State and Federal Legislators' Responses on Social Media to the Mental Health and Burnout of Health Care Workers Throughout the COVID-19 Pandemic: Natural Language Processing and Sentiment Analysis.

Background: Burnout and the mental health burden of the COVID-19 pandemic have disproportionately impacted health care workers. The links between state policies, federal regulations, COVID-19 case counts, strains on health care systems, and the mental health of health care workers continue to evolve. The language used by state and federal legislators in public-facing venues such as social media is important, as it impacts public opinion and behavior, and it also reflects current policy-leader opinions and planned legislation. Objective: The objective of this study was to examine legislators' social media content on Twitter and Facebook throughout the COVID-19 pandemic to thematically characterize policy makers' attitudes and perspectives related to mental health and burnout in the health care workforce. Methods: Legislators' social media posts about mental health and burnout in the health care workforce were collected from January 2020 to November 2021 using Quorum, a digital database of policy-related documents. The total number of relevant social media posts per state legislator per calendar month was calculated and compared with COVID-19 case volume. Differences between themes expressed in Democratic and Republican posts were estimated using the Pearson chi-square test. Words within social media posts most associated with each political party were determined. Machine-learning was used to evaluate naturally occurring themes in the burnout- and mental health-related social media posts. Results: A total of 4165 social media posts (1400 tweets and 2765 Facebook posts) were generated by 2047 unique state and federal legislators and 38 government entities. The majority of posts (n=2319, 55.68%) were generated by Democrats, followed by Republicans (n=1600, 40.34%). Among both parties, the volume of burnout-related posts was greatest during the initial COVID-19 surge. However, there was significant variation in the themes expressed by the 2 major political parties. Themes most correlated with Democratic posts were (1) frontline care and burnout, (2) vaccines, (3) COVID-19 outbreaks, and (4) mental health services. Themes most correlated with Republican social media posts were (1) legislation, (2) call for local action, (3) government support, and (4) health care worker testing and mental health. Conclusions: State and federal legislators use social media to share opinions and thoughts on key topics, including burnout and mental health strain among health care workers. Variations in the volume of posts indicated that a focus on burnout and the mental health of the health care workforce existed early in the pandemic but has waned. Significant differences emerged in the content posted by the 2 major US political parties, underscoring how each prioritized different aspects of the crisis.

Association Between Online Reviews of Substance Use Disorder Treatment Facilities and Drug-Induced Mortality Rates: Cross-Sectional Analysis.

BACKGROUND: Drug-induced mortality across the United States has continued to rise. To date, there are limited measures to evaluate patient preferences and priorities regarding substance use disorder (SUD) treatment, and many patients do not have access to evidence-based treatment options. Patients and their families seeking SUD treatment may begin their search for an SUD treatment facility online, where they can find information about individual facilities, as well as a summary of patient-generated web-based reviews via popular platforms such as Google or Yelp. Web-based reviews of health care facilities may reflect information about factors associated with positive or negative patient satisfaction. The association between patient satisfaction with SUD treatment and drug-induced mortality is not well understood. OBJECTIVE: The objective of this study was to examine the association between online review content of SUD treatment facilities and drug-induced state mortality. METHODS: A cross-sectional analysis of online reviews and ratings of Substance Abuse and Mental Health Services Administration (SAMHSA)-designated SUD treatment facilities listed between September 2005 and October 2021 was conducted. The primary outcomes were (1) mean online rating of SUD treatment facilities from 1 star (worst) to 5 stars (best) and (2) average drug-induced mortality rates from the Centers for Disease Control and Prevention (CDC) WONDER Database (2006-2019). Clusters of words with differential frequencies within reviews were identified. A 3-level linear model was used to estimate the association between online review ratings and drug-induced mortality. RESULTS: A total of 589 SAMHSA-designated facilities (n=9597 reviews) were included in this study. Drug-induced mortality was compared with the average. Approximately half (24/47, 51%) of states had below average ("low") mortality rates (mean 13.40, SD 2.45 deaths per 100,000 people), and half (23/47, 49%) had above average ("high") drug-induced mortality rates (mean 21.92, SD 3.69 deaths per 100,000 people). The top 5 themes associated with low drug-induced mortality included detoxification and addiction rehabilitation services (r=0.26), gratitude for recovery (r=-0.25), thankful for treatment (r=-0.32), caring staff and amazing experience (r=-0.23), and individualized recovery programs (r=-0.20). The top 5 themes associated with high mortality were care from doctors or providers (r=0.24), rude and insensitive care (r=0.23), medication and prescriptions (r=0.22), front desk and reception experience (r=0.22), and dissatisfaction with communication (r=0.21). In the multilevel linear model, a state with a 10 deaths per 100,000 people increase in mortality was associated with a 0.30 lower average Yelp rating (P=.005). CONCLUSIONS: Lower online ratings of SUD treatment facilities were associated with higher drug-induced mortality at the state level. Elements of patient experience may be associated with state-level mortality. Identified themes from online, organically derived patient content can inform efforts to improve high-quality and patient-centered SUD care.

