BioCaster in 2021: automatic disease outbreaks detection from global news media.

SUMMARY: BioCaster was launched in 2008 to provide an ontology-based text mining system for early disease detection from open news sources. Following a 6-year break, we have re-launched the system in 2021. Our goal is to systematically upgrade the methodology using state-of-the-art neural network language models, whilst retaining the original benefits that the system provided in terms of logical reasoning and automated early detection of infectious disease outbreaks. Here, we present recent extensions such as neural machine translation in 10 languages, neural classification of disease outbreak reports and a new cloud-based visualization dashboard. Furthermore, we discuss our vision for further improvements, including combining risk assessment with event semantics and assessing the risk of outbreaks with multi-granularity. We hope that these efforts will benefit the global public health community. AVAILABILITY AND IMPLEMENTATION: BioCaster web-portal is freely accessible at http://biocaster.org.

Changes in quality of life from a homelessness intervention: true change, response shift, or random variation.

BACKGROUND AND OBJECTIVE: Participants experiencing homelessness and mental illness who received housing and support through the At Home/Chez Soi trial showed modest gains in quality of life (QOL) compared to treatment as usual participants. Participants' QOL ratings over time may have been affected by either response shift triggered by new life circumstances or by random variation in the meaning of QOL ratings. This study seeks to identify both phenomena to estimate the intervention's effect on true change in QOL. METHODS: Using the residuals from a regression model of the global item of Lehman's 20-item quality of life interview (QOLI-20), latent trajectory analysis was used to identify response shift, while a measure of overall variability in residuals identified random variation of QOL. The latter was used to adjust group comparisons of QOLI-20 total scores and the global item. RESULTS: Equivalent distributions of both groups' participants across latent trajectory classes (χ2 = 2.97, p = .397) suggest that the intervention did not trigger response shift. However, random variation interacted significantly with the treatment effect on global item ratings. For every increase of one standard deviation of residuals, treatment odds ratios decreased by a factor of 0.70 (SE 1.18, p = .036, 95% CI 0.50-0.98). CONCLUSIONS: Measuring random variation in QOL ratings from the standard deviation of residuals offers the ability to approximate, although indirectly, how participants' QOL is truly affected by a housing intervention. Specifically, we found that QOL improvement is more evident when QOL ratings have a consistent meaning over time.

Inferring global-scale temporal latent topics from news reports to predict public health interventions for COVID-19.

The COVID-19 pandemic has highlighted the importance of non-pharmacological interventions (NPIs) for controlling epidemics of emerging infectious diseases. Despite their importance, NPIs have been monitored mainly through the manual efforts of volunteers. This approach hinders measurement of the NPI effectiveness and development of evidence to guide their use to control the global pandemic. We present EpiTopics, a machine learning approach to support automation of NPI prediction and monitoring at both the document level and country level by mining the vast amount of unlabeled news reports on COVID-19. EpiTopics uses a 3-stage, transfer-learning algorithm to classify documents according to NPI categories, relying on topic modeling to support result interpretation. We identified 25 interpretable topics under 4 distinct and coherent COVID-related themes. Importantly, the use of these topics resulted in significant improvements over alternative automated methods in predicting the NPIs in labeled documents and in predicting country-level NPIs for 42 countries.

EpiTopics: A dynamic machine learning model to predict and inform non-pharmacological public health interventions from global news reports.

Non-pharmacological interventions (NPIs) are important for controlling infectious diseases such as COVID-19, but their implementation is currently monitored in an ad hoc manner. To address this issue, we present a three-stage machine learning framework called EpiTopics to facilitate the surveillance of NPI. In this protocol, we outline the use of transfer-learning to address the limited number of NPI-labeled documents and topic modeling to support interpretation of the results. For complete details on the use and execution of this protocol, please refer to Wen et al. (2022).

Cost-Effectiveness of Housing First With Assertive Community Treatment: Results From the Canadian At Home/Chez Soi Trial.

