Temporally Varying Relative Risks for Infectious Diseases: Implications for Infectious Disease Control

Understanding the sources of efficacy dilution in a trial of a monthly dapivirine vaginal ring for HIV-1 prevention

INTRODUCTION

Women-initiated HIV - 1 prevention products are key to reducing women's HIV-1 risk. Clinical trials of vaginal microbicides have shown limited to no efficacy in intention-to-treat (ITT) analyses. It is hypothesized that these negative results are partly due to efficacy dilution.

METHODS

We developed a microsimulation model of MTN-020/ASPIRE, a phase 3 trial that evaluated monthly use of a dapivirine vaginal ring for HIV-1 prevention. We evaluated four sources of efficacy dilution: trial-level factors: (i) an imbalance in the number of monthly sex acts between study arms and (ii) heterogeneity in risk emergent over time; and individual-level factors: (iii) product non-adherence and (iv) receptive anal intercourse.

RESULTS

Assuming 70% per-vaginal exposure efficacy (consistent with the ITT estimate of 27%), heterogeneity in risk accounted for the largest proportion of efficacy dilution, at 42% (90% CrI: 38, 45), followed by non-adherence (33%; 90% CrI: 27, 39), an imbalance in arms (18%; 90% CrI: 16, 21) and lastly, anal intercourse with less than 10% of efficacy dilution.

CONCLUSION

Our results suggest that heterogeneity in risk was the most important source of efficacy dilution in the ASPIRE trial. Future trials of HIV-1 prevention products for women should consider alternative trial designs and analytic approaches that minimize bias introduced by heterogeneity in risk.

Spatial scales of COVID-19 transmission in Mexico

During outbreaks of emerging infectious diseases, internationally connected cities often experience large and early outbreaks, while rural regions follow after some delay. This hierarchical structure of disease spread is influenced primarily by the multiscale structure of human mobility. However, during the COVID-19 epidemic, public health responses typically did not take into consideration the explicit spatial structure of human mobility when designing nonpharmaceutical interventions (NPIs). NPIs were applied primarily at national or regional scales. Here, we use weekly anonymized and aggregated human mobility data and spatially highly resolved data on COVID-19 cases at the municipality level in Mexico to investigate how behavioral changes in response to the pandemic have altered the spatial scales of transmission and interventions during its first wave (March-June 2020). We find that the epidemic dynamics in Mexico were initially driven by exports of COVID-19 cases from Mexico State and Mexico City, where early outbreaks occurred. The mobility network shifted after the implementation of interventions in late March 2020, and the mobility network communities became more disjointed while epidemics in these communities became increasingly synchronized. Our results provide dynamic insights into how to use network science and epidemiological modeling to inform the spatial scale at which interventions are most impactful in mitigating the spread of COVID-19 and infectious diseases in general.

The role of place-based factors and other social determinants of health on adverse post-sepsis outcomes: a review of the literature

Sepsis remains a common and costly disease. With early recognition and guideline-based treatment, more patients are surviving to hospital discharge. Many survivors experience adverse health events in the months following discharge, while others suffer long-term physical and cognitive decline. Social, biological, and environmental factors affect all aspects of the disease process, from what pathogens one is exposed to, how/if disease develops, what avenues are available for treatment, as well as short- and long-term sequelae of survival. Disparities in sepsis care exist at all stages of a patient's clinical course, but increased survivorship has highlighted the extent to which Social Determinants of Health (SDoH) influence post-discharge adverse events. Despite increased interest in the last decade, a nuanced understanding of causal relationships remains elusive. This is due to several factors: the narrow range of social determinants of health (SDoH) variables typically studied, the inconsistent and non-standardized methods of documenting and reporting SDoH, and the inadequate acknowledgment of how social, environmental, and biological factors interact. Lack of clear understanding of how SDoH influence post- discharge outcomes is an obstacle to development and testing of strategies to mitigate their harms. This paper reviews the literature pertaining to the effects of SDoH on post-discharge outcomes in sepsis, highlights gaps therein, and identifies areas of greatest need for improving the quality and impact of future investigations.

Modeling the impact of racial and ethnic disparities on COVID-19 epidemic dynamics

Background

The impact of variable infection risk by race and ethnicity on the dynamics of SARS-CoV-2 spread is largely unknown.

Methods

Here, we fit structured compartmental models to seroprevalence data from New York State and analyze how herd immunity thresholds (HITs), final sizes, and epidemic risk change across groups.

Results

A simple model where interactions occur proportionally to contact rates reduced the HIT, but more realistic models of preferential mixing within groups increased the threshold toward the value observed in homogeneous populations. Across all models, the burden of infection fell disproportionately on minority populations: in a model fit to Long Island serosurvey and census data, 81% of Hispanics or Latinos were infected when the HIT was reached compared to 34% of non-Hispanic whites.

Conclusions

Our findings, which are meant to be illustrative and not best estimates, demonstrate how racial and ethnic disparities can impact epidemic trajectories and result in unequal distributions of SARS-CoV-2 infection.

