Projecting the impact of variable MDR-TB transmission efficiency on long-term epidemic trends in South Africa and Vietnam

Abstract

Whether multidrug-resistant tuberculosis (MDR-TB) is less transmissible than drug-susceptible (DS-)TB on a population level is uncertain. Even in the absence of a genetic fitness cost, the transmission potential of individuals with MDR-TB may vary by infectiousness, frequency of contact, or duration of disease. We used a compartmental model to project the progression of MDR-TB epidemics in South Africa and Vietnam under alternative assumptions about the relative transmission efficiency of MDR-TB. Specifically, we considered three scenarios: consistently lower transmission efficiency for MDR-TB than for DS-TB; equal transmission efficiency; and an initial deficit in the transmission efficiency of MDR-TB that closes over time. We calibrated these scenarios with data from drug resistance surveys and projected epidemic trends to 2040. The incidence of MDR-TB was projected to expand in most scenarios, but the degree of expansion depended greatly on the future transmission efficiency of MDR-TB. For example, by 2040, we projected absolute MDR-TB incidence to account for 5% (IQR: 4-9%) of incident TB in South Africa and 14% (IQR: 9-26%) in Vietnam assuming consistently lower MDR-TB transmission efficiency, versus 15% (IQR: 8-27%)and 41% (IQR: 23-62%), respectively, assuming shrinking transmission efficiency deficits. Given future uncertainty, specific responses to halt MDR-TB transmission should be prioritized.

Figure 1

Model Structure. States of TB infection and possible transitions between them are represented in panel A. Compartment colors correspond to the respective infectiousness and TB-associated mortality of each state. In addition to TB natural history, populations are also classified by treatment history (treatment-naïve or previously-treated, not shown), and HIV status (represented in panel B). Following HIV infection, populations transition through states of increasing immunosuppression or to antiretroviral therapy (ART) at defined rates. Populations returning to an uninfected state through self-cure remain classified as treatment-naïve; other distinctions between uninfected compartments are shown only for illustrative purposes.

Figure 2

Transmission Efficiency Scenarios. The assumed transmission efficiency (transmission events per 1,000 infectious person-years) of DS-TB over time is drawn in green; the downward slope recapitulates reductions in TB transmission efficiency due to secular trends unrelated to MDR-TB diagnosis and treatment (for example, reductions in crowding, improved socioeconomic conditions, etc.). In our three model scenarios, we assume either that the transmission efficiency of MDR-TB is at a perpetual deficit compared to that of DS-TB (Constant Deficit Scenario, drawn in orange); that the transmission efficiency of MDR-TB is consistently the same as that of DS-TB (No Deficit Scenario, drawn in magenta); or that MDR-TB has lower transmission efficiency than DS-TB initially but gradually converges towards that of DS-TB over time (Shrinking Deficit Scenario, drawn in red). Years are shown for illustrative purposes; dates of MDR-TB emergence and rates of increase/decrease in transmission efficiency are sampled from defined ranges; see Sampling & Calibration for further details.

Figure 3

Projections of MDR-TB Incidence in South Africa and Vietnam. Simulated MDR-TB epidemics in South Africa and Vietnam were projected from 2010 to 2040. Panels A and B illustrate the projections of each scenario in South Africa, while panels C and D illustrate the projections of each scenario in Vietnam. The 2040 projected median (IQR) values are included in the upper right of each panel. IQR represents 25th to 75th percentiles and the 90% range represents the 5th to 95th percentiles of posterior simulations.

Figure 4

Calibration Performance for South Africa. Simulated epidemics are weighted according to how well each reproduced empirical calibration targets (historical estimates of MDR in new and previously-treated TB cases). Recent diagnoses are defined as any populations transitioning from active, untreated TB into any diagnosis/treatment state. Red points represent median and 95% confidence intervals for calibration targets drawn from national survey data. IQR represents 25th to 75th percentiles and the 90% range represents the 5th to 95th percentiles of posterior simulations.

Figure 5

Calibration Performance for Vietnam. Simulated epidemics are weighted according to how well each reproduced empirical calibration targets (historical estimates of MDR in new and previously-treated TB cases). Recent TB diagnoses were used (instead of incident TB cases) to better represent the sampling methodologies used in national drug resistant surveys which were used for calibration. (The 1996 prevalence survey in Vietnam measured the proportion MDR in new cases only.) Recent diagnoses are defined as any populations transitioning from active, untreated TB into any diagnosis/treatment state. Red points represent median and 95% confidence intervals for calibration targets drawn from national survey data. IQR represents 25th to 75th percentiles and the 90% range represents the 5th to 95th percentiles of posterior simulations.

Figure 6

Replication of Previous Findings. The calibration results of an alternative (“Delayed Treatment”) scenario in South Africa – in which we assumed a median delay of 10 years prior to treatment initiation –are represented in panel A. Points represent median and 95% confidence intervals for calibration targets drawn from national survey data. For this scenario only, calibration excluded data on previously-treated TB patients, consistent with methods used in the replicated publication. Simulated MDR-TB epidemics in South Africa under the Delayed Treatment scenario projected to 2040 are represented in panel B. IQR represents 25th to 75th percentiles and the 90% range represents the 5th to 95th percentiles of posterior simulations.

Figure 7

Sensitivity Analysis – Influence of Key Model Parameters on Projections of MDR-TB Incidence in South Africa. The top 5 parameters which most strongly impact the distributions of MDR-TB incidence in 2040 in projections of the epidemic in South Africa are displayed. Each boxplot represents the distribution of values for the primary outcome (the incidence of MDR-TB in 2040) within a given set of simulations. Pairs of boxplots represent groups of simulations categorized by values of a single input parameter: red boxplots represent the outcomes of those simulations with parameter values in the upper 20% of all simulations; blue boxplots represent the outcomes of those simulations with parameter values in the lower 20% of all simulations. More influential parameters demonstrate a greater separation of the distributions of outcome between simulations in the upper quintile and simulations in the lower quintile of parameter values. To the left of each panel are included the input parameter values corresponding to the accompanying quintile. In black is represented the overall distribution of the outcome across all simulations and the median estimate is drawn as a vertical dotted line. Boxes represent the median, 25th, and 75th percentiles of the distribution of outcomes; whiskers represent the 5th and 95th percentiles of the distribution of outcomes. In the Constant Deficit model, parameters involving the reduction in MDR-TB transmission efficiency deficit are excluded by definition.