The Brief Addiction Science Information Source (BASIS)

STASH, Vol. 8(4) – Does the approach bias predict escalation of cannabis use?

When people use drugs for a prolonged period of time, they can develop an automatic tendency to approach rather than avoid drug-related cues – a trait named approach bias (Mogg & Bradley, 2005). This week’s STASH reviews a Dutch study that investigates if the approach bias toward cannabis-related stimuli predicts future escalation of cannabis use among heavy cannabis users (Cousijn, Goudriaan, & Wiers, 2011).

Method

• Participants were 32 heavy cannabis users (34% female) and 41 non-using controls (37% female) aged 18 – 25 recruited through advertisements on the Internet and in the coffee-shops in Amsterdam¹.
  • Heavy cannabis use was defined as use on 10 or more days during the last month and 240 days during the past 2 years.
  • Participants who met two criteria—fewer than 50 life-time use occasions and no use during the last year—were considered to be non-users.  
  • People who did not meet these definitions were not eligible to participate in the study. The authors do not report the number of these people.
• Researchers used an Approach Avoidance Task to assess whether participants had an approach bias towards cannabis-related images. Participants viewed slightly rotated cannabis and neutral images and pulled or pushed a joystick as quickly as they could to indicate the rotation direction (i.e., clockwise or counter-clockwise)¹. Pulling the joystick increased the image size and pushing the joystick out decreased the image size, emphasizing the sense of approach or avoidance, respectively.
• The researcher calculated median approach and avoidance response times (RTs) for cannabis-related and neutral images. Next, they calculated the approach bias for each image category by subtracting the median approach RT from median avoid RT. Positive scores indicate an approach bias (i.e., quicker approach responses).
• Researchers assessed cannabis use and related problems (Cannabis Use Disorder Identification Test, Adamson, & Sellman, 2003) and other measures during the baseline and 6-month follow-up sessions. They achieved a 97% 6-month follow-up rate. 

Results

• The researchers conducted a 2 (user status) X 2 (image type) ANOVA test. Table 1 demonstrated significant (F (1,69) = 4.53; p < .05) image-type/cannabis use interaction effect on the approach bias. Then, the researchers conducted follow-up t-tests to indicate the simple effects of user type within each level of image category. Heavy cannabis users demonstrated a larger approach bias toward cannabis-related images than controls (t [69] = 2.33, p< .03, d = 0.55). Heavy cannabis users and controls did not differ on approach bias toward neutral images (p = .91).

Table 1. Mean (sd) approach bias for cannabis-related and neutral images  

	Cannabis image	Neutral image
Heavy cannabis  users	803.7 (80.1)	827.5 (79.7)
Nonusers	831.4 (72.5)	837.3 (75.1)

• Among heavy cannabis users, the approach bias toward cannabis-related images was positively correlated with weekly cannabis use at 6-month follow-up (R =.42, p<.03).
• The researchers attempted to predict change in weekly cannabis use (in grams) from baseline to 6-month follow-up. Baseline cannabis use variables explained 63% of the variance in change in use. The approach bias toward cannabis-related images explained an additional 13% of the variance (F change [1,20] = 10.65, p<. 01). Cannabis users with a stronger approach bias toward cannabis-related images at baseline were more likely to escalate their cannabis use.
• The approach bias did not predict changes in cannabis-related problems and dependence.

Limitations

• Limited sample size combined with the convenience sample might limit the generalizability of the findings. It also is unclear whether the same patterns would hold in a region that is characterized by less tolerant cannabis use policies.
• A greater proportion of cannabis users than non-users were tobacco smokers. Tobacco use might have contributed to the approach bias among cannabis users. 

Conclusions

The study demonstrated that motivational shifts, as reflected in approach-related movements, are related to cannabis use. The strength of the initial approach bias contributed to the predicted of an escalation in cannabis use. This task might be useful in providing personal feedback to individuals wishing to maintain or cut back on their cannabis use. However, it is important to note that escalating use did not correspond to cannabis-related problems and dependence. Therefore, based on the results of this study, we cannot conclude that approach bias toward cannabis-related cues predicts the cannabis induced problems.

-Julia Braverman

What do you think?  Please use the comment link below to provide feedback on this article.

References

Adamson, S. J., & Sellman, J. D. (2003). A prototype screening instrument for cannabis use disorder: the Cannabis Use Disorders  Identification Test (CUDIT) in an alcohol- dependent clinical sample. Drug Alcohol Review, 22, 309 - 315.

Cousijn, J., Goudriaan, A. E., & Wiers, R. W. (2011). Reaching out toward cannabis: Approach- bias in heavy cannabis users predicts changes in cannabis use. Addiction, 106, 1667 - 1674.