Funding

K.C.M. was supported by National Science Foundation GRFP grant DGE1745303. Y.H.G. and M.L. were funded by the Morris-Singer Foundation. M.L. was supported by SeroNet cooperative agreement U01 CA261277.

Waning of Measured Influenza Vaccine Effectiveness Over Time: The Potential Contribution of Leaky Vaccine Effect

BACKGROUND

Several observational studies have shown decreases in measured influenza vaccine effectiveness (mVE) during influenza seasons. One study found decreases of 6-11%/month during the 2011-2012 to 2014-2015 seasons. These findings could indicate waning immunity but could also occur if vaccine effectiveness is stable and vaccine provides partial protection in all vaccinees ("leaky") rather than complete protection in a subset of vaccinees. Since it is unknown whether influenza vaccine is leaky, we simulated the 2011-2012 to 2014-2015 influenza seasons to estimate the potential contribution of leaky vaccine effect to the observed decline in mVE.

METHODS

We used available data to estimate daily numbers of vaccinations and infections with A/H1N1, A/H3N2, and B viruses. We assumed that vaccine effect was leaky, calculated mVE as 1 minus the Mantel-Haenszel relative risk of vaccine on incident cases, and determined the mean mVE change per 30 days since vaccination. Because change in mVE was highly dependent on infection rates, we performed simulations using low (15%) and high (31%) total (including symptomatic and asymptomatic) seasonal infection rates.

RESULTS

For the low infection rate, decreases (absolute) in mVE per 30 days after vaccination were 2% for A/H1N1 and 1% for A/H3N2and B viruses. For the high infection rate, decreases were 5% for A/H1N1, 4% for A/H3, and 3% for B viruses.

CONCLUSIONS

The leaky vaccine bias could account for some, but probably not all, of the observed intraseasonal decreases in mVE. These results underscore the need for strategies to deal with intraseasonal vaccine effectiveness decline.

Potential Biases Arising From Epidemic Dynamics in Observational Seroprotection Studies

The extent and duration of immunity following infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are critical outstanding questions about the epidemiology of this novel virus, and studies are needed to evaluate the effects of serostatus on reinfection. Understanding the potential sources of bias and methods for alleviating biases in these studies is important for informing their design and analysis. Confounding by individual-level risk factors in observational studies like these is relatively well appreciated. Here, we show how geographic structure and the underlying, natural dynamics of epidemics can also induce noncausal associations. We take the approach of simulating serological studies in the context of an uncontrolled or controlled epidemic, under different assumptions about whether prior infection does or does not protect an individual against subsequent infection, and using various designs and analytical approaches to analyze the simulated data. We find that in studies assessing whether seropositivity confers protection against future infection, comparing seropositive persons with seronegative persons with similar time-dependent patterns of exposure to infection by stratifying or matching on geographic location and time of enrollment is essential in order to prevent bias.

Transmissibility of influenza during the 21st-century epidemics, Spain, influenza seasons 2001/02 to 2017/18

BackgroundUnderstanding influenza seasonality is necessary for determining policies for influenza control.AimWe characterised transmissibility during seasonal influenza epidemics, including one influenza pandemic, in Spain during the 21th century by using the moving epidemic method (MEM) to calculate intensity levels and estimate differences across seasons and age groups.MethodsWe applied the MEM to Spanish Influenza Sentinel Surveillance System data from influenza seasons 2001/02 to 2017/18. A modified version of Goldstein's proxy was used as an epidemiological-virological parameter. We calculated the average starting week and peak, the length of the epidemic period and the length from the starting week to the peak of the epidemic, by age group and according to seasonal virus circulation.ResultsIndividuals under 15 years of age presented higher transmissibility, especially in the 2009 influenza A(H1N1) pandemic. Seasons with dominance/co-dominance of influenza A(H3N2) virus presented high intensities in older adults. The 2004/05 influenza season showed the highest influenza-intensity level for all age groups. In 12 seasons, the epidemic started between week 50 and week 3. Epidemics started earlier in individuals under 15 years of age (-1.8 weeks; 95% confidence interval (CI):-2.8 to -0.7) than in those over 64 years when influenza B virus circulated as dominant/co-dominant. The average time from start to peak was 4.3 weeks (95% CI: 3.6-5.0) and the average epidemic length was 8.7 weeks (95% CI: 7.9-9.6).ConclusionsThese findings provide evidence for intensity differences across seasons and age groups, and can be used guide public health actions to diminish influenza-related morbidity and mortality.

On the increasing role of older adolescents and younger adults during the SARS-CoV-2 epidemic in Mexico

Background

During the first months of the SARS-CoV-2 pandemic, Mexico implemented a national lockdown followed by post-lockdown mitigation.

Methods

We used daily number of SARS-CoV-2-confirmed hospitalizations (by date of symptom onset) to assess the changes in the incidence of individuals between the age of 10-59 years during the epidemic in Mexico. For each age group g, we computed the proportion E(g) of individuals in that age group among all cases aged 10-59y during the early lockdown period (April 20-May 3, 2020), and the corresponding proportion L(g) during the late lockdown period (May 18-31, 2020) and post-lockdown mitigation (June 15-28, 2020). For each later period (late lockdown or post-lockdown), we computed the proportion ratios relative to the early lockdown period PR(g)=L(g)/E(g). For each pair of age groups g1,g2, PR(g1)> PR(g2) is interpreted as a relative increase in SARS-CoV-2 infections in the age group g1 compared to g2 for the late lockdown and post-lockdown periods vs. the early lockdown period.

