A randomized trial of extended telephone-based continuing care for alcohol dependence: within-treatment substance use outcomes

Abstract

Objective: The study tested whether adding up to 18 months of telephone continuing care, either as monitoring and feedback (TM) or longer contacts that included counseling (TMC), to intensive outpatient programs (IOPs) improved outcomes for alcohol-dependent patients.

Method: Participants (N = 252) who completed 3 weeks of IOP were randomized to up to 36 sessions of TM (M = 11.5 sessions), TMC (M = 9.1 sessions), or IOP only (treatment as usual [TAU]). Quarterly assessment of alcohol use (79.9% assessed at 18 months) was corroborated with available collateral reports (N = 63 at 12 months). Participants with cocaine dependence (N = 199) also provided urine samples.

Results: Main effects favored TMC over TAU on any alcohol use (odds ratio [OR] = 1.88, CI [1.13, 3.14]) and any heavy alcohol use (OR = 1.74, CI [1.03, 2.94]). TMC produced fewer days of alcohol use during Months 10-18 and heavy alcohol use during Months 13-18 than TAU (ds = 0.46-0.65). TMC also produced fewer days of any alcohol use and heavy alcohol use than TM during Months 4-6 (ds = 0.39 and 0.43). TM produced lower percent days alcohol use than TAU during Months 10-12 and 13-15 (ds = 0.41 and 0.39). There were no treatment effects on rates of cocaine-positive urines.

Conclusions: Adding telephone continuing care to IOP improved alcohol use outcomes relative to IOP alone. Conversely, shorter calls that provided monitoring and feedback but no counseling generally did not improve outcomes over IOP.

Figure 1

Consort Diagram

Figure 2

Frequency of Alcohol Use During Follow-up. Mean percent days alcohol use within each three-month period of the follow-up, presented for each treatment condition. Error bars indicate +/− 1 standard error (SE). There is a significant treatment condition × time interaction [chi-square (10)= 20.45, p= .025].

Figure 3

Frequency of Heavy Alcohol Use During Follow-up. Mean percent days heavy alcohol use within each three-month period of the follow-up, presented for each treatment condition. Error bars indicate +/− 1 standard error (SE). There is a significant treatment condition × time interaction [chi-square (10)= 21.87, p= .016].

Figure 4

Rates of Any Alcohol Use During Follow. Percentage of participants who report any use of alcohol within a given three-month period of the follow-up, presented for each treatment condition. There was a significant treatment condition main effect [chi-square (2)= 6.32, p= .048]. TMC produced lower rates of any alcohol use than TAU [chi-square (1) = 5.85, p= .016]. Comparisons of TM and TAU and TM and TMC were not significant.

Figure 5

Rates of Any Heavy Alcohol Use During Follow. Percentage of participants who report any heavy alcohol use within a given three-month period of the follow-up, presented for each treatment condition. There was a trend toward a treatment condition main effect [chi-square (2)= 5.37, p= .068]. TMC produced lower rates of heavy alcohol use than TAU [chi-square (1) = 4.29, p= .038], and there was a trend favoring TMC over TM [chi-square (1)= 3.47, p= .063]. TM and TAU did not differ.

Figure 6

Rates of Cocaine Positive Urine Samples. Percentage of participants with a cocaine positive urine sample at each follow-up for each treatment condition, in those with cocaine dependence (N=199). Follow-up rates ranged from 60.2% to 69.8% of those with cocaine dependence. Treatment condition main effects were not significant.