Spatial correlations in geographical spreading of COVID-19 in the United States.

The global spread of the COVID-19 pandemic has followed complex pathways, largely attributed to the high virus infectivity, human travel patterns, and the implementation of multiple mitigation measures. The resulting geographic patterns describe the evolution of the epidemic and can indicate areas that are at risk of an outbreak. Here, we analyze the spatial correlations of new active cases in the USA at the county level and characterize the extent of these correlations at different times. We show that the epidemic did not progress uniformly and we identify various stages which are distinguished by significant differences in the correlation length. Our results indicate that the correlation length may be large even during periods when the number of cases declines. We find that correlations between urban centers were much more significant than between rural areas and this finding indicates that long-range spreading was mainly facilitated by travel between cities, especially at the first months of the epidemic. We also show the existence of a percolation transition in November 2020, when the largest part of the country was connected to a spanning cluster, and a smaller-scale transition in January 2021, with both times corresponding to the peak of the epidemic in the country.

Trends and Regional Variation in Prevalence of Cardiovascular Risk Factors and Association With Socioeconomic Status in Canada, 2005-2016.

Importance: Cardiovascular disease remains the second leading cause of death in Canada. Monitoring and tracking the trends and disparities in major cardiovascular risk factors could provide benchmarks for future cardiovascular health strategies. Objective: To investigate the temporal trends, regional variations, and socioeconomic disparities in major cardiovascular risk factors in Canada from 2005 to 2016. Design, Setting, and Participants: This repeated cross-sectional survey study included adults aged 20 years and older from 6 Canadian Community Health Survey cycles between 2005 and 2016. Cardiovascular risk factors included hypertension, diabetes, obesity, and current smoking. Socioeconomic status was measured using equivalized household income. Data analysis was performed from September 2019 to April 2020. Exposures: A total of 112 health regions and socioeconomic status. Main Outcomes and Measures: Age- and sex-adjusted prevalence of hypertension, diabetes, obesity, and current smoking by year; health regions; and socioeconomic status. Absolute numbers were rounded to base 100 for confidentiality purposes, and percentages were based on weighted numbers. Slope index of inequality (SII) and relative index of inequality (RII) were calculated to assess absolute and relative socioeconomic inequalities, respectively. Results: A total of 670 000 respondents (329 000 [49.1%] men; 341 000 [50.9%] women) aged 20 years and older from 6 survey cycles were enrolled for this study. The largest age group was those aged 40 to 59 years (eg, 2005 cycle: 40.2% [95% CI, 39.9%-40.6%]). In the 2015/2016 cycle, the overall age- and sex-adjusted prevalence rates of hypertension, diabetes, obesity, and current smoking were 20.7% (95% CI, 20.4%-21.1%), 7.2% (95% CI, 7.0%-7.5%), 20.1% (95% CI, 19.7%-20.6%), and 17.8% (95% CI, 17.4%-18.2%), respectively. From 2005 to 2016, there was a significant increase in the prevalence of hypertension, diabetes, and obesity (eg, prevalence of diabetes in both sexes, 2005: 5.8% [95% CI, 5.6%-6.0%]; 2015/2016: 7.2% [95% CI, 7.0%-7.5%]; P < .001) but a significant decrease in the prevalence of current smoking (both sexes, 2005: 22.1% [95% CI, 21.7%-22.5%]; 2015/2016: 17.8% [95% CI, 17.4%-18.2%]; P < .001). The prevalence of all the risk factors varied widely across health regions (eg, obesity, Vancouver Health Service Delivery Area: 6.7% [95% CI, 4.5%-9.0%]; Miramichi Area: 36.8% [95% CI, 27.3%-46.3%]). In addition to obesity among men, all risk factors tended to be more common among those with lower income (eg, prevalence of hypertension in both sexes, 2015/2016, lowest income group: 23.2% [95% CI, 22.4%-24.0%]; highest income group: 18.4% [95% CI, 17.7%-19.1%]). The SII and RII indicated consistent absolute and relative socioeconomic inequalities in hypertension, diabetes, and current smoking over time (eg, RII for hypertension in both sexes, 2005: 1.25; 95% CI, 1.18-1.33; 2015/2016: 1.34; 95% CI, 1.26-1.43). However, the phenomenon of absolute and relative socioeconomic inequalities in obesity was only observed among women (eg, RII for 2015/2016 for obesity in women; 1.74 (95% CI, 1.56-1.93); men: 1.09; 95% CI, 0.99-1.21). Conclusions and Relevance: During the study period, the prevalence of hypertension, diabetes, and obesity significantly increased, while the prevalence of current smoking significantly decreased. Geographic and socioeconomic gaps should be considered and addressed in future interventions and policies targeted at reducing these cardiovascular risk factors in Canada.

