Scenario design for infectious disease projections: Integrating concepts from decision analysis and experimental design.

Across many fields, scenario modeling has become an important tool for exploring long-term projections and how they might depend on potential interventions and critical uncertainties, with relevance to both decision makers and scientists. In the past decade, and especially during the COVID-19 pandemic, the field of epidemiology has seen substantial growth in the use of scenario projections. Multiple scenarios are often projected at the same time, allowing important comparisons that can guide the choice of intervention, the prioritization of research topics, or public communication. The design of the scenarios is central to their ability to inform important questions. In this paper, we draw on the fields of decision analysis and statistical design of experiments to propose a framework for scenario design in epidemiology, with relevance also to other fields. We identify six different fundamental purposes for scenario designs (decision making, sensitivity analysis, situational awareness, horizon scanning, forecasting, and value of information) and discuss how those purposes guide the structure of scenarios. We discuss other aspects of the content and process of scenario design, broadly for all settings and specifically for multi-model ensemble projections. As an illustrative case study, we examine the first 17 rounds of scenarios from the U.S. COVID-19 Scenario Modeling Hub, then reflect on future advancements that could improve the design of scenarios in epidemiological settings.

Scenario Design for Infectious Disease Projections: Integrating Concepts from Decision Analysis and Experimental Design.

Across many fields, scenario modeling has become an important tool for exploring long-term projections and how they might depend on potential interventions and critical uncertainties, with relevance to both decision makers and scientists. In the past decade, and especially during the COVID-19 pandemic, the field of epidemiology has seen substantial growth in the use of scenario projections. Multiple scenarios are often projected at the same time, allowing important comparisons that can guide the choice of intervention, the prioritization of research topics, or public communication. The design of the scenarios is central to their ability to inform important questions. In this paper, we draw on the fields of decision analysis and statistical design of experiments to propose a framework for scenario design in epidemiology, with relevance also to other fields. We identify six different fundamental purposes for scenario designs (decision making, sensitivity analysis, value of information, situational awareness, horizon scanning, and forecasting) and discuss how those purposes guide the structure of scenarios. We discuss other aspects of the content and process of scenario design, broadly for all settings and specifically for multi-model ensemble projections. As an illustrative case study, we examine the first 17 rounds of scenarios from the U.S. COVID-19 Scenario Modeling Hub, then reflect on future advancements that could improve the design of scenarios in epidemiological settings.

Estimating the health impact of vaccination against ten pathogens in 98 low-income and middle-income countries from 2000 to 2030: a modelling study.

BACKGROUND: The past two decades have seen expansion of childhood vaccination programmes in low-income and middle-income countries (LMICs). We quantify the health impact of these programmes by estimating the deaths and disability-adjusted life-years (DALYs) averted by vaccination against ten pathogens in 98 LMICs between 2000 and 2030. METHODS: 16 independent research groups provided model-based disease burden estimates under a range of vaccination coverage scenarios for ten pathogens: hepatitis B virus, Haemophilus influenzae type B, human papillomavirus, Japanese encephalitis, measles, Neisseria meningitidis serogroup A, Streptococcus pneumoniae, rotavirus, rubella, and yellow fever. Using standardised demographic data and vaccine coverage, the impact of vaccination programmes was determined by comparing model estimates from a no-vaccination counterfactual scenario with those from a reported and projected vaccination scenario. We present deaths and DALYs averted between 2000 and 2030 by calendar year and by annual birth cohort. FINDINGS: We estimate that vaccination of the ten selected pathogens will have averted 69 million (95% credible interval 52-88) deaths between 2000 and 2030, of which 37 million (30-48) were averted between 2000 and 2019. From 2000 to 2019, this represents a 45% (36-58) reduction in deaths compared with the counterfactual scenario of no vaccination. Most of this impact is concentrated in a reduction in mortality among children younger than 5 years (57% reduction [52-66]), most notably from measles. Over the lifetime of birth cohorts born between 2000 and 2030, we predict that 120 million (93-150) deaths will be averted by vaccination, of which 58 million (39-76) are due to measles vaccination and 38 million (25-52) are due to hepatitis B vaccination. We estimate that increases in vaccine coverage and introductions of additional vaccines will result in a 72% (59-81) reduction in lifetime mortality in the 2019 birth cohort. INTERPRETATION: Increases in vaccine coverage and the introduction of new vaccines into LMICs have had a major impact in reducing mortality. These public health gains are predicted to increase in coming decades if progress in increasing coverage is sustained. FUNDING: Gavi, the Vaccine Alliance and the Bill & Melinda Gates Foundation.