OBJECTIVE: The At Home/Chez Soi trial for homeless individuals with mental illness showed scattered-site Housing First with Assertive Community Treatment (ACT) to be more effective than treatment as usual. This study evaluated the cost-effectiveness of Housing First with ACT and treatment as usual. METHODS: Between October 2009 and June 2011, a total of 950 homeless individuals with serious mental illness were recruited in five Canadian cities: Vancouver, Winnipeg, Toronto, Montreal, and Moncton. Participants were randomly assigned to Housing First (N=469) or treatment as usual (N=481) and followed up for up to 24 months. The intervention consisted of scattered-site Housing First, using rent supplements, with ACT. The treatment-as-usual group had access to all other services. The perspective of society was adopted for the cost-effectiveness analysis. Days of stable housing served as the outcome measure. Retrospective questionnaires captured service use data. RESULTS: Most (69%) of the costs of the intervention were offset by savings in other costs, such as emergency shelters, reducing the net annual cost of the intervention to about Can$6,311 per person. The incremental cost-effectiveness ratio was Can$41.73 per day of stable housing (95% confidence interval=Can$1.96-$83.70). At up to Can$60 per day, Housing First had more than an 80% chance of being cost-effective, compared with treatment as usual. Cost-effectiveness did not vary by participant characteristics. CONCLUSIONS: Housing First with ACT appeared about as cost-effective as Housing First with intensive case management for people with moderate needs. The optimal mix between the two remains to be determined.

Modeling Chronic Obstructive Pulmonary Disease Progression Using Continuous-Time Hidden Markov Models.

Understanding the progression of chronic diseases, such as chronic obstructive pulmonary disease (COPD), is important to inform early diagnosis, personalized care, and health system management. Data from clinical and administrative systems have the potential to advance this understanding, but traditional methods for modelling disease progression are not well-suited to analyzing data collected at irregular intervals, such as when a patient interacts with a healthcare system. We applied a continuous-time hidden Markov model to irregularly-spaced healthcare utilization events and patient-level characteristics in order to analyze the progression through discrete states of 76,888 patients with COPD. A 4-state model allowed classification of patients into interpretable states of disease progression and generated insights about the role of comorbidities, such as cardiovascular diseases, in accelerating severe trajectories. These results can improve the understanding of the evolution of COPD and point to new hypotheses about chronic disease management and comorbidity.

Cost-effectiveness of Housing First Intervention With Intensive Case Management Compared With Treatment as Usual for Homeless Adults With Mental Illness: Secondary Analysis of a Randomized Clinical Trial.

Importance: In the At Home/Chez Soi trial for homeless individuals with mental illness, the scattered-site Housing First (HF) with Intensive Case Management (ICM) intervention proved more effective than treatment as usual (TAU). Objective: To evaluate the cost-effectiveness of the HF plus ICM intervention compared with TAU. Design, Setting, and Participants: This is an economic evaluation study of data from the At Home/Chez Soi randomized clinical trial. From October 2009 through July 2011, 1198 individuals were randomized to the intervention (n = 689) or TAU (n = 509) and followed up for as long as 24 months. Participants were recruited in the Canadian cities of Vancouver, Winnipeg, Toronto, and Montreal. Participants with a current mental disorder who were homeless and had a moderate level of need were included. Data were analyzed from 2013 through 2019, per protocol. Interventions: Scattered-site HF (using rent supplements) with off-site ICM services was compared with usual housing and support services in each city. Main Outcomes and Measures: The analysis was performed from the perspective of society, with days of stable housing as the outcome. Service use was ascertained using questionnaires. Unit costs were estimated in 2016 Canadian dollars. Results: Of 1198 randomized individuals, 795 (66.4%) were men and 696 (58.1%) were aged 30 to 49 years. Almost all (1160 participants, including 677 in the HF group and 483 in the TAU group) contributed data to the economic analysis. Days of stable housing were higher by 140.34 days (95% CI, 128.14-153.31 days) in the HF group. The intervention cost $14 496 per person per year; reductions in costs of other services brought the net cost down by 46% to $7868 (95% CI, $4409-$11 405). The incremental cost-effectiveness ratio was $56.08 (95% CI, $29.55-$84.78) per additional day of stable housing. In sensitivity analyses, adjusting for baseline differences using a regression-based method, without altering the discount rate, caused the largest change in the incremental cost-effectiveness ratio with an increase to $60.18 (95% CI, $35.27-$86.95). At $67 per day of stable housing, there was an 80% chance that HF was cost-effective compared with TAU. The cost-effectiveness of HF appeared to be similar for all participants, although possibly less for those with a higher number of previous psychiatric hospitalizations. Conclusions and Relevance: In this study, the cost per additional day of stable housing was similar to that of many interventions for homeless individuals. Based on these results, expanding access to HF with ICM appears to be warranted from an economic standpoint. Trial Registration: isrctn.org Identifier: ISRCTN42520374.