Sociodemographic associations of geographic variation in paediatric tonsillectomy and adenoidectomy.

Geographic variation of paediatric tonsillectomy, with or without adenoidectomy, (A/T) has been described since the 1930s until today but no studies have investigated the factors associated with this variation. This study described the geographical distribution of paediatric A/T across the state of Victoria, Australia, and investigated area-level factors associated with this variation. We used linked administrative datasets capturing all paediatric A/T performed between 2010 and 2015 in Victoria. Surgery data were collapsed by patient residence to the level of Local Government Area. Regression models were used to investigate the association between likelihood of surgery and area-level factors. We found a 10.2-fold difference in A/T rates across the state, with areas of higher rates more in regional than metropolitan areas. Area-level factors associated with geographic variation of A/T were percentage of children aged 5-9 years (IRR 1.07, 95%CI 1.01-1.14, P = 0.03) and low English language proficiency (IRR 0.95, 95% CI 0.90-0.99, P = 0.03). In a sub-population analysis of surgeries in the public sector, these factors were low maternal educational attainment (IRR 1.09, 95% CI 1.02-1.16, P < 0.001) and surgical waiting time (IRR 0.99635 95% CI 0.99273-0.99997, P = 0.048). Identifying areas of focus for improvement and factors associated with geographic variation will assist in improving equitable provision of paediatric A/T and decrease variability within regions.

Identifying gaps in the photographic record of the vascular plant flora of the Americas.

Field photographs of plant species are crucial for research and conservation, but the lack of a centralized database makes them difficult to locate. We surveyed 25 online databases of field photographs and found that they harboured only about 53% of the approximately 125,000 vascular plant species of the Americas. These results reflect the urgent need for a centralized database that can both integrate and complete the photographic record of the world's flora.

Social Determinants of Health and Geographic Variation in Medicare per Beneficiary Spending.

Importance: Despite substantial geographic variation in Medicare per beneficiary spending in the US, little is known about the extent to which social determinants of health (SDoH) are associated with this variation. Objective: To determine the associations between SDoH and county-level price-adjusted Medicare per beneficiary spending. Design, Setting, and Participants: This cross-sectional study used county-level data on 2017 Medicare fee-for-service (FFS) spending, patient demographic characteristics (eg, age and gender) and clinical risk score, supply of health care resources (eg, number of hospital beds), and SDoH measures (eg, median income and unemployment rate) from multiple sources. Multivariable regressions were used to estimate the association of the variation in spending across quintiles with SDoH. Main Outcomes and Measures: 2017 county-level price-adjusted Medicare Parts A and B spending per beneficiary. SDoH measures included socioeconomic position, race/ethnicity, social relationships, and residential and community context. Results: Among 3038 counties with 33 495 776 Medicare FFS beneficiaries (18 352 336 [54.8%] women; mean [SD] age, 72 [1.5] years), mean Medicare price-adjusted per beneficiary spending for counties in the highest spending quintile was $3785 (95% CI, $3706-$3862) higher, or 49% higher, than spending for bottom-quintile counties (mean [SD] spending per beneficiary, $11 464 [735] vs $7679 [522]; P < .001). The total contribution (including through both direct and indirect pathways) of SDoH was 37.7% ($1428 of $3785) of this variation, compared with 59.8% ($2265 of $3785) by patient clinical risk, 14.5% ($549 of $3785) by supply of health care resources, and 19.8% ($751 of $3785) by patient demographic characteristics. When all factors were included within the same model, the direct contribution of SDoH was associated with 5.8% of the variation, compared with 4.6% by supply, 4.7% by patient demographic characteristics, and 62.0% by patient clinical risk. Conclusions and Relevance: These findings suggest social determinants of health are associated with considerable proportions of geographic variation in Medicare spending. Policies addressing SDoH for disadvantaged patients in certain regions have the potential to contain health care spending and improve the value of health care; patient SDoH may need to be accounted for in publicly reported physician performance, and in value-based purchasing incentive programs for health care professionals.