Association Between Crowdsourced Health Care Facility Ratings and Mortality in US Counties.

Importance: Mortality across US counties varies considerably, from 252 to 1847 deaths per 100 000 people in 2018. Although patient satisfaction with health care is associated with patient- and facility-level health outcomes, the association between health care satisfaction and community-level health outcomes is not known. Objective: To examine the association between online ratings of health care facilities and mortality across US counties and to identify language specific to 1-star (lowest rating) and 5-star (highest rating) reviews in counties with high vs low mortality. Design, Setting, and Participants: This retrospective population-based cross-sectional study examined reviews and ratings of 95 120 essential health care facilities across 1301 US counties. Counties that had at least 1 essential health care facility with reviews available on Yelp, an online review platform, were included. Essential health care was defined according to the 10 essential health benefits covered by Affordable Care Act insurance plans. Main Outcomes and Measures: The mean rating of essential health care facilities was calculated by county from January 1, 2015, to December 31, 2019. Ratings were on a scale of 1 to 5 stars, with 1 being the worst rating and 5 the best. County-level composite measures of health behaviors, clinical care, social and economic factors, and physical environment were obtained from the University of Wisconsin School of Medicine and Public Health County Health Rankings database. The 2018 age-adjusted mortality by county was obtained from the Centers for Disease Control and Prevention Wide-ranging Online Data for Epidemiological Research database. Multiple linear regression analysis was used to estimate the association between mean facility rating and mortality, adjusting for county health ranking variables. Words with frequencies of use that were significantly different across 1-star and 5-star reviews in counties with high vs low mortality were identified. Results: The 95 120 facilities meeting inclusion criteria were distributed across 1301 of 3142 US counties (41.4%). At the county level, a 1-point increase in mean rating was associated with a mean (SE) age-adjusted decrease of 18.05 (3.68) deaths per 100 000 people (P < .001). Words specific to 1-star reviews in high-mortality counties included told, rude, and wait, and words specific to 5-star reviews in low-mortality counties included Dr, pain, and professional. Conclusions and Relevance: This study found that, at the county level, higher online ratings of essential health care facilities were associated with lower mortality. Equivalent online ratings did not necessarily reflect equivalent experiences of care across counties with different mortality levels, as evidenced by variations in the frequency of use of key words in reviews. These findings suggest that online ratings and reviews may provide insight into unequal experiences of essential health care.

Online Reviews of Mental Health Treatment Facilities: Narrative Themes Associated With Positive and Negative Ratings.

OBJECTIVE: Previous studies indicate that patients' satisfaction with mental health care is correlated with both treatment outcomes and quality of life. The aims of this study were to describe online reviews of mental health treatment facilities, including key themes in review content, and to evaluate the correlation between narrative review themes, facility characteristics, and review ratings. METHODS: United States National Mental Health Services Survey (N-MHSS) facilities were linked to corresponding Yelp pages, created between March 2007 and September 2019. Correlations between review ratings and both machine learning-generated latent Dirichlet allocation topics and N-MHSS-reported facility characteristics were measured by using Spearman's rank-order correlation coefficient. Significance was defined by a Bonferroni-adjusted p<0.001. RESULTS: Of 10,191 unique mental health treatment facilities, 1,383 (13.6%) had relevant Yelp pages with 8,133 corresponding reviews. The number of newly reviewed facilities and the number of new reviews increased throughout the study period. Narrative topics positively correlated with review ratings included caring staff (Spearman's ρ=0.39) and nonpharmacologic treatment (ρ=0.16). Topics negatively correlated with review ratings included rude staff (ρ=-0.14) and safety and abuse (ρ=-0.14). Of 126 N-MHSS survey items, 11 were positively correlated with review rating, including "outpatient mental health facility" (ρ=0.13), and 33 were negatively correlated with review rating, including accepting Medicare (ρ=-0.21). CONCLUSIONS: Narrative topics provide information beyond what is currently collected through the N-MHSS. Topics associated with positive and negative reviews, such as staff attitude toward patients, can guide improvement in patients' satisfaction and engagement with mental health care.

