Estimating the contribution of HIV-infected adults to household pneumococcal transmission in South Africa, 2016-2018: A hidden Markov modelling study.

Human immunodeficiency virus (HIV) infected adults are at a higher risk of pneumococcal colonisation and disease, even while receiving antiretroviral therapy (ART). To help evaluate potential indirect effects of vaccination of HIV-infected adults, we assessed whether HIV-infected adults disproportionately contribute to household transmission of pneumococci. We constructed a hidden Markov model to capture the dynamics of pneumococcal carriage acquisition and clearance observed during a longitudinal household-based nasopharyngeal swabbing study, while accounting for sample misclassifications. Households were followed-up twice weekly for approximately 10 months each year during a three-year study period for nasopharyngeal carriage detection via real-time PCR. We estimated the effect of participant's age, HIV status, presence of a HIV-infected adult within the household and other covariates on pneumococcal acquisition and clearance probabilities. Of 1,684 individuals enrolled, 279 (16.6%) were younger children (<5 years-old) of whom 4 (1.5%) were HIV-infected and 726 (43.1%) were adults (≥18 years-old) of whom 214 (30.4%) were HIV-infected, most (173, 81.2%) with high CD4+ count. The observed range of pneumococcal carriage prevalence across visits was substantially higher in younger children (56.9-80.5%) than older children (5-17 years-old) (31.7-50.0%) or adults (11.5-23.5%). We estimate that 14.4% (95% Confidence Interval [CI]: 13.7-15.0) of pneumococcal-negative swabs were false negatives. Daily carriage acquisition probabilities among HIV-uninfected younger children were similar in households with and without HIV-infected adults (hazard ratio: 0.95, 95%CI: 0.91-1.01). Longer average carriage duration (11.4 days, 95%CI: 10.2-12.8 vs 6.0 days, 95%CI: 5.6-6.3) and higher median carriage density (622 genome equivalents per millilitre, 95%CI: 507-714 vs 389, 95%CI: 311.1-435.5) were estimated in HIV-infected vs HIV-uninfected adults. The use of ART and antibiotics substantially reduced carriage duration in all age groups, and acquisition rates increased with household size. Although South African HIV-infected adults on ART have longer carriage duration and density than their HIV-uninfected counterparts, they show similar patterns of pneumococcal acquisition and onward transmission.

Trachoma Prevalence After Discontinuation of Mass Azithromycin Distribution.

BACKGROUND: As the World Health Organization seeks to eliminate trachoma by 2020, countries are beginning to control the transmission of trachomatous inflammation-follicular (TF) and discontinue mass drug administration (MDA) with oral azithromycin. We evaluated the effect of MDA discontinuation on TF1-9 prevalence at the district level. METHODS: We extracted from the available data districts with an impact survey at the end of their program cycle that initiated discontinuation of MDA (TF1-9 prevalence <5%), followed by a surveillance survey conducted to determine whether TF1-9 prevalence remained below the 5% threshold, warranting discontinuation of MDA. Two independent analyses were performed, 1 regression based and 1 simulation based, that assessed the change in TF1-9 from the impact survey to the surveillance survey. RESULTS: Of the 220 districts included, TF1-9 prevalence increased to >5% from impact to surveillance survey in 9% of districts. Regression analysis indicated that impact survey TF1-9 prevalence was a significant predictor of surveillance survey TF1-9 prevalence. The proportion of simulations with >5% TF1-9 prevalence in the surveillance survey was 2%, assuming the survey was conducted 4 years after MDA. CONCLUSION: An increase in TF1-9 prevalence may represent disease resurgence but could also be due to measurement error. Improved diagnostic tests are crucial to elimination of TF1-9 as a public health problem.

Estimating the contribution of different age strata to vaccine serotype pneumococcal transmission in the pre vaccine era: a modelling study.