Potential geographic distribution of the tiger mosquito Aedes albopictus (Skuse, 1894) (Diptera: Culicidae) in current and future conditions for Colombia.

In Colombia, little is known on the distribution of the Asian mosquito Aedes albopictus, main vector of dengue, chikungunya, and Zika in Asia and Oceania. Therefore, this work sought to estimate its current and future potential geographic distribution under the Representative Concentration Paths (RCP) 2.6 and 8.5 emission scenarios by 2050 and 2070, using ecological niche models. For this, predictions were made in MaxEnt, employing occurrences of A. albopictus from their native area and South America and bioclimatic variables of these places. We found that, from their invasion of Colombia to the most recent years, A. albopictus is present in 47% of the country, in peri-urban (20%), rural (23%), and urban (57%) areas between 0 and 1800 m, with Antioquia and Valle del Cauca being the departments with most of the records. Our ecological niche modelling for the currently suggests that A. albopictus is distributed in 96% of the Colombian continental surface up to 3000 m (p < 0.001) putting at risk at least 48 million of people that could be infected by the arboviruses that this species transmits. Additionally, by 2050 and 2070, under RCP 2.6 scenario, its distribution could cover to nearly 90% of continental extension up to 3100 m (≈55 million of people at risk), while under RCP 8.5 scenario, it could decrease below 60% of continental extension, but expand upward to 3200 m (< 38 million of people at risk). These results suggest that, currently in Colombia, A. albopictus is found throughout the country and climate change could diminish eventually its area of distribution, but increase its altitudinal range. In Colombia, surveillance and vector control programs must focus their attention on this vector to avoid complications in the national public health setting.

The universal visitation law of human mobility.

Human mobility impacts many aspects of a city, from its spatial structure1-3 to its response to an epidemic4-7. It is also ultimately key to social interactions8, innovation9,10 and productivity11. However, our quantitative understanding of the aggregate movements of individuals remains incomplete. Existing models-such as the gravity law12,13 or the radiation model14-concentrate on the purely spatial dependence of mobility flows and do not capture the varying frequencies of recurrent visits to the same locations. Here we reveal a simple and robust scaling law that captures the temporal and spatial spectrum of population movement on the basis of large-scale mobility data from diverse cities around the globe. According to this law, the number of visitors to any location decreases as the inverse square of the product of their visiting frequency and travel distance. We further show that the spatio-temporal flows to different locations give rise to prominent spatial clusters with an area distribution that follows Zipf's law15. Finally, we build an individual mobility model based on exploration and preferential return to provide a mechanistic explanation for the discovered scaling law and the emerging spatial structure. Our findings corroborate long-standing conjectures in human geography (such as central place theory16 and Weber's theory of emergent optimality10) and allow for predictions of recurrent flows, providing a basis for applications in urban planning, traffic engineering and the mitigation of epidemic diseases.

Sub-national analysis and determinants of numbers of antenatal care contacts in Nigeria: assessing the compliance with the WHO recommended standard guidelines.

