%. Main limitations were the lower than anticipated prevalence of TB and short participant follow-up period; cost and quality of life benefits of this screening approach may accrue over a longer time horizon.

CONCLUSIONS

DCXR-CAD with universal HIV screening significantly increased the timeliness and completeness of HIV and TB diagnosis. If implemented at scale, this has potential to rapidly and efficiently improve TB and HIV diagnosis and treatment.

TRIAL REGISTRATION

clinicaltrials.gov NCT03519425.

Pathways to care and preferences for improving tuberculosis services among tuberculosis patients in Zambia: A discrete choice experiment

BACKGROUND

Delays in the diagnosis of tuberculosis (TB) contribute to a substantial proportion of TB-related mortality, especially among people living with HIV (PLHIV). We sought to characterize the diagnostic journey for HIV-positive and HIV-negative patients with a new TB diagnosis in Zambia, to understand drivers of delay, and characterize their preferences for service characteristics to inform improvements in TB services.

METHODS

We assessed consecutive adults with newly microbiologically-confirmed TB at two public health treatment facilities in Lusaka, Zambia. We administered a survey to document critical intervals in the TB care pathway (time to initial care-seeking, diagnosis and treatment initiation), identify bottlenecks and their reasons. We quantified patient preferences for a range of characteristics of health services using a discrete choice experiment (DCE) that assessed 7 attributes (distance, wait times, hours of operation, confidentiality, sex of provider, testing incentive, TB test speed and notification method).

RESULTS

Among 401 patients enrolled (median age of 34 years, 68.7% male, 46.6% HIV-positive), 60.9% and 39.1% were from a first-level and tertiary hospital, respectively. The median time from symptom onset to receipt of TB treatment was 5.0 weeks (IQR: 3.6-8.0) and was longer among HIV-positive patients seeking care at a tertiary hospital than HIV-negative patients (6.4 vs. 4.9 weeks, p = 0.002). The time from symptom onset to initial presentation for evaluation accounted for the majority of time until treatment initiation (median 3.0 weeks, IQR: 1.0-5.0)-an important minority of 11.0% of patients delayed care-seeking ≥8 weeks. The DCE found that patients strongly preferred same-day TB test results (relative importance, 37.2%), facilities close to home (18.0%), and facilities with short wait times (16.9%). Patients were willing to travel to a facility up to 7.6 kilometers further away in order to access same-day TB test results. Preferences for improving current TB services did not differ according to HIV status.

CONCLUSIONS

Prolonged intervals from TB symptom onset to treatment initiation were common, especially among PLHIV, and were driven by delayed health-seeking. Addressing known barriers to timely diagnosis and incorporating patients' preferences into TB services, including same-day TB test results, may facilitate earlier TB care engagement in high burden settings.

Tuberculosis detection from chest x-rays for triaging in a high tuberculosis-burden setting: an evaluation of five artificial intelligence algorithms

BACKGROUND

Artificial intelligence (AI) algorithms can be trained to recognise tuberculosis-related abnormalities on chest radiographs. Various AI algorithms are available commercially, yet there is little impartial evidence on how their performance compares with each other and with radiologists. We aimed to evaluate five commercial AI algorithms for triaging tuberculosis using a large dataset that had not previously been used to train any AI algorithms.

METHODS

Individuals aged 15 years or older presenting or referred to three tuberculosis screening centres in Dhaka, Bangladesh, between May 15, 2014, and Oct 4, 2016, were recruited consecutively. Every participant was verbally screened for symptoms and received a digital posterior-anterior chest x-ray and an Xpert MTB/RIF (Xpert) test. All chest x-rays were read independently by a group of three registered radiologists and five commercial AI algorithms: CAD4TB (version 7), InferRead DR (version 2), Lunit INSIGHT CXR (version 4.9.0), JF CXR-1 (version 2), and qXR (version 3). We compared the performance of the AI algorithms with each other, with the radiologists, and with the WHO's Target Product Profile (TPP) of triage tests (≥90% sensitivity and ≥70% specificity). We used a new evaluation framework that simultaneously evaluates sensitivity, proportion of Xpert tests avoided, and number needed to test to inform implementers' choice of software and selection of threshold abnormality scores.