BACKGROUND: Herd protection through interruption of transmission has contributed greatly to the impact of pneumococcal conjugate vaccines (PCVs) and may enable the use of cost-saving reduced dose schedules. To aid PCV age targeting to achieve herd protection, we estimated which population age groups contribute most to vaccine serotype (VT) pneumococcal transmission. METHODS: We used transmission dynamic models to mirror pre-PCV epidemiology in England and Wales, Finland, Kilifi in Kenya and Nha Trang in Vietnam where data on carriage prevalence in infants, pre-school and school-aged children and adults as well as social contact patterns was available. We used Markov Chain Monte Carlo methods to fit the models and then extracted the per capita and population-based contribution of different age groups to VT transmission. RESULTS: We estimated that in all settings, < 1-year-old infants cause very frequent secondary vaccine type pneumococcal infections per capita. However, 1-5-year-old children have the much higher contribution to the force of infection at 51% (28, 73), 40% (27, 59), 37% (28, 48) and 67% (41, 86) of the total infection pressure in E&W, Finland, Kilifi and Nha Trang, respectively. Unlike the other settings, school-aged children in Kilifi were the dominant source for VT infections with 42% (29, 54) of all infections caused. Similarly, we estimated that the main source of VT infections in infants are pre-school children and that in Kilifi 39% (28, 51) of VT infant infections stem from school-aged children whereas this was below 15% in the other settings. CONCLUSION: Vaccine protection of pre-school children is key for PCV herd immunity. However, in high transmission settings, school-aged children may substantially contribute to transmission and likely have waned much of their PCV protection under currently recommended schedules.

Models of Trachoma Transmission and Their Policy Implications: From Control to Elimination.

Despite great progress in eliminating trachoma from the majority of worldwide districts, trachoma control seems to have stalled in some endemic districts. Can mathematical models help suggest the way forward? We review specific achievements of models in trachoma control in the past. Models showed that, even with incomplete coverage, mass drug administration could eliminate disease through a spillover effect, somewhat analogous to how incomplete vaccine campaigns can eliminate disease through herd protection. Models also suggest that elimination can always be achieved if enough people are treated often enough with an effective enough drug. Other models supported the idea that targeting ages at highest risk or continued improvements in hygiene and sanitation can contribute meaningfully to trachoma control. Models of intensive targeting of a core group may point the way to final eradication even in areas with substantial transmission and within-community heterogeneity.

The persistence of multiple strains of avian influenza in live bird markets.

Multiple subtypes of avian influenza (AI) and novel reassortants are frequently isolated from live bird markets (LBMs). However, our understanding of the drivers of persistence of multiple AI subtypes is limited. We propose a stochastic model of AI transmission within an LBM that incorporates market size, turnover rate and the balance of direct versus environmental transmissibility. We investigate the relationship between these factors and the critical community size (CCS) for the persistence of single and multiple AI strains within an LBM. We fit different models of seeding from farms to two-strain surveillance data collected from Shantou, China. For a single strain and plausible estimates for continuous turnover rates and transmissibility, the CCS was approximately 11 800 birds, only a 4.2% increase in this estimate was needed to ensure persistence of the co-infecting strains (two strains in a single host). Precise values of CCS estimates were sensitive to changes in market turnover rate and duration of the latent period. Assuming a gradual daily sell rate of birds the estimated CCS was higher than when an instantaneous selling rate was assumed. We were able to reproduce prevalence dynamics similar to observations from a single market in China with infection seeded every 5-15 days, and a maximum non-seeding duration of 80 days. Our findings suggest that persistence of co-infections is more likely to be owing to sequential infection of single strains rather than ongoing transmission of both strains concurrently. In any given system for a fixed set of ecological and epidemiological conditions, there is an LBM size below which the risk of sustained co-circulation is low and which may suggest a clear policy opportunity to reduce the frequency of influenza co-infection in poultry.

Enhanced antibiotic distribution strategies and the potential impact of facial cleanliness and environmental improvements for the sustained control of trachoma: a modelling study.

