Table 3 Summary of modelling techniques used, IDM diseases
Paper | Model | Fitted to data from: | Predictions tested? | Technical advances | Model accounts for | Next steps | ||
|---|---|---|---|---|---|---|---|---|
Vector/environment dynamics | Heterogeneity in risk | Access to interventions | ||||||
Chagas disease | ||||||||
Peterson et al. | Deterministic | Parameter values were set according to the literature | No | Formulating a transmission model and analysing the consequences of varying standard assumptions on the transmission cycle | Deterministic vector dynamics with animal hosts in some modelling scenarios | None | Not applicable - vector control only | Develop two independent transmission models. Estimation of changes in transmission rates |
Human African trypanosomiasis, Gambian form | ||||||||
Pandey et al. | Deterministic | Boffa, Guinea | Yes | Data cannot identify whether there is an animal reservoir. But in the presence of animal reservoir, there is high risk of re-emergence of HAT as public health problem. | Includes tsetse and animal compartments | None | All individuals have a probability of receiving treatment | Evaluating 2020 goal in other foci and impact of heterogeneity in human exposure to tsetse. |
Rock et al. | Deterministic | Bandundu, DRC | No | Data supports the existence of an unscreened, high-risk population, but cannot identify whether there is an animal reservoir | Includes tsetse and animal compartments | High risk and low risk human compartments | Randomly participating and non-participating human compartments | Projecting impact of vector control in DRC |
Leprosy | ||||||||
Blok et al. | Stochastic individual | India, Brazil and Indonesia | Yes | Applied SIMCOLEP to predict future leprosy incidence in India, Brazil and Indonesia | Not applicable | Susceptibility: 20Â % of population is susceptible; Type of leprosy: MB vs PB; Contact structure: general population vs within households | All individuals that have been diagnosed with leprosy receive MDT treatment. Probability of being diagnosed is determined by passive case detection delays and possible active case finding activities. | Assess which additional interventions are needed to meet the goals |
Brook et al. | Statistical | 604 analytic districts in India | No | Enhanced active case finding was associated with a higher case detection rate | Not applicable | Not applicable | Not applicable | Develop independent stochastic compartmental transmission model |
Crump & Medley | Statistical | Thailand | Yes | Back-calculation can estimate the number of undiagnosed cases from diagnosed incidence rates | Not applicable | Not applicable | Not applicable | Consideration of gender and age. Analysis of other countries. |
Visceral leishmaniasis in the Indian sub-continent | ||||||||
Chapman et al. | Statistical | Bangladesh | No | Estimating durations of asymptomatic and symptomatic infection | Not applicable | Proportional hazards model for different risk factors including age, sex and bed net use | Not applicable | Developing a transmission model. |
Le Rutte et al. | Deterministic | India and Nepal (KalaNet) | Yes | Developed three model structures, each with a different reservoir of infection, all fitting the data. | Vector population, deterministic. | Age-dependent sandfly exposure. | All individuals have a probability of receiving diagnosis, treatment, and vector control (IRS). | Implement best model structure in stochastic individual based model. Explore effect of additional interventions. |
Added heterogeneity in sandfly exposure. | ||||||||
Applied models to predict future VL incidence with current interventions. | ||||||||