Results

For the late lockdown period, the highest PR estimates belong to persons aged 15-19y (PR=1.69(95%CI(1.05, 2.72))) and 20-24y (PR=1.43(1.10,1.86)). For the post-lockdown period, the highest PR estimates were also in age groups 15-19y (PR=2.05(1.30, 3.24)) and 20-24y (PR=1.49(1.15,1.93)). These estimates were higher in persons 15-24y compared to those ≥30y.

Conclusions

Our results suggest that adolescents and younger adults had an increased relative incidence during late lockdown and the post-lockdown mitigation periods. The role of these age groups during the epidemic should be considered when implementing future pandemic response efforts.

Lockdown measures and relative changes in the age-specific incidence of SARS-CoV-2 in Spain

During the first months of the severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) epidemic in 2020, Spain implemented an initial lockdown period on 15 March followed by a strengthened lockdown period on 30 March when only essential workers continued to commute to work. However, little is known about the epidemic dynamics in different age groups during these periods.We used the daily number of coronavirus 2019 cases (by date of symptom onset) reported to the National Epidemiological Surveillance Network among individuals aged 15-19 years through 65-69 years. For each age group g, we computed the proportion PrE(g) of individuals in age group g among all reported cases aged 15-69 years during the pre-lockdown period (1-10 March 2020) and the corresponding proportion PrL(g) during two lockdown periods (initial: 25 March-3 April; strengthened: 8-17 April 2020). For each lockdown period, we computed the proportion ratios PR(g) = PrL(g)/PrE(g). For each pair of age groups g1, g2, PR(g1)>PR(g2) implies a relative increase in the incidence of detected SARS-CoV-2 infection in the age group g1 compared with g2 for the lockdown period vs. the pre-lockdown period.For the initial lockdown period, the highest PR values were in age groups 50-54 years (PR = 1.21; 95% CI: 1.12,1.30) and 55-59 years (PR = 1.19; 1.11,1.27). For the second lockdown period, the highest PR values were in age groups 15-19 years (PR = 1.26; 0.95,1.68) and 50-54 years (PR = 1.20; 1.09,1.31).Our results suggest that different outbreak control measures led to different changes in the relative incidence by age group. During the initial lockdown period, when non-essential work was allowed, individuals aged 40-64 years, particularly those aged 50-59 years, had a higher relative incidence compared with the pre-lockdown period. Younger adults/older adolescents had an increased relative incidence during the later, strengthened lockdown. The role of different age groups during the epidemic should be considered when implementing future mitigation efforts.

Birth Cohort Studies Assessing Norovirus Infection and Immunity in Young Children: A Review

Globally, noroviruses are among the foremost causes of acute diarrheal disease, yet there are many unanswered questions on norovirus immunity, particularly following natural infection in young children during the first 2 years of life when the disease burden is highest. We conducted a literature review on birth cohort studies assessing norovirus infections in children from birth to early childhood. Data on infection, immunity, and risk factors are summarized from 10 community-based birth cohort studies conducted in low- and middle-income countries. Up to 90% of children experienced atleast one norovirus infection and up to 70% experienced norovirus-associated diarrhea, most often affecting children 6 months of age and older. Data from these studies help to fill critical knowledge gaps for vaccine development, yet study design and methodological differences limit comparison between studies, particularly for immunity and risk factors for disease. Considerations for conducting future birth cohort studies on norovirus are discussed.

Lockdown measures and relative changes in the age-specific incidence of SARS-CoV-2 in Spain

BACKGROUND

The first months of the SARS-CoV-2 epidemic in Spain resulted in high incidence and mortality. A national sero-epidemiological survey suggests higher cumulative incidence of infection in older individuals than in younger individuals. However, little is known about the epidemic dynamics in different age groups, including the relative effect of the lockdown measures introduced on March 15, and strengthened on March 30 to April 14, 2020 when only essential workers continued to work.

METHODS

We used data from the National Epidemiological Surveillance Network (RENAVE in Spanish) on the daily number of reported COVID-19 cases (by date of symptom onset) in eleven 5-year age groups: 15-19y through 65-69y. For each age group g, we computed the proportion E(g) of individuals in age group g among all reported cases aged 15-69y during the pre-lockdown period (March 1-10, 2020) and the corresponding proportion L(g) during two lockdown periods (March 25-April 3 and April 8-17, 2020). For each lockdown period, we computed the proportion ratios PR(g)= L(g)/E(g). For each pair of age groups g1,g2, PR(g1)>PR(g2) implies a relative increase in the incidence of detected SARS-CoV-2 infection in the age group g1 compared with g2 for the later vs. early period.