FINDINGS

Chest x-rays from 23 954 individuals were included in the analysis. All five AI algorithms significantly outperformed the radiologists. The areas under the receiver operating characteristic curve were 90·81% (95% CI 90·33-91·29) for qXR, 90·34% (89·81-90·87) for CAD4TB, 88·61% (88·03-89·20) for Lunit INSIGHT CXR, 84·90% (84·27-85·54) for InferRead DR, and 84·89% (84·26-85·53) for JF CXR-1. Only qXR (74·3% specificity [95% CI 73·3-74·9]) and CAD4TB (72·9% specificity [72·3-73·5]) met the TPP at 90% sensitivity. All five AI algorithms reduced the number of Xpert tests required by 50% while maintaining a sensitivity above 90%. All AI algorithms performed worse among older age groups (>60 years) and people with a history of tuberculosis.

INTERPRETATION

AI algorithms can be highly accurate and useful triage tools for tuberculosis detection in high-burden regions, and outperform human readers.

FUNDING

Government of Canada.

Computer-aided interpretation of chest radiography reveals the spectrum of tuberculosis in rural South Africa

Computer-aided digital chest radiograph interpretation (CAD) can facilitate high-throughput screening for tuberculosis (TB), but its use in population-based active case-finding programs has been limited. In an HIV-endemic area in rural South Africa, we used a CAD algorithm (CAD4TBv5) to interpret digital chest x-rays (CXR) as part of a mobile health screening effort. Participants with TB symptoms or CAD4TBv5 score above the triaging threshold were referred for microbiological sputum assessment. During an initial pilot phase, a low CAD4TBv5 triaging threshold of 25 was selected to maximize TB case finding. We report the performance of CAD4TBv5 in screening 9,914 participants, 99 (1.0%) of whom were found to have microbiologically proven TB. CAD4TBv5 was able to identify TB cases at the same sensitivity but lower specificity as a blinded radiologist, whereas the next generation of the algorithm (CAD4TBv6) achieved comparable sensitivity and specificity to the radiologist. The CXRs of people with microbiologically confirmed TB spanned a range of lung field abnormality, including 19 (19.2%) cases deemed normal by the radiologist. HIV serostatus did not impact CAD4TB's performance. Notably, 78.8% of the TB cases identified during this population-based survey were asymptomatic and therefore triaged for sputum collection on the basis of CAD4TBv5 score alone. While CAD4TBv6 has the potential to replace radiologists for triaging CXRs in TB prevalence surveys, population-specific piloting is necessary to set the appropriate triaging thresholds. Further work on image analysis strategies is needed to identify radiologically subtle active TB.

Evaluation of Triage Tests When Existing Test Capacity Is Constrained: Application to Rapid Diagnostic Testing in COVID-19

Objectives

A triage test is used to determine which patients will undergo an existing or “reference” test. This article explores the potential value of using triage tests before reference tests when the capacity of the reference test is constrained.

Methods

We developed a simple model with inputs: prevalence, sensitivity, specificity, and reference test capacity. We included a case study of rapid diagnostic tests for SARS-CoV-2 antigens used as triage tests before a reference polymerase chain reaction test. Performance data were obtained from an evaluation performed by an academic center on 425 samples from testing centers in the United Kingdom and Germany.

Results

When reference test capacity is constrained, the use of a triage test leads to a relative expansion of the population tested and cases identified; both are higher with a high-specificity triage test. When reference test capacity is not constrained, the potential advantages of introducing a triage test can be assessed using a standard cost-utility framework, balancing the utility of the reduction in the number of reference tests required against the disutility of missed cases associated with the use of a lower-sensitivity triage test. In the constrained case, the advantage of a triage testing strategy in terms of population covered and cases identified is reduced as the prevalence increases. In the unconstrained case, the reduction in reference tests required is reduced and the number of cases missed increase as the prevalence rises.

Conclusion

When the availability of the reference test is constrained, tests added in a triage position do not need high levels of accuracy to increase the number of cases diagnosed. This has implications in many disease areas, including COVID-19.

