Assessing the interruption of the transmission of human helminths with mass drug administration alone: optimizing the design of cluster randomized trials

Abstract

Background: A method is outlined for the use of an individual-based stochastic model of parasite transmission dynamics to assess different designs for a cluster randomized trial in which mass drug administration (MDA) is employed in attempts to eliminate the transmission of soil-transmitted helminths (STH) in defined geographic locations. The hypothesis to be tested is: Can MDA alone interrupt the transmission of STH species in defined settings? Clustering is at a village level and the choice of clusters of villages is stratified by transmission intensity (low, medium and high) and parasite species mix (either Ascaris, Trichuris or hookworm dominant).

Results: The methodological approach first uses an age-structured deterministic model to predict the MDA coverage required for treating pre-school aged children (Pre-SAC), school aged children (SAC) and adults (Adults) to eliminate transmission (crossing the breakpoint in transmission created by sexual mating in dioecious helminths) with 3 rounds of annual MDA. Stochastic individual-based models are then used to calculate the positive and negative predictive values (PPV and NPV, respectively, for observing elimination or the bounce back of infection) for a defined prevalence of infection 2 years post the cessation of MDA. For the arm only involving the treatment of Pre-SAC and SAC, the failure rate is predicted to be very high (particularly for hookworm-infected villages) unless transmission intensity is very low (R₀, or the effective reproductive number R, just above unity in value).

Conclusions: The calculations are designed to consider various trial arms and stratifications; namely, community-based treatment and Pre-SAC and SAC only treatment (the two arms of the trial), different STH transmission settings of low, medium and high, and different STH species mixes. Results are considered in the light of the complications introduced by the choice of statistic to define success or failure, varying adherence to treatment, migration and parameter uncertainty.

Keywords: Cluster randomized trial design; Interrupting transmission; Mass drug administration impact; Soil-transmitted helminths; Stochastic models of transmission.

Fig. 1

True demography and sampled demography. Demography of India 2015 (source http://www.census.gov/population/international/data/idb/informationGateway.php) (blue bars are the full population; orange bars are the sampled population)

Fig. 2

qPRC compared with Kato Katz. qPCR diagnostics test results for a Ascaris lumbricoides and b Necator americanus compared with eggs per gram of faeces determined by Kato Katz (from [17])

Fig. 3

Flow chart summary of the steps in the suggested procedures and the epidemiological calculations required to arrive at the initial estimates of MDA coverage in each age grouping prior to performing the stochastic simulations of the trial

Fig. 4

Stochastic simulations. Two stochastic simulations of three annual rounds of chemotherapy to control Ascaris (R₀ = 2.2; k = 0.9, L = 1 year, ɣ = 0.07) (Infants 0% coverage, 90% coverage of Pre-SAC and SAC, 80% coverage of Adults) showing elimination and bounce back. The mean proportion infected (from 300 replicate runs) are shown for those simulations that result in elimination and those that result in bounce back (parameters as in Table 3)

Fig. 5

Prevalence of Ascaris after MDA. Illustrative example of the prevalence of Ascaris infection in a village at year 5 after three rounds of annual treatment measured followed by 2 years of no treatment. The graph records the frequency distribution of the proportion of a population of a village infected in villages in which transmission is eliminated (blue bars) and those in which bounce back occurs (green bars). In Figs. 6, 7, 8 and 9 all probabilities are conditional with the blue bars and green bars each summing to 1

Fig. 6

Ascaris prevalence distributions 1. Predicted distribution of Ascaris prevalences after three annual rounds of treatment and 2 years after the end of MDA for villages in which elimination occurred and villages in which bounce back resulted. The insert text box gives the PPV and NPV for a prevalence of 30%. These simulations were conducted for 300 villages each with a population of 500 people with MDA coverage of 0% infants - 90% pre-SAC - 90% SAC - 80% adults. Parameter values as defined in Table 3. The predicted percentage elimination in the replicates is 88%

Fig. 7

Ascaris prevalence distributions 2. Identical to Fig. 6 but with the observation period after the end of MDA set at 3 months and the prevalence threshold set at 30%

Fig. 8

Ascaris prevalence distributions 3. Identical to Fig. 7 but with replication in 100 villages with 500 people per village. The percentage elimination is 92%

Fig. 9

Ascaris prevalence distributions 4. Identical to Fig. 7 but with replication in 300 villages and 250 people per village. The percentage elimination is 82.0%

Fig. 10

Effects of immigration. Immigration of infected people between village movements: basic data on demography and parasite distributions per person (k = 0.65, from [38])

Fig. 11

Adherence to treatment. The impact of various assumptions on adherence to drug treatment on the deterministic predictions the number of years of MDA at coverage levels of Pre-SAC and SAC at 90% and Adults at 80% before the breakpoint in transmission is crossed [39]

Fig. 12

Effects of immigration. Immigration of people 3 months post end of 3 years of MDA - from outside the treated village - who are drawn at random from a worm distribution per person and human age distribution data identical to that pertaining in the village prior to treatment

Fig. 13

Parameter sensitivity. Some simulation experiments showing the sensitivity of the two distributions of elimination and bounce back 2 years after cessation of three annual rounds of chemotherapy to control Ascaris (MDA coverage as 0% infants - 90% pre-SAC - 90% SAC - 80% adults) with different parameters sets. Note several parameters vary simultaneously since the sets are taken from the Monte Carlo Markov Chain fitting of the model to age intensity data. Three different set of parameters chosen from one chain of an MCMC run are recorded. A stochastic model is used to convert epg counts for Ascaris to worm counts with density dependence in egg production built in

Fig. 14

Flow chart summary of the steps (following on from Fig. 3) in the suggested procedures and the calculations required to arrive at the PPV and NPV values using the stochastic trial simulation model