Heishman S. J., Evans, R.J., Singleton, E. G., Levin, K. H., Copersino M. L., Gorelick, D. A. (2009). Reliability and validity of a short form of the Marijuana Craving Questionnaire.Drug alcohol Dependence, 102, 35 - 40. 

Mogg, K. & Bradley, F. M. (2005). Attentional and approach biases for smoking cues in smokers: An investigation of copeting theoretical views of addiction. Psychopharmacology, 180, 333 – 341.

STASH, Vol. 8(3) – Compounding the wounds of war: High-risk opioid pain treatment among veterans with mental health issues

Many of the veterans returning from Afghanistan and Iraq have survived injuries typically fatal during past wars. As such, larger portions of the veteran population are experiencing high levels of physical and psychological pain. Opioid medication is very effective at alleviating pain symptoms, but it carries risks of dependence, injury, overdose, and death. Thus, it is crucial to understand how veterans experience their pain relief treatment. This week’s STASH reviews a study that examines opioid prescription practices for veterans, with a focus on how co-morbid mental health conditions impact opioid treatment experience (Seal, Shi, Cohen, Cohen, Maguen, Krebs et al., 2012).

Methods

The study cohort originated from the Veterans Affairs (VA) roster of Operation Enduring Freedom and Operation Iraqi Freedom veterans.

• The study retrospectively examined veterans who entered the VA healthcare system between Oct 1, 2005 and Dec 31, 2008 (N=291,205)
• For the results reported here, the authors focused on veterans who received a new non-cancer diagnosis that could result in pain within 1 year of VA entry (N=141,029)
• Using data obtained from VA databases, the authors examined medical care and outcomes for 1 year following non-cancer diagnosis.
•  The researchers divided the cohort into three groups:
  • Veterans with no mental health diagnosis (49 % of the sample)
  • Veterans with mental health diagnoses not including PTSD (19% of the sample)
  • Veterans with a PTSD diagnosis with or without another mental health diagnosis (32% of the sample).
• The authors measured opioid prescriptions, including duration and dosage, as well as adverse clinical outcomes (e.g., accidents resulting in wounds or injuries, opioid-related accidents and overdoses, and self-inflicted injuries).

Results

• Although 6.5% of veterans without a mental health diagnosis were prescribed opioids, 11.7% of those with a mental health diagnosis other than PTSD and 17.8% of those with a PTSD diagnosis were prescribed opioids. Of the sub-group of patients with a PTSD diagnosis, those with co-morbid substance use disorders were especially likely to be prescribed opioids. (See Figure.)
• Among those who were prescribed opioids, veterans with PTSD experienced higher-risk treatment compared to veterans without mental health diagnoses. Compared to those without a mental health diagnosis, those with PTSD were
  • More likely to be part of the group prescribed the highest quintile dose of opioids (22.7% vs. 15.9%; adjusted relative risk = 1.42)
  • More likely to be prescribed opioids for longer than 2 months (63.2% vs. 42.7%; adjusted relative risk, 1.47)
  • More likely to receive multiple opioids (19.8% vs. 10.7%; adjusted relative risk, 1.87).
• Veterans who were prescribed opioids had a higher rate of experiencing any adverse clinical outcome than those who were not prescribed opioids (9.5% vs. 4.1%; adjusted relative risk: 2.33).

Figure: Rates of receiving an opioid prescription among the veteran groups.

 

* = adjusted relative risk compared with the group with no mental health diagnosis

Note: The group of veterans with PTSD and substance use disorder is a sub-group of the group of veterans with PTSD.

Limitations

• The VA administrative databases lacked variables such as education and level of combat exposure. As a result, potential confounding variables could be missing.
• The study did not have information concerning non-prescription opioid use or prescriptions outside of the VA system.
• Some of the adverse clinical outcomes might have occurred after veterans discontinued their prescription opioid use.
• The specific inclusion criterion of VA veterans with a non-cancer pain diagnosis means that the sample might not be representative of all veterans.

Conclusions

Veterans with a mental health diagnosis experienced higher-risk opioid treatment compared to those without such a diagnosis; this was especially true for veterans with PTSD. This treatment experience was associated with adverse clinical outcomes. The authors speculate that veterans with a mental health condition, especially PTSD, present to their primary care providers with greater pain levels, potentially because they experience psychological pain as physical pain, or because their endogenous pain control systems are not functioning well.  These results suggest that it might be useful to provide more training to VA primary care providers and revisit opioid treatment practices for veterans who present with mental health issues.

-Jed Jeng

What do you think? Please use the comment link below to provide feedback on this article.

References

Seal, K. H., Shi, Y., Cohen, G., Cohen, B. E., Maguen, S., Krebs, E. E., et al. (2012). Association of mental health disorders with prescription opioids and high-risk opioid use in US veterans of Iraq and Afghanistan. Journal of the American Medical Association, 307(9), 940-947.