Estimating the burden of SARS-CoV-2 in France.

France has been heavily affected by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic and went into lockdown on 17 March 2020. Using models applied to hospital and death data, we estimate the impact of the lockdown and current population immunity. We find that 2.9% of infected individuals are hospitalized and 0.5% of those infected die (95% credible interval: 0.3 to 0.9%), ranging from 0.001% in those under 20 years of age to 8.3% in those 80 years of age or older. Across all ages, men are more likely to be hospitalized, enter intensive care, and die than women. The lockdown reduced the reproductive number from 2.90 to 0.67 (77% reduction). By 11 May 2020, when interventions are scheduled to be eased, we project that 3.5 million people (range: 2.1 million to 6.0 million), or 5.3% of the population (range: 3.3 to 9.3%), will have been infected. Population immunity appears to be insufficient to avoid a second wave if all control measures are released at the end of the lockdown.

Epidemiology and transmission of COVID-19 in 391 cases and 1286 of their close contacts in Shenzhen, China: a retrospective cohort study.

BACKGROUND: Rapid spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Wuhan, China, prompted heightened surveillance in Shenzhen, China. The resulting data provide a rare opportunity to measure key metrics of disease course, transmission, and the impact of control measures. METHODS: From Jan 14 to Feb 12, 2020, the Shenzhen Center for Disease Control and Prevention identified 391 SARS-CoV-2 cases and 1286 close contacts. We compared cases identified through symptomatic surveillance and contact tracing, and estimated the time from symptom onset to confirmation, isolation, and admission to hospital. We estimated metrics of disease transmission and analysed factors influencing transmission risk. FINDINGS: Cases were older than the general population (mean age 45 years) and balanced between males (n=187) and females (n=204). 356 (91%) of 391 cases had mild or moderate clinical severity at initial assessment. As of Feb 22, 2020, three cases had died and 225 had recovered (median time to recovery 21 days; 95% CI 20-22). Cases were isolated on average 4·6 days (95% CI 4·1-5·0) after developing symptoms; contact tracing reduced this by 1·9 days (95% CI 1·1-2·7). Household contacts and those travelling with a case were at higher risk of infection (odds ratio 6·27 [95% CI 1·49-26·33] for household contacts and 7·06 [1·43-34·91] for those travelling with a case) than other close contacts. The household secondary attack rate was 11·2% (95% CI 9·1-13·8), and children were as likely to be infected as adults (infection rate 7·4% in children <10 years vs population average of 6·6%). The observed reproductive number (R) was 0·4 (95% CI 0·3-0·5), with a mean serial interval of 6·3 days (95% CI 5·2-7·6). INTERPRETATION: Our data on cases as well as their infected and uninfected close contacts provide key insights into the epidemiology of SARS-CoV-2. This analysis shows that isolation and contact tracing reduce the time during which cases are infectious in the community, thereby reducing the R. The overall impact of isolation and contact tracing, however, is uncertain and highly dependent on the number of asymptomatic cases. Moreover, children are at a similar risk of infection to the general population, although less likely to have severe symptoms; hence they should be considered in analyses of transmission and control. FUNDING: Emergency Response Program of Harbin Institute of Technology, Emergency Response Program of Peng Cheng Laboratory, US Centers for Disease Control and Prevention.

The projected impact of geographic targeting of oral cholera vaccination in sub-Saharan Africa: A modeling study.

