Transcript Open slides - CTN Dissemination Library

Effects of Participant Compensation Amounts on Missing Data and Urine Screen Results
Among Adolescent and Young Adult Opioid Dependent Clinical Trial Participants
NA’NIZHOOZHI
CENTER
C. E. Wilcox1, M. P. Bogenschutz1,2, M. Nakazawa2, G. E. Woody3
1University
of New Mexico Department of Psychiatry, Albuquerque, NM
2University of New Mexico Center on Alcoholism, Substance Abuse and Addictions (CASAA)
3University of Pennsylvania Department of Psychiatry, Philadelphia, PA
ABSTRACT
TABLE 1: Summary of GLMM Analysis for Predicting Non-Missing
Values (N=152)
A secondary analysis of a study of 152 subjects aged 15-21 seeking treatment for opioid
dependence were randomized to 2 week detoxification with buprenorphine/naloxone
(DETOX) or 12 weeks buprenorphine/naloxone (BUP) (Woody et al., 2008). Higher
compensation amounts were associated with improved retention rates, especially in the
DETOX group. These findings, and others, suggest that the amount of financial
compensation given for completing assessments can minimize bias when treatment
condition is associated with differential dropout rates.
INTRODUCTION
Notes: Results from Model 4 applied to the Missing vs. Non-Missing Variable; *p<0.05, **p<0.01.
Non-Missing values are defined as positive or negative urine screen values.
Attrition in substance abuse research is a major problem, with rates of data collection
often falling well below 70% (Lavori et al., 1999). More than 30% dropout can result in •High compensation amount was associated with lower rates of missing data.
considerable bias when analyzing outcomes data (Scott, 2004). Financial compensation
•DETOX assignment increased overall likelihood of missing data.
can increase retention in research studies, raising collection rates of outcomes data to
•The effect of compensation amount on missing data was greater in the DETOX
acceptable levels (Festinger et al., 2008). Follow-up rates in adolescent studies have
group.
been similarly problematic (Stinchfield et al., 1994; Winters et al., 2000). Studies in
youth have shown that intensive follow up protocols improve research adherence
TABLE 2: Summary of GLMM Analysis for Predicting Positive Opioid
(Meyers et al., 2003), and that financial compensation improves treatment adherence
and outcomes (Lott and Jencius, 2009). However, to our knowledge, there has been no Urine Screen with Missing Values Assigned as Non-Positive (N=152)
research on the effects of financial compensation amounts on adherence to research
assessments in substance abusing adolescents and young adults.
METHODS
152 subjects aged 15-21 seeking treatment for opioid dependence were randomized to 2
week detoxification with buprenorphine/naloxone (DETOX) or 12 weeks
buprenorphine/naloxone (BUP) with a dose taper beginning in week 9 and ending in
week 12, each with weekly individual and group drug counseling. Urine drug screens
and self reported drug use were obtained weekly. Patients were paid $5 for completing
the weekly assessments except for weeks 4, 8, and 12, where more extensive
assessments were done and participants were reimbursed $75. GLMM was chosen as
the statistical model for this secondary analysis of incentive effects, as it is effective for
analysis of longitudinal dichotomous variables. Separate models were constructed for
predicting missing, positive, and negative urine drug screen results. Compensation
(High versus Low), Treatment (BUP versus DETOX) and Time (Time 1 = Weeks 3,4,5;
Time 2 = Weeks 7,8,9; Time 3 = Weeks 11,12) were chosen as predictor variables.
RESULTS
FIGURE 1
•Attrition increased over time, with highest attrition at week 11.
•Dropout during the first two weeks was pronounced, especially among DETOX
subjects.
A BRIDGE TO RECOVERY
DISCUSSION
•Higher financial compensation amounts increase retention for follow-up assessments.
•Patients in control groups and actively using patients may be especially vulnerable to
dropout if adequate financial compensation is not provided.
•Attrition increases over time, especially in the case of low financial compensation
amounts. Outcomes data may be affected differently over time by financial
compensation amount, depending on treatment group assignment, introducing the
potential for further bias.