RESULTS

For the first lockdown period, the highest PR values were in age groups 50-54y (PR=1.21; 95% CI: 1.12,1.30) and 55-59y (PR=1.19; 1.11,1.27). For the second lockdown period, the highest PR values were in age groups 15-19y (PR=1.26; 0.95,1.68) and 50-54y (PR=1.20; 1.09,1.31).

CONCLUSIONS

Our results suggest that different outbreak control measures led to different changes in the relative incidence by age group. During the first lockdown period, when non-essential work was allowed, individuals aged 40-64y, particularly those aged 50-59y presented with higher COVID-19 relative incidence compared to pre-lockdown period, while younger adults/older adolescents (together with persons aged 50-59y) had increased relative incidence during the later, strengthened lockdown. The role of different age groups during the epidemic should be considered when implementing future mitigation efforts.

Targeted surveillance strategies for efficient detection of novel antibiotic resistance variants

Genotype-based diagnostics for antibiotic resistance represent a promising alternative to empiric therapy, reducing inappropriate antibiotic use. However, because such assays infer resistance based on known genetic markers, their utility will wane with the emergence of novel resistance. Maintenance of these diagnostics will therefore require surveillance to ensure early detection of novel resistance variants, but efficient strategies to do so remain undefined. We evaluate the efficiency of targeted sampling approaches informed by patient and pathogen characteristics in detecting antibiotic resistance and diagnostic escape variants in Neisseria gonorrhoeae, a pathogen associated with a high burden of disease and antibiotic resistance and the development of genotype-based diagnostics. We show that patient characteristic-informed sampling is not a reliable strategy for efficient variant detection. In contrast, sampling informed by pathogen characteristics, such as genomic diversity and genomic background, is significantly more efficient than random sampling in identifying genetic variants associated with resistance and diagnostic escape.

Potential biases arising from epidemic dynamics in observational seroprotection studies.. [PMID: 32511485 PMCID: PMC7273253 DOI: 10.1101/2020.05.02.20088765]

The extent and duration of immunity following SARS-CoV-2 infection are critical outstanding questions about the epidemiology of this novel virus, and studies are needed to evaluate the effects of serostatus on reinfection. Understanding the potential sources of bias and methods to alleviate biases in these studies is important for informing their design and analysis. Confounding by individual-level risk factors in observational studies like these is relatively well appreciated. Here, we show how geographic structure and the underlying, natural dynamics of epidemics can also induce noncausal associations. We take the approach of simulating serologic studies in the context of an uncontrolled or a controlled epidemic, under different assumptions about whether prior infection does or does not protect an individual against subsequent infection, and using various designs and analytic approaches to analyze the simulated data. We find that in studies assessing the efficacy of serostatus on future infection, comparing seropositive individuals to seronegative individuals with similar time-dependent patterns of exposure to infection, by stratifying or matching on geographic location and time of enrollment, is essential to prevent bias.

The effect of human mobility and control measures on the COVID-19 epidemic in China

The ongoing coronavirus disease 2019 (COVID-19) outbreak expanded rapidly throughout China. Major behavioral, clinical, and state interventions were undertaken to mitigate the epidemic and prevent the persistence of the virus in human populations in China and worldwide. It remains unclear how these unprecedented interventions, including travel restrictions, affected COVID-19 spread in China. We used real-time mobility data from Wuhan and detailed case data including travel history to elucidate the role of case importation in transmission in cities across China and to ascertain the impact of control measures. Early on, the spatial distribution of COVID-19 cases in China was explained well by human mobility data. After the implementation of control measures, this correlation dropped and growth rates became negative in most locations, although shifts in the demographics of reported cases were still indicative of local chains of transmission outside of Wuhan. This study shows that the drastic control measures implemented in China substantially mitigated the spread of COVID-19.

The effect of human mobility and control measures on the COVID-19 epidemic in China

The ongoing coronavirus disease 2019 (COVID-19) outbreak expanded rapidly throughout China. Major behavioral, clinical, and state interventions were undertaken to mitigate the epidemic and prevent the persistence of the virus in human populations in China and worldwide. It remains unclear how these unprecedented interventions, including travel restrictions, affected COVID-19 spread in China. We used real-time mobility data from Wuhan and detailed case data including travel history to elucidate the role of case importation in transmission in cities across China and to ascertain the impact of control measures. Early on, the spatial distribution of COVID-19 cases in China was explained well by human mobility data. After the implementation of control measures, this correlation dropped and growth rates became negative in most locations, although shifts in the demographics of reported cases were still indicative of local chains of transmission outside of Wuhan. This study shows that the drastic control measures implemented in China substantially mitigated the spread of COVID-19.

The effect of human mobility and control measures on the COVID-19 epidemic in China

The ongoing coronavirus disease 2019 (COVID-19) outbreak expanded rapidly throughout China. Major behavioral, clinical, and state interventions were undertaken to mitigate the epidemic and prevent the persistence of the virus in human populations in China and worldwide. It remains unclear how these unprecedented interventions, including travel restrictions, affected COVID-19 spread in China. We used real-time mobility data from Wuhan and detailed case data including travel history to elucidate the role of case importation in transmission in cities across China and to ascertain the impact of control measures. Early on, the spatial distribution of COVID-19 cases in China was explained well by human mobility data. After the implementation of control measures, this correlation dropped and growth rates became negative in most locations, although shifts in the demographics of reported cases were still indicative of local chains of transmission outside of Wuhan. This study shows that the drastic control measures implemented in China substantially mitigated the spread of COVID-19.