BACKGROUND: Cholera causes an estimated 100,000 deaths annually worldwide, with the majority of burden reported in sub-Saharan Africa. In May 2018, the World Health Assembly committed to reducing worldwide cholera deaths by 90% by 2030. Oral cholera vaccine (OCV) plays a key role in reducing the near-term risk of cholera, although global supplies are limited. Characterizing the potential impact and cost-effectiveness of mass OCV deployment strategies is critical for setting expectations and developing cholera control plans that maximize the chances of success. METHODS AND FINDINGS: We compared the projected impacts of vaccination campaigns across sub-Saharan Africa from 2018 through 2030 when targeting geographically according to historical cholera burden and risk factors. We assessed the number of averted cases, deaths, and disability-adjusted life years and the cost-effectiveness of these campaigns with models that accounted for direct and indirect vaccine effects and population projections over time. Under current vaccine supply projections, an approach optimized to targeting by historical burden is projected to avert 828,971 (95% CI 803,370-859,980) cases (equivalent to 34.0% of projected cases; 95% CI 33.2%-34.8%). An approach that balances logistical feasibility with targeting historical burden is projected to avert 617,424 (95% CI 599,150-643,891) cases. In contrast, approaches optimized for targeting locations with limited access to water and sanitation are projected to avert 273,939 (95% CI 270,319-277,002) and 109,817 (95% CI 103,735-114,110) cases, respectively. We find that the most logistically feasible targeting strategy costs US$1,843 (95% CI 1,328-14,312) per DALY averted during this period and that effective geographic targeting of OCV campaigns can have a greater impact on cost-effectiveness than improvements to vaccine efficacy and moderate increases in coverage. Although our modeling approach does not project annual changes in baseline cholera risk or directly incorporate immunity from natural cholera infection, our estimates of the relative performance of different vaccination strategies should be robust to these factors. CONCLUSIONS: Our study suggests that geographic targeting substantially improves the cost-effectiveness and impact of oral cholera vaccination campaigns. Districts with the poorest access to improved water and sanitation are not the same as districts with the greatest historical cholera incidence. While OCV campaigns can improve cholera control in the near term, without rapid progress in developing water and sanitation services or dramatic increases in OCV supply, our results suggest that vaccine use alone is unlikely to allow us to achieve the 2030 goal.

A methodological comparison of risk scores versus decision trees for predicting drug-resistant infections: A case study using extended-spectrum beta-lactamase (ESBL) bacteremia.

BACKGROUND: Timely identification of multidrug-resistant gram-negative infections remains an epidemiological challenge. Statistical models for predicting drug resistance can offer utility where rapid diagnostics are unavailable or resource-impractical. Logistic regression-derived risk scores are common in the healthcare epidemiology literature. Machine learning-derived decision trees are an alternative approach for developing decision support tools. Our group previously reported on a decision tree for predicting ESBL bloodstream infections. Our objective in the current study was to develop a risk score from the same ESBL dataset to compare these 2 methods and to offer general guiding principles for using each approach. METHODS: Using a dataset of 1,288 patients with Escherichia coli or Klebsiella spp bacteremia, we generated a risk score to predict the likelihood that a bacteremic patient was infected with an ESBL-producer. We evaluated discrimination (original and cross-validated models) using receiver operating characteristic curves and C statistics. We compared risk score and decision tree performance, and we reviewed their practical and methodological attributes. RESULTS: In total, 194 patients (15%) were infected with ESBL-producing bacteremia. The clinical risk score included 14 variables, compared to the 5 decision-tree variables. The positive and negative predictive values of the risk score and decision tree were similar (>90%), but the C statistic of the risk score (0.87) was 10% higher. CONCLUSIONS: A decision tree and risk score performed similarly for predicting ESBL infection. The decision tree was more user-friendly, with fewer variables for the end user, whereas the risk score offered higher discrimination and greater flexibility for adjusting sensitivity and specificity.

Comparison of US County-Level Public Health Performance Rankings With County Cluster and National Rankings: Assessment Based on Prevalence Rates of Smoking and Obesity and Motor Vehicle Crash Death Rates.

Importance: Health departments can be grouped together based on sociodemographic characteristics of the population served. Comparisons within these groups can then help with monitoring and improving the health of their populations. Objective: To compare county-level percentile rankings on outcomes of smoking, motor vehicle crash deaths, and obesity within sociodemographic peer clusters vs nationwide rankings. Design, Setting, and Participants: This cross-sectional, population-based study of demographic and health data from the 2014 Behavioral Risk Factor Surveillance System and the 2016 Robert Wood Johnson Foundation County Health Rankings data set was conducted at 3139 of 3143 US counties and county-equivalents. Four locations were excluded due to incomplete data. Data analysis was conducted between January and August 2017. Exposures: Random forest algorithms were used to identify sociodemographic characteristics most associated with the outcomes of interest. These characteristics were race and ethnicity, educational attainment, age, marital status, employment status, sex, and health insurance status. k-means clustering was used to cluster counties based on these sociodemographic characteristics and the percentage of the county classified as rural. Main Outcomes and Measures: County-level smoking prevalence, motor vehicle crash death rate, and obesity prevalence. County percentile rankings on the outcomes of interest were compared in the national context and the within-cluster context. Results: A total of 318 856 967 individuals (mean [SD] individuals per county, 101 579.2 [326 315]; 161 911 910 women [50.8%]) were represented by the 3139 counties used in this analysis. Eight distinct sociodemographic clusters throughout the United States were found. Cluster-specific percentile rankings for both smoking prevalence and motor vehicle crash death rates improved more than 70 percentile points for several counties in the rural, American Indian cluster compared with the nationwide percentiles. Conversely, the young, urban, middle to high socioeconomic status cluster included counties with cluster-specific percentile rankings that declined by 60 percentile points or more compared with the nationwide rankings for all 3 outcomes of interest. Conclusions and Relevance: Comparing county health outcomes on a nationwide or statewide basis fails to adequately account for sociodemographic context. Clustering counties by sociodemographic factors related to the outcome of interest allows a better understanding of other factors that may be shaping the prevalence of health outcomes. These groupings may also aid learning exchange.

