Estimating national and state-level suicide deaths using a novel online symptom search data source.

BACKGROUND: Suicide mortality data are a critical source of information for understanding suicide-related trends in the United States. However, official suicide mortality data experience significant delays. The Google Symptom Search Dataset (SSD), a novel population-level data source derived from online search behavior, has not been evaluated for its utility in predicting suicide mortality trends. METHODS: We identified five mental health related variables (suicidal ideation, self-harm, depression, major depressive disorder, and pain) from the SSD. Daily search trends for these symptoms were utilized to estimate national and state suicide counts in 2020, the most recent year for which data was available, via a linear regression model. We compared the performance of this model to a baseline autoregressive integrated moving average (ARIMA) model and a model including all 422 symptoms (All Symptoms) in the SSD. RESULTS: Our Mental Health Model estimated the national number of suicide deaths with an error of -3.86 %, compared to an error of 7.17 % and 28.49 % for the ARIMA baseline and All Symptoms models. At the state level, 70 % (N = 35) of states had a prediction error of <10 % with the Mental Health Model, with accuracy generally favoring larger population states with higher number of suicide deaths. CONCLUSION: The Google SSD is a new real-time data source that can be used to make accurate predictions of suicide mortality monthly trends at the national level. Additional research is needed to optimize state level predictions for states with low suicide counts.

Development of a Machine Learning Model to Estimate US Firearm Homicides in Near Real Time.

Importance: Firearm homicides are a major public health concern; lack of timely mortality data presents considerable challenges to effective response. Near real-time data sources offer potential for more timely estimation of firearm homicides. Objective: To estimate near real-time burden of weekly and annual firearm homicides in the US. Design, Setting, and Participants: In this prognostic study, anonymous, longitudinal time series data were obtained from multiple data sources, including Google and YouTube search trends related to firearms (2014-2019), emergency department visits for firearm injuries (National Syndromic Surveillance Program, 2014-2019), emergency medical service activations for firearm-related injuries (biospatial, 2014-2019), and National Domestic Violence Hotline contacts flagged with the keyword firearm (2016-2019). Data analysis was performed from September 2021 to September 2022. Main Outcomes and Measures: Weekly estimates of US firearm homicides were calculated using a 2-phase pipeline, first fitting optimal machine learning models for each data stream and then combining the best individual models into a stacked ensemble model. Model accuracy was assessed by comparing predictions of firearm homicides in 2019 to actual firearm homicides identified by National Vital Statistics System death certificates. Results were also compared with a SARIMA (seasonal autoregressive integrated moving average) model, a common method to forecast injury mortality. Results: Both individual and ensemble models yielded highly accurate estimates of firearm homicides. Individual models' mean error for weekly estimates of firearm homicides (root mean square error) varied from 24.95 for emergency department visits to 31.29 for SARIMA forecasting. Ensemble models combining data sources had lower weekly mean error and higher annual accuracy than individual data sources: the all-source ensemble model had a weekly root mean square error of 24.46 deaths and full-year accuracy of 99.74%, predicting the total number of firearm homicides in 2019 within 38 deaths for the entire year (compared with 95.48% accuracy and 652 deaths for the SARIMA model). The model decreased the time lag of reporting weekly firearm homicides from 7 to 8 months to approximately 6 weeks. Conclusions and Relevance: In this prognostic study of diverse secondary data on machine learning, ensemble modeling produced accurate near real-time estimates of weekly and annual firearm homicides and substantially decreased data source time lags. Ensemble model forecasts can accelerate public health practitioners' and policy makers' ability to respond to unanticipated shifts in firearm homicides.

Injury-Related Emergency Department Visits During the COVID-19 Pandemic.