STASH, Vol. 8(2) - A risky inheritance? Brain abnormalities among drug-dependent individuals and their non-addicted siblings

Researchers have observed brain abnormalities among people who are dependent on stimulant drugs (Ersche et al., 2011). More specifically, these individuals show abnormalities in fronto-striatal brain systems known to be involved in self-control (Ersche et al., 2011). It is also known that self-control deficits increase the risk for substance use disorders (Nigg et al., 2006), and that brain structure is largely inherited (Wright et al., 2002). Consequently, inherited patterns of brain structure and function that increase an individual’s risk for impaired self-control likely increase the risks of stimulant dependence. Ersche and colleagues (2012) recently explored this hypothesis by studying the brain structure and functioning of drug-dependent individuals and their non-dependent siblings.

Methods

• Researchers recruited 50 biological sibling pairs. In each pair, one sibling met DSM-IV criteria for current dependence on cocaine or amphetamine, and the other sibling had no history of substance dependence (except nicotine).
• The researchers also recruited 50 unrelated, non-drug dependent volunteers matched with the siblings on age and intelligence quotient.
• All participants performed a button-press behavior repeatedly and quickly except whenever they heard an auditory signal. This task that measured inhibitory control, an aspect of self-control. The researchers measured how long it took participants to refrain from pressing the button during these “stop” trials. The resulting measure, the stop-signal reaction time (i.e., SSRT), predicts the onset of substance abuse among vulnerable individuals (Nigg et al., 2006).
• All participants underwent magnetic resonance brain scans, which allowed the researchers to quantify and compare the integrity of white matter fiber tracts and the volume of gray matter throughout the brain.

Results

• Both drug-dependent individuals (t (95) = 3.83, p < .001) and their non-dependent siblings (t (97) = 3.73, p < .001) had longer SSRTs than the control group participants, suggesting impairments in the regulation of behavior.
• Looking specifically at neural networks used during inhibitory control (i.e., the basal ganglia, inferior frontal gyrus, and presupplementary motor area), the biological siblings showed less white matter integrity than the control group (group F (2, 141) = 26.2, p < .001). See Figure 1.  
• Reduced white matter integrity in this neural network was associated with worse inhibitory performance (i.e., longer SSRTs) (r = 0.24), suggesting that white matter integrity within this neural network is an objective marker of poor inhibitory control, a risk factor for stimulant drug dependence.
• Certain sections of gray matter were significantly smaller among the biological siblings as compared to the controls, including areas known to be involved in addiction (e.g., medial temporal lobe, basal ganglia, orbitofrontal cortex, putamen).

Figure 1: A depiction of white matter integrity.

The blue shading represents areas in which the participant groups differed in terms of white matter integrity. The orange shading represents the inferior frontal gyrus and presupplementary motor area.

From Ersche et al. (2012) 

 

Limitations

• Although some drug-dependent participants were recruited from treatment centers, the authors do not provide information regarding the potential association between treatment history and brain function or structure.
• Because the authors focused on the effects of stimulant dependence, it is unclear if people who are dependent on other substances or experience behavioral expressions of addiction (e.g., gambling) exhibit similar patterns.  
• The non-dependent siblings showed deficits in neural networks known to be involved in self-control, as well as impaired self-control. However, this study is not positioned to explain why these individuals did not go on to develop drug dependence.

Conclusion

A thorough family history is crucial in studying inherited brain structure and determining illness for which individuals are at risk. This study’s findings suggest two conclusions. First, the results lend support to the idea that some risk for substance dependence is inherited with brain structure and functioning and suggests that there might be some merit in either monitoring or providing more specialized treatment to those with a family history of substance use. At the same time, the comparison of biologically related siblings implies that some people, whether through personality differences, lack of drug exposure, brain chemistry, or some other set of protective factors, are able to work against an apparent predisposition to substance dependence. Future work should explore, among other issues, these individuals’ marked ability to resist addiction.

-Ingrid Maurice & Heather Gray

What do you think?  Please use the comment link below to provide feedback on this article.

References

Ersche, K. D., Barnes, A., Jones, P. S., Morein-Zamir, S., Robbins, T. W. & Bullmore, E. T. (2011) Abnormal structure of frontostriatal brain systems is associated with aspects of impulsivity and compulsivity in cocaine dependence. Brain, 134, 2013-24.

Ersche, K. D., Jones, P. S., Williams, G. B., Turton, A. J., Robbins, T. W. & Bullmore, E. T. (2012) Abnormal brain structure implicated in stimulant drug addiction. Science, 335, 601-4.

