Influenza-like Illness Reports Research Articles

Decreased Seasonal Influenza Rates Detected in a Crowdsourced Influenza-Like Illness Surveillance System During the COVID-19 Pandemic: Prospective Cohort Study.

Seasonal respiratory viruses had lower incidence during their 2019-2020 and 2020-2021 seasons, which overlapped with the COVID-19 pandemic. The widespread implementation of precautionary measures to prevent transmission of SARS-CoV-2 has been seen to also mitigate transmission of seasonal influenza. The COVID-19 pandemic also led to changes in care seeking and access. Participatory surveillance systems have historically captured mild illnesses that are often missed by surveillance systems that rely on encounters with a health care provider for detection. This study aimed to assess if a crowdsourced syndromic surveillance system capable of detecting mild influenza-like illness (ILI) also captured the globally observed decrease in ILI in the 2019-2020 and 2020-2021 influenza seasons, concurrent with the COVID-19 pandemic. Flu Near You (FNY) is a web-based participatory syndromic surveillance system that allows participants in the United States to report their health information using a brief weekly survey. Reminder emails are sent to registered FNY participants to report on their symptoms and the symptoms of household members. Guest participants may also report. ILI was defined as fever and sore throat or fever and cough. ILI rates were determined as the number of ILI reports over the total number of reports and assessed for the 2016-2017, 2017-2018, 2018-2019, 2019-2020, and 2020-2021 influenza seasons. Baseline season (2016-2017, 2017-2018, and 2018-2019) rates were compared to the 2019-2020 and 2020-2021 influenza seasons. Self-reported influenza diagnosis and vaccination status were captured and assessed as the total number of reported events over the total number of reports submitted. CIs for all proportions were calculated via a 1-sample test of proportions. ILI was detected in 3.8% (32,239/848,878) of participants in the baseline seasons (2016-2019), 2.58% (7418/287,909) in the 2019-2020 season, and 0.27% (546/201,079) in the 2020-2021 season. Both influenza seasons that overlapped with the COVID-19 pandemic had lower ILI rates than the baseline seasons. ILI decline was observed during the months with widespread implementation of COVID-19 precautions, starting in February 2020. Self-reported influenza diagnoses decreased from early 2020 through the influenza season. Self-reported influenza positivity among ILI cases varied over the observed time period. Self-reported influenza vaccination rates in FNY were high across all observed seasons. A decrease in ILI was detected in the crowdsourced FNY surveillance system during the 2019-2020 and 2020-2021 influenza seasons, mirroring trends observed in other influenza surveillance systems. Specifically, the months within seasons that overlapped with widespread pandemic precautions showed decreases in ILI and confirmed influenza. Concerns persist regarding respiratory pathogens re-emerging with changes to COVID-19 guidelines. Traditional surveillance is subject to changes in health care behaviors. Systems like FNY are uniquely situated to detect disease across disease severity and care seeking, providing key insights during public health emergencies.

A combination of annual and nonannual forces drive respiratory disease in the tropics

IntroductionIt is well known that influenza and other respiratory viruses are wintertime-seasonal in temperate regions. However, respiratory disease seasonality in the tropics is less well understood. In this study, we...

Comparison of Influenza-Like Illness (ILI) incidence data from the novel LeCellPHIA participatory surveillance system with COVID-19 case count data, Lesotho, July 2020 – July 2021

BackgroundWhile laboratory testing for infectious diseases such as COVID-19 is the surveillance gold standard, it is not always feasible, particularly in settings where resources are scarce. In the small country of Lesotho, located in sub-Saharan Africa, COVID-19 testing has been limited, thus surveillance data available to local authorities are limited. The goal of this study was to compare a participatory influenza-like illness (ILI) surveillance system in Lesotho with COVID-19 case count data, and ultimately to determine whether the participatory surveillance system adequately estimates the case count data.MethodsA nationally-representative sample was called on their mobile phones weekly to create an estimate of incidence of ILI between July 2020 and July 2021. Case counts from the website Our World in Data (OWID) were used as the gold standard to which our participatory surveillance data were compared. We calculated Spearman’s and Pearson’s correlation coefficients to compare the weekly incidence of ILI reports to COVID-19 case count data.ResultsOver course of the study period, an ILI symptom was reported 1,085 times via participatory surveillance for an average annual cumulative incidence of 45.7 per 100 people (95% Confidence Interval [CI]: 40.7 – 51.4). The cumulative incidence of reports of ILI symptoms was similar among males (46.5, 95% CI: 39.6 – 54.4) and females (45.1, 95% CI: 39.8 – 51.1). There was a slightly higher annual cumulative incidence of ILI among persons living in peri-urban (49.5, 95% CI: 31.7 – 77.3) and urban settings compared to rural areas. The January peak of the participatory surveillance system ILI estimates correlated significantly with the January peak of the COVID-19 case count data (Spearman’s correlation coefficient = 0.49; P < 0.001) (Pearson’s correlation coefficient = 0.67; P < 0.0001).ConclusionsThe ILI trends captured by the participatory surveillance system in Lesotho mirrored trends of the COVID-19 case count data from Our World in Data. Public health practitioners in geographies that lack the resources to conduct direct surveillance of infectious diseases may be able to use cell phone-based data collection to monitor trends.