BACKGROUND: Nigeria has unimpressive maternal and child health indicators. Compliance with the WHO guidelines on the minimum number of antenatal care (ANC) contacts could improve these indicators. We assessed the compliance with WHO recommended standards on ANC contacts in Nigeria and identify the associated factors. METHODS: Nationally representative cross-sectional data during pregnancy of 21,785 most recent births within five years preceding the 2018 Nigeria Demographic Health Survey was used. The number of ANC contacts was categorised into "None", "1-3", "4-7" and "8 or more" contacts based on subsequent WHO guidelines. Descriptive statistics, bivariable and multivariable multinomial logistic regression was used at p = 0.05. RESULTS: About 25 % of the women had no ANC contact, 58 % had at least 4 contacts while only 20 % had 8 or more ANC contacts. The highest rate of 8 or more ANC contacts was in Osun (80.2 %), Lagos (76.8 %), and Imo (72.0 %) while the lowest rates were in Kebbi (0.2 %), Zamfara (1.1 %) and Yobe (1.3 %). Respondents with higher education were twelve times (adjusted relative risk (aRR): 12.46, 95 % CI: 7.33-21.2), having secondary education was thrice (aRR: 2.91, 95 % CI: 2.35-3.60), and having primary education was twice (aRR: 2.17, 95 % CI: 1.77-2.66) more likely to make at least 8 contacts than those with no education. Respondents from households in the richest and middle wealth categories were 129 and 67 % more likely to make 8 or more ANC contacts compared to those from households in the lowest wealth category respectively. The likelihood of making 8 ANC contacts was 89 and 47 % higher among respondents from communities in the least and middle disadvantaged groups, respectively,  compared to the most disadvantaged group. Other significant variables were spouse education, health care decision making, media access, ethnicity, religion, and other community factors. CONCLUSIONS: Compliance with WHO guidelines on the minimum number of ANC contacts in Nigeria is poor. Thus, Nigeria has a long walk to attaining sustainable development goal's targets on child and maternal health. We recommend that the maternal and child health programmers should review existing policies and develop new policies to adopt, implement and tackle the challenges of adherence to the WHO recommended minimum of 8 ANC contacts. Women's education, socioeconomic status and adequate mobilization of families should be prioritized. There is a need for urgent intervention to narrow the identified inequalities and substantial disparities in the characteristics of pregnant women across the regions and states.

Trends and Opportunities in Tick-Borne Disease Geography.

Tick-borne diseases are a growing problem in many parts of the world, and their surveillance and control touch on challenging issues in medical entomology, agricultural health, veterinary medicine, and biosecurity. Spatial approaches can be used to synthesize the data generated by integrative One Health surveillance systems, and help stakeholders, managers, and medical geographers understand the current and future distribution of risk. Here, we performed a systematic review of over 8,000 studies and identified a total of 303 scientific publications that map tick-borne diseases using data on vectors, pathogens, and hosts (including wildlife, livestock, and human cases). We find that the field is growing rapidly, with the major Ixodes-borne diseases (Lyme disease and tick-borne encephalitis in particular) giving way to monitoring efforts that encompass a broader range of threats. We find a tremendous diversity of methods used to map tick-borne disease, but also find major gaps: data on the enzootic cycle of tick-borne pathogens is severely underutilized, and mapping efforts are mostly limited to Europe and North America. We suggest that future work can readily apply available methods to track the distributions of tick-borne diseases in Africa and Asia, following a One Health approach that combines medical and veterinary surveillance for maximum impact.

Modeling the spatial distribution of anthrax in southern Kenya.

BACKGROUND: Anthrax is an important zoonotic disease in Kenya associated with high animal and public health burden and widespread socio-economic impacts. The disease occurs in sporadic outbreaks that involve livestock, wildlife, and humans, but knowledge on factors that affect the geographic distribution of these outbreaks is limited, challenging public health intervention planning. METHODS: Anthrax surveillance data reported in southern Kenya from 2011 to 2017 were modeled using a boosted regression trees (BRT) framework. An ensemble of 100 BRT experiments was developed using a variable set of 18 environmental covariates and 69 unique anthrax locations. Model performance was evaluated using AUC (area under the curve) ROC (receiver operating characteristics) curves. RESULTS: Cattle density, rainfall of wettest month, soil clay content, soil pH, soil organic carbon, length of longest dry season, vegetation index, temperature seasonality, in order, were identified as key variables for predicting environmental suitability for anthrax in the region. BRTs performed well with a mean AUC of 0.8. Areas highly suitable for anthrax were predicted predominantly in the southwestern region around the shared Kenya-Tanzania border and a belt through the regions and highlands in central Kenya. These suitable regions extend westwards to cover large areas in western highlands and the western regions around Lake Victoria and bordering Uganda. The entire eastern and lower-eastern regions towards the coastal region were predicted to have lower suitability for anthrax. CONCLUSION: These modeling efforts identified areas of anthrax suitability across southern Kenya, including high and medium agricultural potential regions and wildlife parks, important for tourism and foreign exchange. These predictions are useful for policy makers in designing targeted surveillance and/or control interventions in Kenya. We thank the staff of Directorate of Veterinary Services under the Ministry of Agriculture, Livestock and Fisheries, for collecting and providing the anthrax historical occurrence data.