Nigg, J. T., Wong, M. M., Martel, M. M., Jester, J. M., Puttler, L. I., Glass, J. M., Adams, K. M., Fitzgerald, H. E. & Zucker, R. A. (2006) Poor response inhibition as a predictor of problem drinking and illicit drug use in adolescents at risk for alcoholism and other substance use disorders. Journal of the American Academy of Child & Adolescent Psychiatry, 45, 468-475.

Wright, I. C., Sham, P., Murray, R. M., Weinberger, D. R. & Bullmore, E. T. (2002) Genetic contributions to regional variability in human brain structure: methods and preliminary results. Neuroimage, 17, 256-71.

STASH, Vol. 8(1) – Who are dual alcohol and marijuana users?

When people drink alcohol, they are more likely also to use marijuana than any other illicit drug (Degenhardt, Hall et al. 2001). However, little is known about this group of dual alcohol/marijuana users, or about how different levels of alcohol and marijuana use affect other risky behaviors. This week’s STASH reviews a study that investigates behavioral and psychosocial characteristics of those who use both alcohol and marijuana (Harrington, Baird et al. 2012). The investigators sought to determine if there are distinct clusters of alcohol/marijuana users and examine the range of potential negative consequences associated with this pattern of substance use.

Method

• Investigators conducted a secondary data analysis of data previously collected for a randomized control trial focusing on brief intervention at an emergency service within a  trauma center. 
• Participants were 427 individuals who (1) agreed to participate in a study about risk behavior reduction among alcohol and marijuana users and (2) reported using both alcohol and marijuana during the past 30 days.
• Participants reported how frequently they used alcohol and how frequently they used marijuana during the past 30 days.
• The researchers used cluster analysis to identify sub-types of dual users based on frequency of alcohol and marijuana use.In addition, participants completed the following self-report measures of negative consequences:
  • Alcohol Use Disorders Inventory Test (AUDIT): A measure of alcohol use frequency, binge drinking episodes, and alcohol-related negative consequences for the past year.
  • Marijuana Problem Scale (MPS): A measure of negative psychological, social, occupational, and legal consequences of marijuana use.
  • Noteworthy Index of Problems (NIP): A measure of frequency of psychosocial events related to alcohol and/or marijuana use during the past three months (e.g., saying harsh or cruel things to someone, getting in a physical fight, saying embarrassing things). The researchers calculated overall scores as well as separate alcohol-related and marijuana-related scores.
  • Injury Behavior Scale (IBC): A measure of injuries during the past year, with separate scores for total injuries, alcohol-related injuries, and marijuana-related injuries. The researchers calculated overall scores as well as separate alcohol-related and marijuana-related scores.
  • High Risk Behavior Scale (HRB): A measure of 13 risky behaviors, some related to alcohol and/or marijuana use (e.g., driving after marijuana use) and others not necessarily related to alcohol and/or marijuana use (e.g., speeding, being involved in a physical fight).
  • For each of these measures, except the HRB, items were combined to form a single score. For the HRB, items were kept separate.
  • The researchers used Analysis of Variance (ANOVA) and Tukey post-hoc tests to compare how participants in different clusters differed in terms of psychosocial and behavioral characteristics. 

Results

• The analyses revealed four distinct clusters of drug users based on the frequency of alcohol and marijuana use. The clusters were replicated independently using two sub-samples. 

Table 1. Identified clusters

Frequency of Alcohol Use
Frequency of Marijuana Use		Daily	Weekly
	Daily	Cluster 3 (N = 56)	Cluster 1 (N = 93)
	Weekly	Cluster 4 (N = 55)	Cluster 2 (N = 223)

• The researchers observed statistically significant differences among the four clusters for the AUDIT (F (3, 420)=39.92, p < .001), the MPS (F (3, 417)= 6.01, p < .001), the overall NIP (F (3, 410)=10.63, p < .001), the overall IBC (F (3, 423)=6.77, p < .001), and several of the HRB items.
• Figure 1a and Figure 1b show AUDIT, MPS, overall IBC, and overall NIP scores as a function of the four clusters of drug users.  

Figure 1a. AUDIT scores as a function of clusters of participants.

Figure 1b. MPS, total IBC and total NIP scores as a function of clusters of participants

• Post-hoc Tukey tests revealed the following statistically significant specific group differences:
  • For the AUDIT, Cluster 3 scores and Cluster 4 scores were both higher than Cluster 1 and Cluster 2 scores 
  • For the MPS Cluster 3 scores were higher than Cluster 2 and Cluster 4 scores
  • For the NIP, Cluster 3 scores were higher than Cluster 1 and Cluster 2 scores
  • For the IBS, Cluster 3 scores were higher than Cluster 1 and Cluster 2 score

Limitations

• This study uses a self-reported data.
• The study measured only frequency of alcohol and marijuana use, but not quantity.