Characterization of oral swab samples for diagnosis of pulmonary tuberculosis

Oral swab analysis (OSA) has been shown to detect Mycobacterium tuberculosis (MTB) DNA in patients with pulmonary tuberculosis (TB). In previous analyses, qPCR testing of swab samples collected from tongue dorsa was up to 93% sensitive relative to sputum GeneXpert, when 2 swabs per patient were tested. The present study modified sample collection methods to increase sample biomass and characterized the viability of bacilli present in tongue swabs. A qPCR targeting conserved bacterial ribosomal rRNA gene (rDNA) sequences was used to quantify bacterial biomass in samples. There was no detectable reduction in total bacterial rDNA signal over the course of 10 rapidly repeated tongue samplings, indicating that swabs collect only a small portion of the biomass available for testing. Copan FLOQSwabs collected ~2-fold more biomass than Puritan PurFlock swabs, the best brand used previously (p = 0.006). FLOQSwabs were therefore evaluated in patients with possible TB in Uganda. A FLOQSwab was collected from each patient upon enrollment (Day 1) and, in a subset of sputum GeneXpert Ultra-positive patients, a second swab was collected on the following day (Day 2). Swabs were tested for MTB DNA by manual IS6110-targeted qPCR. Relative to sputum GeneXpert Ultra, single-swab sensitivity was 88% (44/50) on Day 1 and 94.4% (17/18) on Day 2. Specificity was 79.2% (42/53). Among an expanded sample of Ugandan patients, 62% (87/141) had colony-forming bacilli in their tongue dorsum swab samples. These findings will help guide further development of this promising TB screening method.

Validation of Differentially Expressed Immune Biomarkers in Latent and Active Tuberculosis by Real-Time PCR

Tuberculosis (TB) remains a major global threat and diagnosis of active TB ((ATB) both extra-pulmonary (EPTB), pulmonary (PTB)) and latent TB (LTBI) infection remains challenging, particularly in high-burden countries which still rely heavily on conventional methods. Although molecular diagnostic methods are available, e.g., Cepheid GeneXpert, they are not universally available in all high TB burden countries. There is intense focus on immune biomarkers for use in TB diagnosis, which could provide alternative low-cost, rapid diagnostic solutions. In our previous gene expression studies, we identified peripheral blood leukocyte (PBL) mRNA biomarkers in a non-human primate TB aerosol-challenge model. Here, we describe a study to further validate select mRNA biomarkers from this prior study in new cohorts of patients and controls, as a prerequisite for further development. Whole blood mRNA was purified from ATB patients recruited in the UK and India, LTBI and two groups of controls from the UK (i) a low TB incidence region (CNTRLA) and (ii) individuals variably-domiciled in the UK and Asia ((CNTRLB), the latter TB high incidence regions). Seventy-two mRNA biomarker gene targets were analyzed by qPCR using the Roche Lightcycler 480 qPCR platform and data analyzed using GeneSpring™ 14.9 bioinformatics software. Differential expression of fifty-three biomarkers was confirmed between MTB infected, LTBI groups and controls, seventeen of which were significant using analysis of variance (ANOVA): CALCOCO2, CD52, GBP1, GBP2, GBP5, HLA-B, IFIT3, IFITM3, IRF1, LOC400759 (GBP1P1), NCF1C, PF4V1, SAMD9L, S100A11, TAF10, TAPBP, and TRIM25. These were analyzed using receiver operating characteristic (ROC) curve analysis. Single biomarkers and biomarker combinations were further assessed using simple arithmetic algorithms. Minimal combination biomarker panels were delineated for primary diagnosis of ATB (both PTB and EPTB), LTBI and identifying LTBI individuals at high risk of progression which showed good performance characteristics. These were assessed for suitability for progression against the standards for new TB diagnostic tests delineated in the published World Health Organization (WHO) technology product profiles (TPPs).

Tuberculosis diagnosis cascade in Blantyre, Malawi: a prospective cohort study

BACKGROUND

Tuberculosis (TB) control relies on early diagnosis and treatment. International guidelines recommend systematic TB screening at health facilities, but implementation is challenging. We investigated completion of recommended TB screening steps in Blantyre, Malawi.

METHODS

A prospective cohort recruited adult outpatients attending Bangwe primary clinic. Entry interviews were linked to exit interviews. The proportion of participants progressing through each step of the diagnostic pathway were estimated. Factors associated with request for sputum were investigated using multivariable logistic regression.

RESULTS

Of 5442 clinic attendances 2397 (44%) had exit interviews. In clinically indicated participants (n = 445) 256 (57.5%) were asked about cough, 36 (8.1%) were asked for sputum, 21 (4.7%) gave sputum and 1 (0.2%) received same-day results. Significant associations with request for sputum were: any TB symptom (aOR:3.20, 95%CI:2.02-5.06), increasing age (aOR:1.02, 95%CI:1.01-1.04 per year) and for HIV-negative participants only, a history of previous TB (aOR:3.37, 95%CI:1.45-7.81). Numbers requiring sputum tests (26/day) outnumbered diagnostic capacity (8-12/day).