Mapping the burden of cholera in sub-Saharan Africa and implications for control: an analysis of data across geographical scales.

BACKGROUND: Cholera remains a persistent health problem in sub-Saharan Africa and worldwide. Cholera can be controlled through appropriate water and sanitation, or by oral cholera vaccination, which provides transient (∼3 years) protection, although vaccine supplies remain scarce. We aimed to map cholera burden in sub-Saharan Africa and assess how geographical targeting could lead to more efficient interventions. METHODS: We combined information on cholera incidence in sub-Saharan Africa (excluding Djibouti and Eritrea) from 2010 to 2016 from datasets from WHO, Médecins Sans Frontières, ProMED, ReliefWeb, ministries of health, and the scientific literature. We divided the study region into 20 km × 20 km grid cells and modelled annual cholera incidence in each grid cell assuming a Poisson process adjusted for covariates and spatially correlated random effects. We combined these findings with data on population distribution to estimate the number of people living in areas of high cholera incidence (>1 case per 1000 people per year). We further estimated the reduction in cholera incidence that could be achieved by targeting cholera prevention and control interventions at areas of high cholera incidence. FINDINGS: We included 279 datasets covering 2283 locations in our analyses. In sub-Saharan Africa (excluding Djibouti and Eritrea), a mean of 141 918 cholera cases (95% credible interval [CrI] 141 538-146 505) were reported per year. 4·0% (95% CrI 1·7-16·8) of districts, home to 87·2 million people (95% CrI 60·3 million to 118·9 million), have high cholera incidence. By focusing on the highest incidence districts first, effective targeted interventions could eliminate 50% of the region's cholera by covering 35·3 million people (95% CrI 26·3 million to 62·0 million), which is less than 4% of the total population. INTERPRETATION: Although cholera occurs throughout sub-Saharan Africa, its highest incidence is concentrated in a small proportion of the continent. Prioritising high-risk areas could substantially increase the efficiency of cholera control programmes. FUNDING: The Bill & Melinda Gates Foundation.

The potential impact of case-area targeted interventions in response to cholera outbreaks: A modeling study.