Conclusion

These analyses identified four distinct groups of alcohol and marijuana users according to the frequency of using both substances. The results showed that individuals who used alcohol daily experienced more alcohol-related problems than individuals who used it weekly, independent of marijuana use. The pattern of marijuana use was less clear. For instance, people who used marijuana daily reported more instances of driving under the influence of marijuana than did weekly marijuana users, regardless of how frequently they used alcohol. However, for the number of marijuana related problems, daily marijuana users were not significantly different from weekly marijuana users, if the marijuana user also used alcohol weekly. The daily users of both substances reported more marijuana-related problems than weekly marijuana users. The authors speculate that individuals who use both substances daily might need targeted intervention and treatment efforts because of the highest number of associated problems.

-Julia Braverman

What do you think?  Please use the comment link below to provide feedback on this article.

References

Degenhardt, L., W. Hall, et al. (2001). "The relationship between cannabis use and other substance use in the general population." Drug and Alcohol Dependence 64(3): 319-327.

Harrington, M., J. Baird, et al. (2012). "Identifying subtypes of dual alcohol and marijuana users: A methodological approach using cluster analysis." Addictive Behaviors 37(1): 119-123.

STASH, Vol. 7(11): Do smart kids become adult users of illegal drugs?

Children who score relatively high on intelligence tests become relatively healthy adults, in some important respects. For instance, White and Batty (2011) noted that they are less likely to smoke, more likely to engage in physical activity, and more likely to eat healthy foods than their peers. However, research is equivocal regarding a link between childhood IQ scores and adult use of illegal drugs (as reviewed by White & Batty, 2011). In this week’s STASH, we review a recent longitudinal, population-based study investigating the connection between childhood IQ and adult use of illegal drugs (White & Batty, 2011).                   

Methods

• The authors used data from the 1970 British Cohort Study, an ongoing longitudinal study of children born in Great Britain during April, 1970.
• The children in the study were enrolled at birth and were invited to participate at ages 5, 10, 16, 26, and 29-30.
• At age 5, participants completed four tests purported to measure general cognitive ability (e.g., the Human Figure Drawing test, shown at right). The researchers transformed scores to the standard IQ distribution (mean = 100, SD = 15). Also at these assessments, the researchers made note of parents’ social class.
• At age 16, participants indicated their lifetime use of cannabis and cocaine (yes/no).
• At age 30, participants indicated their past-year use of several illegal drugs (yes/no).
• The researchers completed logistic regression to estimate the relation between childhood IQ and participant gender with illegal drug use at ages 16 and 30. They modeled childhood IQ as tertiles, with the lowest IQ group as the reference group. Positive odds ratios therefore indicate that a higher risk of illegal drug use is associated with moderate or high childhood IQ.*  In the results presented here, researchers statistically controlled for parental social class at age 5. 

Results

• The researchers present drug use outcomes separately for men and women at both 16 and 30 years.**
• The strongest effects emerged from comparisons between the highest childhood IQ tertile and to the lowest IQ tertile. Figure 1 illustrates only these comparisons.*** 
• Male and female participants with relatively high childhood IQ were more likely to report having used cannabis but not cocaine at age 16.
• Male participants with relatively high childhood IQ were more likely to report past-year use of cannabis and cocaine at age 30.
• Female participants with relatively high childhood IQ were more likely to report past-year polydrug use and use of cannabis and cocaine at age 30.
• Associations between childhood IQ and adult illegal drug use were stronger for women than for men.

Figure 1
Odds ratios and 95% confidence intervals for illegal drug use outcomes at 16 and 30 years, separately for men and women. Adapted from White and Batty, 2011. (Click to enlarge.)

Limitations

• Participants did not report important features of their drug use, such as frequency, quantity, and consequences. Accordingly, these results do not speak to the potential association between childhood IQ and substance use disorders (e.g., dependence, abuse).
• As with most longitudinal studies, this research was marked by some attrition. For instance, only 46% of the participants at the 30-year follow-up completed all earlier surveys. Indeed, participants who completed the 30-year follow-up had slightly higher childhood IQ scores than participants who did not. It is unclear how this selective attrition might have influenced observed relationships between childhood IQ and adult drug use. 

Conclusion
In this large, population-based cohort study, the authors observed a consistent pattern: higher childhood IQ predicted greater likelihood of illegal drug use during adulthood, independent of parental social class. The authors speculate about a potential pathway involving personality. According to this reasoning, high IQ individuals use drugs because they are characterized by traits linked with drug use, such as stimulation seeking and openness to experience. The presence of stronger associations for women than men requires further exploration, as does the potential association between childhood IQ and adult substance dependence and abuse.

-Heather Gray

What do you think?  Please use the comment link below to provide feedback on this article.