The winding road to health: A systematic scoping review on the effect of geographical accessibility to health care on infectious diseases in low- and middle-income countries.

BACKGROUND: Geographical accessibility to healthcare is an important component of infectious disease dynamics. Timely access to health facilities can prevent disease progression and enables disease notification through surveillance systems. The importance of accounting for physical accessibility in response to infectious diseases is increasingly recognized. Yet, there is no comprehensive review of the literature available on infectious diseases in relation to geographical accessibility to care. Therefore, we aimed at evaluating the current state of knowledge on the effect of geographical accessibility to health care on infectious diseases in low- and middle-income countries. METHODS AND FINDINGS: A search strategy was developed and conducted on Web of Science and PubMed on 4 March 2019. New publications were checked until May 28, 2020. All publication dates were eligible. Data was charted into a tabular format and descriptive data analyses were carried out to identify geographical regions, infectious diseases, and measures of physical accessibility among other factors. Search queries in PubMed and Web of Science yielded 560 unique publications. After title and abstract screening 99 articles were read in full detail, from which 64 articles were selected, including 10 manually. Results of the included publications could be broadly categorized into three groups: (1) decreased spatial accessibility to health care was associated with a higher infectious disease burden, (2) decreased accessibility was associated to lower disease reporting, minimizing true understanding of disease distribution, and (3) the occurrence of an infectious disease outbreak negatively impacted health care accessibility in affected regions. In the majority of studies, poor geographical accessibility to health care was associated with higher disease incidence, more severe health outcomes, higher mortality, and lower disease reporting. No difference was seen between countries or infectious diseases. CONCLUSIONS: Currently, policy-makers and scientists rely on data collected through passive surveillance systems, introducing uncertainty on disease estimates for remote communities. Our results highlight the need for increasing integration of geographical accessibility measures in disease risk modelling, allowing more realistic disease estimates and enhancing our understanding of true disease burden. Additionally, disease risk estimates could be used in turn to optimize the allocation of health services in the prevention and detection of infectious diseases.

Extracting and modeling geographic information from scientific articles.

Scientific articles often contain relevant geographic information such as where field work was performed or where patients were treated. Most often, this information appears in the full-text article contents as a description in natural language including place names, with no accompanying machine-readable geographic metadata. Automatically extracting this geographic information could help conduct meta-analyses, find geographical research gaps, and retrieve articles using spatial search criteria. Research on this problem is still in its infancy, with many works manually processing corpora for locations and few cross-domain studies. In this paper, we develop a fully automatic pipeline to extract and represent relevant locations from scientific articles, applying it to two varied corpora. We obtain good performance, with full pipeline precision of 0.84 for an environmental corpus, and 0.78 for a biomedical corpus. Our results can be visualized as simple global maps, allowing human annotators to both explore corpus patterns in space and triage results for downstream analysis. Future work should not only focus on improving individual pipeline components, but also be informed by user needs derived from the potential spatial analysis and exploration of such corpora.

Illuminating Racial Inequity in Diabetes Control: Differences Based on Gender and Geography.