BACKGROUND: Despite some success in controlling trachoma with repeated mass drug administration (MDA), some hyperendemic regions are not responding as fast as anticipated. Available data suggests that individuals with higher bacterial infection loads are less likely to resolve infection following a single dose of treatment, and thus remain a source of re-emergent infection following treatment. We assessed the potential impact of a new double-dose antibiotic distribution strategy in addition to enhanced facial cleanliness (F) and environmental improvements (E). METHODS: Using a within-community mathematical model of trachoma transmission we assessed the impact of a new double-dose antibiotic distribution strategy given 2 weeks apart, with and without enhanced F&E. We compared the annual double-dose strategy to single-dose annual MDA treatment in hyper-, meso- and hypoendemic settings, and to biannual MDA at 6-monthly intervals in hyperendemic communities. RESULTS: The findings from our mathematical model suggest that implementing the new double-dose strategy for 5 years or less was predicted to control infection more successfully than annual or 6-monthly treatment. Infection was controlled more readily if treatment was combined with enhanced F&E. The results appeared robust to variation in a number of key epidemiological parameters. To have long-term impact on transmission, enhanced F&E is essential for high transmission settings. CONCLUSION: Our current findings are based on simualtion modelling only, due to lack of epidemilogical data, however they do suggest that the  annual double-dose treatment strategy is encouraging for trachoma control. In high transmission settings, both MDA and enhanced F&E are needed for sustained control.

A systematic review of reported reassortant viral lineages of influenza A.

BACKGROUND: Most previous evolutionary studies of influenza A have focussed on genetic drift, or reassortment of specific gene segments, hosts or subtypes. We conducted a systematic literature review to identify reported claimed reassortant influenza A lineages with genomic data available in GenBank, to obtain 646 unique first-report isolates out of a possible 20,781 open-access genomes. RESULTS: After adjusting for correlations, only: swine as host, China, Europe, Japan and years between 1997 and 2002; remained as significant risk factors for the reporting of reassortant viral lineages. For swine H1, more reassortants were observed in the North American H1 clade compared with the Eurasian avian-like H1N1 clade. Conversely, for avian H5 isolates, a higher number of reported reassortants were observed in the European H5N2/H3N2 clade compared with the H5N2 North American clade. CONCLUSIONS: Despite unavoidable biases (publication, database choice and upload propensity) these results synthesize a large majority of the current literature on novel reported influenza A reassortants and are a potentially useful prerequisite to inform further algorithmic studies.

Risk factors for UK Plasmodium falciparum cases.

BACKGROUND: An increasing proportion of malaria cases diagnosed in UK residents with a history of travel to malaria endemic areas are due to Plasmodium falciparum. METHODS: In order to identify travellers at most risk of acquiring malaria a proportional hazards model was used to estimate the risk of acquiring malaria stratified by purpose of travel and age whilst adjusting for entomological inoculation rate (EIR) and duration of stay in endemic countries. RESULTS: Travellers visiting friends and relatives and business travellers were found to have significantly higher hazard of acquiring malaria (adjusted hazard ratio (HR) relative to that of holiday makers 7.4, 95% CI 6.4-8.5, p < 0. 0001 and HR 3.4, 95% CI 2.9-3.8, p < 0. 0001, respectively). All age-groups were at lower risk than children aged 0-15 years. CONCLUSIONS: These estimates of the increased risk for business travellers and those visiting friends and relatives should be used to inform programmes to improve awareness of the risks of malaria when travelling.

Surveillance of low pathogenic novel H7N9 avian influenza in commercial poultry barns: detection of outbreaks and estimation of virus introduction time.