Reference

White, J., & Batty, G. D. (2011). Intelligence across childhood in relation to illegal drug use in adulthood: 1970 British Cohort Study. Journal of Epidemiology and Community Health, Online first: November 14, 2011.



* Preliminary analyses suggested that modeling IQ as a linear term was inappropriate.
** The interaction term between childhood IQ and participant gender for drug use outcomes was significant at 30 years but not at 16 years.
*** Comparisons between the moderate IQ tertile and the lowest IQ tertile were generally in the same direction. Readers interested in more details can see White and Batty (2011).

 

STASH, Vol. 7(10) – How does marijuana smoking impact driving? A randomized placebo- controlled study

About 10 million Americans report that they have driven under the influence of an illicit drug at least once during the past year (US Department of Health and Human Services, Substance and Mental Health Services Association, Office of Applied Studies, 2007). Marijuana is the most commonly used illicit drug (ibid.). This week’s STASH reviews a randomized control study that examines the effect of smoking marijuana versus a placebo on a driving simulator (Anderson, Rizzo, Block, Pearlson & O’Leary, 2010). 

Method

• Investigators recruited 50 men and 35 women 18  - 31 y/o who used marijuana at least once but fewer than 10 times per month during the past year. The researchers excluded participants whose urine-screening test detected any drug other than tetrahydrocannablnol (THC) the psychoactive ingredient in marijuana.
• During the baseline session, participants became familiarized with a math task that would serve as a distracting task during the driving session.
• During the smoking session, participants smoked a cigarette that contained 2.9% (active) or 0% (placebo) of THC. The instructions encouraged participants to consume the entire cigarette, but participants could stop anytime if they felt uncomfortable. 
• Using a driving simulator, participants then performed a driving assessment (see Figure 1). Each participant drove for approximately 15 miles. An uneventful section of the drive lasted one minute. Then, the driving was interrupted by the following events:
  • Multitasking. Drivers completed a math test designed to distract them from the driving. The investigators measured the number of math errors made compared to the baseline, as well as the speed and the steering wheel position.
  • Response to emergency vehicle.  The task assessed attention to the appearance of a police car. The dependent measures were speed, steering position and reaction time.  
  • Go/NoGo. The task measured safe driving through the yellow light. Safe driving was operationalized as no hesitation in making a decision.
  • Dog incursion avoidance. The task measured safe avoidance of a dog, measured as the ability to stop the car or steer clear of the dog.
  • Intersection incursion avoidance. The task measured the speed at first contact and avoidance tactic.

Figure 1. View of a driving simulator (copied from Anderson, Rizzo, Block, Pearlson & O’Leary, 2010)



• The researchers assessed participants’ heart rate, self-reported level of “highness” (0: no effect – 10:highest) and sleepiness (Stanford Sleepiness Scale; Hoddesetal. 1973) at baseline, after smoking and after driving. 

Results

• The results include only those participants who completed the entire cigarette. The analytic sample included 49 men (25 in the active group) and 24 women (9 in the active group)
• As expected, marijuana smoking significantly increased heart rate (F = 66.4, p < .001) immediately after smoking, and subjective feeling of “highness” compared to placebo (F = 65.1, p < .001) immediately after smoking.
• Women rated themselves as being “higher” than men for both active marijuana and placebo (F = 4.6, p < .05).
• Men were less sleepy than women after driving (F = 4.6, p < .04). Women who smoked active marijuana reported to be more sleepy than men after driving (F = 6.0, p < .02), but not immediately after smoking.
• Both marijuana and placebo groups performed similarly on all driving tasks; there were no sex differences.

Limitations

• The driving simulator results are not necessary directly applicable to real life driving situations.
• This study only investigated the first 15 miles of driving under certain conditions. It is still possible that marijuana affects prolonged driving, or driving under conditions that were not investigated in the study (e.g., slippery road or obscure vision).
• This study uses a small sample, especially in the active marijuana female group. This might be a reason for a failure to find sex X drug effect interaction for all measures, but sleepiness.

Conclusion

A meta analytic study concluded that there was a subtle effect of marijuana on driving performance (Berghaus, Sheer, & Shmidt, 1995). Anderson et al. (2010) provided results that did not support these earlier findings. Using a modern driving simulator might have influenced these results; in addition, the different finding might have emerged because of Anderson et al. investigated different driving tasks and driving evaluation methods.

Results obtained in a lab are not always generalizable to real life. Specifically, previous studies showed that marijuana influenced driving performance impairments are more likely to be manifest within a driver stimulator test compared to on road settings (US Department of Transportation, 1993). The present study did not find any marijuana effect on simulated driving. However, we should interpret this null finding with caution because there are many methodological (e.g., large measurement error) and analytical (e.g., small sample size; small effect size) reasons for failing to finding differences between groups. In addition, it is possible that smaller undetected effects exist or that marijuana impacts actual driving. 

