Deep-NCA: A deep learning methodology for performing noncompartmental analysis of pharmacokinetic data.

Noncompartmental analysis (NCA) is a model-independent approach for assessing pharmacokinetics (PKs). Although the existing NCA algorithms are very well-established and widely utilized, they suffer from low accuracies in the setting of sparse PK samples. In response, we developed Deep-NCA, a deep learning (DL) model to improve the prediction of key noncompartmental PK parameters. Our methodology utilizes synthetic PK data for model training and uses an innovative patient-specific normalization method for data preprocessing. Deep-NCA demonstrated adequate performance across six previously unseen simulated drugs under multiple dosing, showcasing effective generalization. Compared to traditional NCA, Deep-NCA exhibited superior performance for sparse PK data. This study advances the application of DL to PK studies and introduces an effective method for handling sparse PK data. With further validation and refinement, Deep-NCA could significantly enhance the efficiency of drug development by providing more accurate NCA estimates while requiring fewer PK samples.

gPKPDviz: A flexible R shiny tool for pharmacokinetic/pharmacodynamic simulations using mrgsolve.

GPKPDviz is a Shiny application (app) dedicated to real-time simulation, visualization, and assessment of the pharmacokinetic/pharmacodynamic (PK/PD) models. Within the app, gPKPDviz is capable of generating virtual populations and complex dosing and sampling scenarios, which, together with the streamlined workflow, is designed to efficiently assess the impact of covariates and dosing regimens on PK/PD end points. The actual population data from clinical trials can be loaded into the app for simulation if desired. The app-generated dosing regimens include single or multiple dosing, and more complex regimens, such as loading doses or intermittent dosing. When necessary, the dosing regimens can be defined externally and loaded to the app for simulation. Using mrgsolve as the simulation engine, gPKPDviz is typically used for population simulation, however, with a slight modification of the mrgsolve model, gPKPDviz is capable of performing individual simulations with individual post hoc parameters, individual dosing logs, and individual sampling timepoints through an external dataset. A built-in text editor has a debugging feature for the mrgsolve model, providing the same error messages as model compilation in R. GPKPDviz has had stringent validation by comparing simulation results between the app and using mrgsolve in R. GPKPDviz is a member of the suite of Modeling and Simulation Shiny apps developed at Genentech to facilitate the typical modeling work in Clinical Pharmacology. For broader access to the Pharmacometric community, gPKPDviz has been published as an open-source application in GitHub under the terms of GNU General Public License.

Archway Phase 3 Trial of the Port Delivery System with Ranibizumab for Neovascular Age-Related Macular Degeneration 2-Year Results.

PURPOSE: To report 2-year results from the Archway clinical trial of the Port Delivery System with ranibizumab (PDS) for treatment of neovascular age-related macular degeneration (nAMD). DESIGN: Phase 3, randomized, multicenter, open-label, active-comparator-controlled trial. PARTICIPANTS: Patients with previously treated nAMD diagnosed within 9 months of screening and responsive to anti-vascular endothelial growth factor therapy. METHODS: Patients were randomized 3:2 to PDS with ranibizumab 100 mg/ml with fixed refill-exchanges every 24 weeks (PDS Q24W) or intravitreal ranibizumab 0.5 mg injections every 4 weeks (monthly ranibizumab). Patients were followed through 4 complete refill-exchange intervals (∼2 years). MAIN OUTCOME MEASURES: Change in best-corrected visual acuity (BCVA) Early Treatment Diabetic Retinopathy Study (ETDRS) letter score from baseline averaged over weeks 44 and 48, weeks 60 and 64, and weeks 88 and 92 (noninferiority margin, -3.9 ETDRS letters). RESULTS: The PDS Q24W was noninferior to monthly ranibizumab, with differences in adjusted mean change in BCVA score from baseline averaged over weeks 44/48, 60/64 and 88/92 of -0.2 (95% confidence interval [CI], -1.8 to +1.3), +0.4 (95% CI, -1.4 to +2.1) and -0.6 ETDRS letters (95% CI, -2.5 to +1.3), respectively. Anatomic outcomes were generally comparable between arms through week 96. Through each of 4 PDS refill-exchange intervals, 98.4%, 94.6%, 94.8%, and 94.7% of PDS Q24W patients assessed did not receive supplemental ranibizumab treatment. The PDS ocular safety profile was generally unchanged from primary analysis. Prespecified ocular adverse events of special interest (AESI) were reported in 59 (23.8%) PDS and 17 (10.2%) monthly ranibizumab patients. The most common AESI reported in both arms was cataract (PDS Q24W, 22 [8.9%]; monthly ranibizumab, 10 [6.0%]). Events in the PDS Q24W arm included (patient incidence) 10 (4.0%) conjunctival erosions, 6 (2.4%) conjunctival retractions, 4 (1.6%) endophthalmitis cases, and 4 (1.6%) implant dislocations. Serum ranibizumab sampling showed that the PDS continuously released ranibizumab over the 24-week refill-exchange interval and ranibizumab serum concentrations were within the range experienced with monthly ranibizumab. CONCLUSIONS: The PDS Q24W showed noninferior efficacy to monthly ranibizumab through approximately 2 years, with approximately 95% of PDS Q24W patients not receiving supplemental ranibizumab treatment in each refill-exchange interval. The AESIs were generally manageable, with learnings continually implemented to minimize PDS-related AEs. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.

