Underdiagnosis and overtreatment: provider practices for tuberculosis in India

Tuberculosis (TB) is the second leading infectious disease killer worldwide after COVID-19; accounting for more than a quarter of the global TB burden, India shoulders the greatest TB burden in the world. High-quality care in TB includes early and accurate diagnosis, followed by rapid initiation of the correct drug regimen, patient support, and management of relevant comorbidities. The Indian private healthcare sector is large and unregulated; it treats half of Indian TB patients, with evidence suggesting that the quality of care they receive is suboptimal. Standardized patients (SPs) are increasingly used in low-income countries to assess quality of medical care. SPs presented in one of four case presentations; Case 1 consisted of a classic symptomatic case of presumptive TB who has had 2-3 weeks of cough and fever; Case 2, additionally, presents with a history of completed chest CXR and 1 week of broad-spectrum antibiotic treatment ordered by another provider, with no improvement; Case 3 also carries a positive microbiological test (sputum microscopy); Case 4 presented as an adult multidrug resistant (MDR) TB suspect with 4 weeks of cough and fever. In this manuscript-based thesis, I assess quality of TB care and provider practices across outpatient primary care settings in India’s private sector. I study if providers empirically prescribe anti-TB medications on the basis of clinical and chest X-ray (CXR) findings, and their propensity to prescribe potentially harmful medications or treatment regimes. Among SP cases who presented to providers with typical TB symptoms and an abnormal CXR (Case 2), 182 of 795 (25%; 95% CI: 21–28) of interactions resulted in a ideal correct management, where the provider prescribed a microbiological test (sputum smear, Xpert MTB/RIF or culture) AND did not offer a concurrent prescription for a steroid or antibiotic (more specifically of fluoroquinolone or anti-TB medication-ATT). Of all Case 2 presentations, 210 of 795 (23%; 95% CI: 19–26) resulted in potential empiric TB treatment, with a prescription or dispensation of ATT. 140 of 795 (13%; 95% CI: 10–16) of interactions resulted in a prescriptions/dispensation of ATT AND additionally, a prescription indicated for confirmatory microbiological testing, either via sputum smear microscopy, Gene Xpert MTB/RIF, or culture. Across all case presentations, 749 of 6685 (12%; 95% CI: 11–14) SPs were prescribed fluoroquinolones and 512 of 6685 (6%; 95% CI: 5–7), steroids. When ATT was prescribed, a majority of first-line ATT regimens were a four drug fixed-dose combination – some 236 of 418 (64%; 95% CI: 57–70%). In a repeat cross-sectional study in Patna, I looked at two standardised patient cases presenting at pharmacies: first, a patient presenting with 2-3 weeks of pulmonary tuberculosis symptoms (Case 1); and second, a patient with microbiologically confirmed pulmonary tuberculosis (Case 2). Ideal management for both cases were defined a priori as referral to a health-care provider without dispensing antibiotics or steroids or both. Only some pharmacies correctly managed patients with presumed tuberculosis, but most correctly managed a case of confirmed tuberculosis. Across both rounds of data collection, 331 of 936 (35%; 95% CI: 32–38%) of interactions were correctly managed; at baseline, 215 of 500 (43%; 95% CI: 39–47%) of interactions were correctly managed whereas 116 of 436 (27%; 95% CI: 23–31%) were correctly managed in the second round of data collection. We saw a decrease in correct case management between Case 1 and Case 2 by 20 percentage points from baseline to round 2 in our difference-in-difference analysis. Pharmacies did not dispense anti-tuberculosis drugs for either case or round. Overall, this body of work contributes to fill some knowledge gaps regarding private provider behaviors in urban India, thus helping to design and implement tailored interventions aimed to promote the rational use of medicines and diagnostics