Changes in epidemic intensity of influenza during 2014-2020 in Shanghai

Objective: To evaluate the performance of the influenza surveillance network and compare the epidemic intensity of influenza during 2014-2020 in Shanghai. Methods: Based on the weekly reports of influenza-like illness (ILI) and laboratory-confirmed influenza cases from January 1, 2014 to December 31, 2020. This study first evaluated the data reporting and specimen collection of ILI cases for each sentinel hospital, and then calculated the percentage of ILI (ILI%), the proportion of specimens tested positive for influenza, and the incidence of influenza among all ILI outpatient and emergency visits to measure the epidemic intensity of influenza. Finally, seasonal autoregressive integrated moving average (ARIMA) model was applied to quantify the changes in epidemic intensity of influenza in 2020. Results: The proportion of influenza surveillance sentinel hospitals with a score of less than 5 in the evaluation of ILI data reporting and samples collection were 9.68% and 21.05% in 2020 in Shanghai, respectively. ILI% was estimated to be 1.51% (95%CI: 1.50%-1.51%) and 2.31% (95%CI: 2.30%-2.32%), respectively for 2014-2019 and 2020; the proportion of specimens tested positive was 24.27% (95%CI: 24.02%- 24.51%) and 7.15% (95%CI: 6.78%-7.54%), respectively; and the incidence of influenza was 3.66‰ (95%CI: 3.62‰-3.70‰) and 1.65‰ (95%CI: 1.57‰-1.74‰), respectively. ARIMA model showed that ILI% was increased by 45.25% in 2020 in Shanghai, and the proportion of specimens tested positive and the incidence of influenza were reduced by 78.45% and 51.80%, respectively. Conclusions: In 2020, the performance of influenza surveillance system has changed, ILI% has increased, the proportion of specimens tested positive and the incidence of influenza has decreased in Shanghai. The change in the quality of influenza surveillance is also a potential factor affecting the epidemic intensity of influenza. In the future, the quality control of influenza surveillance network still needs to be further strengthened.

Time Series Analysis of Congestive Heart Failure Discharges in Florida (USA) Post Tropical Cyclones.

The aim of this study was to analyze congestive heart failure (CHF) discharges in Florida (USA) post tropical cyclones from 2007 through 2017. This was a retrospective longitudinal time series analysis of hospital CHF quarterly discharges across Florida using the Healthcare Cost and Utilization Project (HCUP) database. The autoregressive integrated moving average (ARIMA) model was used with correlated seasonal regressor variables such as cyclone frequency, maximum cyclone wind speed, average temperature, and reports of influenza-like illness (ILI). A total of 3,372,993 patients were identified, with average age in each quarter ranging 72.2 to 73.9 years and overall mortality ranging 4.3% to 6.4%. The CHF discharges within each year peaked from October through December and nadired from April through June with an increasing overall time trend. Significant correlation was found between CHF discharge and the average temperature (P <.001), with approximately 331.8 less CHF discharges (SE = 91.7) per degree of increase in temperature. However, no significant correlation was found between CHF discharges and frequency of cyclones, the maximum wind speed, and reported ILI. This study suggests that with the current methods and the HCUP dataset, there is no significant increase in overall CHF discharges in Florida as a result of recent previous cyclone occurrences.