Concentration QTc analysis of giredestrant: Overcoming QT/heart rate confounding in the presence of drug-induced heart rate changes.

Concentration-QTc (C-QTc) analysis has become a common approach for evaluating proarrhythmic risk and delayed cardiac repolarization of oncology drug candidates. Significant heart rate (HR) change has been associated with certain classes of oncology drugs and can result in over- or underestimation of the true QT prolongation risk. Because oncology early clinical trials typically lack a placebo control arm or time-matched, treatment-free baseline electrocardiogram collection, significant HR change brings additional challenges to C-QTc analysis in the oncology setting. In this work, a spline-based correction method (QTcSPL) was explored to mitigate the impact of HR changes in giredestrant C-QTc analysis. Giredestrant is a selective estrogen receptor degrader being developed for the treatment of patients with estrogen receptor-positive (ER+) breast cancer. A dose-related HR decrease has been observed in patients under giredestrant treatment, with significant reductions (>10 bpm) observed at supratherapeutic doses. The QTcSPL method demonstrated superior functionality to reduce the correlation between QTc and HR as compared with the Fridericia correction (QTcF). The effect of giredestrant exposure on QTc was evaluated at the clinical dose of 30 mg and supratherapeutic dose of 100 mg based on a prespecified linear mixed effect model. The upper 90% confidence interval of ΔQTcSPL and ΔQTcF were below the 10 ms at both clinical and supratherapeutic exposures, suggesting giredestrant has a low risk of QT prolongation at clinically relevant concentrations. This work demonstrated the use case of QTcSPL to address HR confounding challenges in the context of oncology drug development for the first time.

Comparing trained and untrained probabilistic ensemble forecasts of COVID-19 cases and deaths in the United States.

The U.S. COVID-19 Forecast Hub aggregates forecasts of the short-term burden of COVID-19 in the United States from many contributing teams. We study methods for building an ensemble that combines forecasts from these teams. These experiments have informed the ensemble methods used by the Hub. To be most useful to policymakers, ensemble forecasts must have stable performance in the presence of two key characteristics of the component forecasts: (1) occasional misalignment with the reported data, and (2) instability in the relative performance of component forecasters over time. Our results indicate that in the presence of these challenges, an untrained and robust approach to ensembling using an equally weighted median of all component forecasts is a good choice to support public health decision-makers. In settings where some contributing forecasters have a stable record of good performance, trained ensembles that give those forecasters higher weight can also be helpful.

Astegolimab or Efmarodocokin Alfa in Patients With Severe COVID-19 Pneumonia: A Randomized, Phase 2 Trial.

