Engagement of private healthcare providers for case finding of Tuberculosis and Diabetes Mellitus in Pakistan

Abstract Background: The rising co-epidemic of tuberculosis (TB) and diabetes mellitus (DM) is a challenge for constrained health systems in low and middle-income countries. Diabetes is a known risk factor for tuberculosis and associated with poor tuberculosis treatment outcomes, while tuberculosis is associated with worsening glycemic control. We investigated the performance of bi-directional TB and DM case finding approaches through a private-sector engagement model in Karachi, Pakistan. Methods: Between July 2016 and July 2018, private health care providers were engaged to generate referrals for bi-directional TB and DM screening at private diagnostic and treatment centers in Karachi, Pakistan. Individuals diagnosed with TB underwent glycated hemoglobin (HbA1c) testing at the time of anti-tuberculous treatment initiation and at three -month follow up stage. All individuals with a history of diabetes or random blood sugar of greater than 200 mg/dl were screened for TB using a chest X-ray and Xpert MTB/RIF. Results: A total of 6,312 persons with tuberculosis were tested on HbA1c at treatment initiation, of whom 1,516 (24%) were newly diagnosed with DM. About one third of those with HbA1c in the diabetic range (≥ 6.5%) at baseline were found to have a normal HbA1c (200mg/dl, underwent chest x-ray and Xpert MTB/RIF testing, with 321 (13.4%) known and 54 (3.8%) new diabetics respectively identified with tuberculosis. Conclusion: This study demonstrates a high yield of TB and DM through bidirectional screening and the feasibility of engagement of private sector in finding missing cases of tuberculosis and diabetes. Given the high prevalence of undiagnosed DM in individuals with TB tuberculosis patients, there is a need to scale-up DM screening within TB programmes. Increased awareness of the high risk of TB among individuals with DM is needed among private health providers and screening for TB among diabetics should be strongly considered.