CONCLUSIONS

Patients were lost at every stage of the TB care cascade, with same day sputum submission following all steps of the diagnosis cascade achieved in only 4.7% if clinically indicated. Infection control strategies should be implemented, with reporting on early steps of the TB care cascade formalised. High-throughput screening interventions, such as digital CXR, that can achieve same-day TB diagnosis are urgently needed to meet WHO End TB goals.

Chest x-ray analysis with deep learning-based software as a triage test for pulmonary tuberculosis: a prospective study of diagnostic accuracy for culture-confirmed disease

BACKGROUND

Deep learning-based radiological image analysis could facilitate use of chest x-rays as triage tests for pulmonary tuberculosis in resource-limited settings. We sought to determine whether commercially available chest x-ray analysis software meet WHO recommendations for minimal sensitivity and specificity as pulmonary tuberculosis triage tests.

METHODS

We recruited symptomatic adults at the Indus Hospital, Karachi, Pakistan. We compared two software, qXR version 2.0 (qXRv2) and CAD4TB version 6.0 (CAD4TBv6), with a reference of mycobacterial culture of two sputa. We assessed qXRv2 using its manufacturer prespecified threshold score for chest x-ray classification as tuberculosis present versus not present. For CAD4TBv6, we used a data-derived threshold, because it does not have a prespecified one. We tested for non-inferiority to preset WHO recommendations (0·90 for sensitivity, 0·70 for specificity) using a non-inferiority limit of 0·05. We identified factors associated with accuracy by stratification and logistic regression.

FINDINGS

We included 2198 (92·7%) of 2370 enrolled participants. 2187 (99·5%) of 2198 were HIV-negative, and 272 (12·4%) had culture-confirmed pulmonary tuberculosis. For both software, accuracy was non-inferior to WHO-recommended minimum values (qXRv2 sensitivity 0·93 [95% CI 0·89-0·95], non-inferiority p=0·0002; CAD4TBv6 sensitivity 0·93 [0·90-0·96], p<0·0001; qXRv2 specificity 0·75 [0·73-0·77], p<0·0001; CAD4TBv6 specificity 0·69 [0·67-0·71], p=0·0003). Sensitivity was lower in smear-negative pulmonary tuberculosis for both software, and in women for CAD4TBv6. Specificity was lower in men and in those with previous tuberculosis, and reduced with increasing age and decreasing body mass index. Smoking and diabetes did not affect accuracy.

INTERPRETATION

In an HIV-negative population, these software met WHO-recommended minimal accuracy for pulmonary tuberculosis triage tests. Sensitivity will be lower when smear-negative pulmonary tuberculosis is more prevalent.

FUNDING

Canadian Institutes of Health Research.

Evaluation of multi-antigen serological screening for active tuberculosis among people living with HIV

Better triage tests for screening tuberculosis (TB) disease are needed for people living with HIV (PLHIV). We performed the first evaluation of a previously-validated 8-antigen serological panel to screen PLHIV for pulmonary TB in Kampala, Uganda. We selected a random 1:1 sample with and without TB (defined by sputum culture) from a cohort of PLHIV initiating antiretroviral therapy. We used a multiplex microbead immunoassay and an ensemble machine learning classifier to determine the area under the receiver operating characteristic curve (AUC) for Ag85A, Ag85B, Ag85C, Rv0934-P38, Rv3881, Rv3841-BfrB, Rv3873, and Rv2878c. We then assessed the performance with the addition of four TB-specific antigens ESAT-6, CFP-10, Rv1980-MPT64, and Rv2031-HSPX, and every antigen combination. Of 262 participants (median CD4 cell-count 152 cells/μL [IQR 65-279]), 138 (53%) had culture-confirmed TB. The 8-antigen panel had an AUC of 0.53 (95% CI 0.40-0.66), and the additional 4 antigens did not improve performance (AUC 0.51, 95% CI 0.39-0.64). When sensitivity was restricted to ≥90% for the 8- and 12-antigen panel, specificity was 2.2% (95% CI 0-17.7%) and 8.1% (95% CI 0-23.9%), respectively. A three-antigen combination (Rv0934-P38, Ag85A, and Rv2031-HSPX) outperformed both panels, with an AUC of 0.60 (95% CI 0.48-0.73), 90% sensitivity (95% CI 78.2-96.7%) and 29.7% specificity (95% CI 15.9-47%). The multi-antigen panels did not achieve the target accuracy for a TB triage test among PLHIV. We identified a new combination that improved performance for TB screening in an HIV-positive sample compared to an existing serological panel in Uganda, and suggests an approach to identify novel antigen combinations specifically for screening TB in PLHIV.