INTRODUCTION: On March 13, 2020, the U.S. declared COVID-19 to be a national emergency. As communities adopted mitigation strategies, there were potential changes in the trends of injuries treated in emergency department. This study provides national estimates of injury-related emergency department visits in the U.S. before and during the pandemic. METHODS: A secondary retrospective cohort study was conducted using trained, on-site hospital coders collecting data for injury-related emergency department cases from medical records from a nationally representative sample of 66 U.S. hospital emergency departments. Injury emergency department visit estimates in the year before the pandemic (January 1, 2019-December 31, 2019) were compared with estimates of the year of pandemic declaration (January 1, 2020-December 31, 2020) for overall nonfatal injury-related emergency department visits, motor vehicle, falls-related, self-harm-, assault-related, and poisoning-related emergency department visits. RESULTS: There was an estimated 1.7 million (25%) decrease in nonfatal injury-related emergency department visits during April through June 2020 compared with those of the same timeframe in 2019. Similar decreases were observed for emergency department visits because of motor vehicle‒related injuries (199,329; 23.3%) and falls-related injuries (497,971; 25.1%). Monthly 2020 estimates remained relatively in line with 2019 estimates for self-harm‒, assault-, and poisoning-related emergency department visits. CONCLUSIONS: These findings provide updates for clinical and public health practitioners on the changing profile of injury-related emergency department visits during the COVID-19 pandemic. Understanding the short- and long-term impacts of the pandemic is important to preventing future injuries.

Characterizing and Identifying the Prevalence of Web-Based Misinformation Relating to Medication for Opioid Use Disorder: Machine Learning Approach.

BACKGROUND: Expanding access to and use of medication for opioid use disorder (MOUD) is a key component of overdose prevention. An important barrier to the uptake of MOUD is exposure to inaccurate and potentially harmful health misinformation on social media or web-based forums where individuals commonly seek information. There is a significant need to devise computational techniques to describe the prevalence of web-based health misinformation related to MOUD to facilitate mitigation efforts. OBJECTIVE: By adopting a multidisciplinary, mixed methods strategy, this paper aims to present machine learning and natural language analysis approaches to identify the characteristics and prevalence of web-based misinformation related to MOUD to inform future prevention, treatment, and response efforts. METHODS: The team harnessed public social media posts and comments in the English language from Twitter (6,365,245 posts), YouTube (99,386 posts), Reddit (13,483,419 posts), and Drugs-Forum (5549 posts). Leveraging public health expert annotations on a sample of 2400 of these social media posts that were found to be semantically most similar to a variety of prevailing opioid use disorder-related myths based on representational learning, the team developed a supervised machine learning classifier. This classifier identified whether a post's language promoted one of the leading myths challenging addiction treatment: that the use of agonist therapy for MOUD is simply replacing one drug with another. Platform-level prevalence was calculated thereafter by machine labeling all unannotated posts with the classifier and noting the proportion of myth-indicative posts over all posts. RESULTS: Our results demonstrate promise in identifying social media postings that center on treatment myths about opioid use disorder with an accuracy of 91% and an area under the curve of 0.9, including how these discussions vary across platforms in terms of prevalence and linguistic characteristics, with the lowest prevalence on web-based health communities such as Reddit and Drugs-Forum and the highest on Twitter. Specifically, the prevalence of the stated MOUD myth ranged from 0.4% on web-based health communities to 0.9% on Twitter. CONCLUSIONS: This work provides one of the first large-scale assessments of a key MOUD-related myth across multiple social media platforms and highlights the feasibility and importance of ongoing assessment of health misinformation related to addiction treatment.

Identifying Incidents of Public Health Significance Using the National Poison Data System, 2013-2018.

Data System. The American Association of Poison Control Centers (AAPCC) and the Centers for Disease Control and Prevention (CDC) jointly monitor the National Poison Data System (NPDS) for incidents of public health significance (IPHSs).Data Collection/Processing. NPDS is the data repository for US poison centers, which together cover all 50 states, the District of Columbia, and multiple territories. Information from calls to poison centers is uploaded to NPDS in near real time and continuously monitored for specific exposures and anomalies relative to historic data.Data Analysis/Dissemination. AAPCC and CDC toxicologists analyze NPDS-generated anomalies for evidence of public health significance. Presumptive results are confirmed with the receiving poison center to correctly identify IPHSs. Once verified, CDC notifies the state public health department.Implications. During 2013 to 2018, 3.7% of all NPDS-generated anomalies represented IPHSs. NPDS surveillance findings may be the first alert to state epidemiologists of IPHSs. Data are used locally and nationally to enhance situational awareness during a suspected or known public health threat. NPDS improves CDC's national surveillance capacity by identifying early markers of IPHSs.