The objective of this study is to describe racial differences in type 2 diabetes mellitus "diabetes" control among the adults in the USA, and to examine attributes that may exacerbate racial differences. Secondary analyses of data from the National Ambulatory Medical Care Survey (NAMCS) collected in years 2012-2014 in the USA. Study sample was limited to White or African American patients aged 25 or older and living with diabetes (n = 4106). Outcome measure, poor diabetes control, was based on lab values for HbA1c (> 7%). Covariates include demographics, insurance, comorbid conditions, and continuity of care and location (urban vs. rural). Overall, African Americans have 33% higher odds of poor diabetes control compared with Whites. Adjusted probability of poor diabetes control was 48% overall, 65% for African American women and 69% for African Americans living in rural areas. African Americans continue to have poorer diabetes control compared to Whites. This difference is exacerbated for African American women, and for all African Americans living in rural areas. Policy should include concentrated screening and treatment resources for African Americans in rural settings.

Sociodemographic characteristics and exposure patterns of pesticide-related cases reported to a poison service center in Brazil between 2012 and 2016.

Pesticide poisoning is a significant public health problem responsible for an estimated three million poisoning cases per year and more than 250,000 deaths, most of which occur in underdeveloped countries. We evaluated pesticide exposure cases reported to a toxicological service center in Brazil, between 2012 and 2016. There were 3211 cases of pesticide exposure, with a high prevalence in adults aged 20-39 years (41.2%). Attempted suicide was the leading cause of pesticide cases (48%). Occupational exposure to pesticides of agricultural use was more frequently observed among men. Accidental exposure and attempted suicide were more frequently observed in urban areas while occupational exposure was more prevalent in rural areas. A higher exposure rate was observed among men in counties with higher agricultural activities. Establishing prevalence and cause of pesticide exposure is important to provide subsidy for evidence-based interventions in the field.

Cost and benefit of military quarantine policies.

The initial response to COVID-19 included quarantine policies. This study aims to determine the infection containment proportions and cost of two variations of quarantine policies based on geographic travel and close contact with infected individuals within deployed US military populations. Special Operations Command Africa (SOCAF) records of individuals quarantined between March 1, 2020 and June 1, 2020 were examined. The infection containment proportion and cost in containment hours were compared between types of quarantine and between geographic areas. Geographic quarantine contained 2 cases out of 63 quarantined individuals in West Africa (3.2%) compared to 0 out of 221 in East Africa (p = 0.0486). Close contact quarantine contained 3 cases out of 31 quarantined individuals in West Africa compared to 4 out of 55 in East Africa (7.3%, p = 0.6989). Total confinement was 42,048 h for each contained infection using geographic quarantine compared to 4076 h using close contact quarantine. In the US military population deployed to Africa for COVID-19, quarantining based on geographic movement is an order of magnitude more costly in terms of time for each contained infection then quarantining based on close contact with infected individuals. There is not a statistical difference between East and West Africa. The associated costs of quarantine must be carefully weighed against the risk of disease spread.

Vaccination against seasonal influenza and socio-economic and environmental factors as determinants of the geographic variation of COVID-19 incidence and mortality in the Italian elderly.

BACKGROUND: A possible protective effect of seasonal influenza vaccination against the spread of the COVID-19 epidemic has been suggested. METHODS: We used publicly available data bases to explore the hypothesis as well as the effect of multiple social and environmental factors in the 20 Italian regions. RESULTS: Our results suggest that vaccination against seasonal influenza might beneficially impact on incidence and severity of the novel corona virus epidemic. Population density and vehicular traffic were also moderately associated with cumulative incidence of COVID-19. None of the other variables we considered showed an effect on cumulative incidence, case fatality rate or mortality from COVID-19. CONCLUSIONS: Extending influenza vaccination coverage particularly among the elderly, vulnerable individuals with specific chronic medical conditions, health care workers, and workers in other essential services, early in the upcoming 2020 influenza season, might help reduce the health impact of a second epidemic wave of COVID-19.

Voicing Jordanian Adolescents' Suicide.