2204. Pathogen Co-infections and Trends in Influenza-like Illness in PAIVED

Abstract Background The Pragmatic Assessment of Influenza Vaccine Effectiveness in the Department of Defense (DoD) (PAIVED) is a multicenter, multiservice study assessing influenza vaccine effectiveness in active-duty service members, retirees, and dependents. In its fourth season (2021/22), PAIVED offers a unique opportunity to examine influenza-like illness (ILI) trends prior to and during the COVID-19 pandemic in a prospectively followed, well-defined cohort. Methods Over the past 4 influenza seasons, PAIVED has enrolled DoD beneficiaries who were randomized to receive egg-based, cell-based, or recombinant-derived influenza vaccine. Participants provided some basic demographic information and were then sent a weekly text or email that inquired about ILI symptoms, defined as 1) having cough or sore throat, plus 2) feeling feverish/having chills or having body aches/fatigue. Participants with ILI completed a daily symptom diary for one week and submitted a nasal swab for PCR-based pathogen detection. Table 1.Demographic characteristics of PAIVED participants over four seasonsFigure 1.Percent of PAIVED participants with influenza-like illness, SARS-CoV-2, influenza, and rhinovirus identified in swab samples collected over four seasons. Results Over the 4 seasons, 15,449 participants were followed for ILI (Table 1) with 3,407 participants reporting a total of 3,985 ILIs. For the 2021/22 season, ILI reports peaked in January (Figure 1). Overall, 4.7% of episodes had more than one pathogen identified (Table 2). Among the 122 coinfections identified to date, most were coinfections with rhinoviruses (91/122, 75%), including rhinovirus coinfections with seasonal coronaviruses (29, 24%), metapneumovirus (18, 15%), SARS-CoV-2 (17, 14%), and influenza (14, 11%). SARS-CoV-2 and influenza were found together in one sample. The lab data will continue to be processed for the current season (2021/22). Table 2.Pathogens identified in PAIVED nasal swabs over four seasons Conclusion ILI rates were lowest during the third year (2020/21), consistent with national influenza surveillance reports of influenza and outpatient ILI activity, suggesting that measures taken to reduce transmission of SARS-CoV-2 reduced the spread of other respiratory viruses. The emergence of the SARS-CoV-2 omicron variant in December 2021 was associated with higher ILI rates. Among those individuals for whom a sample was collected, coinfections were highest in 2018/19. Data collection and specimen analysis are ongoing for 2021/22. Disclosures Jitendrakumar Modi, MD, GlaxoSmithKline: I am a paid speaker for GSK. I do not speak for their flu brand. Timothy Burgess, MD, MPH, AstraZeneca: The HJF, in support of the USU IDCRP, was funded to conduct or augment unrelated Phase III Mab and vaccine trials as part of US Govt. COVID19 response.

Addressing delayed case reporting in infectious disease forecast modeling.

Infectious disease forecasting is of great interest to the public health community and policymakers, since forecasts can provide insight into disease dynamics in the near future and inform interventions. Due to delays in case reporting, however, forecasting models may often underestimate the current and future disease burden.In this paper, we propose a general framework for addressing reporting delay in disease forecasting efforts with the goal of improving forecasts. We propose strategies for leveraging either historical data on case reporting or external internet-based data to estimate the amount of reporting error. We then describe several approaches for adapting general forecasting pipelines to account for under- or over-reporting of cases. We apply these methods to address reporting delay in data on dengue fever cases in Puerto Rico from 1990 to 2009 and to reports of influenza-like illness (ILI) in the United States between 2010 and 2019. Through a simulation study, we compare method performance and evaluate robustness to assumption violations. Our results show that forecasting accuracy and prediction coverage almost always increase when correction methods are implemented to address reporting delay. Some of these methods required knowledge about the reporting error or high quality external data, which may not always be available. Provided alternatives include excluding recently-reported data and performing sensitivity analysis. This work provides intuition and guidance for handling delay in disease case reporting and may serve as a useful resource to inform practical infectious disease forecasting efforts.

Report of the National Influenza Surveillance Scheme, 2011 to 2018.

