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. MethodsWe 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.

Background Automated radiologic analysis using computer-aided detection software (CAD) could facilitate chest X-ray (CXR) use in tuberculosis diagnosis. There is little to no evidence on the accuracy of commercially-available deep learning-based CAD in different populations, including patients with smear-negative tuberculosis and people living with HIV (PLWH). Methods We collected CXRs and individual patient data (IPD) from studies evaluating CAD in patients self-referring for tuberculosis symptoms with culture or nucleic acid amplification testing as the reference. We re-analyzed CXRs with three CAD (CAD4TB version (v) 6, Lunit v3.1.0.0, and qXR v2). We estimated sensitivity and specificity within each study and pooled using IPD meta-analysis. We used multivariable meta-regression to identify characteristics modifying accuracy. Results We included CXRs and IPD of 3727/3967 participants from 4/7 eligible studies. 17% (621/3727) were PLWH. 17% (645/3727) had microbiologically-confirmed tuberculosis. Despite using the same threshold score for classifying CXR in every study, sensitivity and specificity varied from study to study. The software had similar unadjusted accuracy (at 90% pooled sensitivity, pooled specificities were: CAD4TBv6, 56.9% [95%CI:51.7-61.9]; Lunit, 54.1% [44.6-63.3]; qXRv2, 60.5% [51.7-68.6]). Adjusted absolute differences in pooled sensitivity between PLWH and HIV-uninfected participants was: CAD4TBv6, -13.4% [-21.1, -6.9]; Lunit, +2.2% [-3.6, +6.3]; qXRv2: -13.4% [-21.5, -6.6]); between smear-negative and smear-positive tuberculosis was: CAD4TBv6, -12.3% [-19.5, -6.1]; Lunit, -17.2% [-24.6, -10.5]; qXRv2, -16.6% [-24.4, -9.9]. Accuracy was similar to human readers. Conclusions For CAD CXR analysis to be implemented as a high-sensitivity tuberculosis rule-out test, users will need threshold scores identified from their own patient populations, and stratified by HIV- and smear-status.

No abstract

Background Successful delivery and completion of tuberculosis preventive treatment delivery is necessary for tuberculosis elimination. Shorter preventive treatment regimens currently have higher medication costs, but patients spend less time in care and are more likely to complete treatment. It is unknown how economic costs of successful delivery differ between longer and shorter regimens in high-tuberculosis-burden settings. Methods We developed survey instruments to collect costs from program and patient sources, considering costs incurred from when household contacts first entered the health system. We compared the cost per completed course of preventive treatment with either 6 months of daily isoniazid (6H) or 3 months of weekly isoniazid and rifapentine (3HP), delivered by the Indus tuberculosis program in Karachi, Pakistan, between October 2016 and February 2018. Results During this period, 459 individuals initiated 6H, and 643 initiated 3HP; 39% and 61% completed treatment, respectively. Considering costs to both the program and care recipients, the cost per completed course was 394 USD for 6H and 333 USD for 3HP. Using a new 2020 price for rifapentine reduced the cost per completed course of 3HP to 290 USD. Under varying assumptions about drug prices and costs incurred by care recipients, the cost per completed course was lower for 3HP in all scenarios, and the largest cost drivers were the salaries of clinical staff. Conclusions In a high-burden setting, the cost of successful delivery of 3HP was lower than that of 6H, driven by higher completion.

INTRODUCTION Tobacco smoking among tuberculosis (TB) patients leads to poorer treatment outcomes. Smoking cessation support should be integrated into routine TB care. We measured healthcare providers’ fidelity to a smoking cessation intervention integrated into routine TB care, in Bangladesh and Pakistan. We aimed to understand the role of providers and settings in the implementation of behavior support (BS) messages for TB and smoking cessation. METHODS The integrated BS intervention was implemented in TB clinics (24 public and 1 private). Cross-sectional data were collected on the fidelity of delivery of the BS intervention using a predefined fidelity index based on an existing validated method of measuring intervention fidelity. Audio-recordings of patient-provider BS sessions were coded using the fidelity index. Intervention fidelity was presented as the proportion of sessions that implemented BS messages. RESULTS A total of 96 sessions were conducted, 37 in Bangladesh and 59 in Pakistan. In public settings, TB medication advice was offered in 91.9% (95% CI: 78.7– 97.2) of sessions in Bangladesh, and in 75.5% (95% CI: 62.4–85.1) of sessions in Pakistan; whilst it was offered in 83.3% (95% CI: 43.7–97.0) of sessions in the private setting in Pakistan. Patients’ smoking status was assessed in 70.3% (95% CI: 54.2–82.5) of sessions in Bangladesh, and in 34.0% (95% CI: 22.7–47.4) of sessions in the public setting and in 66.7% (95% CI: 30.0–90.3) of sessions in the private setting in Pakistan. A quit date was set in 32.4% (95% CI: 19.6–48.5) of all sessions in Bangladesh, and in 33.3% (95% CI: 9.6–70.0) of all sessions in the public setting in Pakistan. CONCLUSIONS Fidelity to the intended delivery of the intervention was found to be high for TB-related messages but not for smoking cessation messages. Clinic contexts may play a mediating role in health workers’ opportunities to deliver the intervention as planned. TRIAL REGISTRATION International Standard Randomized Clinical Trial Number (ISRCTN43811467). Registered 23 March 2016, https://doi.org/10.1186/ISRCTN43811467