Chapter 6 – Clinical Trials in Substance using populations

Intro:
- what is a trial, types of trials, primary and secondary objectives, participant recruitment
informed consent.
- IRB (institutional review board): protects the rights and welfare of human subjects.

Types
- Efficacy trials/phase II – compare an intervention to a treatment, designed to be sensitive
to defect a “signal” or effect
 Participants have the illness of interest but no other comorbidities, no other
medications or active treatments, usually fixed doses, randomized, sensitive scales.
 Ex. Buprenorphine 16mg or 24mg in patients with OUD, no other SUD or psychiatric
disorders, no other medications, 100 pts.
- Efficacy trials/phase III – larger sample size, less stringent exclusion criteria, population
more similar to target clinical population, detect more ADRs
 Ex. Buprenorphine titrated up to 24mg, pts with other psych medications and SUDs,
600pts.
- Effectiveness trials – assess the impact of an effective therapy in real-world setting.
 Uses real-world population
 Usually an intervention and “treatment as usual” group
 More flexible dosing regimens
 Allow concurrent treatments
 Ex. Buprenorphine vs placebo, both groups get IOP

Comparative effectiveness study - the generation of and synthesis of evidence that compares
the benefits and harms of alternative methods to prevent, diagnose, treat, and monitor a
clinical condition or to improve the delivery of care
 Similar to effectiveness trials, although comparator may be a different modality of
treatment
 Ex. Buprenorphine versus extended release naltrexone, looking at outcome
measures, ADRs, healthcare utilization, co-morbidities like HIV, legal issues etc.

Features of Clinical Trials

Randomization – different methods, but the goal is to have the groups be as similar as can be to
avoid selection bias

Blinding
Single-blind – patient only
Double-blind – patient & investigator
Type I error: finding a difference when one does not exist,
false positive

Type II error: finding no difference when one does in fact

exist, false negative

Power is the ability to detect an effect

Confidence interval – measure of statistical significance

Statistical Analysis Plans

- Specify the proposed analyses for the primary, secondary, and exploratory outcomes
- Intention-to-treat - all randomized participants whether they received the intervention or
not and whether they may or may not have had data assessments
- As-treated/per protocol - population that completed the protocol with minimal or no
missing data

What to do with missing data points?

This is inherent to all areas or research but especially so in SUD work. Options are treating
missing data points as positive, negative, neutral. Some trials use a last result carried-forward
approach. This needs to be considered and planned for beforehand.

Statistical Significance & Effect Sizes

Effect size – measure of the magnitude of an effect
Cohen’s d: 0.2 small, 0.5 medium, 0.8 large

Trial Designs
- Superiority designs: the outcome of interest is whether one intervention produces a
superior outcome versus the comparator.
 One method of showing this is with dose-response studies, where a statistically
significant positive slope is considered to be evidence of efficacy of a medication.
- Non-inferiority designs: compares two active treatments with the purpose of determining
whether the effectiveness of one treatment is not worse than the standard therapy.
- Adaptive Designs: sequential assignments of participants to new treatments are made
following predetermined decisions rules
 more closely replicate clinical practice
  In the simplest model, participants are randomized to a treatment group and are
assessed for response/nonresponse at a decision point. Those patients who do not
respond can then be assigned an alternate treatment assignment while responders
may continue with their treatment. Study participants can also be randomized twice,
initially to a treatment group and then following a decision point, to a second
randomized assignment.
 STAR*D

Outcome Metrics
- Common primary outcomes are:
 Abstinence - can be defined as 90 days with no substance use, or no positive UDS
screens
 Days of use
 Heavy drinking days (4 men and 3 women)
- FDA guidance on outcome measures: a measure that predicts ongoing abstinence and/or
good psychosocial functioning and physical functioning.
 FDA specifies that for trials in AUD, must be at least 6 months because abstinence at
6 months predicts abstinence at 5 years
- Other groups and experts contend that a 50% reduction in drug use is clinically meaningful.

- Secondary measures
 Quantification of substance use, usually self-reported
 Biological assays – serum levels, urine levels
 Scales – cravings, withdrawal
- CIWA
- Clinical opioid withdrawal scale (COWS)
- The Structured Clinical Interview for DSM-5 (SCID-5) is used to
systematically evaluate psychiatric diagnoses during screening.
- Beck Depression inventory – self-rated
- Hamilton Depression rating scale (HAM-D)
- Hamilton anxiety rating scale
- Wisconsin card sorting test – cognitive flexibility test
- Addiction severity index (ASI) – a 1-hour semi-structured interview
designed to address multiple problem domains in substance-abusing
patients
 Reduction in HIV incidence
 Healthcare utilization
 Quality of life measures – housing, mobility, employment, pain, legal issues
 Serum assays
In one study, pts were given PO buprenorphine 8mg daily and another group was
given 16mg every other day. The Trough levels were 0.80ng/mL and 0.77ng/mL
with no difference in withdrawal symptoms between the two groups; so the
 inference is that serum levels of 0.7ng/mL are enough to suppress withdrawal
symptoms.
 Treatment adherence – assays, diaries, direct observation, pill counts
2 trials looking at the effectiveness for Bupropion for methamphetamine use
were negative, however treatment adherence rates were as low as 32%; this
would be considered a failed trial. However, looking at end-of-treatment
abstinence in adherent vs non-adherent participants showed rates of 54% vs 18%
respectively.

Cost-analysis
- Cost-effective analysis: quality-adjusted life years – measure the benefit of an intervention
per unit cost
- Cost-benefit analysis: whether the benefit of an intervention exceed its cost

Conclusions
There are lots of issues that need to be considered when designing a trial, and these are guided,
in part, by an explicit understanding of the purpose of the trail.

Moreover, there is no consensus on what constitutes an adequate duration of abstinence or

what level of improvement in psychosocial functioning is acceptable in those who do not
achieve full abstinence. Other inherent issues include high dropout rates, missing data points
and treatment adherence.