Group-based randomized trial of contingencies for health and abstinence in HIV patients

Abstract

Objective: Contingency management (CM) treatments are usually applied individually for drug abstinence, but CM can also be targeted toward health behaviors and implemented in groups. This study evaluated effects of a group-based CM intervention that focused on reinforcing health behaviors.

Method: HIV-positive patients with cocaine or opioid use disorders (n = 170) were randomized to weekly CM or 12-step (TS) groups for 24 weeks (mean attendance was 10.8 +/- 8.1 sessions for CM participants and 9.0 +/- 6.9 session for TS participants). During the treatment period, both groups received compensation for attendance ($10 per session) and submission of urine samples (about $2 per sample). In addition, participants received $25 for submitting samples and completing evaluations at Months 1, 3, 6, 9, and 12; 65-75 of the 81 participants assigned to TS and 71-80 of the 89 participants assigned to CM completed these evaluations. During the treatment period, patients in the CM group received chances to win prizes contingent upon completing health activities and submitting substance-free specimens (M = $260, SD = $267).

Results: Mean attendance was 10.8 +/- 8.1 sessions for CM participants and 9.0 +/- 6.9 sessions for TS participants. CM participants submitted a significantly greater number of consecutive drug-free specimens than did TS participants (5.2 +/- 6.0 vs. 3.7 +/- 5.6), but proportions of negative samples did not differ between groups during treatment or at follow-up evaluations. From pre- to posttreatment, CM participants showed greater reductions in viral loads and HIV-risk behaviors than did TS participants, but these effects were not maintained throughout the follow-up period.

Conclusions: These data suggest the efficacy of group-based CM for HIV-positive substance abusers, but more research is needed to extend the long-term benefits.

Figure 1

Flow chart of participants in study.

Figure 2

Proportion of patients submitting various numbers of consecutive samples negative for alcohol, opioids, and cocaine during the 24-week treatment period. Participants assigned to the 12-step facilitation condition (n = 81) are shown in empty bars, and those assigned to contingency management (n = 89) are shown in filled bars.

Figure 3

Viral loads results between baseline and month 6. Data represent estimated means from hierarchical linear models of participants assigned to 12-step treatment (dashed lines; n = 81) or contingency management treatment (solid lines; n = 89) and take into account data collected between baseline and month 6 (end of treatment). The lower bound on the y-axis of 2.3 reflects non-detectable viral loads as data were log-transformed, and values of 2.3 were included for non-detectable viral loads.

Figure 4

Changes in HIV-Risk Behavior Scale scores between baseline and month 6 and between baseline and month 12. Changes are shown for total scores, as well as for the sexual and drug risk behavioral sub-scores. Participants assigned to the 12-step facilitation condition are shown in empty bars, and those assigned to contingency management are shown in filled bars. The asterisks indicate statistically significant differences between conditions based on univariate analysis of variance (p<.05).