Geographical targeting of active case finding for tuberculosis in Pakistan using hotspots identified by artificial intelligence software (SPOT-TB): study protocol for a pragmatic stepped wedge cluster randomised control trial.

INTRODUCTION: Pakistan has significantly strengthened its capacity for active case finding (ACF) for tuberculosis (TB) that is being implemented at scale in the country. However, yields of ACF have been lower than expected, raising concerns on its effectiveness in the programmatic setting. Distribution of TB in communities is likely to be spatially heterogeneous and targeting of ACF in areas with higher TB prevalence may help improve yields. The primary aim of SPOT-TB is to investigate whether a policy change to use a geographically targeted approach towards ACF supported by an artificial intelligence (AI) software, MATCH-AI, can improve yields in Pakistan. METHODS AND ANALYSIS: SPOT-TB will use a pragmatic, stepped wedge cluster randomised design. A total of 30 mobile X-ray units and their field teams will be randomised to receive the intervention. Site selection for ACF in the intervention areas will be guided primarily through the use of MATCH-AI software that models subdistrict TB prevalence and identifies potential disease hotspots. Control areas will use existing approaches towards site selection that are based on staff knowledge, experience and analysis of historical data. The primary outcome measure is the difference in bacteriologically confirmed incident TB detected in the intervention relative to control areas. All remaining ACF-related procedures and algorithms will remain unaffected by this trial. ETHICS AND DISSEMINATION: Ethical approval has been obtained from the Health Services Academy, Islamabad, Pakistan (7-82/IERC-HSA/2022-52) and from the Common Management Unit for TB, HIV and Malaria, Ministry of Health Services, Regulation and Coordination, Islamabad, Pakistan (26-IRB-CMU-2023). Findings from this study will be disseminated through publications in peer-reviewed journals and stakeholder meetings in Pakistan with the implementing partners and public-sector officials. Findings will also be presented at local and international medical and public health conferences. TRIAL REGISTRATION NUMBER: NCT06017843.

Comparative effectiveness of individualized longer and standardized shorter regimens in the treatment of multidrug resistant tuberculosis in a high burden country.

Objective: To compare the effectiveness of second line injectables containing shorter (duration 9-12 months) and longer treatment regimens (LTR, duration ≥ 20 months) among multidrug-resistant tuberculosis (MDR-TB) patients with no documented resistance and history of treatment with any second-line anti-TB drug (SLD) for ≥ 1 month. Methods: This was an observational cohort study of MDR-TB patients treated at eight PMDT units in Pakistan. Patients' data from baseline until treatment outcomes were collected from Electronic Nominal Recording and Reporting System. The treatment outcomes of "cured" and "treatment completed" were grouped together as successful, whereas "death," "treatment failure," and "lost to follow-up" were collectively grouped as unsuccessful outcomes. Time to sputum culture conversion (SCC) was analyzed using the Kaplan-Meier method and the differences between groups were compared through the log-rank test. Multivariate Cox proportional hazards and binary logistic regression analyses were used to find predictors of time to SCC and unsuccessful treatment outcomes. A p-value < 0.05 was considered statistically significant. Results: A total 701 eligible MDR-TB patients [313 treated with shorter treatment regimen (STR) and 388 treated with LTR at eight centres in Pakistan were evaluated]. Time to achieve SCC was significantly shorter in STR group [mean: 2.03 months, 95% confidence interval (CI):1.79-2.26] than in LTR group (mean: 2.69 months, 95% CI: 2.35-3.03) (p-value<0.001, Log-rank test). Treatment success was higher in STR (83.7%) than in LTR (73.2%) group (p-value <0.001) due to high cure (79.9% vs. 70.9%, p-value = 0.006) and low death (9.9% vs. 18.3%, p-value = 0.002) rates with STR. Treatment with STR emerged the only predictor of early SCC [adjusted Hazards ratio (aHR) = 0.815, p-value = 0.014], whereas, patient's age of 41-60 (OR = 2.62, p-value<0.001) and >60 years (OR = 5.84, p-value<0.001), baseline body weight of 31-60 (OR = 0.36, p-value = 0.001) and >60 kg (OR = 0.23, p-value <0.001), and treatment with LTR (OR = 1.88, p-value = 0.001) had statistically significant association with unsuccessful treatment outcomes. Conclusion: STR exhibited superior anti-microbial activity against MDR-TB. When compared LTR, treatment with STR resulted in significantly early SCC, high cure, and lower death rates among MDR-TB patients who had no documented resistance and history of treatment with any SLD ≥ 1 month.