BACKGROUND: Cholera prevention and control interventions targeted to neighbors of cholera cases (case-area targeted interventions [CATIs]), including improved water, sanitation, and hygiene, oral cholera vaccine (OCV), and prophylactic antibiotics, may be able to efficiently avert cholera cases and deaths while saving scarce resources during epidemics. Efforts to quickly target interventions to neighbors of cases have been made in recent outbreaks, but little empirical evidence related to the effectiveness, efficiency, or ideal design of this approach exists. Here, we aim to provide practical guidance on how CATIs might be used by exploring key determinants of intervention impact, including the mix of interventions, "ring" size, and timing, in simulated cholera epidemics fit to data from an urban cholera epidemic in Africa. METHODS AND FINDINGS: We developed a micro-simulation model and calibrated it to both the epidemic curve and the small-scale spatiotemporal clustering pattern of case households from a large 2011 cholera outbreak in N'Djamena, Chad (4,352 reported cases over 232 days), and explored the potential impact of CATIs in simulated epidemics. CATIs were implemented with realistic logistical delays after cases presented for care using different combinations of prophylactic antibiotics, OCV, and/or point-of-use water treatment (POUWT) starting at different points during the epidemics and targeting rings of various radii around incident case households. Our findings suggest that CATIs shorten the duration of epidemics and are more resource-efficient than mass campaigns. OCV was predicted to be the most effective single intervention, followed by POUWT and antibiotics. CATIs with OCV started early in an epidemic focusing on a 100-m radius around case households were estimated to shorten epidemics by 68% (IQR 62% to 72%), with an 81% (IQR 69% to 87%) reduction in cases compared to uncontrolled epidemics. These same targeted interventions with OCV led to a 44-fold (IQR 27 to 78) reduction in the number of people needed to target to avert a single case of cholera, compared to mass campaigns in high-cholera-risk neighborhoods. The optimal radius to target around incident case households differed by intervention type, with antibiotics having an optimal radius of 30 m to 45 m compared to 70 m to 100 m for OCV and POUWT. Adding POUWT or antibiotics to OCV provided only marginal impact and efficiency improvements. Starting CATIs early in an epidemic with OCV and POUWT targeting those within 100 m of an incident case household reduced epidemic durations by 70% (IQR 65% to 75%) and the number of cases by 82% (IQR 71% to 88%) compared to uncontrolled epidemics. CATIs used late in epidemics, even after the peak, were estimated to avert relatively few cases but substantially reduced the number of epidemic days (e.g., by 28% [IQR 15% to 45%] for OCV in a 100-m radius). While this study is based on a rigorous, data-driven approach, the relatively high uncertainty about the ways in which POUWT and antibiotic interventions reduce cholera risk, as well as the heterogeneity in outbreak dynamics from place to place, limits the precision and generalizability of our quantitative estimates. CONCLUSIONS: In this study, we found that CATIs using OCV, antibiotics, and water treatment interventions at an appropriate radius around cases could be an effective and efficient way to fight cholera epidemics. They can provide a complementary and efficient approach to mass intervention campaigns and may prove particularly useful during the initial phase of an outbreak, when there are few cases and few available resources, or in order to shorten the often protracted tails of cholera epidemics.

High resolution age-structured mapping of childhood vaccination coverage in low and middle income countries.

BACKGROUND: The expansion of childhood vaccination programs in low and middle income countries has been a substantial public health success story. Indicators of the performance of intervention programmes such as coverage levels and numbers covered are typically measured through national statistics or at the scale of large regions due to survey design, administrative convenience or operational limitations. These mask heterogeneities and 'coldspots' of low coverage that may allow diseases to persist, even if overall coverage is high. Hence, to decrease inequities and accelerate progress towards disease elimination goals, fine-scale variation in coverage should be better characterized. METHODS: Using measles as an example, cluster-level Demographic and Health Surveys (DHS) data were used to map vaccination coverage at 1 km spatial resolution in Cambodia, Mozambique and Nigeria for varying age-group categories of children under five years, using Bayesian geostatistical techniques built on a suite of publicly available geospatial covariates and implemented via Markov Chain Monte Carlo (MCMC) methods. RESULTS: Measles vaccination coverage was found to be strongly predicted by just 4-5 covariates in geostatistical models, with remoteness consistently selected as a key variable. The output 1 × 1 km maps revealed significant heterogeneities within the three countries that were not captured using province-level summaries. Integration with population data showed that at the time of the surveys, few districts attained the 80% coverage, that is one component of the WHO Global Vaccine Action Plan 2020 targets. CONCLUSION: The elimination of vaccine-preventable diseases requires a strong evidence base to guide strategies and inform efficient use of limited resources. The approaches outlined here provide a route to moving beyond large area summaries of vaccination coverage that mask epidemiologically-important heterogeneities to detailed maps that capture subnational vulnerabilities. The output datasets are built on open data and methods, and in flexible format that can be aggregated to more operationally-relevant administrative unit levels.

Quantifying the Risk and Cost of Active Monitoring for Infectious Diseases.

During outbreaks of deadly emerging pathogens (e.g., Ebola, MERS-CoV) and bioterror threats (e.g., smallpox), actively monitoring potentially infected individuals aims to limit disease transmission and morbidity. Guidance issued by CDC on active monitoring was a cornerstone of its response to the West Africa Ebola outbreak. There are limited data on how to balance the costs and performance of this important public health activity. We present a framework that estimates the risks and costs of specific durations of active monitoring for pathogens of significant public health concern. We analyze data from New York City's Ebola active monitoring program over a 16-month period in 2014-2016. For monitored individuals, we identified unique durations of active monitoring that minimize expected costs for those at "low (but not zero) risk" and "some or high risk": 21 and 31 days, respectively. Extending our analysis to smallpox and MERS-CoV, we found that the optimal length of active monitoring relative to the median incubation period was reduced compared to Ebola due to less variable incubation periods. Active monitoring can save lives but is expensive. Resources can be most effectively allocated by using exposure-risk categories to modify the duration or intensity of active monitoring.

