Skip to main content
Figure 4 | BMC Public Health

Figure 4

From: Problem drinking as a risk factor for tuberculosis: a propensity score matched analysis of a national survey

Figure 4

Sensitivity of the estimated prevalence odds ratio of TB to unmeasured confounders. Estimated percent reduction of the prevalence odds ratio of TB in problem drinkers vs. moderate drinkers/abstainers when hypothetical, randomly generated, binary confounders are introduced in the models. Axes represent the adjusted odds ratio of the association — irrespective of the direction — of the hypothetical confounder with problem drinking (OR Problem ) and TB (OR TB ). Colours represent couples of values for OR Problem and OR TB producing the same reduction in the prevalence odds ratio. (N = 1000 simulated confounders. Contour plot, Shepard interpolation [27]).

Back to article page