Search

OBJECTIVES: To develop proof-of-concept algorithms using alternative approaches to capture provider sentiment in ICU notes. DESIGN: Retrospective observational cohort study. SETTING: The Multiparameter Intelligent Monitoring of Intensive Care III (MIMIC-III) and the University of California, San Francisco (UCSF) deidentified notes databases. PATIENTS: Adult (≥18 yr old) patients admitted to the ICU. MEASUREMENTS AND MAIN RESULTS: We developed two sentiment models: 1) a keywords-based approach using a consensus-based clinical sentiment lexicon comprised of 72 positive and 103 negative phrases, including negations and 2) a Decoding-enhanced Bidirectional Encoder Representations from Transformers with disentangled attention-v3-based deep learning model (keywords-independent) trained on clinical sentiment labels. We applied the models to 198,944 notes across 52,997 ICU admissions in the MIMIC-III database. Analyses were replicated on an external sample of patients admitted to a UCSF ICU from 2018 to 2019. We also labeled sentiment in 1,493 note fragments and compared the predictive accuracy of our tools to three popular sentiment classifiers. Clinical sentiment terms were found in 99% of patient visits across 88% of notes. Our two sentiment tools were substantially more predictive (Spearman correlations of 0.62-0.84, p values < 0.00001) of labeled sentiment compared with general language algorithms (0.28-0.46). CONCLUSION: Our exploratory healthcare-specific sentiment models can more accurately detect positivity and negativity in clinical notes compared with general sentiment tools not designed for clinical usage.

We investigated the content of liberal and conservative news media Facebook posts on race and ethnic health disparities. A total of 3,327,360 liberal and conservative news Facebook posts from the United States (US) from January 2015 to May 2022 were collected from the Crowd Tangle platform and filtered for race and health-related keywords. Qualitative content analysis was conducted on a random sample of 1750 liberal and 1750 conservative posts. Posts were analyzed for a continuum of hate speech using a newly developed method combining faceted Rasch item response theory with deep learning. Across posts referencing Asian, Black, Latinx, Middle Eastern, and immigrants/refugees, liberal news posts had lower hate scores compared to conservative posts. Liberal news posts were more likely to acknowledge and detail the existence of racial/ethnic health disparities, while conservative news posts were more likely to highlight the negative consequences of protests, immigration, and the disenfranchisement of Whites. Facebook posts from liberal and conservative news focus on different themes with fewer discussions of racial inequities in conservative news. Investigating the discourse on race and health in social media news posts may inform our understanding of the public's exposure to and knowledge of racial health disparities, and policy-level support for ameliorating these disparities.

Suicide risk is elevated among military service members who recently transitioned to civilian life. Identifying high-risk service members before this transition could facilitate provision of targeted preventive interventions. We investigated the feasibility of doing this by attempting to develop a prediction model for self-reported suicide attempts (SAs) after leaving or being released from active duty in the Study to Assess Risk and Resilience in Servicemembers-Longitudinal Study (STARRS-LS). This study included two self-report panel surveys (LS1: 2016-2018, LS2: 2018-2019) administered to respondents who previously participated while on active duty in one of three Army STARRS 2011-2014 baseline self-report surveys. We focus on respondents who left active duty >12 months before their LS survey (n = 8899). An ensemble machine learning model using predictors available prior to leaving active duty was developed in a 70% training sample and validated in a 30% test sample. The 12-month self-reported SA prevalence (SE) was 1.0% (0.1). Test sample AUC (SE) was 0.74 (0.06). The 15% of respondents with highest predicted risk included nearly two-thirds of 12-month SAs and over 80% of medically serious 12-month SAs. These results show that it is possible to identify soldiers at high post-transition self-report SA risk before the transition. Future model development is needed to examine prediction of SAs assessed by administrative data and using surveys administered closer to the time of leaving active duty.