BACKGROUND: Both high and low pathogenic subtype A avian influenza remain ongoing threats to the commercial poultry industry globally. The emergence of a novel low pathogenic H7N9 lineage in China presents itself as a new concern to both human and animal health and may necessitate additional surveillance in commercial poultry operations in affected regions. METHODS: Sampling data was simulated using a mechanistic model of H7N9 influenza transmission within commercial poultry barns together with a stochastic observation process. Parameters were estimated using maximum likelihood. We assessed the probability of detecting an outbreak at time of slaughter using both real-time polymerase chain reaction (rt-PCR) and a hemagglutinin inhibition assay (HI assay) before considering more intense sampling prior to slaughter. The day of virus introduction and R0 were estimated jointly from weekly flock sampling data. For scenarios where R0 was known, we estimated the day of virus introduction into a barn under different sampling frequencies. RESULTS: If birds were tested at time of slaughter, there was a higher probability of detecting evidence of an outbreak using an HI assay compared to rt-PCR, except when the virus was introduced <2 weeks before time of slaughter. Prior to the initial detection of infection N sample = 50 (1%) of birds were sampled on a weekly basis once, but after infection was detected, N sample = 2000 birds (40%) were sampled to estimate both parameters. We accurately estimated the day of virus introduction in isolation with weekly and 2-weekly sampling. CONCLUSIONS: A strong sampling effort would be required to infer both the day of virus introduction and R0. Such a sampling effort would not be required to estimate the day of virus introduction alone once R0 was known, and sampling N sample = 50 of birds in the flock on a weekly or 2 weekly basis would be sufficient.

An economic evaluation of two cervical screening algorithms in Belgium: HR-HPV primary compared to HR-HPV and liquid-based cytology co-testing.

OBJECTIVE: To assess the costs and benefits of two algorithms for cervical cancer screening in Belgium (1) high-risk human papillomavirus (HR-HPV) primary screening and (2) HR-HPV and liquid-based cytology (LBC) co-testing. METHODS: A decision tree was adapted from published work and parameterised using HORIZON study data and Belgian cost and population data. The theoretical model represents two different screening algorithms for a cohort of 577 846 women aged 25-64 attending routine cervical screening. Scenario analyses were used to explore the impact of including vaccinated women and alternative pricing approaches. Uncertainty analyses were conducted. RESULTS: The cost per woman screened was €113.50 for HR-HPV primary screening and €101.70 for co-testing, representing a total cost of €65 588 573 and €58 775 083, respectively, for the cohort; a 10% difference. For one screening cycle, compared to HR-HPV primary, co-testing resulted in 13 173 more colposcopies, 67 731 more HR-HPV tests and 477 020 more LBC tests. Co-testing identified 2351 more CIN2+ cases per year (27% more than HR-HPV primary) and 1602 more CIN3+ cases (24% more than HR-HPV primary) than HR-HPV primary. CONCLUSION: In Belgium, a co-testing algorithm could increase cervical pre-cancer detection rates compared to HR-HPV primary. Co-testing would cost less than HR-HPV primary if the cost of the HPV test and LBC were cost-neutral compared to the current cost of LBC screening but would cost more if the cost per HPV test and LBC were the same in both co-testing and HR-HPV primary strategies.

Determining the Relationship Between Seizure-Free Days and Other Predictors of Quality of Life in Patients with Dravet Syndrome and Their Carers from FFA Registration Studies.

INTRODUCTION: Dravet syndrome (DS) is a rare, lifelong epileptic encephalopathy characterised by frequent and severe seizures associated with premature mortality. Typically diagnosed in infancy, patients also experience progressive behavioural, motor-function and cognitive decline. Twenty percent of patients do not reach adulthood. Quality of life (QoL) is impaired for both patients and their carers. Reducing convulsive seizure frequency, increasing convulsive seizure-free days (SFDs) and improving patient/carer QoL are primary treatment goals in DS. This study explored the relationship between SFDs and patients' and carers' QoL to inform a cost-utility analysis of fenfluramine (FFA). METHODS: In FFA registration studies, patients (or their carer proxies) completed the Paediatric QoL inventory (PedsQL). These data were mapped to EuroQol-5 Dimensions Youth version (EQ-5D-Y) to provide patient utilities. Carer utilities were collected using EQ-5D-5L and mapped to EQ-5D-3L to align patient and carer QoL on the same scale. Linear mixed-effects and panel regression models were tested and Hausman tests identified the most appropriate approach for each group. On this basis, a linear mixed-effects regression model was used to examine the relationships between patient EQ-5D-Y and clinically relevant variables (age, frequency of SFDs per 28 days, motor impairments and treatment dose). A linear panel regression model examined the relationship between SFDs and carer QoL. RESULTS: After adjustment for age and underlying comorbidities, the patient regression model showed that SFDs per 28 days was a significant predictor of QoL. Each additional patient-SFD increased utility by 0.005 (p < 0.001). The carer linear panel model also showed that increasing SFDs per 28 days was a significant predictor of improved QoL. Each additional SFD increased carer utility by 0.014 (p < 0.001). CONCLUSION: This regression framework highlights that SFDs are significantly correlated with both patients' and carers' QoL. Treatment with effective antiseizure medications that increase SFDs directly improves QoL for patients and their carers.