This report describes influenza surveillance activities in Australia for the period 2011 to 2018. Data were extracted from several sources constituting the National Influenza Surveillance Scheme (NISS). Laboratory-confirmed influenza notification rates (per 100,000 population) increased from 122 in 2011 to 1,021 in 2017, before declining to 235 in 2018. The highest laboratory-confirmed notification rates during the eight-year period were from the smaller jurisdictions (South Australia and the Northern Territory), except in 2016 when Queensland reported the highest rate. Similar trends were observed in community reports of influenza-like illness (ILI), presentations of ILI to sentinel general practice (GP) sites, and influenza hospitalisations. Children aged 14 years or younger, and adults 65 years of age or older, had the highest notification rates of laboratory-confirmed influenza. Adults aged 65 years or older and patients with comorbidities had higher rates of influenza-associated hospitalisations and mortality. Over half of eligible patients admitted to sentinel hospitals (57%) received oseltamivir treatment, with 17% receiving the treatment within 48 hours of symptom onset. Influenza type A predominated over the eight years, except in 2015 when type B predominated. This trend was consistent with Australian World Health Organization Collaborating Centre (WHOCC) data for influenza isolates tested. Of influenza viruses circulating during the reporting period, A(H1N1) viruses were mostly antigenically similar to the vaccine strain A/California/7/2009 (H1N1), except in 2017 and 2018 when they were mostly similar to A/Michigan/45/2015. The A(H3N2) strains varied over the years but included the vaccine strains A/Perth/16/2009, A/Switzerland/9715293/2013, A/Hong Kong/4801/2014, and A/Singapore/INFIMH/2016. The B/Victoria/2/87 lineage (represented by the 2016 vaccine strain B/Brisbane/60/2008) and the B/Yamagata/16/88 lineage (represented by the southern hemisphere 2015 vaccine strain B/Phuket/3073/2013) were the circulating influenza B viruses during the reporting period. Influenza B accounted for just under a third of notifications (32%) from 2011 to 2018. Fifty-four per cent of influenza B viruses characterised by the Australian WHOCC site during the eight-year period were from the Yamagata lineage, although the proportion was higher (67%) when analysing the most recent four years (2015 to 2018). In contrast, during the 2010 season, 99% of all influenza B viruses characterised by the Australian WHOCC site were in the B-Victoria lineage. During the 2018 season the Australian WHOCC site detected, for the first time, swine A(H3N2)v virus from a human patient in Australia, highlighting the need to maintain vigilance for zoonotic infections.

Estimating the cumulative incidence of COVID-19 in the United States using influenza surveillance, virologic testing, and mortality data: Four complementary approaches.

Effectively designing and evaluating public health responses to the ongoing COVID-19 pandemic requires accurate estimation of the prevalence of COVID-19 across the United States (US). Equipment shortages and varying testing capabilities have however hindered the usefulness of the official reported positive COVID-19 case counts. We introduce four complementary approaches to estimate the cumulative incidence of symptomatic COVID-19 in each state in the US as well as Puerto Rico and the District of Columbia, using a combination of excess influenza-like illness reports, COVID-19 test statistics, COVID-19 mortality reports, and a spatially structured epidemic model. Instead of relying on the estimate from a single data source or method that may be biased, we provide multiple estimates, each relying on different assumptions and data sources. Across our four approaches emerges the consistent conclusion that on April 4, 2020, the estimated case count was 5 to 50 times higher than the official positive test counts across the different states. Nationally, our estimates of COVID-19 symptomatic cases as of April 4 have a likely range of 2.3 to 4.8 million, with possibly as many as 7.6 million cases, up to 25 times greater than the cumulative confirmed cases of about 311,000. Extending our methods to May 16, 2020, we estimate that cumulative symptomatic incidence ranges from 4.9 to 10.1 million, as opposed to 1.5 million positive test counts. The proposed combination of approaches may prove useful in assessing the burden of COVID-19 during resurgences in the US and other countries with comparable surveillance systems.

Associations between weather-related data and influenza reports: A pilot study and related policy implications

AimThe purpose of this retrospective, correlational pilot study was to explore the relationship between historical weekly weather data including temperature, dew point, humidity, barometric pressure, visibility, and cloud cover compared to weekly influenza-like illness reports over a four year period. BackgroundClimate and weather-related conditions may affect the viral activity and transmission of influenza, although this relationship has not been widely studied in nursing. Some research suggests that there are causal links between cold temperatures, low indoor humidity, minimal sun exposure, and influenza outbreaks. Additionally, rapid weather variability in a warming climate can increase influenza epidemic risk. MethodsData from a local public health district were extracted and used to correlate with weekly weather averages for the area. ResultsFindings showed that current influenza reports are significantly associated with temperature and visibility, both lagged two weeks. ConclusionsThough more research is needed, nurses must understand, recognize, and act upon weather and climate factors that affect the health of populations. With a greater understanding of the relationship between weather and influenza-like illness, nurses and other healthcare providers can potentially work to respond to and mitigate the consequences of weather-related illness as well as anticipate and prepare for increased flu burden. Furthermore, nurses can remain engaged in climate protective initiatives and policy development at their local community and/or organizational levels to underscore and advocate for the needs of populations and groups they serve.

Differences in Regional Patterns of Influenza Activity Across Surveillance Systems in the United States: Comparative Evaluation.