BACKGROUND: A recently developed obstetrical comorbidity scoring system enables the comparison of severe maternal morbidity rates independent of health status at the time of birth hospitalization. However, the scoring system has not been evaluated in racial-ethnic and socioeconomic groups or used to assess disparities in severe maternal morbidity. OBJECTIVE: This study aimed to evaluate the performance of an obstetrical comorbidity scoring system when applied across racial-ethnic and socioeconomic groups and to determine the effect of comorbidity score risk adjustment on disparities in severe maternal morbidity. STUDY DESIGN: We analyzed a population-based cohort of live births that occurred in California during 2011 through 2017 with linked birth certificates and birth hospitalization discharge data (n=3,308,554). We updated a previously developed comorbidity scoring system to include the International Classification of Diseases, Ninth and Tenth Revisions, Clinical Modifications diagnosis codes and applied the scoring system to subpopulations (groups) defined by race-ethnicity, nativity, payment method, and educational attainment. We then calculated the risk-adjusted rates of severe maternal morbidity (including and excluding blood transfusion-only cases) for each group and estimated the disparities for these outcomes before and after adjustment for the comorbidity score using logistic regression. RESULTS: The obstetric comorbidity scores performed consistently across groups (C-statistics ranged from 0.68 to 0.76; calibration curves demonstrated overall excellent prediction of absolute risk). All non-White groups had significantly elevated rates of severe maternal morbidity before and after risk adjustment for comorbidities when compared with the White group (1.3% before, 1.3% after) (American Indian-Alaska Native: 2.1% before, 1.8% after; Asian: 1.5% before, 1.7% after; Black: 2.5% before, 2.0% after; Latinx: 1.6% before, 1.7% after; Pacific Islander: 2.2% before, 1.9% after; and multi-race groups: 1.7% before, 1.6% after). Risk adjustment also modestly increased disparities for the foreign-born group and government insurance groups. Higher educational attainment was associated with decreased severe maternal morbidity rates, which was largely unaffected by comorbidity risk adjustment. The pattern of results was the same whether or not transfusion-only cases were included as severe maternal morbidity. CONCLUSION: These results support the use of an updated comorbidity scoring system to assess disparities in severe maternal morbidity. Disparities in severe maternal morbidity decreased in magnitude for some racial-ethnic and socioeconomic groups and increased in magnitude for other groups after adjustment for the comorbidity score.

BackgroundIncreasing influenza vaccination coverage in school-aged children may substantially reduce community transmission. School-located influenza vaccinations (SLIV) aim to promote vaccinations by increasing accessibility, which may be especially beneficial to race/ethnicity groups that face high barriers to preventative care. Here, we evaluate the effectiveness of a city-wide SLIV program by race/ethnicity from 2014 to 2018.MethodsWe used multivariate matching to pair schools in the intervention district in Oakland, CA with schools in a comparison district in West Contra Costa County, CA. We distributed cross-sectional surveys to measure caregiver-reported student vaccination status and estimated differences in vaccination coverage levels and reasons for non-vaccination between districts stratifying by race/ethnicity. We estimated difference-in-differences (DID) of laboratory confirmed influenza hospitalization incidence between districts stratified by race/ethnicity using surveillance data.ResultsDifferences in influenza vaccination coverage in the intervention vs. comparison district were larger among White (2017-18: 21.0% difference [95% CI: 9.7%, 32.3%]) and Hispanic/Latino (13.4% [8.8%, 18.0%]) students than Asian/Pacific Islander (API) (8.9% [1.3%, 16.5%]), Black (5.9% [-2.2%, 14.0%]), and multiracial (6.3% [-1.8%, 14.3%)) students. Concerns about vaccine effectiveness or safety were more common among Black and multiracial caregivers. Logistical barriers were less common in the intervention vs. comparison district, with the largest difference among White students. In both districts, hospitalizations in 2017-18 were higher in Blacks (Intervention: 111.5 hospitalizations per 100,00; Comparison: 134.1 per 100,000) vs. other races/ethnicities. All-age influenza hospitalization incidence was lower in the intervention site vs. comparison site among White/API individuals in 2016-17 (DID -25.14 per 100,000 [95% CI: -40.14, -10.14]) and 2017-18 (-36.6 per 100,000 [-52.7, -20.5]) and Black older adults in 2017-18 (-282.2 per 100,000 (-508.4, -56.1]), but not in other groups.ConclusionsSLIV was associated with higher vaccination coverage and lower influenza hospitalization, but associations varied by race/ethnicity. SLIV alone may be insufficient to ensure equitable influenza outcomes.