Providing Incentive for Emergency Physician X-Waiver Training: An Evaluation of Program Success and Postintervention Buprenorphine Prescribing.

STUDY OBJECTIVE: Emergency department (ED) initiation of buprenorphine for patients with opioid use disorder increases treatment engagement but remains an uncommon practice. One important barrier to ED-initiated buprenorphine is the additional training requirement (X waiver). Our objective is to evaluate the influence of a financial incentive program on emergency physician completion of X-waiver training. Secondary objectives are to evaluate the program's effect on buprenorphine prescribing and to explore physician attitudes toward the incentive. METHODS: We conducted a prospective, observational cohort study set in 3 urban academic EDs before and after implementation of a financial incentive program providing $750 for completion of X-waiver training. We describe program participation as well as rates of buprenorphine prescribing per opioid use disorder-related encounter before and after the intervention period, using electronic health record data. We also completed a postintervention physician survey assessing attitudes about the incentive program. RESULTS: Overall, 89% of eligible emergency physicians (56/63) completed the X-waiver training during the 6-week incentive period. In the 5 months after the incentive, buprenorphine prescribing per opioid use disorder-related encounter increased from 0.5% to 16% (Δ 15%; 95% confidence interval 10.6% to 19.9%), with substantial variability across sites (range 8% to 22% of opioid use disorder-related encounters). In a postintervention survey, 67% of participating physicians indicated that they would have completed the training for a lower amount. CONCLUSION: A financial incentive paying approximately half the clinical rate was effective in promoting emergency physician X-waiver training. The effect on ED-based buprenorphine prescribing was positive but variable across sites, and likely dependent on the availability of additional supports.

Characterizing Tweet Volume and Content About Common Health Conditions Across Pennsylvania: Retrospective Analysis.

BACKGROUND: Tweets can provide broad, real-time perspectives about health and medical diagnoses that can inform disease surveillance in geographic regions. Less is known, however, about how much individuals post about common health conditions or what they post about. OBJECTIVE: We sought to collect and analyze tweets from 1 state about high prevalence health conditions and characterize the tweet volume and content. METHODS: We collected 408,296,620 tweets originating in Pennsylvania from 2012-2015 and compared the prevalence of 14 common diseases to the frequency of disease mentions on Twitter. We identified and corrected bias induced due to variance in disease term specificity and used the machine learning approach of differential language analysis to determine the content (words and themes) most highly correlated with each disease. RESULTS: Common disease terms were included in 226,802 tweets (174,381 tweets after disease term correction). Posts about breast cancer (39,156/174,381 messages, 22.45%; 306,127/12,702,379 prevalence, 2.41%) and diabetes (40,217/174,381 messages, 23.06%; 2,189,890/12,702,379 prevalence, 17.24%) were overrepresented on Twitter relative to disease prevalence, whereas hypertension (17,245/174,381 messages, 9.89%; 4,614,776/12,702,379 prevalence, 36.33%), chronic obstructive pulmonary disease (1648/174,381 messages, 0.95%; 1,083,627/12,702,379 prevalence, 8.53%), and heart disease (13,669/174,381 messages, 7.84%; 2,461,721/12,702,379 prevalence, 19.38%) were underrepresented. The content of messages also varied by disease. Personal experience messages accounted for 12.88% (578/4487) of prostate cancer tweets and 24.17% (4046/16,742) of asthma tweets. Awareness-themed tweets were more often about breast cancer (9139/39,156 messages, 23.34%) than asthma (1040/16,742 messages, 6.21%). Tweets about risk factors were more often about heart disease (1375/13,669 messages, 10.06%) than lymphoma (105/4927 messages, 2.13%). CONCLUSIONS: Twitter provides a window into the Web-based visibility of diseases and how the volume of Web-based content about diseases varies by condition. Further, the potential value in tweets is in the rich content they provide about individuals' perspectives about diseases (eg, personal experiences, awareness, and risk factors) that are not otherwise easily captured through traditional surveys or administrative data.