Temporal rise in the proportion of both younger adults and older adolescents among COVID-19 cases in Germany: evidence of lesser adherence to social distancing practices?

Background

There is uncertainty about the role of different age groups in propagating the SARS-CoV-2 epidemics in different countries, particularly under current social distancing practices.

Methods

We used the Robert Koch Institute data on weekly COVID-19 cases in different age groups in Germany. To minimize the effect of changes in healthcare seeking behavior (e.g. for older adults) and testing practices, we included the following eight 5-year age groups in the analyses: 10-14y through 45-49y. For each age group g, we considered the proportion PL(g) of individuals in age group g among all detected cases aged 10-49y during weeks 13-14, 2020 (later period), as well as corresponding proportion PE(g) for weeks 10-11, 2020 (early period), and defined the relative risk RR(g) for the age group g to be the ratio RR(g) = PL(g)/PE(g). For each pair of age groups g1, g2, a higher value of RR(g1) compared to RR(g2), or, alternatively, a value above 1 for the odds ratio OR(g1, g2) = RR(g1)/RR(g2) for a COVID-19 case to be in group g1 vs. g2 for the later vs. early periods is interpreted as the relative increase in the population incidence of SARS-Cov-2 in the age group g1 compared to g2 for the later vs. early period.

Results

The relative risk RR(g) was highest for individuals aged 20-24y (RR=1.4(95% CI (1.27,1.55))), followed by individuals aged 15-19y (RR=1.14(0.99,1.32)), aged 30-34y (RR=1.07(0.99,1.16)), aged 25-29y (RR= 1.06(0.98,1.15)), aged 35-39y (RR=0.95(0.87,1.03)), aged 40-44y (RR=0.9(0.83,0.98)), aged 45-49y (RR=0.83(0.77,0.89)) and aged 10-14y (RR=0.78(0.64,0.95)). For the age group 20-24y, the odds ratio relative to any other age group for a case to be during the later vs. early period was significantly above 1. For the age group 15-19y, the odds ratio relative to any other age group either above 35y or 10-14y for a case to be during the later vs. early period was significantly above 1.

Conclusions

The observed relative increase with time in the prevalence of individuals aged 15-34y (particularly those aged 20-24y) among detected COVID-19 cases in Germany is unlikely to be explained by increases in the likelihood of seeking medical care or the likelihood of being tested for individuals in those age groups compared to individuals aged 35-49y or 10-14y, and should be indicative of the actual increase in the prevalence of individuals aged 15-34y among SARS-CoV-2 infections in the German population. That increase likely reflects elevated mixing among individuals aged 15-34y (particularly those aged 20-24y) compared to other age groups, possibly due to lesser adherence to social distancing practices.

Temporal rise in the proportion of younger adults and older adolescents among coronavirus disease (COVID-19) cases following the introduction of physical distancing measures, Germany, March to April 2020

Using data on coronavirus disease (COVID-19) cases in Germany from the Robert Koch Institute, we found a relative increase with time in the prevalence in 15-34 year-olds (particularly 20-24-year-olds) compared with 35-49- and 10-14-year-olds (we excluded older and younger ages because of different healthcare seeking behaviour). This suggests an elevated role for that age group in propagating the epidemic following the introduction of physical distancing measures.

Depletion-of-susceptibles Bias in Analyses of Intra-season Waning of Influenza Vaccine Effectiveness

Bias arises in studies of waning vaccine effectiveness when higher-risk individuals are depleted from the at-risk population at different rates between study groups. We examined how this bias arises and how to avoid it. A reanalysis of data from California confirmed a finding of intra-season waning of influenza vaccine effectiveness.

The effect of human mobility and control measures on the COVID-19 epidemic in China

The ongoing COVID-19 outbreak has expanded rapidly throughout China. Major behavioral, clinical, and state interventions are underway currently to mitigate the epidemic and prevent the persistence of the virus in human populations in China and worldwide. It remains unclear how these unprecedented interventions, including travel restrictions, have affected COVID-19 spread in China. We use real-time mobility data from Wuhan and detailed case data including travel history to elucidate the role of case importation on transmission in cities across China and ascertain the impact of control measures. Early on, the spatial distribution of COVID-19 cases in China was well explained by human mobility data. Following the implementation of control measures, this correlation dropped and growth rates became negative in most locations, although shifts in the demographics of reported cases are still indicative of local chains of transmission outside Wuhan. This study shows that the drastic control measures implemented in China have substantially mitigated the spread of COVID-19.