BackgroundWe measured the effectiveness of a city-wide school-located influenza vaccination (SLIV) program implemented in over 102 elementary schools in Oakland, California.MethodsWe conducted a retrospective cohort study among Kaiser Permanente Northern California (KPNC) members of all ages residing in either the intervention or a multivariate-matched comparison site from September 2011 - August 2017. Outcomes included medically attended acute respiratory illness (MAARI), influenza hospitalization, and Oseltamivir prescriptions. We estimated difference-in-differences (DIDs) in 2014-15, 2015-16, and 2016-17 using generalized linear models and adjusted for race, ethnicity, age, sex, health plan, and language.ResultsPre-intervention member characteristics were similar between sites. The proportion of KPNC members vaccinated for influenza by KPNC or the SLIV program was 8-11% higher in the intervention site than the comparison site during the intervention period. Among school-aged children, SLIV was associated with lower Oseltamivir prescriptions per 1,000 (DIDs: -3.5 (95% CI -5.5, -1.5) in 2015-16; -4.0 (95% CI -6.5, -1.6) in 2016-17) but not with other outcomes. SLIV was associated with lower MAARI per 1,000 in adults 65 + years (2014-15: -13.2, 95% CI -23.2, -3.2; 2015-16: -21.5, 95% CI -31.1, -11.9; 2016-17: -13.0, 95% CI -23.2, -2.9). There were few significant associations with other outcomes among adults.ConclusionsA city-wide SLIV intervention was associated with higher influenza vaccination coverage, lower Oseltamivir prescriptions in school-aged children, and lower MAARI among people over 65 years, suggesting possible indirect effects of SLIV among older adults.

BACKGROUND:It is estimated that vaccinating 50%-70% of school-aged children for influenza can produce population-wide indirect effects. We evaluated a city-wide school-located influenza vaccination (SLIV) intervention that aimed to increase influenza vaccination coverage. The intervention was implemented in ≥95 preschools and elementary schools in northern California from 2014 to 2018. Using a matched cohort design, we estimated intervention impacts on student influenza vaccination coverage, school absenteeism, and community-wide indirect effects on laboratory-confirmed influenza hospitalizations. METHODS AND FINDINGS:We used a multivariate matching algorithm to identify a nearby comparison school district with pre-intervention characteristics similar to those of the intervention school district and matched schools in each district. To measure student influenza vaccination, we conducted cross-sectional surveys of student caregivers in 22 school pairs (2017 survey, N = 6,070; 2018 survey, N = 6,507). We estimated the incidence of laboratory-confirmed influenza hospitalization from 2011 to 2018 using surveillance data from school district zip codes. We analyzed student absenteeism data from 2011 to 2018 from each district (N = 42,487,816 student-days). To account for pre-intervention differences between districts, we estimated difference-in-differences (DID) in influenza hospitalization incidence and absenteeism rates using generalized linear and log-linear models with a population offset for incidence outcomes. Prior to the SLIV intervention, the median household income was $51,849 in the intervention site and $61,596 in the comparison site. The population in each site was predominately white (41% in the intervention site, 48% in the comparison site) and/or of Hispanic or Latino ethnicity (26% in the intervention site, 33% in the comparison site). The number of students vaccinated by the SLIV intervention ranged from 7,502 to 10,106 (22%-28% of eligible students) each year. During the intervention, influenza vaccination coverage among elementary students was 53%-66% in the comparison district. Coverage was similar between the intervention and comparison districts in influenza seasons 2014-2015 and 2015-2016 and was significantly higher in the intervention site in seasons 2016-2017 (7%; 95% CI 4, 11; p < 0.001) and 2017-2018 (11%; 95% CI 7, 15; p < 0.001). During seasons when vaccination coverage was higher among intervention schools and the vaccine was moderately effective, there was evidence of statistically significant indirect effects: The DID in the incidence of influenza hospitalization per 100,000 in the intervention versus comparison site was -17 (95% CI -30, -4; p = 0.008) in 2016-2017 and -37 (95% CI -54, -19; p < 0.001) in 2017-2018 among non-elementary-school-aged individuals and -73 (95% CI -147, 1; p = 0.054) in 2016-2017 and -160 (95% CI -267, -53; p = 0.004) in 2017-2018 among adults 65 years or older. The DID in illness-related school absences per 100 school days during the influenza season was -0.63 (95% CI -1.14, -0.13; p = 0.014) in 2016-2017 and -0.80 (95% CI -1.28, -0.31; p = 0.001) in 2017-2018. Limitations of this study include the use of an observational design, which may be subject to unmeasured confounding, and caregiver-reported vaccination status, which is subject to poor recall and low response rates. CONCLUSIONS:A city-wide SLIV intervention in a large, diverse urban population was associated with a decrease in the incidence of laboratory-confirmed influenza hospitalization in all age groups and a decrease in illness-specific school absence rate among students in 2016-2017 and 2017-2018, seasons when the vaccine was moderately effective, suggesting that the intervention produced indirect effects. Our findings suggest that in populations with moderately high background levels of influenza vaccination coverage, SLIV programs are associated with further increases in coverage and reduced influenza across the community.