Factors Affecting the Transition from Paper to Digital Data Collection for Mobile Tuberculosis Active Case Finding in Low Internet Access Settings in Pakistan.

Between September 2020 and March 2021, Mercy Corps piloted hybrid digital (CAPI) and paper-based (PAPI) data collection as part of its tuberculosis (TB) active case finding strategy. Data were collected using CAPI and PAPI at 140 TB chest camps in low Internet access areas of Punjab and Khyber Pakhtunkhwa provinces in Pakistan. PAPI data collection was performed primarily during the camp and entered using a tailor-performed CAPI tool after camps. To assess the feasibility of this hybrid approach, quality of digital records were measured against the paper "gold standard", and user acceptance was evaluated through focus group discussions. Completeness of digital data varied by indicator, van screening team, and month of implementation: chest camp attendees and pulmonary TB cases showed the highest CAPI/PAPI completeness ratios (1.01 and 0.96 respectively), and among them, all forms of TB diagnosis and treatment initiation were lowest (0.63 and 0.64 respectively). Vans entering CAPI data with high levels of completeness generally did so for all indicators, and significant differences in mean indicator completeness rates between PAPI and CAPI were observed between vans. User feedback suggested that although the CAPI tool required practice to gain proficiency, the technology was appreciated and will be better perceived once double entry in CAPI and PAPI can transition to CAPI only. CAPI data collection enables data to be entered in a more timely fashion in low-Internet-access settings, which will enable more rapid, evidence-based program steering. The current system in which double data entry is conducted to ensure data quality is an added burden for staff with many activities. Transitioning to a fully digital data collection system for TB case finding in low-Internet-access settings requires substantial investments in M&E support, shifts in data reporting accountability, and technology to link records of patients who pass through separate data collection stages during chest camp events.

TB Hackathon: Development and Comparison of Five Models to Predict Subnational Tuberculosis Prevalence in Pakistan.

Pakistan's national tuberculosis control programme (NTP) is among the many programmes worldwide that value the importance of subnational tuberculosis (TB) burden estimates to support disease control efforts, but do not have reliable estimates. A hackathon was thus organised to solicit the development and comparison of several models for small area estimation of TB. The TB hackathon was launched in April 2019. Participating teams were requested to produce district-level estimates of bacteriologically positive TB prevalence among adults (over 15 years of age) for 2018. The NTP provided case-based data from their 2010-2011 TB prevalence survey, along with data relating to TB screening, testing and treatment for the period between 2010-2011 and 2018. Five teams submitted district-level TB prevalence estimates, methodological details and programming code. Although the geographical distribution of TB prevalence varied considerably across models, we identified several districts with consistently low notification-to-prevalence ratios. The hackathon highlighted the challenges of generating granular spatiotemporal TB prevalence forecasts based on a cross-sectional prevalence survey data and other data sources. Nevertheless, it provided a range of approaches to subnational disease modelling. The NTP's use and plans for these outputs shows that, limitations notwithstanding, they can be valuable for programme planning.

High rate of successful treatment outcomes among childhood rifampicin/multidrug-resistant tuberculosis in Pakistan: a multicentre retrospective observational analysis.

BACKGROUND: There was a complete lack of information about the treatment outcomes of rifampicin/multidrug resistant (RR/MDR) childhood TB patients (age ≤ 14 years) from Pakistan, an MDR-TB 5th high burden country. Therefore, this study evaluated the socio-demographic characteristics, drug resistance pattern, treatment outcomes and factors associated with unsuccessful outcomes among childhood RR/MDR-TB patients in Pakistan. METHODS: This was a multicentre retrospective record review of all microbiologically confirmed childhood RR/MDR-TB patients (age ≤ 14 years) enrolled for treatment at seven units of programmatic management of drug-resistant TB (PMDT) in Pakistan. The baseline and follow-up information of enrolled participants from treatment initiation until the end of treatment were retrieved from electronic nominal recording and reporting system. World Health Organization (WHO) defined criterion was used for deciding treatment outcomes. The outcomes of "cured" and "treatment completed" were collectively grouped as successful, whereas "death", "treatment failure" and "lost to follow-up" were grouped together as unsuccessful outcomes. Multivariable binary logistic regression analysis was used to find factors associated with unsuccessful outcomes. A p-value < 0.05 reflected statistically significant findings. RESULTS: A total of 213 children RR/MDR-TB (84 RR and 129 MDR-TB) were included in the study. Majority of them were females (74%), belonged to the age group 10-14 years (82.2%) and suffered from pulmonary TB (85.9%). A notable proportion (37.1%) of patients had no history of previous TB treatment. Patients were resistant to a median of two drugs (interquartile range: 1-4) and 23% were resistant to any second line anti-TB drug. A total of 174 (81.7%) patients achieved successful treatment outcomes with 144 (67.6%) patients being cured and 30 (14.1%) declared treatment completed. Among the 39 (18.3%) patients with unsuccessful outcomes, 35 (16.4%) died and 4 (1.9%) experienced treatment failure. In multivariable analysis, the use of ethambutol had statistically significant negative association with unsuccessful outcomes (odds ratio = 0.36, p-value = 0.02). CONCLUSIONS: In this study, the WHO target of successful treatment outcomes (≥ 75%) among childhood RR/MDR-TB patients was achieved. The notable proportion of patients with no history of previous TB treatment (37.1%) and the disproportionately high number of female patients (74%) respectively stress for infection control measures and provision of early and high quality care for female drug susceptible TB patients.

