Cost-effectiveness modelling to optimise active screening strategy for gambiense human African trypanosomiasis in endemic areas of the Democratic Republic of Congo

Abstract

Background: Gambiense human African trypanosomiasis (gHAT) has been brought under control recently with village-based active screening playing a major role in case reduction. In the approach to elimination, we investigate how to optimise active screening in villages in the Democratic Republic of Congo, such that the expenses of screening programmes can be efficiently allocated whilst continuing to avert morbidity and mortality.

Methods: We implement a cost-effectiveness analysis using a stochastic gHAT infection model for a range of active screening strategies and, in conjunction with a cost model, we calculate the net monetary benefit (NMB) of each strategy. We focus on the high-endemicity health zone of Kwamouth in the Democratic Republic of Congo.

Results: High-coverage active screening strategies, occurring approximately annually, attain the highest NMB. For realistic screening at 55% coverage, annual screening is cost-effective at very low willingness-to-pay thresholds (<DOLLAR/>20.4 per disability adjusted life year (DALY) averted), only marginally higher than biennial screening (<DOLLAR/>14.6 per DALY averted). We find that, for strategies stopping after 1, 2 or 3 years of zero case reporting, the expected cost-benefits are very similar.

Conclusions: We highlight the current recommended strategy-annual screening with three years of zero case reporting before stopping active screening-is likely cost-effective, in addition to providing valuable information on whether transmission has been interrupted.

Keywords: African sleeping sickness; African trypanosomiasis; Cost-effectiveness; Mathematical model.

Fig. 1

The cost of an active screening strategy. The cost of active screening for a coverage of 55%, a screening interval of 1 year, stopping active screening after 3 screenings when no cases are detected and stopping reactive screening after 1 screening with no cases, under the assumption of WTP_c=0.5. Mean values for all quantities are taken from one million stochastic simulations. a The number of infected people dramatically decreases with time for this coverage (total shaded blue area, with left axis) with the majority of these infections being in the high-risk group (darker blue fraction). The proportion of tsetse that are also infective is reduced with time (green line, with right axis). b The total change in costs of implementing a particular screening strategy (left axis) and the number of DALYs averted from the baseline of only passive surveillance (right axis). c The contribution to the cost from each component of the cost function for years 1, 5, 10, 20 and 30 after starting an active screening programme. Full costs are given in the table in the bottom row of the table. A population size of N_H=1,000 is used. All costs are denominated in 2018 US dollars

Fig. 2

The mean NMB of different active screening strategies. The mean NMB of different active screening strategies for given WTP per DALY averted (given as multiplication factor WTP_c of GDP per capita from the DRC) and the proportion of passive infections that are treated, p_t. The red areas show a negative NMB, whilst blue areas are positive NMB, with white at the boundary of no change. The maximum NMB for each WTP_c and p_t combination is marked by a yellow circle on each heatmap, with a cross if the maximum is for no active screening (only observed here for WTP_c=0 and p_t=27%). A population size of N_H=1,000 is used and we fix z_a=3, the number of consecutive active screening rounds with zero-detections necessary to cease operations

Fig. 3

Theoretical optimum strategy. Theoretical optimum strategy for the mean simulation of infection dynamics given a range of WTP values (horizontal axis). a–c examine the impact of different treatment coverage (p_t) on a the optimal screening coverage, b the optimal screening interval and c the optimal number of zero-detections required to stop screening, to achieve the highest NMB for given WTP. These results assume a village population of 1000 where the disease in endemic. d–f examine the impact of different assumptions about population size and endemicity on d the optimal screening coverage, e the optimal screening interval and f the optimal number of active zero-detections required to stop screening. The demography and disease endemicity assumptions are as follows: a population of population 1,000 where the disease is endemic (‘1000 (E)’), a village of population 1000 where only one person is initially infected (single infection, or ‘1000 (S)’), a village of population 250 where disease is endemic (‘250 (E)’), and a village of population 1000 where disease is endemic and there exists a small probability of infectious importations (‘1000 (I)’). We fix z_r=1 for all simulations

Fig. 4

Optimal strategy given a maximum screening coverage. Optimal strategy given a maximum screening coverage informed by historic averages in the DRC. Results are shown for a WTP_c=0.2,0.5,3. z_r=1, p_t=27%, N_H=1,000 are fixed and the optimum t and z_a is found simultaneously. a Optimal screening interval t. b Optimal number of zero-detections to stop screening z_a

Fig. 5

The cost-effectiveness of active screening strategies. a Cost-effectiveness plane showing the total cost of a strategy and the associated total number of DALYs. Mean values for each strategy are shown by the coloured crosses. b Cost-effectiveness acceptability curves (CEACs) for each strategy are shown by lines, with the cost-effectiveness acceptability frontier (CEAF) shown by the numbered background colour, which demonstrated the values for the ICER. WTP is shown in 2018 USD on the top and as the WTP_c coefficient on the bottom, where the coefficient is the multiplier of the GDP per capita of the DRC. See Table 3 for the full descriptions of the strategies