OBJECTIVES: Severe cases of COVID-19 pneumonia can lead to acute respiratory distress syndrome (ARDS). Release of interleukin (IL)-33, an epithelial-derived alarmin, and IL-33/ST2 pathway activation are linked with ARDS development in other viral infections. IL-22, a cytokine that modulates innate immunity through multiple regenerative and protective mechanisms in lung epithelial cells, is reduced in patients with ARDS. This study aimed to evaluate safety and efficacy of astegolimab, a human immunoglobulin G2 monoclonal antibody that selectively inhibits the IL-33 receptor, ST2, or efmarodocokin alfa, a human IL-22 fusion protein that activates IL-22 signaling, for treatment of severe COVID-19 pneumonia. DESIGN: Phase 2, double-blind, placebo-controlled study (COVID-astegolimab-IL). SETTING: Hospitals. PATIENTS: Hospitalized adults with severe COVID-19 pneumonia. INTERVENTIONS: Patients were randomized to receive IV astegolimab, efmarodocokin alfa, or placebo, plus standard of care. The primary endpoint was time to recovery, defined as time to a score of 1 or 2 on a 7-category ordinal scale by day 28. MEASUREMENTS AND MAIN RESULTS: The study randomized 396 patients. Median time to recovery was 11 days (hazard ratio [HR], 1.01 d; p = 0.93) and 10 days (HR, 1.15 d; p = 0.38) for astegolimab and efmarodocokin alfa, respectively, versus 10 days for placebo. Key secondary endpoints (improved recovery, mortality, or prevention of worsening) showed no treatment benefits. No new safety signals were observed and adverse events were similar across treatment arms. Biomarkers demonstrated that both drugs were pharmacologically active. CONCLUSIONS: Treatment with astegolimab or efmarodocokin alfa did not improve time to recovery in patients with severe COVID-19 pneumonia.

Evaluation of FluSight influenza forecasting in the 2021-22 and 2022-23 seasons with a new target laboratory-confirmed influenza hospitalizations.

Accurate forecasts can enable more effective public health responses during seasonal influenza epidemics. Forecasting teams were asked to provide national and jurisdiction-specific probabilistic predictions of weekly confirmed influenza hospital admissions for one through four weeks ahead for the 2021-22 and 2022-23 influenza seasons. Across both seasons, 26 teams submitted forecasts, with the submitting teams varying between seasons. Forecast skill was evaluated using the Weighted Interval Score (WIS), relative WIS, and coverage. Six out of 23 models outperformed the baseline model across forecast weeks and locations in 2021-22 and 12 out of 18 models in 2022-23. Averaging across all forecast targets, the FluSight ensemble was the 2nd most accurate model measured by WIS in 2021-22 and the 5th most accurate in the 2022-23 season. Forecast skill and 95% coverage for the FluSight ensemble and most component models degraded over longer forecast horizons and during periods of rapid change. Current influenza forecasting efforts help inform situational awareness, but research is needed to address limitations, including decreased performance during periods of changing epidemic dynamics.

A Phase 1 study of GDC-0134, a dual leucine zipper kinase inhibitor, in ALS.

OBJECTIVE: Dual leucine zipper kinase (DLK), which regulates the c-Jun N-terminal kinase pathway involved in axon degeneration and apoptosis following neuronal injury, is a potential therapeutic target in amyotrophic lateral sclerosis (ALS). This first-in-human study investigated safety, tolerability, and pharmacokinetics (PK) of oral GDC-0134, a small-molecule DLK inhibitor. Plasma neurofilament light chain (NFL) levels were explored in GDC-0134-treated ALS patients and DLK conditional knockout (cKO) mice. METHODS: The study included placebo-controlled, single and multiple ascending-dose (SAD; MAD) stages, and an open-label safety expansion (OLE) with adaptive dosing for up to 48 weeks. RESULTS: Forty-nine patients were enrolled. GDC-0134 (up to 1200 mg daily) was well tolerated in the SAD and MAD stages, with no serious adverse events (SAEs). In the OLE, three study drug-related SAEs occurred: thrombocytopenia, dysesthesia (both Grade 3), and optic ischemic neuropathy (Grade 4); Grade ≤2 sensory neurological AEs led to dose reductions/discontinuations. GDC-0134 exposure was dose-proportional (median half-life = 84 h). Patients showed GDC-0134 exposure-dependent plasma NFL elevations; DLK cKO mice also exhibited plasma NFL compared to wild-type littermates. INTERPRETATION: This trial characterized GDC-0134 safety and PK, but no adequately tolerated dose was identified. NFL elevations in GDC-0134-treated patients and DLK cKO mice raised questions about interpretation of biomarkers affected by both disease and on-target drug effects. The safety profile of GDC-0134 was considered unacceptable and led to discontinuation of further drug development for ALS. Further work is necessary to understand relationships between neuroprotective and potentially therapeutic effects of DLK knockout/inhibition and NFL changes in patients with ALS.