Impact of interventions and the incidence of ebola virus disease in Liberia-implications for future epidemics.

To better understand the impact of national and global efforts to contain the Ebola virus disease epidemic of 2014­15 in Liberia, we provide a detailed timeline of the major interventions and relate them to the epidemic curve. In addition to personal experience in the response, we systematically reviewed situation reports from the Liberian government, UN, CDC, WHO, UNICEF, IFRC, USAID, and local and international news reports to create the timeline. We extracted data on the timing and nature of activities and compared them to the timeline of the epidemic curve using the reproduction number­the estimate of the average number of new cases caused by a single case. Interventions were organized around five major strategies, with the majority of resources directed to the creation of treatment beds. We conclude that no single intervention stopped the epidemic; rather, the interventions likely had reinforcing effects, and some were less likely than others to have made a major impact. We find that the epidemic's turning coincided with a reorganization of the response in August­September 2014, the emergence of community leadership in control efforts, and changing beliefs and practices in the population. Ebola Treatment Units were important for Ebola treatment, but the vast majority of these treatment centre beds became available after the epidemic curve began declining. Similarly, the United Nations Mission for Ebola Emergency Response was launched after the epidemic curve had already turned. These findings have significant policy implications for future epidemics and suggest that much of the decline in the epidemic curve was driven by critical behaviour changes within local communities, rather than by international efforts that came after the epidemic had turned. Future global interventions in epidemic response should focus on building community capabilities, strengthening local ownership, and dramatically reducing delays in the response.

How social structures, space, and behaviors shape the spread of infectious diseases using chikungunya as a case study.

Whether an individual becomes infected in an infectious disease outbreak depends on many interconnected risk factors, which may relate to characteristics of the individual (e.g., age, sex), his or her close relatives (e.g., household members), or the wider community. Studies monitoring individuals in households or schools have helped elucidate the determinants of transmission in small social structures due to advances in statistical modeling; but such an approach has so far largely failed to consider individuals in the wider context they live in. Here, we used an outbreak of chikungunya in a rural community in Bangladesh as a case study to obtain a more comprehensive characterization of risk factors in disease spread. We developed Bayesian data augmentation approaches to account for uncertainty in the source of infection, recall uncertainty, and unobserved infection dates. We found that the probability of chikungunya transmission was 12% [95% credible interval (CI): 8-17%] between household members but dropped to 0.3% for those living 50 m away (95% CI: 0.2-0.5%). Overall, the mean transmission distance was 95 m (95% CI: 77-113 m). Females were 1.5 times more likely to become infected than males (95% CI: 1.2-1.8), which was virtually identical to the relative risk of being at home estimated from an independent human movement study in the country. Reported daily use of antimosquito coils had no detectable impact on transmission. This study shows how the complex interplay between the characteristics of an individual and his or her close and wider environment contributes to the shaping of infectious disease epidemics.

Impact on Epidemic Measles of Vaccination Campaigns Triggered by Disease Outbreaks or Serosurveys: A Modeling Study.

BACKGROUND: Routine vaccination supplemented by planned campaigns occurring at 2-5 y intervals is the core of current measles control and elimination efforts. Yet, large, unexpected outbreaks still occur, even when control measures appear effective. Supplementing these activities with mass vaccination campaigns triggered when low levels of measles immunity are observed in a sample of the population (i.e., serosurveys) or incident measles cases occur may provide a way to limit the size of outbreaks. METHODS AND FINDINGS: Measles incidence was simulated using stochastic age-structured epidemic models in settings conducive to high or low measles incidence, roughly reflecting demographic contexts and measles vaccination coverage of four heterogeneous countries: Nepal, Niger, Yemen, and Zambia. Uncertainty in underlying vaccination rates was modeled. Scenarios with case- or serosurvey-triggered campaigns reaching 20% of the susceptible population were compared to scenarios without triggered campaigns. The best performing of the tested case-triggered campaigns prevent an average of 28,613 (95% CI 25,722-31,505) cases over 15 y in our highest incidence setting and 599 (95% CI 464-735) cases in the lowest incidence setting. Serosurvey-triggered campaigns can prevent 89,173 (95% CI, 86,768-91,577) and 744 (612-876) cases, respectively, but are triggered yearly in high-incidence settings. Triggered campaigns reduce the highest cumulative incidence seen in simulations by up to 80%. While the scenarios considered in this strategic modeling exercise are reflective of real populations, the exact quantitative interpretation of the results is limited by the simplifications in country structure, vaccination policy, and surveillance system performance. Careful investigation into the cost-effectiveness in different contexts would be essential before moving forward with implementation. CONCLUSIONS: Serologically triggered campaigns could help prevent severe epidemics in the face of epidemiological and vaccination uncertainty. Hence, small-scale serology may serve as the basis for effective adaptive public health strategies, although, in high-incidence settings, case-triggered approaches are likely more efficient.