Social mixing patterns relevant to infectious diseases spread by close contact in urban Blantyre, Malawi.

INTRODUCTION: Understanding human mixing patterns relevant to infectious diseases spread through close contact is vital for modelling transmission dynamics and optimisation of disease control strategies. Mixing patterns in low-income countries like Malawi are not well known. METHODOLOGY: We conducted a social mixing survey in urban Blantyre, Malawi between April and July 2021 (between the 2nd and 3rd wave of COVID-19 infections). Participants living in densely-populated neighbourhoods were randomly sampled and, if they consented, reported their physical and non-physical contacts within and outside homes lasting at least 5 min during the previous day. Age-specific mixing rates were calculated, and a negative binomial mixed effects model was used to estimate determinants of contact behaviour. RESULTS: Of 1201 individuals enroled, 702 (58.5%) were female, the median age was 15 years (interquartile range [IQR] 5-32) and 127 (10.6%) were HIV-positive. On average, participants reported 10.3 contacts per day (range: 1-25). Mixing patterns were highly age-assortative, particularly those within the community and with skin-to-skin contact. Adults aged 20-49 y reported the most contacts (median:11, IQR: 8-15) of all age groups; 38% (95%CI: 16-63) more than infants (median: 8, IQR: 5-10), who had the least contacts. Household contact frequency increased by 3% (95%CI: 2-5) per additional household member. Unemployed participants had 15% (95%CI: 9-21) fewer contacts than other adults. Among long range (>30 m away from home) contacts, secondary school children had the largest median contact distance from home (257 m, IQR 78-761). HIV-positive status in adults >=18 years-old was not associated with changed contact patterns (rate ratio: 1.01, 95%CI: (0.91-1.12)). During this period of relatively low COVID-19 incidence in Malawi, 301 (25.1%) individuals stated that they had limited their contact with others due to COVID-19 precautions; however, their reported contacts were 8% (95%CI: 1-13) higher. CONCLUSION: In urban Malawi, contact rates, are high and age-assortative, with little reported behavioural change due to either HIV-status or COVID-19 circulation. This highlights the limits of contact-restriction-based mitigation strategies in such settings and the need for pandemic preparedness to better understand how contact reductions can be enabled and motivated.

Strengthening data collection for neglected tropical diseases: What data are needed for models to better inform tailored intervention programmes?

Locally tailored interventions for neglected tropical diseases (NTDs) are becoming increasingly important for ensuring that the World Health Organization (WHO) goals for control and elimination are reached. Mathematical models, such as those developed by the NTD Modelling Consortium, are able to offer recommendations on interventions but remain constrained by the data currently available. Data collection for NTDs needs to be strengthened as better data are required to indirectly inform transmission in an area. Addressing specific data needs will improve our modelling recommendations, enabling more accurate tailoring of interventions and assessment of their progress. In this collection, we discuss the data needs for several NTDs, specifically gambiense human African trypanosomiasis, lymphatic filariasis, onchocerciasis, schistosomiasis, soil-transmitted helminths (STH), trachoma, and visceral leishmaniasis. Similarities in the data needs for these NTDs highlight the potential for integration across these diseases and where possible, a wider spectrum of diseases.

Vaccine strategies to reduce the burden of pneumococcal disease in HIV-infected adults in Africa.