E-cigarette use is increasing dramatically among adolescents as social media marketing portrays "vaping" products as healthier alternatives to conventional cigarettes. In September 2018, the Food and Drug Administration (FDA) launched an anti-vaping campaign, in U.S. high schools, on social media and other platforms, emphasizing "The Real Cost" of e-cigarettes. Using a novel deep learning approach, we assessed changes in vaping-related content on Instagram from 2017 to 2019 and drew an inference about the initial impact of the FDA's Real Cost campaign on Instagram. We collected 245,894 Instagram posts that used vaping-related hashtags (e.g., #vape, #ejuice) in four samples from 2017 to 2019. We compared the "like" count from these posts before and after the FDA campaign. We used deep learning image classification to analyze 49,655 Instagram image posts, separating images of men, women, and vaping devices. We also conducted text analysis and topic modeling to detect the common words and themes in the posted captions. Since September 2018, the FDA-sponsored hashtag #TheRealCost has been used about 50 times per month on Instagram, whereas vaping-related hashtags we tracked were used up to 10,000 times more often. Comparing the pre-intervention (2017, 2018) and post-intervention (2019) samples of vaping-related Instagram posts, we found a three-fold increase in the median "like" count (10 vs. 28) and a 6-fold increase in the proportion of posts with more than 100 likes (2 vs. 15%). Over 70% of Instagram vaping images featured e-juices and devices, with a growing number of images depicting a "pod," the type of discrete vaping device that delivers high concentration of nicotine and is favored by novice e-cigarette users. In addition, the Instagram analytics data shared by the vaping influencers we interviewed showed underage Instagram users among their followers.

AbstractBackgroundIt is estimated that vaccinating 50-70% of school-aged children for influenza can produce population-wide indirect effects. We evaluated a city-wide, school-located influenza vaccination (SLIV) intervention that aimed to increase influenza vaccination coverage. The intervention was implemented in over 95 pre-schools and elementary schools in northern California from 2014 to 2018. Using a matched prospective cohort design, we estimated intervention impacts on student influenza vaccination coverage, school absenteeism, and community-wide indirect effects on laboratory-confirmed influenza hospitalizations.Methods and FindingsWe used a multivariate matching algorithm to identify a nearby comparison school district with similar pre-intervention characteristics and matched schools in each district. To measure student influenza vaccination, we conducted cross-sectional surveys of student caregivers in 22 school pairs (2016 survey N = 6,070; 2017 survey N = 6,507). We estimated the incidence of laboratory-confirmed influenza hospitalization from 2011-2018 using surveillance data from school district zip codes. We analyzed student absenteeism data from 2011-2018 from each district (N = 42,487,816 student-days). To account for pre-intervention differences between districts, we estimated difference-in-differences (DID) in influenza hospitalization incidence and absenteeism rates using generalized linear and log-linear models with a population offset for incidence outcomes.The number of students vaccinated by the SLIV intervention ranged from 7,502 to 10,106 (22-28% of eligible students) each year. During the intervention, influenza vaccination coverage among elementary students was 53-66% in the comparison district. Coverage was similar between the intervention and comparison districts in 2014-15 and 2015-16 and was significantly higher in the intervention site in 2016-17 (7% 95% CI 4, 11) and 2017-18 (11% 95% CI 7, 15). During seasons when vaccination coverage was higher among intervention schools and the vaccine was moderately effective, there was evidence of statistically significant indirect effects: adjusting for pre-intervention differences between districts, the reduction in influenza hospitalizations in the intervention site was 76 (95% CI 20, 133) in 2016-17 and 165 (95% CI 86, 243) in 2017-18 among non-elementary school aged individuals and 327 (5, 659) in 2016-17 and 715 (236, 1195) in 2017-18 among adults 65 years or older. The reduction in illness-related school absences during influenza season was 3,538 (95% CI 709, 6,366) in 2016-17 and 8,249 (95% CI 3,213, 13,285) in 2017-18. Limitations of this study include the use of an observational design, which may be subject to unmeasured confounding, and caregiver-reported vaccination status, which is subject to poor recall and low response rates.ConclusionA city-wide SLIV intervention in a large, diverse urban population decreased the incidence of laboratory-confirmed influenza hospitalization in all age groups and decreased illness-specific school absence rates among students during seasons when the vaccine was moderately effective, suggesting that the intervention produced indirect effects. Our findings suggest that in populations with moderately high background levels of influenza vaccination coverage, SLIV programs can further increase coverage and reduce influenza across communities.