An open repository of real-time COVID-19 indicators.

The COVID-19 pandemic presented enormous data challenges in the United States. Policy makers, epidemiological modelers, and health researchers all require up-to-date data on the pandemic and relevant public behavior, ideally at fine spatial and temporal resolution. The COVIDcast API is our attempt to fill this need: Operational since April 2020, it provides open access to both traditional public health surveillance signals (cases, deaths, and hospitalizations) and many auxiliary indicators of COVID-19 activity, such as signals extracted from deidentified medical claims data, massive online surveys, cell phone mobility data, and internet search trends. These are available at a fine geographic resolution (mostly at the county level) and are updated daily. The COVIDcast API also tracks all revisions to historical data, allowing modelers to account for the frequent revisions and backfill that are common for many public health data sources. All of the data are available in a common format through the API and accompanying R and Python software packages. This paper describes the data sources and signals, and provides examples demonstrating that the auxiliary signals in the COVIDcast API present information relevant to tracking COVID activity, augmenting traditional public health reporting and empowering research and decision-making.

Safety, Tolerability, and Pharmacokinetics of High-Volume Subcutaneous Crenezumab, With and Without Recombinant Human Hyaluronidase in Healthy Volunteers.

Compared with intravenous formulations, subcutaneous (s.c.) formulations of therapeutic monoclonal antibodies may provide increased patient access and more convenient administration options, although historically high-volume s.c. administration (> 10-15 mL) has been challenging. We report results from two phase I studies in healthy participants (GP29523 and GP40201) that evaluated s.c. crenezumab, an anti-Aß monoclonal antibody in development for individuals at risk for autosomal-dominant Alzheimer's disease. GP29523 assessed safety, tolerability, and pharmacokinetics (PK) in 68 participants (aged 50-80 years) who received single ascending doses (600-7,200 mg) of crenezumab or placebo (4-40 mL). GP40201 assessed safety, tolerability, and PK in 72 participants (aged 18-80 years) who received different combinations of dose (1,700-6,800 mg), infusion volume (10-40 mL), and flow rate (2-4 mL/minute), with/without recombinant human hyaluronidase (rHuPH20). There were no serious or dose-limiting adverse events in either study. There were no meaningful differences in pain scores among reference placebo (4 mL), test placebo (4-40 mL), or crenezumab (600-7,200 mg) in GP29523, or across treatments with varying infusion volume, flow rate, dose, or rHuPH20 co-administration or concentration in GP40201. Transient erythema was the most common infusion site reaction in both studies. In GP40201 at volumes of ≥ 20 mL, rHuPH20 co-administration appeared to reduce infusion site swelling incidence, but, in some cases, was associated with larger areas of infusion site erythema. Crenezumab exhibited approximately dose-proportional PK, and s.c. bioavailability was 66% and independent of dose or rHuPH20 co-administration. High-dose, high-concentration, high-volume s.c. crenezumab formulated with/without rHuPH20 was well-tolerated in healthy participants, with an acceptable safety profile.

An open challenge to advance probabilistic forecasting for dengue epidemics.

A wide range of research has promised new tools for forecasting infectious disease dynamics, but little of that research is currently being applied in practice, because tools do not address key public health needs, do not produce probabilistic forecasts, have not been evaluated on external data, or do not provide sufficient forecast skill to be useful. We developed an open collaborative forecasting challenge to assess probabilistic forecasts for seasonal epidemics of dengue, a major global public health problem. Sixteen teams used a variety of methods and data to generate forecasts for 3 epidemiological targets (peak incidence, the week of the peak, and total incidence) over 8 dengue seasons in Iquitos, Peru and San Juan, Puerto Rico. Forecast skill was highly variable across teams and targets. While numerous forecasts showed high skill for midseason situational awareness, early season skill was low, and skill was generally lowest for high incidence seasons, those for which forecasts would be most valuable. A comparison of modeling approaches revealed that average forecast skill was lower for models including biologically meaningful data and mechanisms and that both multimodel and multiteam ensemble forecasts consistently outperformed individual model forecasts. Leveraging these insights, data, and the forecasting framework will be critical to improve forecast skill and the application of forecasts in real time for epidemic preparedness and response. Moreover, key components of this project-integration with public health needs, a common forecasting framework, shared and standardized data, and open participation-can help advance infectious disease forecasting beyond dengue.