-Julia Braverman

What do you think?  Please use the comment link below to provide feedback on this article.

References

Anderson, B.M., Rizzo, M., Block, R. I., Pearlson, G. D., & O’Leary, D.S. (2010) Sex differences in the effects of marijuana on simulated driving performance. Journal of Psychoactive Drugs, 40, 19 – 30.

Berghaus G, Sheer N, & Schmidt P. (1995) Effects of cannabis on psychomotor skills and driving performance–A meta-analysis of experimental studies. In CN Kloeden and AJ McLean (eds.), Proceedings of the 13th International Conference on Alcohol, Drugs and Traffic Safety. Adelaide, Australia: The University of Adelaide, NHMRC Road Accident Research Unit, 403–409.

Hoddes. E.: Zarcone. V.; Smythe. H.: Phillips. R. & Dement. W.C. (1973)

Quantificalion of sleepiness: A new approach. Psvchophysiology 10, 11-36.

US Department of Transportation, National Highway Traffic Safety Administration. (1993) It appears performance is more affected by THC in laboratory (settings) than (in) actual driving tests. Marijuana and Actual Driving Performance: Final Report.

US Department of Health and Human Services, Substance and Mental Health Services Association, Office of Applied Studies. (2007) 2006 National Survey on Drug Use and Health: National Results.

STASH, Vol. 7(9): Taking Matters into Their Own Hands: Cannabis Use Following Inpatient Treatment for PTSD among a Sample of Military Veterans

Military Veterans who have experienced trauma are at increased risk for PTSD symptoms and cannabis use among other disorders (Bremner et al., 1996). In line with Khantzian’s self-medication hypothesis (Khantzian, 1985), people might turn to cannabis as an attempt to cope with PTSD symptoms, including anxious arousal and sleep disturbance. Research suggests that individuals who have experienced trauma are a particularly vulnerable population for relapsing to cannabis use after a period of discontinuation (Bremner et al., 1996).
The study we review in this week’s STASH (Bonn-Miller et al., 2011) was the first to examine cannabis use among American Veterans following inpatient treatment for PTSD. This study prospectively examined the relationship between treatment changes in PTSD symptom severity and cannabis use four months after discharge from residential treatment for PTSD. The researchers hypothesized that lower levels of change in PTSD symptom severity (i.e., PTSD symptoms not improving) would predict more frequent cannabis use at 4-month follow-up post-treatment than their symptom improving counterparts. In addition, the researchers examined patterns of substance use specifically by looking at two additional substance use classes: alcohol and opiates. Use of these substances tends to co-occur among those with PTSD at high rates (Chilcoat & Menard, 2003).

Methods

• The sample was composed of 432 male military Veterans (56.9% white, Mage= 51.06 years)

• All Veterans had a primary diagnosis of PTSD and were inpatients at the Department of Veterans Affairs (VA) Medical Center in Palo Alto.

• The researchers assessed the sample at 3 points:

              1)    At treatment intake
              2)    At treatment discharge
              3)    At 4-month follow-up which was administered by mail

   Measures

• The researchers used the PTSD Checklist—Military Version (PCL-M; Weathers et al., 1993) to measure PTSD symptom severity at treatment intake and treatment discharge.

• The PCL-M asks questions about DSM-IV PTSD symptoms; respondents indicate the degree to which they have been bothered by each of the 17 symptoms within the past month.

• The researchers calculated PTSD symptom severity change as the PCL-M score difference between intake and discharge. Lower levels of change indicate less symptom improvement.

• The researchers used the Northeast Program Evaluation Center (NEPEC) to assess substance use at intake and follow-up. Patients reported the number of days that they used cannabis, alcohol, cocaine, opiates, and amphetamines during the 2 months before intake assessment as well as 1 month before follow-up assessment (i.e., 3 months after discharge)

Results

• Roughly half (54.3%; n=19) of participants who used cannabis pre-treatment relapsed to cannabis use after treatment; 10.1% (n= 40) of participants who did not report cannabis use pre-treatment reported cannabis use after treatment.

• Participants with lower levels of PCL-M score improvement reported significantly greater frequency of cannabis use during the 30 days before the 4-month follow-up, even after controlling for frequency of cannabis use pre-treatment (β = .11; p < .05)

• Using post-hoc analyses, the researchers explored whether changes in certain PTSD symptom clusters were predictive of cannabis use at follow-up after controlling for frequency of cannabis use pre-treatment (see Table 1):

        -    Levels of improvement in PTSD “re-experiencing” severity were not associated             with cannabis use at follow-up (β = -0.09; p= .07),
        -    Lower levels of improvement in PTSD “avoidance/numbing” severity significantly             predicted more cannabis use at follow-up (β =-0.11; p< .05)
        -    Lower levels of improvement in PTSD “hyperarousal” symptoms significantly             predicted more cannabis use at follow-up (β = -0.09; p= .05).