A Clinical Decision Tree to Predict Whether a Bacteremic Patient Is Infected With an Extended-Spectrum ß-Lactamase-Producing Organism.

BACKGROUND: Timely identification of extended-spectrum ß-lactamase (ESBL) bacteremia can improve clinical outcomes while minimizing unnecessary use of broad-spectrum antibiotics, including carbapenems. However, most clinical microbiology laboratories currently require at least 24 additional hours from the time of microbial genus and species identification to confirm ESBL production. Our objective was to develop a user-friendly decision tree to predict which organisms are ESBL producing, to guide appropriate antibiotic therapy. METHODS: We included patients ≥18 years of age with bacteremia due to Escherichia coli or Klebsiella species from October 2008 to March 2015 at Johns Hopkins Hospital. Isolates with ceftriaxone minimum inhibitory concentrations ≥2 µg/mL underwent ESBL confirmatory testing. Recursive partitioning was used to generate a decision tree to determine the likelihood that a bacteremic patient was infected with an ESBL producer. Discrimination of the original and cross-validated models was evaluated using receiver operating characteristic curves and by calculation of C-statistics. RESULTS: A total of 1288 patients with bacteremia met eligibility criteria. For 194 patients (15%), bacteremia was due to a confirmed ESBL producer. The final classification tree for predicting ESBL-positive bacteremia included 5 predictors: history of ESBL colonization/infection, chronic indwelling vascular hardware, age ≥43 years, recent hospitalization in an ESBL high-burden region, and ≥6 days of antibiotic exposure in the prior 6 months. The decision tree's positive and negative predictive values were 90.8% and 91.9%, respectively. CONCLUSIONS: Our findings suggest that a clinical decision tree can be used to estimate a bacteremic patient's likelihood of infection with ESBL-producing bacteria. Recursive partitioning offers a practical, user-friendly approach for addressing important diagnostic questions.

Heterogeneity of the HIV epidemic in agrarian, trading, and fishing communities in Rakai, Uganda: an observational epidemiological study.

BACKGROUND: Understanding the extent to which HIV burden differs across communities and the drivers of local disparities is crucial for an effective and targeted HIV response. We assessed community-level variations in HIV prevalence, risk factors, and treatment and prevention service uptake in Rakai, Uganda. METHODS: The Rakai Community Cohort Study (RCCS) is an open, population-based cohort of people aged 15-49 years in 40 communities. Participants are HIV tested and interviewed to obtain sociodemographic, behavioural, and health information. RCCS data from Aug 10, 2011, to May 30, 2013, were used to classify communities as agrarian (n=27), trading (n=9), or lakeside fishing sites (n=4). We mapped HIV prevalence with Bayesian methods, and characterised variability across and within community classifications. We also assessed differences in HIV risk factors and uptake of antiretroviral therapy and male circumcision between community types. FINDINGS: 17 119 individuals were included, 9215 (54%) of whom were female. 9931 participants resided in agrarian, 3318 in trading, and 3870 in fishing communities. Median HIV prevalence was higher in fishing communities (42%, range 38-43) than in trading (17%, 11-21) and agrarian communities (14%, 9-26). Antiretroviral therapy use was significantly lower in both men and women in fishing communities than in trading (age-adjusted prevalence risk ratio in men 0·64, 95% CI 0·44-0·97; women 0·53, 0·42-0·66) and agrarian communities (men 0·55, 0·42-0·72; women 0·65, 0·54-0·79), as was circumcision coverage among men (vs trading 0·48, 0·42-0·55; vs agrarian 0·64, 0·56-0·72). Self-reported risk behaviours were significantly higher in men than in women and in fishing communities than in other community types. INTERPRETATION: Substantial heterogeneity in HIV prevalence, risk factors, and service uptake in Rakai, Uganda, emphasises the need for local surveillance and the design of targeted HIV responses. High HIV burden, risk behaviours, and low use of combination HIV prevention in fishing communities make these populations a priority for intervention. FUNDING: National Institute of Mental Health, the National Institute of Allergy and Infectious Diseases, the National Institute of Child Health and Development, and the National Institute for Allergy and Infectious Diseases Division of Intramural Research, National Institutes of Health; the Bill & Melinda Gates Foundation; and the Johns Hopkins University Center for AIDS Research.