Accuracy of real-time multi-model ensemble forecasts for seasonal influenza in the U.S.

Seasonal influenza results in substantial annual morbidity and mortality in the United States and worldwide. Accurate forecasts of key features of influenza epidemics, such as the timing and severity of the peak incidence in a given season, can inform public health response to outbreaks. As part of ongoing efforts to incorporate data and advanced analytical methods into public health decision-making, the United States Centers for Disease Control and Prevention (CDC) has organized seasonal influenza forecasting challenges since the 2013/2014 season. In the 2017/2018 season, 22 teams participated. A subset of four teams created a research consortium called the FluSight Network in early 2017. During the 2017/2018 season they worked together to produce a collaborative multi-model ensemble that combined 21 separate component models into a single model using a machine learning technique called stacking. This approach creates a weighted average of predictive densities where the weight for each component is determined by maximizing overall ensemble accuracy over past seasons. In the 2017/2018 influenza season, one of the largest seasonal outbreaks in the last 15 years, this multi-model ensemble performed better on average than all individual component models and placed second overall in the CDC challenge. It also outperformed the baseline multi-model ensemble created by the CDC that took a simple average of all models submitted to the forecasting challenge. This project shows that collaborative efforts between research teams to develop ensemble forecasting approaches can bring measurable improvements in forecast accuracy and important reductions in the variability of performance from year to year. Efforts such as this, that emphasize real-time testing and evaluation of forecasting models and facilitate the close collaboration between public health officials and modeling researchers, are essential to improving our understanding of how best to use forecasts to improve public health response to seasonal and emerging epidemic threats.

Sacroiliac Joint Fusion Methodology - Minimally Invasive Compared to Screw-Type Surgeries: A Systematic Review and Meta-Analysis.

BACKGROUND: Sacroiliac (SI) joint fusion represents a unique area of orthopedic surgery with procedural literature dating to the early 1920s, showing limited innovation in either technique or hardware over the last 90 years. Recent improvements in the diagnosis and treatment of SI joint dysfunction warrant comparisons to older surgical techniques. OBJECTIVE: To evaluate treatment efficacies and patient outcomes associated with minimally invasive joint fusion in comparison to screw-type surgeries. STUDY DESIGN: Systematic review and meta-analysis. SETTING: Electronic databases, EMBASE, Pubmed (Medline), manual bibliography cross-referencing for published works until Dec. 31, 2017. METHODS: A thorough literature search was performed in adherence to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology. Data repositories accessed included Pubmed and EMBASE, until Dec. 31, 2017. All studies evaluating sacroiliac joint fusion and reporting quantifiable outcome data were included. Exclusion criteria included nonhuman studies, qualitative reviews, and meta-analyses. Data compilation, coding, and extraction were performed using MedAware Systems proprietary software. Data from each study were extracted by 2 analysts, using software that allowed automatic comparisons of all data fields. The standardized mean difference (SMD) was used as a summary statistic for pooling outcomes data across studies. Multiple outcome measures were grouped into 3 categories, according to similarity of measurements - Pain, Disability/Physical Function, and Global/QOL. RESULTS: A total of 20 studies had adequate data to calculate a SMD, and were included in the meta-analysis. Results of iFuse trials were compared to screw type trials, pooled in 3 categories of outcomes - Pain, Disability/Physical Function, and Global/QOL. The Pain category showed a statistically significant (P = 0.03) difference in outcomes for patients receiving the iFuse implant compared to screw types (SMD = 2.04 [95%CI: 1.76 to 2.33] vs. 1.28 [95%CI: 0.47 to 2.09]), with iFuse showing significantly better outcomes. The Disability category also showed a statistically significant (P = 0.01) difference in outcomes for patients receiving the iFuse implant compared to screw types (SMD = 1.68 [95%CI: 1.43 to 1.94] vs. 0.26 [95%CI: -1.90 to 2.41]), with iFuse showing significantly better outcomes. For Global/Quality of Life (QOL) outcomes, there was a significant difference (P = 0.04) between iFuse and screw-type procedures (SMD = 0.99 [95%CI: 0.75 to 1.24] vs. 0.60 [95%CI: 0.33 to 0.88]), with iFuse showing significantly better outcomes. There was a statistically significant correlation between lower baseline Oswestry Disability Index (ODI) and Short Form 36 Health Survey (SF-36) values and better post treatment outcomes (r2 = 0.47, P < 0.01, and r2 = 0.30, P < 0.01, respectively). An association was found between pain at baseline and better outcomes (r2 = 0.21, P < 0.01), where worse baseline pain was associated with better outcomes. LIMITATIONS: There was a limited number of studies in this meta-analysis with treatments that could be properly classified as screw-type. CONCLUSION: In this analysis, compared to screw-type surgeries, the iFuse system showed statistically superior outcomes. This was the case when outcome measures were classified into 3 main categories - Pain, Disability/Physical Function, and Global/QOL. KEY WORDS: Meta-analysis, systematic review, sacroiliac joint, sacroiliac joint fusion.

