Increases in Rates of Hospitalizations and Emergency Department Visits with Marijuana Associated Administrative Codes in Colorado from 2010 to 2014

BACKGROUND: The Colorado Hospital Association (CHA) collects data on hospitalizations (HD) and emergency department (ED) discharges from participating hospitals in the state of Colorado. The data includes patient demographics, admit and discharge dates, and ICD-9-CM discharge and procedure codes. The CHA dataset was used to investigate rates of HD and ED visits associated with possible marijuana exposures, diagnoses, and billing codes.

METHODS: HD and ED visits with possible marijuana exposures, diagnoses, or billing codes were determined by reporting of marijuana associated diagnosis codes E854.1, 969.6, 304.3, and 305.2 in the first three listed discharge codes. Rates were calculated using total HD or ED visits in each time period and examined by year and marijuana legalization era annually from 2000 through June 2014 for HD and 2011 through June 2014 for ED visits. Logistic regression was used to test differences in rates across marijuana legalization eras and a Bonferroni correction was use to adjust for multiple comparisons.

RESULTS: Rates of HD with possible marijuana exposures, diagnoses, or billing codes in the first three diagnosis codes by year remain constant from 2000 to 2009 with a increasing trend from 2010 through June 2014 for both rates of HD and ED visits. Rates of HD by marijuana legalization eras showed a significant increase in rates by 29.1% from 2001-2009 to 2010-2013 (χ²(1, N=6,630,988)=297.9, p<0.0001) and by 42.3% from 2010-2013 to 2014 (χ²(1, N=3,576,906)=109.2, p<0.0001). Rates of ED visits by marijuana legalization eras showed a significant increase in rates by 48.5% (χ²(1, N=3,090,829)=565.2, p<0.0001) from 2011-2013 to 2014. The highest rates of HD and ED visits were in January through June 2014.

CONCLUSIONS: The use of these diagnosis codes does not necessarily indicate that marijuana was the primary reason for the HD or ED visit, rather the presence of a marijuana associated code reflects that marijuana use was noted by the treating physician. We hypothesize that this analysis reflects marijuana use despite the limitations; however, it does not necessarily show the healthcare burden of marijuana use. These results indicate that rates of HD and ED visits with marijuana associated codes have increased from 2010 to 2014 after remaining stable for 10 years. It is unclear whether marijuana use, changes to the medical coding system, or awareness of marijuana legalization is driving the increases in rates. Nevertheless, these finding support the need for further investigations into HD and ED visits with marijuana associated coding.