This prognostic study develops a model based on administrative data for regular US Army soldiers that can predict suicides 1 to 120 months after leaving active service. Key Points: Question: Can suicides after leaving active US Army service be predicted from administrative data available prior to leaving? Findings: This prognostic study showed that suicides after leaving active service can be predicted with moderate to good accuracy using administrative data available before leaving service. The 10% of soldiers with highest predicted risk accounted for 30.7% to 46.6% of all suicides across horizons. Meaning: These results demonstrate that this model could facilitate targeted delivery of a high-risk posttransition suicide prevention intervention to soldiers who were identified before leaving active service. Importance: The suicide rate of military servicemembers increases sharply after returning to civilian life. Identifying high-risk servicemembers before they leave service could help target preventive interventions. Objective: To develop a model based on administrative data for regular US Army soldiers that can predict suicides 1 to 120 months after leaving active service. Design, Setting, and Participants: In this prognostic study, a consolidated administrative database was created for all regular US Army soldiers who left service from 2010 through 2019. Machine learning models were trained to predict suicides over the next 1 to 120 months in a random 70% training sample. Validation was implemented in the remaining 30%. Data were analyzed from March 2023 through March 2024. Main outcome and measures: The outcome was suicide in the National Death Index. Predictors came from administrative records available before leaving service on sociodemographics, Army career characteristics, psychopathologic risk factors, indicators of physical health, social networks and supports, and stressors. Results: Of the 800 579 soldiers in the cohort (84.9% male; median [IQR] age at discharge, 26 [23-33] years), 2084 suicides had occurred as of December 31, 2019 (51.6 per 100 000 person-years). A lasso model assuming consistent slopes over time discriminated as well over all but the shortest risk horizons as more complex stacked generalization ensemble machine learning models. Test sample area under the receiver operating characteristic curve ranged from 0.87 (SE = 0.06) for suicides in the first month after leaving service to 0.72 (SE = 0.003) for suicides over 120 months. The 10% of soldiers with highest predicted risk accounted for between 30.7% (SE = 1.8) and 46.6% (SE = 6.6) of all suicides across horizons. Calibration was for the most part better for the lasso model than the super learner model (both estimated over 120-month horizons.) Net benefit of a model-informed prevention strategy was positive compared with intervene-with-all or intervene-with-none strategies over a range of plausible intervention thresholds. Sociodemographics, Army career characteristics, and psychopathologic risk factors were the most important classes of predictors. Conclusions and relevance: These results demonstrated that a model based on administrative variables available at the time of leaving active Army service can predict suicides with meaningful accuracy over the subsequent decade. However, final determination of cost-effectiveness would require information beyond the scope of this report about intervention content, costs, and effects over relevant horizons in relation to the monetary value placed on preventing suicides. [ABSTRACT FROM AUTHOR]

Background and Hypothesis Psychosis-associated diagnostic codes are increasingly being utilized as case definitions for electronic health record (EHR)-based algorithms to predict and detect psychosis. However, data on the validity of psychosis-related diagnostic codes is limited. We evaluated the positive predictive value (PPV) of International Classification of Diseases (ICD) codes for psychosis. Study Design Using EHRs at 3 health systems, ICD codes comprising primary psychotic disorders and mood disorders with psychosis were grouped into 5 higher-order groups. 1133 records were sampled for chart review using the full EHR. PPVs (the probability of chart-confirmed psychosis given ICD psychosis codes) were calculated across multiple treatment settings. Study Results PPVs across all diagnostic groups and hospital systems exceeded 70%: Mass General Brigham 0.72 [95% CI 0.68–0.77], Boston Children's Hospital 0.80 [0.75–0.84], and Boston Medical Center 0.83 [0.79–0.86]. Schizoaffective disorder PPVs were consistently the highest across sites (0.80–0.92) and major depressive disorder with psychosis were the most variable (0.57–0.79). To determine if the first documented code captured first-episode psychosis (FEP), we excluded cases with prior chart evidence of a diagnosis of or treatment for a psychotic illness, yielding substantially lower PPVs (0.08–0.62). Conclusions We found that the first documented psychosis diagnostic code accurately captured true episodes of psychosis but was a poor index of FEP. These data have important implications for the case definitions used in the development of risk prediction models designed to predict or detect undiagnosed psychosis. [ABSTRACT FROM AUTHOR]