Collaborative efforts to forecast seasonal influenza in the United States, 2015-2016.

Since 2013, the Centers for Disease Control and Prevention (CDC) has hosted an annual influenza season forecasting challenge. The 2015-2016 challenge consisted of weekly probabilistic forecasts of multiple targets, including fourteen models submitted by eleven teams. Forecast skill was evaluated using a modified logarithmic score. We averaged submitted forecasts into a mean ensemble model and compared them against predictions based on historical trends. Forecast skill was highest for seasonal peak intensity and short-term forecasts, while forecast skill for timing of season onset and peak week was generally low. Higher forecast skill was associated with team participation in previous influenza forecasting challenges and utilization of ensemble forecasting techniques. The mean ensemble consistently performed well and outperformed historical trend predictions. CDC and contributing teams will continue to advance influenza forecasting and work to improve the accuracy and reliability of forecasts to facilitate increased incorporation into public health response efforts.

A collaborative multiyear, multimodel assessment of seasonal influenza forecasting in the United States.

Influenza infects an estimated 9-35 million individuals each year in the United States and is a contributing cause for between 12,000 and 56,000 deaths annually. Seasonal outbreaks of influenza are common in temperate regions of the world, with highest incidence typically occurring in colder and drier months of the year. Real-time forecasts of influenza transmission can inform public health response to outbreaks. We present the results of a multiinstitution collaborative effort to standardize the collection and evaluation of forecasting models for influenza in the United States for the 2010/2011 through 2016/2017 influenza seasons. For these seven seasons, we assembled weekly real-time forecasts of seven targets of public health interest from 22 different models. We compared forecast accuracy of each model relative to a historical baseline seasonal average. Across all regions of the United States, over half of the models showed consistently better performance than the historical baseline when forecasting incidence of influenza-like illness 1 wk, 2 wk, and 3 wk ahead of available data and when forecasting the timing and magnitude of the seasonal peak. In some regions, delays in data reporting were strongly and negatively associated with forecast accuracy. More timely reporting and an improved overall accessibility to novel and traditional data sources are needed to improve forecasting accuracy and its integration with real-time public health decision making.

Nonmechanistic forecasts of seasonal influenza with iterative one-week-ahead distributions.

Accurate and reliable forecasts of seasonal epidemics of infectious disease can assist in the design of countermeasures and increase public awareness and preparedness. This article describes two main contributions we made recently toward this goal: a novel approach to probabilistic modeling of surveillance time series based on "delta densities", and an optimization scheme for combining output from multiple forecasting methods into an adaptively weighted ensemble. Delta densities describe the probability distribution of the change between one observation and the next, conditioned on available data; chaining together nonparametric estimates of these distributions yields a model for an entire trajectory. Corresponding distributional forecasts cover more observed events than alternatives that treat the whole season as a unit, and improve upon multiple evaluation metrics when extracting key targets of interest to public health officials. Adaptively weighted ensembles integrate the results of multiple forecasting methods, such as delta density, using weights that can change from situation to situation. We treat selection of optimal weightings across forecasting methods as a separate estimation task, and describe an estimation procedure based on optimizing cross-validation performance. We consider some details of the data generation process, including data revisions and holiday effects, both in the construction of these forecasting methods and when performing retrospective evaluation. The delta density method and an adaptively weighted ensemble of other forecasting methods each improve significantly on the next best ensemble component when applied separately, and achieve even better cross-validated performance when used in conjunction. We submitted real-time forecasts based on these contributions as part of CDC's 2015/2016 FluSight Collaborative Comparison. Among the fourteen submissions that season, this system was ranked by CDC as the most accurate.