BackgroundThe Centers for Disease Control and Prevention (CDC) tracks influenza-like illness (ILI) using information on patient visits to health care providers through the Outpatient Influenza-like Illness Surveillance Network (ILINet). As participation in this system is voluntary, the composition, coverage, and consistency of health care reports vary from state to state, leading to different measures of ILI activity between regions. The degree to which these measures reflect actual differences in influenza activity or systematic differences in the methods used to collect and aggregate the data is unclear.ObjectiveThe objective of our study was to qualitatively and quantitatively compare national and region-specific ILI activity in the United States across 4 surveillance data sources—CDC ILINet, Flu Near You (FNY), athenahealth, and HealthTweets.org—to determine whether these data sources, commonly used as input in influenza modeling efforts, show geographical patterns that are similar to those observed in CDC ILINet’s data. We also compared the yearly percentage of FNY participants who sought health care for ILI symptoms across geographical areas.MethodsWe compared the national and regional 2018-2019 ILI activity baselines, calculated using noninfluenza weeks from previous years, for each surveillance data source. We also compared measures of ILI activity across geographical areas during 3 influenza seasons, 2015-2016, 2016-2017, and 2017-2018. Geographical differences in weekly ILI activity within each data source were also assessed using relative mean differences and time series heatmaps. National and regional age-adjusted health care–seeking percentages were calculated for each influenza season by dividing the number of FNY participants who sought medical care for ILI symptoms by the total number of ILI reports within an influenza season. Pearson correlations were used to assess the association between the health care–seeking percentages and baselines for each surveillance data source.ResultsWe observed consistent differences in ILI activity across geographical areas for CDC ILINet and athenahealth data. ILI activity for FNY displayed little variation across geographical areas, whereas differences in ILI activity for HealthTweets.org were associated with the total number of tweets within a geographical area. The percentage of FNY participants who sought health care for ILI symptoms differed slightly across geographical areas, and these percentages were positively correlated with CDC ILINet and athenahealth baselines.ConclusionsOur findings suggest that differences in ILI activity across geographical areas as reported by a given surveillance system may not accurately reflect true differences in the prevalence of ILI. Instead, these differences may reflect systematic collection and aggregation biases that are particular to each system and consistent across influenza seasons. These findings are potentially relevant in the real-time analysis of the influenza season and in the definition of unbiased forecast models.

Nonannual seasonality of influenza-like illness in a tropical urban setting.

BackgroundIn temperate and subtropical climates, respiratory diseases exhibit seasonal peaks in winter. In the tropics, with no winter, peak timings are irregular.MethodsTo obtain a detailed picture of influenza‐like illness (ILI) patterns in the tropics, we established an mHealth study in community clinics in Ho Chi Minh City (HCMC). During 2009‐2015, clinics reported daily case numbers via SMS, with a subset performing molecular diagnostics for influenza virus. This real‐time epidemiology network absorbs 6000 ILI reports annually, one or two orders of magnitude more than typical surveillance systems. A real‐time online ILI indicator was developed to inform clinicians of the daily ILI activity in HCMC.ResultsFrom August 2009 to December 2015, 63 clinics were enrolled and 36 920 SMS reports were received, covering approximately 1.7M outpatient visits. Approximately 10.6% of outpatients met the ILI case definition. ILI activity in HCMC exhibited strong nonannual dynamics with a dominant periodicity of 206 days. This was confirmed by time series decomposition, stepwise regression, and a forecasting exercise showing that median forecasting errors are 30%‐40% lower when using a 206‐day cycle. In ILI patients from whom nasopharyngeal swabs were taken, 31.2% were positive for influenza. There was no correlation between the ILI time series and the time series of influenza, influenza A, or influenza B (all P > 0.15).ConclusionThis suggests, for the first time, that a nonannual cycle may be an essential driver of respiratory disease dynamics in the tropics. An immunological interference hypothesis is discussed as a potential underlying mechanism.

The influence of statutory holidays on the proportion of the outpatient and emergency visits for influenza-like illness

Objective: To analyze the reasons for the fluctuations in the percentage of outpatient or emergency visits for influenza-like illness (ILI) during the Spring Festival and National Day in 2014-2018 surveillance season. Methods: ILI surveillance data was collected during the period of Spring Festival and National Day in mainland China, and downloaded from Chinese Influenza Surveillance Information System, during the 2014-2018 surveillance season. Results: There was no significant difference noticed in the number of ILI reports in the festival week with weeks before or after in both the southern and northern provinces. The number of outpatient visits was much less than that of the week before and after, but the number of emergency visits was statistically significantly increased. Conclusion: In the holiday peak of ILI%, the major causes was the impact of holiday-off at sentinel hospitals, resulting in a large reduction in the number of outpatient visits in the consulting room during the festivals.

