Articles
Use of standardised patients to assess gender differences in quality of tuberculosis care in urban India: a two-city, cross-sectional study

https://doi.org/10.1016/S2214-109X(19)30031-2Get rights and content
Under a Creative Commons license
open access

Summary

Background

In India, men are more likely than women to have active tuberculosis but are less likely to be diagnosed and notified to national tuberculosis programmes. We used data from standardised patient visits to assess whether these gender differences occur because of provider practice.

Methods

We sent standardised patients (people recruited from local populations and trained to portray a scripted medical condition to health-care providers) to present four tuberculosis case scenarios to private health-care providers in the cities of Mumbai and Patna. Sampling and weighting allowed for city representative interpretation. Because standardised patients were assigned to providers by a field team blinded to this study, we did balance and placebo regression tests to confirm standardised patients were assigned by gender as good as randomly. Then, by use of linear and logistic regression, we assessed correct case management, our primary outcome, and other dimensions of care by standardised patient gender.

Findings

Between Nov 21, 2014, and Aug 21, 2015, 2602 clinical interactions at 1203 private facilities were completed by 24 standardised patients (16 men, eight women). We found standardised patients were assigned to providers as good as randomly. We found no differences in correct management by patient gender (odds ratio 1·05; 95% CI 0·76–1·45; p=0·77) and no differences across gender within any case scenario, setting, provider gender, or provider qualification.

Interpretation

Systematic differences in quality of care are unlikely to be a cause of the observed under-representation of men in tuberculosis notifications in the private sector in urban India.

Funding

Grand Challenges Canada, Bill & Melinda Gates Foundation, World Bank Knowledge for Change Program.

Cited by (0)

*

Contributed equally

Contributed equally