Results from the second year of a collaborative effort to forecast influenza seasons in the United States.

Accurate forecasts could enable more informed public health decisions. Since 2013, CDC has worked with external researchers to improve influenza forecasts by coordinating seasonal challenges for the United States and the 10 Health and Human Service Regions. Forecasted targets for the 2014-15 challenge were the onset week, peak week, and peak intensity of the season and the weekly percent of outpatient visits due to influenza-like illness (ILI) 1-4 weeks in advance. We used a logarithmic scoring rule to score the weekly forecasts, averaged the scores over an evaluation period, and then exponentiated the resulting logarithmic score. Poor forecasts had a score near 0, and perfect forecasts a score of 1. Five teams submitted forecasts from seven different models. At the national level, the team scores for onset week ranged from <0.01 to 0.41, peak week ranged from 0.08 to 0.49, and peak intensity ranged from <0.01 to 0.17. The scores for predictions of ILI 1-4 weeks in advance ranged from 0.02-0.38 and was highest 1 week ahead. Forecast skill varied by HHS region. Forecasts can predict epidemic characteristics that inform public health actions. CDC, state and local health officials, and researchers are working together to improve forecasts.

A human judgment approach to epidemiological forecasting.

Infectious diseases impose considerable burden on society, despite significant advances in technology and medicine over the past century. Advanced warning can be helpful in mitigating and preparing for an impending or ongoing epidemic. Historically, such a capability has lagged for many reasons, including in particular the uncertainty in the current state of the system and in the understanding of the processes that drive epidemic trajectories. Presently we have access to data, models, and computational resources that enable the development of epidemiological forecasting systems. Indeed, several recent challenges hosted by the U.S. government have fostered an open and collaborative environment for the development of these technologies. The primary focus of these challenges has been to develop statistical and computational methods for epidemiological forecasting, but here we consider a serious alternative based on collective human judgment. We created the web-based "Epicast" forecasting system which collects and aggregates epidemic predictions made in real-time by human participants, and with these forecasts we ask two questions: how accurate is human judgment, and how do these forecasts compare to their more computational, data-driven alternatives? To address the former, we assess by a variety of metrics how accurately humans are able to predict influenza and chikungunya trajectories. As for the latter, we show that real-time, combined human predictions of the 2014-2015 and 2015-2016 U.S. flu seasons are often more accurate than the same predictions made by several statistical systems, especially for short-term targets. We conclude that there is valuable predictive power in collective human judgment, and we discuss the benefits and drawbacks of this approach.

Flexible Modeling of Epidemics with an Empirical Bayes Framework.

Seasonal influenza epidemics cause consistent, considerable, widespread loss annually in terms of economic burden, morbidity, and mortality. With access to accurate and reliable forecasts of a current or upcoming influenza epidemic's behavior, policy makers can design and implement more effective countermeasures. This past year, the Centers for Disease Control and Prevention hosted the "Predict the Influenza Season Challenge", with the task of predicting key epidemiological measures for the 2013-2014 U.S. influenza season with the help of digital surveillance data. We developed a framework for in-season forecasts of epidemics using a semiparametric Empirical Bayes framework, and applied it to predict the weekly percentage of outpatient doctors visits for influenza-like illness, and the season onset, duration, peak time, and peak height, with and without using Google Flu Trends data. Previous work on epidemic modeling has focused on developing mechanistic models of disease behavior and applying time series tools to explain historical data. However, tailoring these models to certain types of surveillance data can be challenging, and overly complex models with many parameters can compromise forecasting ability. Our approach instead produces possibilities for the epidemic curve of the season of interest using modified versions of data from previous seasons, allowing for reasonable variations in the timing, pace, and intensity of the seasonal epidemics, as well as noise in observations. Since the framework does not make strict domain-specific assumptions, it can easily be applied to some other diseases with seasonal epidemics. This method produces a complete posterior distribution over epidemic curves, rather than, for example, solely point predictions of forecasting targets. We report prospective influenza-like-illness forecasts made for the 2013-2014 U.S. influenza season, and compare the framework's cross-validated prediction error on historical data to that of a variety of simpler baseline predictors.