Treatment Outcomes of Extensively Drug-Resistant Tuberculosis in Pakistan: A Countrywide Retrospective Record Review.

Background: The current study is conducted with the aim to the fill the gap of information regarding treatment outcomes and variables associated with unsuccessful outcome among XDR-TB patients from Pakistan. Methods: A total of 404 culture confirmed XDR-TB patients who received treatment between 1st May 2010 and June 30, 2017 at 27 treatment centers all over Pakistan were retrospectively followed until their treatment outcomes were reported. A p-value <0.05 reflected a statistical significant association. Results: The patients had a mean age 32.9 ± 14.1 years. The overall treatment success rate was 40.6% (95% confidence interval [CI]:35.80-45.60%). A total of 155 (38.4%) patients were declared cured, 9 (2.2%) completed treatment, 149 (36.9%) died, 60 (14.9%) failed treatment and 31 (7.7%) were lost to follow up (LTFU). The results of the multivariate binary logistic regression analysis revealed that the patients' age of >60 years (OR = 4.69, 95%CI:1.57-15.57) and receiving high dose isoniazid (OR = 2.36, 95%CI:1.14-4.85) had statistically significant positive association with death, whereas baseline body weight >40 kg (OR = 0.43, 95%CI:0.25-0.73) and sputum culture conversion in the initial two months of treatment (OR = 0.33, 95%CI:0.19-0.58) had statistically significant negative association with death. Moreover, male gender had statistically significant positive association (OR = 1.92, 95%CI:1.04-3.54) with LTFU. Conclusion: The treatment success rate (40.6%) of XDR-TB patients in Pakistan was poor. Providing special attention and enhanced clinical management to patients with identified risk factors for death and LTFU in the current cohort may improve the treatment outcomes.

Effectiveness of Shorter Treatment Regimen in Multidrug-Resistant Tuberculosis Patients in Pakistan: A Multicenter Retrospective Record Review.

In Pakistan, the treatment of multidrug-resistant tuberculosis (MDR-TB) with a shorter treatment regimen (STR), that is, 4-6 months of amikacin, moxifloxacin (Mfx), ethionamide, clofazimine (Cfz), pyrazinamide (Z), ethambutol (E), and high-dose isoniazid, followed by 5 months of Mfx, Cfz, Z, and E, was initiated in 2018. However, there is a lack of information about its effectiveness in Pakistani healthcare settings. Therefore, this retrospective record review of MDR-TB patients treated with STR at eight treatment sites in Pakistan aimed to fill this gap. Data were analyzed using SPSS 23. Multivariate binary logistic regression (MVBLR) analysis was conducted to find factors associated with death and treatment failure, and lost to follow-up (LTFU). A P-value < 0.05 was considered statistically significant. Of 912 MDR-TB patients enrolled at the study sites, only 313 (34.3%) eligible patients were treated with STR and included in the current study. Of them, a total of 250 (79.9%) were cured, 12 (3.8%) completed treated, 31 (9.9%) died, 16 (5.1%) were LTFU, and four (1.3%) were declared as treatment failures. The overall treatment success rate was 83.7%. In MVBLR analysis, patients' age of 41-60 (odds ratio [OR] = 4.9, P-value = 0.020) and > 60 years (OR = 3.6, P-value = 0.035), being underweight (OR = 2.7, P-value = 0.042), and previous TB treatment (OR = 0.4, P-value = 0.042) had statistically significant association with death and treatment failure, whereas patients' age of > 60 years (OR = 5.4, P-value = 0.040) and previous TB treatment (OR = 0.2, P-value = 0.008) had statistically significant association with LTFU. The treatment success rate of STR was encouraging. However, to further improve the treatment outcomes, special attention should be paid to the patients with identified risk factors.