INTRODUCTION: Streptococcus pneumoniae is the leading cause of invasive bacterial disease, globally. Despite antiretroviral therapy, adults infected with human immunodeficiency virus (HIV) are also at high risk of pneumococcal carriage and disease. Pneumococcal conjugate vaccines (PCVs) provide effective protection against vaccine serotype (VT) carriage and disease in children, and have been introduced worldwide, including most HIV-affected low- and middle-income countries. Unlike high-income countries, the circulation of VT persists in the PCV era in some low-income countries and results in a continued high burden of pneumococcal disease in HIV-infected adults. Moreover, no routine vaccination that directly protects HIV-infected adults in such settings has been implemented. AREAS COVERED: Nonsystematic review on the pneumococcal burden in HIV-infected adults and vaccine strategies to reduce this burden. EXPERT OPINION: We propose and discuss the relative merit of changing the infant PCV program to use (1a) a two prime plus booster dose schedule, (1b) a two prime plus booster dose schedule with an additional booster dose at school entry, to directly vaccinate (2a) HIV-infected adults or vaccinating (2b) HIV-infected pregnant women for direct protection, with added indirect protection to the high-risk neonates. We identify key knowledge gaps for such an evaluation and propose strategies to overcome them.

Infant contact in day-care centres in Vietnam: A cross-sectional study to understand infant infection risk.

Background: Infant contact information (skin-to-skin contact between infants and others) is important to understand Streptococcus pneumoniae transmission patterns. A few studies have investigated infant contact patterns by asking the mother/guardian to record all contacts a child makes in one day. However, this approach does not capture contact behaviour in day-care. Our study describes the frequency and nature of physical contacts of infants in day-care to understand infant infection risk in day-care in Nha Trang, Vietnam. Methods: This cross-sectional study enrolled infants aged <12 months, attending 10 randomly selected day-care centres in Nha Trang. Physical contacts of each infant for one day at the day-care centre were observed. The mean number of infants' contacts and factors associated with contact numbers were assessed using negative binomial regression. Results: In total 14 infants, aged 6-11 months, were enrolled, and a total of 96 contacts were recorded. The mean number of contacts an infant made in one day was 6.9. Infants who walked independently (age-adjusted rate ratio 1.68, 95% confidence interval 1.06-2.68) and those cared for in a larger group (1.99, 1.42-2.79) had more contacts at day-care. About 50% of infants made contact with at least one person from a commune different from the infant's, and 50% made contact with at least one other infant at day-care. Conclusion: This study found that day-care attendance may be one factor that increases contact rates of infants in Nha Trang and diversifies those contacts in terms of age and geographical spread. In this study, day-care attendance not only increased contact rates beyond those usually experienced by young children cared for at home but specifically increased contact rates with other children and adults from other communes. Day-care may play a key role in the transmission of respiratory pathogens like Streptococcus pneumoniae to infants.

Optimising sampling regimes and data collection to inform surveillance for trachoma control.

It is estimated that 190 million individuals are at risk of blindness from trachoma, and that control by mass drug administration (MDA) is reducing this risk in many populations. Programs are monitored using prevalence of follicular trachoma disease (TF) in children. However, as programs progress to low prevalence there are challenges interpreting this indirect measure of infection. PCR and sero-surveillance are being considered as complementary tools to monitor low-level transmission, but there are questions on how they can be most effectively used. We use a previously-published, mathematical model to explore the dynamic relationship between TF and PCR throughout a control program and a sero-catalytic model to evaluate the utility of two cross-sectional sero-surveys for estimating sero-conversion rates. The simulations show that whilst PCR is more sensitive than TF at detecting infection, the probability of detecting at least one positive individual declines during an MDA program more quickly for PCR than for TF (for the same sample size). Towards the end of a program there is a moderate chance of a random sample showing both low PCR prevalence and higher TF prevalence, which may contribute to the lack of correlation observed in epidemiological studies. We also show that conducting two cross-sectional sero-surveys 10 years apart can provide more precise and accurate estimation of epidemiological parameters than a single survey, supporting previous findings that whilst serology holds great promise, multiple cross-sections from the same community are needed to generate the most valuable information about transmission. These results highlight that the quantitative dynamics of infection and disease should be included alongside the many logistical and practical factors to be considered in designing a monitoring and evaluation strategy at the operational research level, in order to help subsequently inform data collection for individual country programs. Whilst our simulations provide some insight, they also highlight that some level of longitudinal, individual-level data on reinfection and disease may be needed to monitor elimination progress.

