# Table 2 Random effect negative binomial regression on ER visits

IRR SE p 95% CI
Race/ethnicity
Non-Latino Whitea
Black 0.852 0.031 < .001 0.793, 0.915
Latino 0.826 0.025 < .001 0.778, 0.876
Age 1.000 0.001 .836 0.997, 1.002
Male 0.777 0.021 < .001 0.737, 0.819
Education 1.017 0.005 .001 1.007, 1.028
Homeless 1.212 0.035 < .001 1.145, 1.283
History of mental health issues 0.661 0.018 < .001 0.627, 0.698
Days of mental health counseling 1.021 0.003 < .001 1.015, 1.028
Days of psychiatric care 1.032 0.003 < .001 1.025, 1.039
Days of physical problems 1.067 0.001 < .001 1.065, 1.069
Age at first drug use 0.998 0.002 .210 0.995, 1.001
Days of primary drug use 0.996 0.001 .001 0.994, 0.998
Primary drug problem
Alcohola
Cocaine 1.790 0.075 < .001 1.649, 1.942
Heroin 1.113 0.051 .020 1.017, 1.218
Marijuana 1.194 0.059 < .001 1.083, 1.316
Methamphetamine 1.113 0.072 .096 0.981, 1.264
Other 1.620 0.090 < .001 1.454, 1.806
Children younger than 18 1.008 0.004 .036 1.001, 1.016
Program modality
Outpatienta
Methadone 0.964 0.116 .763 0.761, 1.221
Residential 1.606 0.068 < .001 1.479, 1.744
1. Note: ER, emergency room; IRR, incidence rate ratio. IRRs can be interpreted as the estimated rate ratio for a 1-unit increase in the independent variable, given the other variables are held constant in the model. For example, if days of physical problems increased by 1 point, the ratio for number of ER visits would be expected to increase by a factor of IRR = 1.067, while holding all other variables in the model constant.
2. Wald chi-square tests with degrees of freedom (20) = 6693.30. The corresponding p-value is less than 0.0001.
3. aReference category.