# Table 3 Predictors of the first visit interval (number of days between the first and second visits)a

Variableb Est. S.E. p Incidence
Rate Ratio
95% Confidence Interval Predicted
First Interval
Lower Upper
Intercept 2.70 0.08 <.0001 . . . .
Age −0.01 0.00 <.0001 0.99 0.99 0.99 Age 25 = 12.5
Age 40 = 10.8
Receiving social assistance:
Yes 0.07 0.03 0.01 1.07 1.01 1.14 12.5
No (ref) . . 1.00 . . 11.7
Alcohol problems:
Yes 0.09 0.03 0.001 1.09 1.03 1.16 12.6
No (ref) . . 1.00 . . 11.6
Opioid problems c:
Yes 0.10 0.03 0.0005 1.11 1.04 1.17 12.7
No (ref) . . 1.00 . . 11.5
Frequency of substance use, past 30 days:
Binge use −0.13 0.07 0.07 0.87 0.77 1.01 11.4
3 times per week to daily −0.13 0.03 <.0001 0.88 0.83 0.93 11.4
Up to 2 times per week −0.03 0.04 0.41 0.97 0.90 1.05 12.6
No use (ref) . . 1.00 . . 13.0
Past-year injection drug use:
Yes −0.14 0.04 0.001 0.87 0.80 0.94 11.3
No (ref) . . 1.00 . . 13.0
Treatment mandate:
Legal 0.25 0.06 <.0001 1.29 1.14 1.44 14.4
Child protection 0.07 0.03 0.02 1.07 1.07 1.01 12.0
Employer/school −0.01 0.08 0.91 0.99 1.01 1.14 11.1
None (ref) . . 1.00 . . 11.2
1. aNegative binomial regression with random effect for agency; includes one episode per client (first episode selected). N = 4200. Excludes 662 clients (12.8% of 5162) whose first episode consisted of only 1 visit, and excludes 300 clients (5.8% of 5162) with > 1 visit but missing values on predictor variables. AIC = 29,021.6, BIC = 29,040.8
2. bCoefficients were not statistically significant for: education, employment status, marital status, cannabis problem, stimulant problem, legal system involvement, pregnant at admission
3. cIncludes prescription opioids, heroin, opium, and over-the-counter codeine