Epidemiological and Serological Investigation into the Role of Gestational Maternal Influenza Virus Infection and Autism Spectrum Disorders.

The literature concerning gestational maternal influenza virus infection and risk of autism spectrum disorders (ASD) is inconclusive. To address this uncertainty, we obtained information from questionnaires and samples from the Autism Birth Cohort, a prospective birth cohort comprising mothers, fathers, and offspring recruited in Norway in 1999 to 2008. Through questionnaires, referrals, and linkages to the Norwegian National Patient Registry, we identified 338 mothers of children with ASD and 348 frequency-matched controls for whom plasma samples that had been collected midpregnancy and after delivery were available for influenza virus serology via luciferase immunoprecipitation and hemagglutinin inhibition assays for influenza virus strains circulating during the study period. Assay data were combined to define serological status and integrated with self-reports of influenza-like illness to estimate ASD risk. Neither influenza A nor influenza B virus infection was associated with increased ASD risk. Integration of reports of symptoms of influenza-like illness with serology revealed an increase in risk for seropositive women with symptoms, but this increase did not achieve statistical significance (a level of P < 0.05) in the comparison with seronegative women without symptoms (adjusted odds ratio, 1.93; 95% confidence interval, 0.95 to 3.89; P = 0.068). Although chance may explain our findings, the magnitude of the potential association may be of biological importance, and dismissing our findings could result in failure to detect a bona fide association (type II error). If the association is true, we posit that the risk is due to activation of the maternal immune system following infection rather than direct fetal infection. Data on levels of cytokines or other mediators of inflammation would allow us to test the validity of this hypothesis. IMPORTANCE The causes of most cases of autism spectrum disorders (ASD) are unknown. Some epidemiological studies suggest that maternal gestational influenza virus infection may increase the risk of ASD in offspring. Here, we describe an analysis of a large birth cohort with results based on questionnaires that prospectively addressed subjective reports of influenza-like illness and serological assays for objective determination of influenza virus infection. Although serologic evidence of gestational influenza virus infection alone was not associated with risk, positive serology and symptoms of influenza-like illness cannot yet be definitely ruled out as a risk factor.

Socio-environmental and measurement factors drive variation in influenza-like illness

ObjectiveTo assess the use of medical claims records for surveillance andepidemiological inference through a case study that examines howecological and social determinants and measurement error contributeto spatial heterogeneity in reports of influenza-like illness across theUnited States.IntroductionTraditional infectious disease epidemiology is built on thefoundation of high quality and high accuracy data on disease andbehavior. Digital infectious disease epidemiology, on the other hand,uses existing digital traces, re-purposing them to identify patterns inhealth-related processes. Medical claims are an emerging digital datasource in surveillance; they capture patient-level data across an entirepopulation of healthcare seekers, and have the benefits of medicalaccuracy through physician diagnoses, and fine spatial and temporalresolution in near real-time.Our work harnesses the large volume and high specificity ofdiagnosis codes in medical claims to improve our understanding ofthe mechanisms driving spatial variation in reported influenza activityeach year. The mechanisms hypothesized to drive these patterns areas varied as: environmental factors affecting transmission or virussurvival, travel flows between different populations, population agestructure, and socioeconomic factors linked to healthcare access andquality of life. Beyond process mechanisms, the nature of surveillancedata collection may affect our interpretation of spatial epidemiologicalpatterns [1], particularly since influenza is a non-reportable diseasewith non-specific symptoms ranging from asymptomatic to severe.Considering the ways in which medical claims are generated, biasesmay arise from healthcare-seeking behavior, insurance coverage, andmedical claims database coverage in study populations.MethodsUsing aggregated U.S. medical claims for influenza-like illness(ILI) from the 2001-2002 through 2008-2009 flu seasons [2],we developed a Bayesian hierarchical modeling framework toestimate the importance of both ecological and social determinantsand measurement-related factors on observed county-level variationof influenza disease burden across the United States. Integrated NestedLaplace Approximation (INLA) techniques for Bayesian inferencewere used to render our questions computationally tractable due tothe high spatial resolution of our data (Figure 1) and the multiplicityof models in our analysis [3]. Linking data from a variety of publiclyavailable sources, we determined the strength, directionality, andconsistency of these factors over multiple flu seasons.ResultsWe found that measurement-related factors – healthcare-seekingbehavior, insurance coverage, and medical claims database coverage– were strong predictors of greater ILI intensity across seasons.Secondarily, poverty and specific humidity were negatively associatedwith ILI intensity for several seasons. Finally, by incorporatingmechanistic and measurement factors into our model, our modelpredictions present an improved map of influenza-like illness in theUnited States for the flu seasons in our study period.ConclusionsWe present a flexible modeling approach that applies to differentmedical claims diagnosis codes and disease surveillance data anddemonstrates the utility of Bayesian hierarchical models for large-scale ecological analyses. Our results increase our knowledge of thespatial distribution of influenza and the underlying processes thatdrive these patterns, promote finer spatial targeting for differenttypes of interventions, and enable the interpolation of burden in areasdifficult to surveil through traditional public health. Moreover, theyhighlight the relative contributions of surveillance data collectionand ecological processes to spatial variation in disease, and highlightthe importance of considering measurement biases when usingsurveillance data for epidemiological inference.

