Study setting and design
A retrospective cross sectional study in which, bidirectional screening for TB and DM was offered to clients attending private TB diagnostic and treatment centers “Sehatmand Zindagi (SZ) (healthy life) centers and community screening camps in Karachi, Pakistan between July 2016 and July 2018. Karachi is Pakistan’s most populous city and the country’s economic hub with an estimated 75% of all health services availed in the private sector (10). Pakistan has the fourth highest burden of DM globally with a current estimated prevalence of 17.1% in the adult population. An estimated 8.5 million adults are living with undiagnosed DM (2). Pakistan is also ranked fifth among high tuberculosis burden countries with an estimated 36% case notification gap (3).
The SZ centers, located in low middle income neighborhoods of Pakistan operate as a social business, providing free TB diagnostics, chest X-ray and Xpert MTB/RIF, and treatment. Revenue is generated through other laboratory tests and radiology services. All TB cases are notified to the National TB Control Program (NTP). Each center has established linkages with other health providers in the vicinity including those in the informal health sector. The center has a dedicated team who engage private health providers and facilitate referral of individuals with presumptive TB to the centers. Clients are also able to self-refer to centers. Community based screening camps are conducted by trained community health workers employing mobile X- ray vans.
Recruitment of individuals with TB for DM testing
Individuals presenting at SZ Centers and screening camps with a previous history of TB or with signs and symptoms of TB were referred to the TB arm of screening where they were tested for DM if TB diagnosis was established. TB screening was conducted using digital chest X- rays with CAD4TB 5 (version 4.12.0) software for automated scoring Individuals with presumptive TB were defined as those with a threshold CAD4TB score of 70, who then submitted a sputum sample for Xpert MTB/RIF testing. A positive Xpert MTB/RIF testing result or a strong indication at clinical evaluation of the CXR and symptoms lead to the diagnosis of bacteriologically positive or clinical TB respectively. All individuals identified with TB underwent HbA1c testing at anti-tuberculous treatment (ATT) initiation. Those who gave consent also had a 3-month HbA1c follow up. According to the recommendation of 2009 American Diabetes Association (ADA), an HbA1c < 5.7% is classified as normal, 5.7-6.4% as pre-diabetes, and ≥ 6.5% as diabetes (11). Individuals who were diagnosed with diabetes at the SZ Center were counselled and referred to their general practitioner for further management.
Recruitment of individuals with DM for TB testing
Individuals attending the community camps and SZ Centers were verbally screened for history of diabetes. Those identified with DM or those presenting with signs and symptoms of DM were referred to the DM arm of screening where they were subsequently tested for TB if DM was diagnosed. If no known history of DM was found, they were offered a point of care glucose test. Individuals with an RBS>200 mg/dl or history of diabetes, were tested on chest X-ray and Xpert MTB/RIF.
Data management and analysis:
We analyzed retrospective data of 10,136 Individuals who participated in the TB-DM bidirectional screening project funded by the World Diabetes Foundation. Under this project, data was recorded using a custom-built mobile-phone application at community camps, A customized web-based laboratory management system (LMS) software was used to book tests and enter screening data at the centers. Both the mobile and web-based applications were integrated with the Central Management Information System with auto generated reports to track key project metrics. The data recording and reporting systems included several data validation checks to ensure data-accuracy. Field supervisors and project management staff were responsible for overall data-validation and accuracy including reporting to the NTP. Baseline characteristics of participants were described using means and medians and the prevalence of DM and TB were calculated. The association of outcomes (DM and TB) with a prior defined factor was explored using logistic regression. A comparative frequency analysis was conducted for HbA1c values at baseline and after 3 months of ATT for a subset of TB patients. All data was analyzed using Stata version 13.0 (StataCorp, Texas, USA).
Ethical Approval and consent:
An ethical approval was deemed unnecessary for this study by The Institutional Review Board (IRB) at Interactive Research and Development (IRD) under the IRB exemption category 7 under 45 CFR 46.101(b). The IRB is registered with the U.S. Department of Health and Human Services (DHHS), Office for Human Research Protections (IRB#00005148). A Verbal consent was obtained from participants before conducting blood glucose, HbA1c and Xpert MTB/RIF tests. This study was part of a larger study (IRB approval number IRD_IRB_2016_08_001). De-identified data from the project was used for data analysis.