Hospital readmissions and the day of the week.

Objectives Patients discharged from hospitals on a Friday (Friday discharges) are readmitted sooner (a shorter time-to-emergency-readmission) than those discharged on any other day of the week. To evaluate the cost-effectiveness of increasing weekend capacity, the effect estimate of Friday discharge on time-to-emergency-readmission needs to be precise. However, precise effect estimation is complicated by the confounding effect of differing healthcare-seeking behaviour and admission practices, and therefore different admission probability, by day of the week. The objective of this research was to examine how differing admission probability by day of the week influences the effect of discharge day on time-to-emergency-readmission. Methods We used a Markov model to determine how day of the week admission probability would theoretically affect the time-to-emergency-readmission for Friday and Wednesday discharges. We tested this in a cohort of patients who have had a history of respiratory illness, using a Cox proportional hazards model to fit the time-to-emergency-readmission to any Quebec hospital as a function of the day of the week of discharge and admission. We fitted another Cox model with an additional time-varying covariate for the current day of the week, to model differing admission probabilities by day of the week. Results Our Markov model showed that if admission probability is lower on the weekends, Friday discharges will be readmitted later (longer time-to-emergency-readmission) than Wednesday discharges. Using hospital admission data, we found that Friday discharges were readmitted slightly earlier than Wednesday discharges (HR: 1.03, 95% CI: (1.02, 1.05)). After adding a time-varying covariate for the current day of the week, the length of time-to-emergency-readmission for a Friday discharge increased, but it was still earlier than a Wednesday discharge (HR: 1.04, 95% CI: (1.01, 1.07)). Conclusions The lower admission probabilities on the weekend confound the effect of Friday discharge on time-to-emergency-readmission by increasing the time-to-emergency-readmission. This confounding effect causes an underestimate of the effect of Friday discharge on time-to-emergency-readmission.

Initial Usability Evaluation of a Knowledge-Based Population Health Information System: The Population Health Record (PopHR).

We report the baseline usability of a novel web-based application, the Population Health Record (PopHR), designed to facilitate the effective use of population health information by public health professionals and to support evidence-based decision-making. The usability test was conducted with ten potential users who each completed eight tasks using the PopHR system. Participant responses were recorded, including timestamps for each data entry. Overall, the task completion rate was 96% while the success rate was 88%. The average time-on-task was 3.11 minutes, with more time spent on tasks requiring a user to stratify data along multiple dimensions, such as age, sex, or geographical region. Usability scores indicated that the current version of PopHR has good usability. Potential improvements identified included adding supporting information, offering different visualizations, and enhancing system stability. These findings are examples of addressable usability problems encountered in developing a population health record system.

Epidemiology of recurrent traumatic brain injury in the general population: A systematic review.

OBJECTIVE: To comprehensively assess recurrent traumatic brain injury (rTBI) risk and risk factors in the general population. METHODS: We systematically searched MEDLINE, EMBASE, and the references of included studies until January 16, 2017, for general population observational studies reporting rTBI risk or risk factors. Estimates were not meta-analyzed due to significant methodologic heterogeneity between studies, which was evaluated using meta-regression. RESULTS: Twenty-two studies reported recurrence risk and 11 reported on 27 potential risk factors. rTBI risk was heterogeneous and varied from 0.43% (95% confidence interval [CI] 0.19%-0.67%) to 41.92% (95% CI 34.43%-49.40%), with varying follow-up periods (3 days-55 years). Median time to recurrence ranged from 0.5 to 3.8 years. In studies where cases were ascertained from multiple points of care, at least 5.50% (95% CI 4.80%-6.30%) of patients experienced a recurrence after a 1-year follow-up. Studies that used administrative data/self-report surveys to ascertain cases tended to report higher risk. Risk factors measured at time of index traumatic brain injury (TBI) that were significantly associated with rTBI in more than one study were male sex, prior TBI before index case, moderate or severe TBI, and alcohol intoxication. Risk factors reported in a single study that were significantly associated with rTBI were epilepsy, not seeking medical care, and multiple factors indicative of low socioeconomic status. CONCLUSIONS: rTBI is an important contributor to the general population TBI burden. Certain risk factors can help identify individuals at higher risk of these repeated injuries. However, higher quality research that improves on rTBI surveillance methodology is needed.