Using a Bayesian Method to Assess Google, Twitter, and Wikipedia for ILI Surveillance

ObjectiveTo comparatively analyze Google, Twitter, and Wikipedia byevaluating how well change points detected in each web-based sourcecorrespond to change points detected in CDC ILI data.IntroductionTraditional influenza surveillance relies on reports of influenza-like illness (ILI) by healthcare providers, capturing individualswho seek medical care and missing those who may search, post,and tweet about their illnesses instead. Existing research has shownsome promise of using data from Google, Twitter, and Wikipediafor influenza surveillance, but with conflicting findings, studies haveonly evaluated these web-based sources individually or dually withoutcomparing all three of them1-5. A comparative analysis of all threeweb-based sources is needed to know which of the web-based sourcesperforms best in order to be considered to complement traditionalmethods.MethodsWe collected publicly available, de-identified data from the CDCILINet system, Google Flu Trends, HealthTweets.org, and Wikipediafor the 2012-2015 influenza seasons. Bayesian change point analysiswas the method used to detect change points, or seasonal changes,in each of the web-data sources for comparison to change pointsin CDC ILI data. All analyses was conducted using the R package‘bcp’ v4.0.0 in RStudio v0.99.484. Sensitivity and positive predictivevalues (PPV) were then calculated.ResultsDuring the 2012-2015 influenza seasons, a high sensitivity of 92%was found for Google, while the PPV for Google was 85%. A lowsensitivity of 50% was found for Twitter; a low PPV of 43% wasfound for Twitter also. Wikipedia had the lowest sensitivity of 33%and lowest PPV of 40%.ConclusionsGoogle had the best combination of sensitivity and PPV indetecting change points that corresponded with change points found inCDC data. Overall, change points in Google, Twitter, and Wikipediadata occasionally aligned well with change points captured in CDCILI data, yet these sources did not detect all changes in CDC data,which could indicate limitations of the web-based data or signify thatthe Bayesian method is not adequately sensitive. These three web-based sources need to be further studied and compared using otherstatistical methods before being incorporated as surveillance data tocomplement traditional systems.Figure 1. Detection of change points, 2012-2013 influenza seasonFigure 2. Detection of change points, 2013-2014 influenza seasonFigure 3. Detection of change points, 2014-2015 influenza season

Relating calls to US poison centers for potential exposures to medications to Centers for Disease Control and Prevention reporting of influenza-like illness

Context: The Centers for Disease Control (CDC) monitors influenza like illness (ILI) and the National Poison Data System (NPDS) warehouses call data uploaded by US poison centers regarding reported exposures to medication. Objective: We examined the relationship between calls to poison centers regarding reported exposures to medications commonly used to treat ILI and weekly reports of ILI. Materials and Methods: The CDC reports ILI, by age group, for each of 10 Health and Human Services (HHS) regions. We examined NPDS summary data from calls reported to poison centers regarding reported exposures to acetaminophen, cough/cold medications, and promethazine, for the same weeks, age groups, and HHS regions for influenza seasons 2000–2013. ILI and NPDS exposures were examined using graphical plots, descriptive statistics, stepwise regression analysis, and Geographic Information Systems (GIS). Results: About 5,101,841 influenza-like illness cases were reported to the CDC, and 2,122,940 calls regarding reported exposures to medications commonly used to treat ILI, were reported by poison centers to the NPDS over the 13 flu seasons. Analysis of stepwise models of the linear untransformed data involving 24 NPDS data groups and for 60 ILI measures, over the 13 influenza seasons, demonstrated that reported exposures to medications used to treat ILI correlated with reported cases of ILI with a median R2 = 0.489 (min R2 = 0.248, max R2 = 0.717), with mean ± SD of R2 = 0.494 ± 0.121. Median number of parameters used (degrees of freedom – 1) was 7. Conclusions: NPDS data regarding poison center calls for selected ILI medication exposures were highly correlated with CDC ILI data. Since NPDS data are available in real time, it provides complimentary ILI monitoring. This approach may provide public health value in predicting other illnesses which are not currently as thoroughly monitored.