The relationship between seasonality, latitude and tuberculosis notifications in Pakistan.

BACKGROUND: Pakistan ranks amongst the top 20 highest burden tuberculosis (TB) countries in the world. Approximately 369,548 cases of TB (all forms) were notified in 2018, with an estimated incidence of 265 per 100,000 people per year. In other settings, TB has been shown to demonstrate seasonal variation, with higher incidence in the spring/summer months and lower incidence in the autumn/winter; the amplitude of seasonal variation has also been reported to be higher with increasing distance from the equator. METHODS: Notifications of newly-diagnosed pulmonary and extrapulmonary TB cases were obtained for 139 districts in Pakistan from 2011 to 2017. Data were provided by the Pakistan National TB Control Programme, Islamabad, Pakistan. Statistical analyses were performed to determine whether there was seasonal variation in TB notifications in Pakistan; whether the amplitude of seasonal variation in TB notifications varied according to latitude; whether the amplitude of seasonal variation of TB in Pakistan differed between extrapulmonary TB vs. pulmonary TB. To assess the quarterly seasonality of TB, we used the X-13-ARIMA-SEATS seasonal adjustment programme from the United States Census Bureau. The mean difference and corresponding 95% confidence intervals of seasonal amplitudes between different latitudes and clinical phenotype of TB were estimated using linear regression. RESULTS: TB notifications were highest in quarter 2, and lowest in quarter 4. The mean amplitude of seasonal variation was 25.5% (95% CI 25.0 to 25.9%). The mean seasonal amplitude of TB notifications from latitude 24.5°N- < 26.5°N was 29.5% (95% CI 29.3 to 29.7%) whilst the mean seasonal amplitude of TB notifications from latitude 34.5°N - < 36.5°N was 21.7% (95% CI 19.6 to 23.9%). The mean seasonal amplitude of TB notifications across Pakistan between latitudes 24.5°N to 36.5°N reached statistically significant difference (p < 0.001). The amplitude of seasonal variation was greater for extrapulmonary TB (mean seasonal amplitude: 32.6, 95% CI 21.4 to 21.8%) vs. smear positive pulmonary TB mean seasonal amplitude: 21.6, 95% CI 32.1 to 33.1%), p < 0.001. CONCLUSION: TB notifications in Pakistan exhibit seasonal variation with a peak in quarter 2 (April-June) and trough in quarter 4 (October-December). The amplitude of seasonality decreases with increasing latitude, and is more pronounced for extrapulmonary than for pulmonary TB.

Extending contact screening within a 50-m radius of an index tuberculosis patient using Xpert MTB/RIF in urban Pakistan: Did it impact treatment outcomes?

BACKGROUND: Pakistan implemented initiatives to detect tuberculosis (TB) patients through extended contact screening (ECS); it improved case detection but treatment outcomes need assessment. OBJECTIVES: To compare treatment outcomes of pulmonary TB (PTB) patients detected by ECS with those detected by routine passive case finding (PCF). METHODS: A cohort study using secondary program data conducted in Lahore, Faisalabad and Rawalpindi districts and Islamabad in 2013-15. We used log binomial regression models to assess if ECS was associated with unfavorable treatment outcomes (death, loss-to-follow-up, failure, not evaluated) after adjusting for potential confounders. RESULTS: We included 79,431 people with PTB; 4604 (5.8%) were detected by ECS with 4052 (88%) bacteriologically confirmed. In all PTB patients the proportion with unfavorable outcomes was not significantly different in ECS group (9.6%) compared to PCF (9.9%), however, among bacteriologically confirmed patients unfavorable outcomes were significantly lower in ECS (9.9%) than PCF group (11.6%, P = 0.001). ECS was associated with a lower risk of unfavorable outcomes (adjusted relative risk (aRR) 0.90; 95% CI 0.82-0.99) among 'all PTB' patients and bacteriologically confirmed PTB patients (aRR 0.91; 95% CI 0.82-1.00). CONCLUSION: In PTB patients detected by ECS the treatment outcomes were not inferior to those detected by PCF.