Determining post-treatment surveillance criteria for predicting the elimination of Schistosoma mansoni transmission

• Published in: Parasites & vectors 2019-09, Vol.12 (1), p.437-437, Article 437
• Main Authors: Toor, Jaspreet, Truscott, James E, Werkman, Marleen, Turner, Hugo C, Phillips, Anna E, King, Charles H, Medley, Graham F, Anderson, Roy M
• Format: Article
• Language: English
• Subjects: Adolescent; Adult; Adults; Age; Aged; Aged, 80 and over; Analysis; Animals; Anthelmintics - therapeutic use; Care and treatment; Child; Child, Preschool; Communities; Computer simulation; Control; Diagnostic systems; Disease Eradication; Disease transmission; Disease Transmission, Infectious; Elimination of transmission; Epidemiological Monitoring; Female; Humans; Infant; Infant, Newborn; Infection; Infections; Male; Mass Drug Administration; Middle Aged; Models, Theoretical; Organizations; Parasites; Partial differential equations; Positive predictive value; Post-treatment surveillance; Praziquantel; Prevalence; Prevalence studies (Epidemiology); Prevalence threshold; Probability theory; Public health; Risk factors; Schistosoma mansoni; Schistosoma mansoni - isolation & purification; Schistosomiasis; Schistosomiasis mansoni - drug therapy; Schistosomiasis mansoni - epidemiology; Schistosomiasis mansoni - transmission; Stochastic individual-based model; Stochasticity; Surveillance; Thresholds; Treatment Outcome; Tropical diseases; Young Adult
• ISSN: 1756-3305; 1756-3305
• DOI: 10.1186/s13071-019-3611-8

The World Health Organization (WHO) has set elimination (interruption of transmission) as an end goal for schistosomiasis. However, there is currently little guidance on the monitoring and evaluation strategy required once very low prevalence levels have been reached to determine whether elimination or resurgence of the disease will occur after stopping mass drug administration (MDA) treatment. We employ a stochastic individual-based model of Schistosoma mansoni transmission and MDA impact to determine a prevalence threshold, i.e. prevalence of infection, which can be used to determine whether elimination or resurgence will occur after stopping treatment with a given probability. Simulations are run for treatment programmes with varying probabilities of achieving elimination and for settings where adults harbour low to high burdens of infection. Prevalence is measured based on using a single Kato-Katz on two samples per individual. We calculate positive predictive values (PPV) using PPV ≥ 0.9 as a reliable measure corresponding to ≥ 90% certainty of elimination. We analyse when post-treatment surveillance should be carried out to predict elimination. We also determine the number of individuals across a single community (of 500-1000 individuals) that should be sampled to predict elimination. We find that a prevalence threshold of 1% by single Kato-Katz on two samples per individual is optimal for predicting elimination at two years (or later) after the last round of MDA using a sample size of 200 individuals across the entire community (from all ages). This holds regardless of whether the adults have a low or high burden of infection relative to school-aged children. Using a prevalence threshold of 0.5% is sufficient for surveillance six months after the last round of MDA. However, as such a low prevalence can be difficult to measure in the field using Kato-Katz, we recommend using 1% two years after the last round of MDA. Higher prevalence thresholds of 2% or 5% can be used but require waiting over four years for post-treatment surveillance. Although, for treatment programmes where elimination is highly likely, these higher thresholds could be used sooner. Additionally, switching to more sensitive diagnostic techniques, will allow for a higher prevalence threshold to be employed.