Measles elimination: progress, challenges and implications for rubella control.

Measles and rubella are major vaccine-preventable causes of child mortality and disability. They have been eliminated from the Americas and some other regions have also come close to elimination. In this paper, we review regional progress toward measles and rubella control/elimination goals, describe the recent epidemiology of these infections and discuss challenges to achieving the goals. Globally, measles vaccination is estimated to prevent nearly 2 million deaths each year. Despite this remarkable progress, large measles outbreaks have occurred in recent years, often involving older persons who were not vaccinated in earlier years. Such an occurrence would be particularly damaging for rubella control programmes as it could lead to peaks in congenital rubella syndrome. Challenges to achieving and sustaining high vaccination coverage include civil conflict, weak health systems, geographic, cultural and economic barriers to reaching certain population groups and inadequate monitoring and use of data for action. Countries and regions aiming to eliminate measles and control rubella urgently need to improve the implementation and monitoring of both routine and mass vaccination campaign strategies.

Propensity score estimation: neural networks, support vector machines, decision trees (CART), and meta-classifiers as alternatives to logistic regression.

OBJECTIVE: Propensity scores for the analysis of observational data are typically estimated using logistic regression. Our objective in this review was to assess machine learning alternatives to logistic regression, which may accomplish the same goals but with fewer assumptions or greater accuracy. STUDY DESIGN AND SETTING: We identified alternative methods for propensity score estimation and/or classification from the public health, biostatistics, discrete mathematics, and computer science literature, and evaluated these algorithms for applicability to the problem of propensity score estimation, potential advantages over logistic regression, and ease of use. RESULTS: We identified four techniques as alternatives to logistic regression: neural networks, support vector machines, decision trees (classification and regression trees [CART]), and meta-classifiers (in particular, boosting). CONCLUSION: Although the assumptions of logistic regression are well understood, those assumptions are frequently ignored. All four alternatives have advantages and disadvantages compared with logistic regression. Boosting (meta-classifiers) and, to a lesser extent, decision trees (particularly CART), appear to be most promising for use in the context of propensity score analysis, but extensive simulation studies are needed to establish their utility in practice.

The cost of simplifying air travel when modeling disease spread.

BACKGROUND: Air travel plays a key role in the spread of many pathogens. Modeling the long distance spread of infectious disease in these cases requires an air travel model. Highly detailed air transportation models can be over determined and computationally problematic. We compared the predictions of a simplified air transport model with those of a model of all routes and assessed the impact of differences on models of infectious disease. METHODOLOGY/PRINCIPAL FINDINGS: Using U.S. ticket data from 2007, we compared a simplified "pipe" model, in which individuals flow in and out of the air transport system based on the number of arrivals and departures from a given airport, to a fully saturated model where all routes are modeled individually. We also compared the pipe model to a "gravity" model where the probability of travel is scaled by physical distance; the gravity model did not differ significantly from the pipe model. The pipe model roughly approximated actual air travel, but tended to overestimate the number of trips between small airports and underestimate travel between major east and west coast airports. For most routes, the maximum number of false (or missed) introductions of disease is small (<1 per day) but for a few routes this rate is greatly underestimated by the pipe model. CONCLUSIONS/SIGNIFICANCE: If our interest is in large scale regional and national effects of disease, the simplified pipe model may be adequate. If we are interested in specific effects of interventions on particular air routes or the time for the disease to reach a particular location, a more complex point-to-point model will be more accurate. For many problems a hybrid model that independently models some frequently traveled routes may be the best choice. Regardless of the model used, the effect of simplifications and sensitivity to errors in parameter estimation should be analyzed.