Disparities in Zika Virus Testing and Incidence Among Women of Reproductive Age-New York City, 2016

CONTEXT

The New York City Department of Health and Mental Hygiene (NYC DOHMH) performs surveillance for reportable diseases, including Zika virus (ZIKV) infection and disease, to inform public health responses. Incidence rates of other mosquito-borne diseases related to international travel are associated with census tract poverty level in NYC, suggesting that high poverty areas might be at higher risk for ZIKV infections.

OBJECTIVES

We assessed ZIKV testing rates and incidence of travel-associated infection among reproductive age women in NYC to identify areas with high incidence and low testing rates and assess the effectiveness of public health interventions.

DESIGN

We analyzed geocoded ZIKV surveillance data collected by NYC DOHMH. Women aged 15 to 44 years tested during January-July 2016 (n = 4733) were assigned to census tracts, which we grouped by poverty level and quartile of the number of persons born in countries or territories with mosquito-borne ZIKV transmission as a proxy for risk of travel to these areas. We calculated crude ZIKV testing rates, incidence rates, and incidence rate ratios (IRRs).

SETTING

New York City.

RESULTS

Eight percent of patients (n = 376) tested had evidence of ZIKV infection. Cumulative incidence was higher both in areas with higher versus lower poverty levels (IRR = 2.4; 95% confidence interval [CI], 2.0-3.0) and in areas with the largest versus smallest populations of persons born in countries or territories with mosquito-borne ZIKV transmission (IRR = 11.3; 95% CI, 6.2-20.7). Initially, ZIKV testing rates were lowest in higher poverty areas with the largest populations of persons born in countries or territories with mosquito-borne ZIKV transmission (15/100 000), but following targeted interventions, testing rates were highest in these areas (80/100 000).

CONCLUSIONS

Geocoded data enabled us to identify communities with low testing but high ZIKV incidence rates, intervene to promote testing and reduce barriers to testing, and measure changes in testing rates.

How to make more from exposure data? An integrated machine learning pipeline to predict pathogen exposure

Predicting infectious disease dynamics is a central challenge in disease ecology. Models that can assess which individuals are most at risk of being exposed to a pathogen not only provide valuable insights into disease transmission and dynamics but can also guide management interventions. Constructing such models for wild animal populations, however, is particularly challenging; often only serological data are available on a subset of individuals and nonlinear relationships between variables are common. Here we provide a guide to the latest advances in statistical machine learning to construct pathogen-risk models that automatically incorporate complex nonlinear relationships with minimal statistical assumptions from ecological data with missing data. Our approach compares multiple machine learning algorithms in a unified environment to find the model with the best predictive performance and uses game theory to better interpret results. We apply this framework on two major pathogens that infect African lions: canine distemper virus (CDV) and feline parvovirus. Our modelling approach provided enhanced predictive performance compared to more traditional approaches, as well as new insights into disease risks in a wild population. We were able to efficiently capture and visualize strong nonlinear patterns, as well as model complex interactions between variables in shaping exposure risk from CDV and feline parvovirus. For example, we found that lions were more likely to be exposed to CDV at a young age but only in low rainfall years. When combined with our data calibration approach, our framework helped us to answer questions about risk of pathogen exposure that are difficult to address with previous methods. Our framework not only has the potential to aid in predicting disease risk in animal populations, but also can be used to build robust predictive models suitable for other ecological applications such as modelling species distribution or diversity patterns.

Epidemiological analysis of respiratory and intestinal infectious diseases in three counties of Sichuan: the baseline survey of Disaster Mitigation Demonstration Area in western China

Background

Natural disasters can indirectly induce epidemics of infectious diseases through air and water pollution, accelerated pathogen reproduction, and population migration. This study aimed to explore the epidemiological characteristics of the main infectious diseases in Sichuan, a province with a high frequency of natural disasters.

Methods

Data were collected from the local Centers for Disease Control infectious disease reports from Lu, Shifang and Yuexi counties from 2011 to 2015 and from the baseline survey of the Disaster Mitigation Demonstration Area in Western China in 2016. Principal component regression was used to explore the main influencing factors of respiratory infectious diseases (RIDs).

Results

The incidence rates of RIDs and intestinal infectious diseases (IIDs) in 2015 were 78.99/100,000, 125.53/100,000, 190.32/100,000 and 51.70/100,000, 206.00/100,000, 69.16/100,000 in Lu, Shifang and Yuexi respectively. The incidence rates of pulmonary tuberculosis (TB) was the highest among RIDs in the three counties. The main IIDs in Lu and Shifang were hand-foot-mouth disease (HFMD) and other infectious diarrhea; however, the main IIDs in Yuexi was bacillary dysentery. The proportions of illiterate and ethnic minorities and per capita disposable income were the top three influencing factors of RIDs.

Conclusions

TB was the key point of RIDs prevention among the three counties. The key preventable IIDs in Lu and Shifang were HFMD and other infectious diarrhea, and bacillary dysentery was the major IIDs in Yuexi. The incidence rates of RIDs was associated with the population composition, the economy and personal hygiene habits.

