Skip to main content
. Author manuscript; available in PMC: 2014 Aug 1.
Published in final edited form as: Acad Emerg Med. 2013 Aug;20(8):807–815. doi: 10.1111/acem.12188

Table 4.

Multivariable logistic regression analysis identifying predictors of ED assessment of self-harm, clustered by study site*

Characteristic Complete dataset (N=94,354) Dataset with outlying site excluded (n=78,408)
Adjusted Risk Ratio 95% CI Adjusted Risk Ratio 95% CI
Weekend presentation 1.30 1.04–1.62 1.27 0.86–1.88
Presentation on evening/night shift (1500–0659 hours) 1.36 1.01–1.85 0.88 0.63–1.22
Age ≥65 years 0.56 0.35–0.92 0.67 0.36–1.26
Male gender 1.17 1.10–1.26 1.14 1.02–1.27
 Interaction of age x gender NA 0.83 0.70–0.98
Race/ethnicity
 White Referent referent
 Black or African-American 0.89 0.58–1.37 0.63 0.39–1.03
 Asian 0.98 0.70–1.43 0.90 0.67–1.23
 American Indian or Alaska native 2.30 1.53–3.47 0.66 0.22–1.94
 Native Hawaiian or Pacific Islander 0.64 0.21–1.95 0.22 0.02–1.77
 Hispanic or Latino 2.41 1.67–3.48 0.85 0.70–1.03
 Other 0.79 0.41–1.54 1.09 0.63–1.87
 Not documented 1.91 1.21–3.02 0.45 0.20–1.02
*

Controlling for weekend presentation, age, sex, and race/ethnicity

Interpretation of the adjusted risk ratio for age and sex in the setting of an interaction requires consideration of the interaction coefficient. See text and Table 5