Zheyuan Wu

Introduction: Treatment compliance is considered significant for curing tuberculosis (TB). Digital adherence technologies (DATs) have been recommended for improving treatment compliance and outcomes of patients with TB. Whereas the effectiveness of DATs on TB treatment remains indefinite, we estimated whether a mobile application has an effect on treatment adherence and outcomes among patients with TB in an urban area in China. Methods: We conducted a multicentre, two-arm, open-label, equivalent, stratified block randomisation trial from 7 districts involving 4670 eligible new patients with TB in Shanghai, China. Patients in the intervention group were instructed to get access to the app through WeChat. In the control group, patients were treated with standard of care. The primary outcome of this study was a composite outcome, which included death, treatment failure, loss to follow-up and relapse. The secondary outcomes were mainly described as overall poor treatment adherence and monthly poor treatment adherence. A generalised estimating equation is used to estimate the correlation between the two groups’ outcomes after adjusting for concomitant variables. Results: Between 1 March 2021 and 31 December 2022, 741 patients (371 in the intervention group and 370 in the control group) were included in the intention-to-treat population, 430 (58%) were male and the median age was 33 years (IQR 27–49). 80 (22%) patients in the intervention group and 104 (28%) patients in the control group had a primary composite outcome event. The adjusted risk difference for the intervention versus control was −7.04 percentage points (95% CI −13.25 to −1.14, p=0.025); secondary outcome included overall poor treatment adherence with an adjusted risk difference −16.39 percentage points (95% CI −22.13 to −9.62, p<0.001). Conclusions: The mobile health application improved TB treatment adherence and outcomes significantly, especially for loss to follow-up. Future research should be focused on implementation problems in various settings and the improvement of patient-centred management strategies. Trial registration number: ChiCTR2000037575.

Background: Xpert MTB/RIF (Xpert) is an automated molecular test recommended by World Health Organization (WHO) for diagnosis of tuberculosis (TB). This study evaluated the effect of Xpert implementation on the detection of pulmonary TB (PTB) and rifampicin-resistant TB (RR-TB) cases in Shanghai, China. Methods: Xpert was routinely implemented in 2018 for all presumptive PTB patients. All PTB patients above 15 years-old identified within the Provincial TB Control Program during the first half of each of 2017 and 2018, were enrolled to compare the difference in proportions of bacteriological confirmation, patients with drug susceptibility test (DST) results for rifampicin (ie, DST coverage) and RR-TB detection before and after Xpert’s implementation. Results: A total of 6047 PTB patients were included in the analysis with 1691 tested by Xpert in 2018. Percentages of bacteriological confirmation, DST coverage and RR-TB detection in 2017 and 2018 were 50% vs. 59%, 36% vs. 49% and 2% vs. 3%, respectively (all p-values < 0.05). Among 1103 PTB patients who completed sputum smear, culture and Xpert testing in 2018, Xpert detected an additional 121 (11%) PTB patients who were negative by smear and culture, but missed 248 (23%) smear and/or culture positive patients. Besides, it accounted for an increase of 9% in DST coverage and 1% in RR-TB detection. The median time from first visit to a TB hospital to RR-TB detection was 62 days (interquartile range -IQR 48–84.2) in 2017 vs. 9 days (IQR 2–45.7) in 2018 (p-value < 0.001). In the multivariate model, using Xpert was associated with decreased time to RR-TB detection (adjusted hazard ratio = 4.62, 95% confidence interval: 3.18–6.71). Conclusions: Integrating Xpert with smear, culture and culture-based DST in a routine setting significantly increased bacteriological confirmation, DST coverage and RR-TB detection with a dramatic reduction in the time to RR-TB diagnosis in Shanghai, China. Our findings can be useful for other regions that attempt to integrate Xpert into routine PTB and RR-TB case-finding cascade. Further study should focus on the identification and elimination of operational level challenges to fully utilize the benefit of rapid diagnosis by Xpert.

To assess the effect of a contact investigation strategy by assessing the incidence of tuberculosis and the case detection rate among contacts of tuberculosis patients. The pulmonary tuberculosis incidence among contacts was determined retrospectively from a tuberculosis information management system. For each detection method (symptom examination only, symptom examination plus chest radiography or other alternatives), the detection rate of pulmonary tuberculosis patients among contacts was derived from contact investigation form records. Sixty-nine cases of pulmonary tuberculosis were identified among a total of 8137 contacts after an average follow-up of 2.6 years (range: 0.25–5.25) during the period from 2010 to 2014. The incidence density was 329/100,000 person-years (PYs), and the 95% confidence interval (CI) was 256–419/100,000 PYs, which was significantly higher than the notification rate during the same period in the general population (29–30/100,000 PYs). The incidence density was higher (p < 0.0001) among male contacts (462/100,000 PYs) than among female contacts (236/100,000 PYs). The incidence density did not differ (p > 0.05) between contacts whose index case was sputum smear positive and those whose index case was sputum smear negative. Contacts who were biologically related family of the index cases exhibited a higher (p < 0.05) incidence density (475/100,000 PYs) than other contacts (281/100,000 PYs). Fifteen of the 69 incident cases were found through contact investigation, corresponding to a case detection rate via contact investigation of 22% (95% CI: 13–33%). The relevance ratio was 288/100,000 (12/4163) by both chest radiography and symptom survey, which was significantly higher than the rate detected by symptom survey alone, of 57/100,000 (2/3486), p = 0.028. The cumulative incidence in the contacts was 761/100,000 (62/8137) within 3 years from the time that the index cases were diagnosed with pulmonary tuberculosis, which was higher than the incidence rate of 210/100,000 (7/3328) recorded after 3 years (p < 0.001). The contacts were at higher risk of pulmonary tuberculosis than the general population; however, only approximately 22% of the incident cases could be detected through contact investigation. Therefore, the contact investigation strategy must be improved for better detection of potential pulmonary tuberculosis cases.