National surveillance for radiological exposures and intentional potassium iodide and iodine product ingestions in the United States associated with the 2011 Japan radiological incident.

BACKGROUND: In March of 2011, an earthquake struck Japan causing a tsunami that resulted in a radiological release from the damaged Fukushima Daiichi nuclear power plant. Surveillance for potential radiological and any iodine/iodide product exposures was initiated on the National Poison Data System (NPDS) to target public health messaging needs within the United States (US). Our objectives are to describe self-reported exposures to radiation, potassium iodide (KI) and other iodine/iodide products which occurred during the US federal response and discuss its public health impact. METHODS: All calls to poison centers associated with the Japan incident were identified from March 11, 2011 to April 18, 2011 in NPDS. Exposure, demographic and health outcome information were collected. Calls about reported radiation exposures and KI or other iodine/iodide product ingestions were then categorized with regard to exposure likelihood based on follow-up information obtained from the PC where each call originated. Reported exposures were subsequently classified as probable exposures (high likelihood of exposure), probable non-exposures (low likelihood of exposure), and suspect exposure (unknown likelihood of exposure). RESULTS: We identified 400 calls to PCs associated with the incident, with 340 information requests (no exposure reported) and 60 reported exposures. The majority (n = 194; 57%) of the information requests mentioned one or more substances. Radiation was inquired about most frequently (n = 88; 45%), followed by KI (n = 86; 44%) and other iodine/iodide products (n = 47; 24%). Of the 60 reported exposures, KI was reported most frequently (n = 25; 42%), followed by radiation (n = 22; 37%) and other iodine/iodide products (n = 13; 22%). Among reported KI exposures, most were classified as probable exposures (n = 24; 96%); one was a probable non-exposure. Among reported other iodine/iodide product exposures, most were probable exposures (n = 10, 77%) and the rest were suspect exposures (n = 3; 23%). The reported radiation exposures were classified as suspect exposures (n = 16, 73%) or probable non-exposures (n = 6; 27%). No radiation exposures were classified as probable exposures. A small number of the probable exposures to KI and other iodide/iodine products reported adverse signs or symptoms (n = 9; 26%). The majority of probable exposures had no adverse outcomes (n = 28; 82%). These data identified a potential public health information gap regarding KI and other iodine/iodide products which was then addressed through public health messaging activities. CONCLUSION: During the Japan incident response, surveillance activities using NPDS identified KI and other iodine/iodide products as potential public health concerns within the US, which guided CDC's public health messaging and communication activities. Regional PCs can provide timely and additional information during a public health emergency to enhance data collected from surveillance activities, which in turn can be used to inform public health decision-making.

Using poison center data for national public health surveillance for chemical and poison exposure and associated illness.

The National Poison Data System (NPDS) is a national near-real-time surveillance system that improves situational awareness for chemical and poison exposures, according to data from US poison centers. NPDS is the successor to the Toxic Exposure Surveillance System. The Centers for Disease Control and Prevention (CDC) use these data, which are owned and managed by the American Association of Poison Control Centers, to improve public health surveillance for chemical and poison exposures and associated illness, identify early markers of chemical events, and enhance situational awareness during outbreaks. Information recorded in this database is from self-reported calls from the public or health care professionals. In 2009, NPDS detected 22 events of public health significance and CDC used the system to monitor several multistate outbreaks. One of the limitations of the system is that exposures do not necessarily represent a poisoning. Incorporating NPDS data into the public health surveillance network and subsequently using NPDS to rapidly identify chemical and poison exposures exemplifies the importance of the poison centers and NPDS to public health surveillance. This integration provides the opportunity to improve the public health response to chemical and poison exposures, minimizes morbidity and mortality, and serves as an important step forward in surveillance technology and integration.