The COMBINE Study:Design and Methodology

Stephanie S. O’Malley, Ph.D.

for

The COMBINE Study Research Group

JAMA Vol. 295,17. 2003-2017, 2006 (May 3rd)

Study Design

• Randomized clinical trial

• 11 clinical centers

• 1,383 participants with DSM-IV alcohol dependence

• Participants were randomized to one of 9 treatment conditions

Two Types of Counselingduring 16 weeks of treatment

Medical Management (MM)

Combined Behavioral Intervention (CBI)

Medication Conditions: 8 Cells

• Acamprosate + Naltrexone• Acamprosate Only (+ Naltrexone placebo)• Naltrexone Only (+ Acamprosate placebo)• Double Placebo

• All received Medical Management (MM)• Half also got behavioral counseling (CBI)• Medications were administered double blind

Cell 9: No Pills, No MM

• Besides the 8 groups taking medications• One group received:

– Behavioral counseling – No pills– No medical management

Treatment Group Combinations(1383 Randomized participants)

Medical Management (n=607)

Placebo Acamprosate Placebo 153 152 Naltrexone 154 148

Medical Management + CBI (n=619)

Placebo Acamprosate No Pills Placebo 156 151 Naltrexone 155 157 No Pills 157

Study Sample

Inclusion Criteria

• DSM IV alcohol dependence

• At least 4, but no more than 21, abstinence days prior to randomization

• Drinking more than 21 drinks per week with at least 2 heavy drinking days in last 30

• Major Axis 1 Psych diagnosis

• Psych disorder requiring medication

• Other substance abuse (except MJ, nicotine) in last 90 days and clean UDS

• Unstable medical conditions including LFT’s greater than 3x normal

Exclusion Criteria

Participant (n=1383) Characteristics at Baseline

Mean s.d.

Age 44 10.2

N %

Male 955 69.1

Married 581 42.0

Education > H.S. 398

71.2

Employed 1006 72.7

Non-minority 1055 76.3

Drinking and Severity Measures at Baseline

Mean s.d.

Drinking days, % in last 30 days 74.9 24.92

Drinks per drinking day 12.5 7.81

ADS Score 16.7 7.44

DSM-IV symptoms 5.5 1.49

Total OCDS Score 20.0 9.59

Total DrInC Score 47.6 20.45

GGT, % > normal (63 IU/L) 31.4

CDT, % > normal (2.6%) 52.1

Medication Dosing• Target Doses and Procedures

– Naltrexone 100 mg daily or placebo, given as two 50 mg tablets orally in AM

– Acamprosate 3 grams (3000 mg) or placebo given as two 500 mg tablets orally in AM, mid-day, and PM

– Blister packed to enhance compliance– Pill counts for compliance were obtained

• Rationale For Doses– Naltrexone: 100 mg doses may reduce effects of

missed doses and provide greater efficacy than 50 mg– Acamprosate: 3 gram dose found effective in Lipha

U.S. clinical trial– Pilot work demonstrated that the combination was well

tolerated at these doses (Johnson et al., 2003;Combine Research Group, 2003)

Medication Dosing (continued)

– Dose reductions were allowed to try to retain patients in treatment and we re-challenged them when practical

– Patients who discontinued meds (or therapy) were not dropped from the protocol and continued to be assessed 

Behavioral Interventions

• Medical Management Sessions– By a licensed health care professional

• 14 physicians, 28 nurses, 1 physician assistant, 1 clinical pharmacist

– Initial visit 45 minutes modeled after a new patient visit

– Subsequent visits were 20 minutes on average– Up to 9 sessions scheduled weekly for 1 month, bi-

weekly for 3 months and once in the final month– Median number of sessions attended = 9

Behavioral Interventions

• Combined Behavioral Intervention– Licensed behavioral health specialists, all with at least

a masters degree– Up to twenty 50-minute sessions, the frequency and

number negotiated with the patient– Median number of sessions attended =10

Schedule of Assessments

• Baseline• In-treatment at 9 MM visits• In-treatment: 2, 4 months

• Post-treatment follow-up: 8, 12, 16 months

Assessment Domains

• Screening / Eligibility• Medical / Physiological / Laboratory• Treatment-Related Expectancies• Alcohol Consumption• Alcohol / Drug Involvement• Craving• Motivation• Psychological / Psychiatric• Mutual Help Involvement• Quality of Life• Service Utilization• Treatment Compliance and Process

Primary Outcome Measures

• Percent Days Abstinent (not drinking) (PDA)• Time to first heavy drinking day

– Males: 5 drinks in one day– Females: 4 drinks in one day

Statistical Methods

• PDA: Mixed effect linear models– Intention to treat population: all randomized patients

with any post-randomization drinking data

• Time to Relapse to Heavy Drinking: Proportional hazards models – Intention to treat population: all randomized patients

(loss to follow-up treated as relapse)

• Family-wise error control:  = .05, two-tailed for each main effect and interaction

• Bonferroni correction for two co-primary endpoints leading to an = .025 for each primary measure

Statistical Analysis

• Primary analyses– Based on 16 week treatment period

• Secondary analysis for – CBI without pills comparisons– 1 year post-treatment period

• Primary efficacy analysis

– Traditional factorial ANOVA model fitting and evaluating main effects and interactions

Treatment Group CombinationsAcamprosate Main Effect

Medical Management (n=607)

Placebo Acamprosate Placebo 153 152 Naltrexone 154 148

Medical Management + CBI (n=619)

Placebo Acamprosate Placebo 156 151 Naltrexone 155 157

Treatment Group CombinationsNaltrexone Main Effect

Medical Management (n=607)

Placebo Acamprosate Placebo 153 152 Naltrexone 154 148

Medical Management + CBI (n=619)

Placebo Acamprosate Placebo 156 151 Naltrexone 155 157

Treatment Group CombinationsCBI Main Effect

Medical Management (n=607)

Placebo Acamprosate Placebo 153 152 Naltrexone 154 148

Medical Management + CBI (n=619)

Placebo Acamprosate Placebo 156 151 Naltrexone 155 157

Treatment Group CombinationsNaltrexone x CBI Interaction

Medical Management (n=607)

Placebo Acamprosate Placebo 153 152 Naltrexone 154 148

Medical Management + CBI (n=619)

Placebo Acamprosate Placebo 156 151 Naltrexone 155 157