Multivariate and Longitudinal Health System Indicators.

Within population health information systems, indicators are commonly presented as independent, cross-sectional measures, neglecting the multivariate, longitudinal nature of disease progression, health care use, and profiles of performance. We use administrative claims data of Montreal, Canada to identify patterns across indicators and over time in chronic obstructive pulmonary disease patients. We first cluster regions based on four health service indicators. Our second approach discovers individual-level trajectories based on a hidden Markov model using the same four indicators. Both approaches offer additional insights by facilitating the discovery and interpretation of indicators, such as a dual interpretation of low use of general practitioner services. These approaches to the analysis and visualization of health indicators can provide a foundation for information displays that will help decision makers identify areas of concern, predict future disease burden, and implement appropriate policies.

Costs of services for homeless people with mental illness in 5 Canadian cities: a large prospective follow-up study.

BACKGROUND: Limited evidence on the costs of homelessness in Canada is available. We estimated the average annual costs, in total and by cost category, that homeless people with mental illness engender from the perspective of society. We also identified individual characteristics associated with higher costs. METHODS: As part of the At Home/Chez Soi trial of Housing First for homeless people with mental illness, 990 participants were assigned to the usual-treatment (control) group in 5 Canadian cities (Vancouver, Winnipeg, Toronto, Montréal and Moncton) between October 2009 and June 2011. They were followed for up to 2 years. Questionnaires ascertained service use and income, and city-specific unit costs were estimated. We adjusted costs for site differences in sample characteristics. We used generalized linear models to identify individual-level characteristics associated with higher costs. RESULTS: Usable data were available for 937 participants (94.6%). Average annual costs (excluding medications) per person in Vancouver, Winnipeg, Toronto, Montréal and Moncton were $53 144 (95% confidence interval [CI] $46 297-$60 095), $45 565 (95% CI $41 039-$50 412), $58 972 (95% CI $52 237-$66 085), $56 406 (95% CI $50 654-$62 456) and $29 610 (95% CI $24 995-$34 480), respectively. Net costs ranged from $15 530 to $341 535. Distributions of costs across categories varied significantly across cities. Lower functioning and a history of psychiatric hospital stays were the most important predictors of higher costs. INTERPRETATION: Homeless people with mental illness generate very high costs for society. Programs are needed to reorient this spending toward more effectively preventing homelessness and toward meeting the health, housing and social service needs of homeless people.

Media content about vaccines in the United States and Canada, 2012-2014: An analysis using data from the Vaccine Sentimeter.

BACKGROUND: A system for monitoring vaccine-related media content was previously developed and studied from an international perspective. This monitoring approach could also have value at a regional level, but it has yet to be evaluated at this scale. We examined regional patterns of vaccine-related media topics and sentiment in the US and Canada. METHODS: We extracted vaccine-relevant US and Canadian online media reports between June 2012 and October 2014 from the Vaccine Sentimeter, a HealthMap-based automated media monitoring system for news aggregators and blogs. We analyzed regional distributions of reports about vaccines, categories (i.e., topics), sentiment, and measles outbreaks. FINDINGS: The Vaccine Sentimeter captured 10,715 reports during the study period. Negative sentiment was highest in reports about vaccine safety (47%), Hepatitis B (19%), and Vermont (18%). Analyses of measles outbreaks revealed geographical variation in media content. For example, religious beliefs were mentioned in 27% of measles reports in Texas and 22% of British Columbia reports, but there were no references to religion in media on measles from California. INTERPRETATIONS: A regional analysis of online sentiment towards vaccine can provide insights that may give US and Canadian public health practitioners a deeper understanding of media influences on vaccine choices in their regions and consequently lead to more effective public health action.