BACKGROUND: Studies focusing on adolescent suicide in Arab countries are particularly scarce, with the few available undertaken from within an epidemiological paradigm. OBJECTIVE: This study aimed to understand Jordanian adolescents' perceptions, beliefs, and attitudes toward suicide. METHODS: A qualitative design using 12 dual-moderator focus group discussions was conducted in public schools. Participants were selected from the three main regions of the country (rural southern, urban central, and suburban northern). Participants included Jordanian adolescent boys and girls, aged 14-17 years, who reported experiencing mild to moderate depressive symptoms. A relational content analysis approach was used for coding data, and a content analysis was used to identify salient thematic categories. Data were analyzed using NVivo software. RESULTS: Four themes emerged, including perceived risk factors, perceived protective factors, active and passive suicidal ideations, and e-games and Internet influences. Main risk factors were depression; anxiety; stigma, shame, and isolation; family issues; life pressures; and guilt. Conversely, religiosity, perceived positive family functioning, and availability of long-term goals seemed to confer resilience to adolescents' suicidal ideation and behavior. Passive suicidality (having death wishes without any plans to complete suicide) was noticed most among participants who feared jeopardizing the family's reputation if they committed suicide. Several boys with active suicidal ideations used the Blue Whale Challenge e-game to learn how to complete suicide and relinquish their problems. DISCUSSION: Suicide is a multifactorial problem requiring multimodal strategies. Evidence from this research suggests that those most passionate about the outcome of interest are encouraged to redouble efforts to reduce modifiable risk factors, enrich protective factors, target the underlying psychiatric illness that informs suicidal ideations and behavior, and research the effect of social media and Internet activity more deeply. Parents are advised to monitor the online activities of their children and familiarize themselves with the digital applications they use.

Geographic Differences in Cannabis Conversations on Twitter: Infodemiology Study.

BACKGROUND: Infodemiology is an emerging field of research that utilizes user-generated health-related content, such as that found in social media, to help improve public health. Twitter has become an important venue for studying emerging patterns in health issues such as substance use because it can reflect trends in real-time and display messages generated directly by users, giving a uniquely personal voice to analyses. Over the past year, several states in the United States have passed legislation to legalize adult recreational use of cannabis and the federal government in Canada has done the same. There are few studies that examine the sentiment and content of tweets about cannabis since the recent legislative changes regarding cannabis have occurred in North America. OBJECTIVE: To examine differences in the sentiment and content of cannabis-related tweets by state cannabis laws, and to examine differences in sentiment between the United States and Canada between 2017 and 2019. METHODS: In total, 1,200,127 cannabis-related tweets were collected from January 1, 2017, to June 17, 2019, using the Twitter application programming interface. Tweets then were grouped geographically based on cannabis legal status (legal for adult recreational use, legal for medical use, and no legal use) in the locations from which the tweets came. Sentiment scoring for the tweets was done with VADER (Valence Aware Dictionary and sEntiment Reasoner), and differences in sentiment for states with different cannabis laws were tested using Tukey adjusted two-sided pairwise comparisons. Topic analysis to determine the content of tweets was done using latent Dirichlet allocation in Python, using a Java implementation, LdaMallet, with Gensim wrapper. RESULTS: Significant differences were seen in tweet sentiment between US states with different cannabis laws (P=.001 for negative sentiment tweets in fully illegal compared to legal for adult recreational use states), as well as between the United States and Canada (P=.003 for positive sentiment and P=.001 for negative sentiment). In both cases, restrictive state policy environments (eg, those where cannabis use is fully illegal, or legal for medical use only) were associated with more negative tweet sentiment than less restrictive policy environments (eg, where cannabis is legal for adult recreational use). Six key topics were found in recent US tweet contents: fun and recreation (keywords, eg, love, life, high); daily life (today, start, live); transactions (buy, sell, money); places of use (room, car, house); medical use and cannabis industry (business, industry, company); and legalization (legalize, police, tax). The keywords representing content of tweets also differed between the United States and Canada. CONCLUSIONS: Knowledge about how cannabis is being discussed online, and geographic differences that exist in these conversations may help to inform public health planning and prevention efforts. Public health education about how to use cannabis in ways that promote safety and minimize harms may be especially important in places where cannabis is legal for adult recreational and medical use.