The utility of serology for elimination surveillance of trachoma.

Robust surveillance methods are needed for trachoma control and recrudescence monitoring, but existing methods have limitations. Here, we analyse data from nine trachoma-endemic populations and provide operational thresholds for interpretation of serological data in low-transmission and post-elimination settings. Analyses with sero-catalytic and antibody acquisition models provide insights into transmission history within each population. To accurately estimate sero-conversion rates (SCR) for trachoma in populations with high-seroprevalence in adults, the model accounts for secondary exposure to Chlamydia trachomatis due to urogenital infection. We estimate the population half-life of sero-reversion for anti-Pgp3 antibodies to be 26 (95% credible interval (CrI): 21-34) years. We show SCRs below 0.015 (95% confidence interval (CI): 0.0-0.049) per year correspond to a prevalence of trachomatous inflammation-follicular below 5%, the current threshold for elimination of active trachoma as a public health problem. As global trachoma prevalence declines, we may need cross-sectional serological survey data to inform programmatic decisions.

Improving our forecasts for trachoma elimination: What else do we need to know?

The World Health Organization (WHO) has targeted trachoma for elimination as a public health concern by 2020. Mathematical modelling is used for a range of infectious diseases to assess the impact of different intervention strategies on the prevalence of infection or disease. Here we evaluate the performance of four different mechanistic mathematical models that could all realistically represent trachoma transmission. We fit the four different mechanistic models of trachoma transmission to cross-sectional age-specific Polymerase Chain Reaction (PCR) and Trachomatous inflammation, follicular (TF) prevalence data. We estimate 4 or 3 parameters within each model, including the duration of an individual's infection and disease episode using Markov Chain Monte Carlo. We assess the performance of each models fit to the data by calculating the deviance information criterion. We then model the implementation of different interventions for each model structure to assess the feasibility of elimination of trachoma with different model structures. A model structure which allowed some re-infection in the disease state (Model 2) was statistically the most well performing model. All models struggled to fit to the very high prevalence of active disease in the youngest age group. Our simulations suggested that for Model 3, with annual antibiotic treatment and transmission reduction, the chance of reducing active disease prevalence to < 5% within 5 years was very low, while Model 2 and 4 could ensure that active disease prevalence was reduced within 5 years. Model 2 here fitted to the data best of the models evaluated. The appropriate level of susceptibility to re-infection was, however, challenging to identify given the amount and kind of data available. We demonstrate that the model structure assumed can lead to different end points following the implementation of the same interventions. Our findings are likely to extend beyond trachoma and should be considered when modelling other neglected tropical diseases.

Probabilistic forecasts of trachoma transmission at the district level: A statistical model comparison.

The World Health Organization and its partners are aiming to eliminate trachoma as a public health problem by 2020. In this study, we compare forecasts of TF prevalence in 2011 for 7 different statistical and mechanistic models across 9 de-identified trachoma endemic districts, representing 4 unique trachoma endemic countries. We forecast TF prevalence between 1-6 years ahead in time and compare the 7 different models to the observed 2011 data using a log-likelihood score. An SIS model, including a district-specific random effect for the district-specific transmission coefficient, had the highest log-likelihood score across all 9 districts and was therefore the best performing model. While overall the deterministic transmission model was the least well performing model, although it did comparably well to the other models for 8 of 9 districts. We perform a statistically rigorous comparison of the forecasting ability of a range of mathematical and statistical models across multiple endemic districts between 1 and 6 years ahead of the last collected TF prevalence data point in 2011, assessing results against surveillance data. This study is a step towards making statements about likelihood and time to elimination with regard to the WHO GET2020 goals.