On the role of different age groups during pertussis epidemics in California, 2010 and 2014

There is limited information on the roles of different age groups in propagating pertussis outbreaks, and the temporal changes in those roles since the introduction of acellular pertussis vaccines. The relative roles of different age groups in propagating the 2010 and the 2014 pertussis epidemics in California were evaluated using the relative risk (RR) statistic that measures the change in the group's proportion among all detected cases before vs. after the epidemic peak. For the 2010–11 epidemic, evidence for a predominant transmission age group was weak, with the largest RR estimates being 1.26 (95% CI 1.08–1.46) (aged 11–13 years); 1.19 (1.01–1.4) (aged 9–10 years); 1.17 (0.86–1.59) (aged 14–15 years); 1.12 (0.86–1.46) (aged 16–19 years) and 1.1 (0.89–1.36) (aged 7–8 years). The 2014 epidemic showed a strong signal of the role of older adolescents, with the highest RR estimate being in those aged 14–15 years (RR = 1.83, 1.61–2.07), followed by adolescents aged 16–19 years (RR = 1.41, 1.24–1.61) and 11–13 years (RR = 1.26, 1.12–1.41), with lower RR estimates in other age groups. As the time following introduction of acellular pertussis vaccines in California progressed, older adolescents played an increasing role in transmission during the major pertussis outbreaks. Booster pertussis vaccination for older adolescents with vaccines effective against pertussis transmission should be considered with the aim of mitigating future pertussis epidemics in the community.

On the Relative Role of Different Age Groups During Epidemics Associated With Respiratory Syncytial Virus

Background

While circulation of respiratory syncytial virus (RSV) results in high rates of hospitalization, particularly among young children and elderly individuals, little is known about the role of different age groups in propagating annual RSV epidemics.

Methods

We evaluate the roles played by individuals in different age groups during RSV epidemics in the United States between 2001 and 2012, using the previously defined relative risk (RR) statistic estimated from the hospitalization data from the Healthcare Cost and Utilization Project. Transmission modeling was used to examine the robustness of our inference method.

Results

Children aged 3-4 years and 5-6 years each had the highest RR estimate for 5 of 11 seasons included in this study, with RSV hospitalization rates in infants being generally higher during seasons when children aged 5-6 years had the highest RR estimate. Children aged 2 years had the highest RR estimate during one season. RR estimates in infants and individuals aged ≥11 years were mostly lower than in children aged 1-10 years. Highest RR values aligned with groups for which vaccination had the largest impact on epidemic dynamics in most model simulations.

Conclusions

Our estimates suggest the prominent relative roles of children aged ≤10 years (particularly among those aged 3-6 years) in propagating RSV epidemics. These results, combined with further modeling work, should help inform RSV vaccination policies.

Real-time analysis of the diphtheria outbreak in forcibly displaced Myanmar nationals in Bangladesh

BACKGROUND

Between August and December 2017, more than 625,000 Rohingya from Myanmar fled into Bangladesh, settling in informal makeshift camps in Cox's Bazar district and joining 212,000 Rohingya already present. In early November, a diphtheria outbreak hit the camps, with 440 reported cases during the first month. A rise in cases during early December led to a collaboration between teams from Médecins sans Frontières-who were running a provisional diphtheria treatment centre-and the London School of Hygiene and Tropical Medicine with the goal to use transmission dynamic models to forecast the potential scale of the outbreak and the resulting resource needs.

METHODS

We first adjusted for delays between symptom onset and case presentation using the observed distribution of reporting delays from previously reported cases. We then fit a compartmental transmission model to the adjusted incidence stratified by age group and location. Model forecasts with a lead time of 2 weeks were issued on 12, 20, 26 and 30 December and communicated to decision-makers.

RESULTS

The first forecast estimated that the outbreak would peak on 19 December in Balukhali camp with 303 (95% posterior predictive interval 122-599) cases and would continue to grow in Kutupalong camp, requiring a bed capacity of 316 (95% posterior predictive interval (PPI) 197-499). On 19 December, a total of 54 cases were reported, lower than forecasted. Subsequent forecasts were more accurate: on 20 December, we predicted a total of 912 cases (95% PPI 367-2183) and 136 (95% PPI 55-327) hospitalizations until the end of the year, with 616 cases actually reported during this period.

CONCLUSIONS

Real-time modelling enabled feedback of key information about the potential scale of the epidemic, resource needs and mechanisms of transmission to decision-makers at a time when this information was largely unknown. By 20 December, the model generated reliable forecasts and helped support decision-making on operational aspects of the outbreak response, such as hospital bed and staff needs, and with advocacy for control measures. Although modelling is only one component of the evidence base for decision-making in outbreak situations, suitable analysis and forecasting techniques can be used to gain insights into an ongoing outbreak.

Effects of the rotavirus vaccine program across age groups in the United States: analysis of national claims data, 2001-2016

BACKGROUND

The direct effectiveness of infant rotavirus vaccination implemented in 2006 in the United States has been evaluated extensively, however, understanding of population-level vaccine effectiveness (VE) is still incomplete.

