BACKGROUND: Following legalization of marijuana for recreational use there has been growing concern among policymakers that greater exposure and access may lead to increased healthcare utilization (e.g. emergency department (ED) visits) connected to marijuana use. Recent reports suggest marijuana retail outlets are often located in disadvantaged neighborhoods leading to potentially higher risk for vulnerable groups. We assessed the association between retail outlet exposure(s) along with census tract sociodemographic factors and marijuana-related emergency department (MJED) visits.
METHODS: Potentially associated MJED visits for 2013 and 2014 were extracted from the Colorado Hospital Association dataset using ICD-9-CM codes (304.3, 304.30, 304.31, 304.32, 305.2, 305.20, 305.21, 305.22, 969.6, and E854.1) for Denver County residents. Sociodemographic characteristics (i.e. percent of residents 15-40 years of age, in poverty, black, Hispanic, and owning a home) for Denver’s 143 census tracts were obtained from American Community Survey (2010-2014). Addresses of marijuana retail outlets, as of December 2014, were obtained from Colorado Department of Revenue, geocoded and assigned to a census tract. Using a pre/post strategy, we examined 2013 (pre-period) and 2014 (post-period) MJED visits in relation to census tract sociodemographic characteristics and presence of retail outlets. A negative binomial regression model was used to estimate the relationship between retail outlets and MJED visits controlling for all-cause ED utilization and census tract characteristics.
RESULTS: During 2013 and 2014 there were 2,748 MJED visits, 424 among children < 18 years of age. Almost all census tracts, (98% or 140/143), had at least one MJED visit. In 2014 (post-period only), 46% (66/143) of census tracts had at least one retail outlet. MJED visits in 2014 did not significantly increase compared to 2013 (1389 vs. 1359). Higher rates of census tract-level poverty were significantly associated with more MJED visits. Presence of a retail outlet, home ownership, and younger population (ages 15-40) had no significant impact on MJED visits.
CONCLUSIONS: MJED visits among Denver residents remained unchanged following the opening of retail outlets. Higher community levels of poverty were associated with higher numbers of MJED visits. Classifying MJED using diagnosis codes may result in some misclassification however, monitoring MJED utilization can identify areas of greater need to inform public health departments and policymakers to prioritize prevention and intervention strategies.