•Non-active users (those with negative urine screens) may be less influenced towards
dropout overall by the length of time in the study than active users (those with positive
urine screens).
•BUP assignment independently increased the rate of negative urine screen results,
supporting the findings of the primary study showing that BUP assignment was
associated with better outcomes.
•The especially strong effect of compensation amount on follow-up rates in active
users could be understood to reflect coercion, and the possibility that high
compensation biases analyses by preferentially retaining subjects who are more ‘down
and out’ (Festinger 2008). However, given the markedly low retention rates during low
financial compensation weeks, it is more likely that the increased retention rates
favored by higher financial compensation minimizes bias by decreasing attrition to
acceptable levels.
CONCLUSIONS
Higher participant compensation amounts are essential for improving retention in
research studies of substance abusing youth and likely minimize bias in outcomes
analyses, especially when there are differences in retention as a function of treatment
group assignment.
REFERENCES
Notes: Results from Model 4 applied to the Positive vs. Non-Positive Variable; +p<0.10, **p<0.01.
Non-Positive values are defined as missing or negative urine screen results.
Festinger, D.S., Marlowe, D.B., Dugosh, K.L., Croft, J.R., Arabia, P.L., 2008. Higher
magnitude cash payments improve research follow-up rates without increasing drug use
•Probability of positive urine screens was significantly higher during the high
or perceived coercion. Drug Alcohol Depend 96, 128-135.
compensation weeks.
Lavori, P.W., Bloch, D.A., Bridge, P.T., Leiderman, D.B., LoCastro, J.S., Somoza, E.,
•The effect of compensation on the probability of positive urine screens was
1999. Plans, designs, and analyses for clinical trials of anti-cocaine medications: where
greater in DETOX subjects in comparison to BUP subjects.
we are today. NIDA/VA/SU Working Group on Design and Analysis. J Clin
•The rate of urine screens being positive significantly decreased over time.
Psychopharmacol 19, 246-256.
Lott, D.C., Jencius, S., 2009. Effectiveness of very low-cost contingency management
in a community adolescent treatment program. Drug Alcohol Depend 102, 162-165.
TABLE 3: Summary of GLMM Analysis for Predicting Negative Opioid Meyers, K., Webb, A., Frantz, J., Randall, M., 2003. What does it take to retain
Urine Screen with Missing Values Assigned as Non-Negative (N=152) substance-abusing adolescents in research protocols? Delineation of effort required,
strategies undertaken, costs incurred, and 6-month post-treatment differences by
retention difficulty. Drug Alcohol Depend 69, 73-85.
Scott, C.K., 2004. A replicable model for achieving over 90% follow-up rates in
longitudinal studies of substance abusers. Drug Alcohol Depend 74, 21-36.
Stinchfield, R.D., Niforopulos, L., Feder, S.H., 1994. Follow-up contact bias in
adolescent substance abuse treatment outcome research. J Stud Alcohol 55, 285-289.
Winters, K.C., Stinchfield, R.D., Opland, E., Weller, C., Latimer, W.W., 2000. The
effectiveness of the Minnesota Model approach in the treatment of adolescent drug
abusers. Addiction 95, 601-612.
Notes: Results from Model 4 applied to the Negative vs. Non-Negative Variable; *p<0.05, **p<0.01.
Woody, G.E., Poole, S.A., Subramaniam, G., Dugosh, K., Bogenschutz, M., Abbott, P.,
Non-Negative values are defined as missing or positive urine screen results.
Patkar, A., Publicker, M., McCain, K., Potter, J.S., Forman, R., Vetter, V., McNicholas,
•Probability of negative urine screens was significantly higher during the high
L., Blaine, J., Lynch, K.G., Fudala, P., 2008. Extended vs short-term buprenorphinecompensation weeks.
naloxone for treatment of opioid-addicted youth: a randomized trial. JAMA 300, 2003•BUP assignment increased the rate of urine screens being negative.
2011.
•The effect of compensation amount on the probability of negative urine screens
ACKNOWLEDGEMENTS
did not differ between DETOX and BUP subjects.
•There was no change in the rate of urine screens being negative over time.
This research was supported by NIDA’s Clinical Trials Network