Time and geographic variations in human papillomavirus vaccine uptake in Washington state.

This study examines geographic variations of human papillomavirus (HPV) vaccine uptake, the most significant disparity in HPV vaccination, in Washington State. We evaluated Washington State Immunization Information System (WA-IIS) data on target age (11-12 year old adolescents) between 2008 and 2018. A Bayesian spatio-temporal analysis was conducted to examine uptake at the census tract level. Urban-rural disparities in vaccine rates were assessed using t-tests. Persistently high and low vaccine areas and their contributing sociodemographic factors were then identified using a multinomial logistic regression. HPV vaccine uptake gradually increased after 2010, but remained persistently low. Average vaccine uptake rates from 2010 through 2018 in urban areas were 11%-34% for initiation and 4-19% for completion. These rates were 9-22% initiation and 3-11% completion in rural areas. We observed statistically significant (p < 0.05) differences between the estimated vaccine rates for urban and rural census tracts. Race/ethnicity and socioeconomic status were associated with this urban-rural disparity. The odds of being in low vaccine rural areas increased with increase in Area Deprivation Index (ADI) (OR = 1.14, CI = (1.10, 1.19)), and decreased with percentage increase in Black (OR = 0.43, CI = (0.02, 0.85)) and Hispanic (OR = 0.97, CI = (0.94, 1.00)) population. Bayesian spatial analysis was effective in capturing spatio-temporal patterns in HPV vaccine rates and identifying areas with persistently low vaccination over time. This analytic approach can be used to guide public health policies and geographically target interventions to reduce HPV vaccine disparities and to prevent future HPV-related cancers.

A multi-modal approach towards mining social media data during natural disasters - a case study of Hurricane Irma.

Streaming social media provides a real-time glimpse of extreme weather impacts. However, the volume of streaming data makes mining information a challenge for emergency managers, policy makers, and disciplinary scientists. Here we explore the effectiveness of data learned approaches to mine and filter information from streaming social media data from Hurricane Irma's landfall in Florida, USA. We use 54,383 Twitter messages (out of 784K geolocated messages) from 16,598 users from Sept. 10 - 12, 2017 to develop 4 independent models to filter data for relevance: 1) a geospatial model based on forcing conditions at the place and time of each tweet, 2) an image classification model for tweets that include images, 3) a user model to predict the reliability of the tweeter, and 4) a text model to determine if the text is related to Hurricane Irma. All four models are independently tested, and can be combined to quickly filter and visualize tweets based on user-defined thresholds for each submodel. We envision that this type of filtering and visualization routine can be useful as a base model for data capture from noisy sources such as Twitter. The data can then be subsequently used by policy makers, environmental managers, emergency managers, and domain scientists interested in finding tweets with specific attributes to use during different stages of the disaster (e.g., preparedness, response, and recovery), or for detailed research.