METHODS

We analyzed time series data on rotavirus gastroenteritis (RVGE) and all-cause acute gastroenteritis (AGE) hospitalization rates in the United States from the MarketScan® Research Databases for July 2001-June 2016. Individuals were grouped into ages 0-4, 5-9, 10-14, 15-24, 25-44, and 45-64 years. Negative binomial regression models were fitted to monthly RVGE and AGE data to estimate the direct, indirect, overall, and total VE.

RESULTS

A total of 9211 RVGE and 726,528 AGE hospitalizations were analyzed. Children 0-4 years of age had the largest declines in RVGE hospitalizations with direct VE of 87% (95% CI: 83, 90%). Substantial indirect effects were observed across age groups and generally declined in each older group. Overall VE against RVGE hospitalizations for all ages combined was 69% (95% CI: 62, 76%). Total VE was highest among young children; a vaccinated child in the post-vaccine era has a 95% reduced risk of RVGE hospitalization compared to a child in the pre-vaccine era. We observed higher direct VE in odd post-vaccine years and an opposite pattern for indirect VE.

CONCLUSIONS

Vaccine benefits extended to unvaccinated individuals in all age groups, suggesting infants are important drivers of disease transmission across the population. Imperfect disease classification and changing disease incidence may lead to bias in observed direct VE.

TRIAL REGISTRATION

Not applicable.

Periods of high dengue transmission defined by rainfall do not impact efficacy of dengue vaccine in regions of endemic disease

OBJECTIVE

To evaluate the association of rainy season with overall dengue disease incidence and with the efficacy of the Sanofi Pasteur recombinant, live, attenuated, tetravalent vaccine (CYD-TDV) in two randomized, controlled multicenter phase III clinical trials in Asia and Latin America.

METHODS

Rainy seasons were defined for each study site using climatological information from the World Meteorological Organization. The dengue attack rate in the placebo group for each study month was calculated as the number of symptomatic, virologically-confirmed dengue events in a given month divided by the number of participants at risk in the same month. Time-dependent Cox proportional hazard models were used to test whether rainy season was associated with dengue disease and whether it modified vaccine efficacy in each of the two trials and in both of the trials combined.

FINDINGS

Rainy season, country, and age were all significantly associated with dengue disease in both studies. Vaccine efficacy did not change during the rainy season in any of the analyses.

CONCLUSIONS

Although dengue transmission and exposure are expected to increase during the rainy season, our results indicate that CYD-TDV vaccine efficacy remains constant throughout the year in endemic regions.

On the Role of Different Age Groups and Pertussis Vaccines During the 2012 Outbreak in Wisconsin

Background

There is limited information on the roles of different age groups in propagating pertussis outbreaks, and on the impact of vaccination on pertussis transmission in the community.

Methods

The relative roles of different age groups in propagating the 2012 pertussis outbreak in Wisconsin were evaluated using the relative risk (RR) statistic that measures the change in the group’s proportion among all detected cases before vs after the epidemic peak. The impact of vaccination in different age groups against infection (that is potentially different from the protective effect against detectable disease) was evaluated using the odds ratios (ORs), within each age group, for being vaccinated vs undervaccinated before vs after the outbreak’s peak.

Results

The RR statistic suggests that children aged 13–14 years played the largest relative role during the outbreak’s ascent (with estimates consistent across the 3 regions in Wisconsin that were studied), followed by children aged 7–8, 9–10, and 11–12 years. Young children and older teenagers and adults played more limited relative roles during the outbreak. Results of the vaccination status analysis for the fifth dose of DTaP (for children aged 7–8 years: OR, 0.44; 95% confidence interval [CI], 0.23–0.86; for children aged 9–10 years: OR, 0.51; 95% CI, 0.27–0.95); and for Tdap for children aged 13–14 years (OR, 0.38, 95% CI, 0.16–0.89) are consistent with protective effect against infection.

Conclusions

While our epidemiological findings for the fifth dose of DTaP and for Tdap are consistent with protective effect against infection, further studies, including those estimating vaccine effectiveness against infection/transmission to others particularly for pertussis vaccines for adolescents, are needed to evaluate the impact of vaccination on the spread of pertussis in the community.

Risk ratios for contagious outcomes

Epidemiologists commonly use the risk ratio to summarize the relationship between a binary covariate and outcome, even when outcomes may be dependent. Investigations of transmissible diseases in clusters-households, villages or small groups-often report risk ratios. Epidemiologists have warned that risk ratios may be misleading when outcomes are contagious, but the nature of this error is poorly understood. In this study, we assess the meaning of the risk ratio when outcomes are contagious. We provide a mathematical definition of infectious disease transmission within clusters, based on the canonical stochastic susceptible-infective model. From this characterization, we define the individual-level ratio of instantaneous infection risks as the inferential target, and evaluate the properties of the risk ratio as an approximation of this quantity. We exhibit analytically and by simulation the circumstances under which the risk ratio implies an effect whose direction is opposite that of the true effect of the covariate. In particular, the risk ratio can be greater than one even when the covariate reduces both individual-level susceptibility to infection, and transmissibility once infected. We explain these findings in the epidemiologic language of confounding and Simpson's paradox, underscoring the pitfalls of failing to account for transmission when outcomes are contagious.