Combining syndromic surveillance and ILI data using particle filter for epidemic state estimation

Designing effective mitigation strategies against influenza outbreak requires an accurate prediction of a disease’s future course of spreading. Real time information such as syndromic surveillance data and influenza-like-illness (ILI) reports by clinicians can be used to generate estimates of the current state of spreading of a disease. Syndromic surveillance data are immediately available, in contrast to ILI reports that require data collection and processing. On the other hand, they are less credible than ILI data because they are essentially behavioral responses from a community. In this paper, we present a method to combine immediately-available-but-less-reliable syndromic surveillance data with reliable-but-time-delayed ILI data. This problem is formulated as a non-linear stochastic filtering problem, and solved by a particle filtering method. Our experimental results from hypothetical pandemic scenarios show that state estimation is improved by utilizing both sets of data compared to when using only one set. However, the amount of improvement depends on the relative credibility and length of delay in ILI data. An analysis for a linear, Gaussian case is presented to support the results observed in the experiments.

Automated Influenza-like Illness Reporting—An Efficient Adjunct to Traditional Sentinel Surveillance

We compared an electronic health record-based influenza-like illness (ILI) surveillance system with manual sentinel surveillance and virologic data to evaluate the utility of the automated system for routine ILI surveillance. We obtained weekly aggregate ILI reports from the Electronic medical record Support for Public Health (ESP) disease-detection and reporting system, which used an automated algorithm to identify ILI visits among a patient population of about 700,000 in Eastern Massachusetts. The percentage of total visits for ILI ("percent ILI") in ESP, percent ILI in the Massachusetts Department of Public Health's sentinel surveillance system, and percentage of laboratory specimens submitted to participating Massachusetts laboratories that tested positive for influenza were compared for the period October 2007-September 2011. We calculated Spearman's correlation coefficients and compared ESP and sentinel surveillance systems qualitatively, in terms of simplicity, flexibility, data quality, acceptability, timeliness, and usefulness. ESP and sentinel surveillance percent ILI always peaked within one week of each other. There was 80% correlation between the two and 71%-73% correlation with laboratory data. Sentinel surveillance percent ILI was higher than ESP percent ILI during influenza seasons. The amplitude of variation in ESP percent ILI was greatest for 5- to 49-year-olds and typically peaked for the 5- to 24-year-old age group before the others. The ESP system produces percent ILI data of similar quality to sentinel surveillance and offers the advantages of shifting disease reporting burden from clinicians to information systems, allowing tracking of disease by age group, facilitating efficient surveillance for very large populations, and producing consistent and timely reports.

Intensive care unit admissions for community-acquired pneumonia are seasonal but are not associated with weather or reports of influenza-like illness in the community

PurposeThe aims of this study were to determine if there is seasonal variation in the number of intensive care unit (ICU) admissions for community-acquired pneumonia (CAP) and if there is a relationship between these admissions and weather or reports of influenza-like illness in the community. Materials and MethodsIn this time series analysis in 3 medical-surgical ICUs (8, 13, and 20 beds) in the Vancouver region, we included patients admitted to adult ICUs for CAP between January 2002 and March 2006. We used Poisson regression to analyze the association between weekly number of ICU admissions for CAP, and average temperature, range in temperature, total precipitation, and cases of influenza-like illness/100 physician visits reported by sentinel physicians in the community. ResultsIn 740 patients admitted to ICUs for CAP, admissions peaked each year in the winter-spring months. In multivariate models, a sine function with a single annual peak was independently associated with number of patients admitted to ICU for CAP (rate ratio [95% confidence interval], 1.12 [1.00, 1.26]), but neither the weather measurements nor the weekly rate of reported influenza-like illness was significantly associated. ConclusionIntensive care unit admissions for CAP are seasonal, but neither weather measurements nor weekly rate of reported influenza-like illness in the community is associated with these admissions.