Vietnamese Americans have high rates of hepatitis B virus (HBV) infection but low rates of knowledge and screening. A population-based survey conducted in 2011 of Vietnamese Americans in two geographic areas (n = 1666) was analyzed. The outcome variables were having heard of HBV and a score summarizing knowledge of HBV transmission. Most respondents (86.0%) had heard of HBV. Correct knowledge of transmission ranged from 59.5% for sex, 68.1% for sharing toothbrushes, 78.6% for during birth, and 85.0% for sharing needles. In multivariable analyses, factors associated with having heard of HBV and higher knowledge included Northern California residence, longer U.S. residence, higher education, family history of HBV, and discussing HBV with family/friends. Higher income was associated with having heard of HBV. English fluency and being U.S.-born were associated with higher knowledge. Interventions to increase knowledge of HBV transmission are needed to decrease this health disparity among Vietnamese Americans.

Outreach and engagement that connects the Patton College at Ohio University with P-12 schools has been a strong tradition in the Southeastern Ohio/Appalachian region. In the mid-1980s, a partnership aligned with the Coalition of Essential Schools and 9 "Common Principles"' was one of the first. Alignment with 19 "Postulates" of the National Network for Educational Renewal followed. The seeds had been planted, Ohio University established a Center for Partnerships, and several Professional Development Schools began to emerge and flourish. In 2008 a new alignment with NAPDS, incorporating the "9 Essentials" into the PDS practices was welcomed. Most recently, the Blue Ribbon Panel Report commissioned by NCATE (2010) brought "10 Design Principles for Clinically Based Preparation." In 2013 full membership in the National Network for Education Renewal (NNER) was initiated. In 2014 PDS Partnerships with the Athens City Schools received the NAPDS Award for Exemplary Professional Development School Achievement. The purpose in this article is to share highlights drawn from the application for this award.

Introduction Instagram and TikTok, video-based social media platforms popular among adolescents, contain tobacco-related content despite the platforms' policies prohibiting substance-related posts. Prior research identified themes in e-cigarette-related social media posts using qualitative or text-based machine learning methods. We developed an image-based computer vision model to identify e-cigarette products in social media images and videos. Aims and Methods We created a data set of 6999 Instagram images labeled for 8 object classes: mod or pod devices, e-juice containers, packaging boxes, nicotine warning labels, e-juice flavors, e-cigarette brand names, and smoke clouds. We trained a DyHead object detection model using a Swin-Large backbone, evaluated the model's performance on 20 Instagram and TikTok videos, and applied the model to 14 072 e-cigarette-related promotional TikTok videos (2019–2022; 10 276 485 frames). Results The model achieved the following mean average precision scores on the image test set: e-juice container: 0.89; pod device: 0.67; mod device: 0.54; packaging box: 0.84; nicotine warning label: 0.86; e-cigarette brand name: 0.71; e-juice flavor name: 0.89; and smoke cloud: 0.46. The prevalence of pod devices in promotional TikTok videos increased by 15% from 2019 to 2022. The prevalence of e-juices increased by 33% from 2021 to 2022. The prevalence of e-juice flavor names and e-cigarette brand names increased by about 100% from 2019 to 2022. Conclusions Deep learning-based object detection technology enables automated analysis of visual posts on social media. Our computer vision model can detect the presence of e-cigarettes products in images and videos, providing valuable surveillance data for tobacco regulatory science (TRS). Implications Prior research identified themes in e-cigarette-related social media posts using qualitative or text-based machine learning methods. We developed an image-based computer vision model to identify e-cigarette products in social media images and videos. We trained a DyHead object detection model using a Swin-Large backbone, evaluated the model's performance on 20 Instagram and TikTok videos featuring at least two e-cigarette objects, and applied the model to 14 072 e-cigarette-related promotional TikTok videos (2019–2022; 10 276 485 frames). The deep learning model can be used for automated, scalable surveillance of image- and video-based e-cigarette-related promotional content on social media, providing valuable data for TRS. Social media platforms could use computer vision to identify tobacco-related imagery and remove it promptly, which could reduce adolescents' exposure to tobacco content online. [ABSTRACT FROM AUTHOR]