• Levels of improvement in PTSD symptoms were not associated with the frequency of any other non-cannabis substance use, suggesting a substance specificity effect for cannabis.

Table 1: Relationship between PCL-M Symptom Improvement during Treatment and Follow-Up Cannabis Use

PCL-M Change (Intake to Follow-Up)	β	P
Re-experiencing	- .09	.07
Avoidance/numbing	- .11	<.05
Hyperarousal	- .09	.05

Limitations

• The researchers did not collect information regarding co-occurring Axis-I or Axis-II diagnoses. As a result, they could not study the relationships between substance use and co-morbid conditions.

• Because the VA Palo Alto Medical Center was responsible for collection of data and determination of benefits, participants might have been reluctant to disclose some patterns of drug use. A third party assessment (both in terms of person and institution) might have helped to prevent this potential bias.

• This study used a limited sample, only male military Veterans. This might mean that these findings will not generalize to women or victims of other kinds of trauma.

Discussion
Veterans who experienced less improvement in their PTSD symptoms used cannabis more frequently after treatment. This was particularly true for the PTSD symptoms avoidance/numbing and hyperarousal.  These results support the idea that Veterans with PTSD might turn to cannabis to cope with hyperarousal symptoms like irritability and sleep disturbances, and numbing symptoms like emotional detachment, in line with with Khantzian’s self-medication theory (1997). The substance specificity effect suggests that cannabis might be particularly helpful in alleviating these PTSD symptoms, at least temporarily.
Future research should clarify the perceived and actual benefits of cannabis use among the Veteran population, as well as the detrimental effects of cannabis use.  Future research also could seek to identify the perceived and actual benefits of other kinds of coping mechanisms among those who have PTSD.

References
Bonn-Miller, M.O, Vujanovic, A.A., Drescher, K.D. (2011) Cannabis use among military veterans after residential treatment for posttraumatic stress disorder. Psychology of Addictive Behaviors, 25(3), 485-491.

Bremner, J. D., Southwick, S. M., Darnell, A., & Charney, D. S. (1996).Chronic PTSD in Vietnam combat Veterans: Course of illness and substance abuse. The American Journal of Psychiatry, 153, 369–375.

Chilcoat, H. D., &Mennard, C. (2003). Epidemiological investigations: Comorbidity of posttraumatic stress disorder and substance use disorder. In P. Ouimette& P. J. Brown (Eds.), Trauma and substance abuse: Causes, consequences, and treatment of comorbid disorders. Washington, DC: American Psychological Association.

Khantzian, E. J. (1985). The self-medication hypothesis of addictive disorders: Focus on heroin and cocaine dependence. American Journal of Psychiatry, 142, 1259–1264.

Weathers, F., Litz, B., Herman, D., Huska, J., & Keane, T. (October, 1993). The PTSD checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at the Annual Convention of the International Society for Traumatic Stress Studies, San Antonio, TX.

STASH, Vol. 7(8) - Patterns of cocaine use among male and female undergraduates

Recent studies reveal that cocaine use is more prevalent among young adults compared to other user population groups (Substance Abuse and Mental Health Administration (SAMHSA), 2008). Estimates of past-year cocaine use among young adults have steadily increased over the last ten years (Kasperski et al., 2011). However, male and female college students might be different in terms of cocaine use in some important ways.  While some studies cite higher prevalence rates of cocaine use in young males, others cite riskier behavior in female cocaine users, suggesting that young females might be more susceptible than males to developing serious problems with cocaine (Kasperski et al., 2011). This week’s STASH features a study that analyzed gender differences in prevalence of cocaine exposure and actual cocaine use in a longitudinal sample of 1,253 college students.

Methods

• The researchers administered annual questionnaires from the National Survey on Drug Use and Health to track a sample of college students throughout their four years at a large, mid-Atlantic university.
• The analyses examined frequency of cocaine use and DSM-IV criteria for cocaine dependence in the sample of students. Analyses covered two additional kinds of data: age at first offer of cocaine and age at first use of cocaine.
  • Age at first offer of cocaine was operationalized as the participant’s age during the first assessment in which an offer was reported. For individuals who were offered cocaine prior to Year 1, age at first offer was captured by self-report.
  • Age of first use of cocaine was computed similarly.
• 1,253 college students took part in the study (70.8% White; 48.5% Male). Follow-up rates ranged from 91.1% in Year 2 to 87.6% in Year 4. Researchers excluded 192 students because of missing data.
• Data from the first four years of college were analyzed to detect changes over time and possible gender differences.

Results
Gender differences appeared in opportunities to use and patterns of actual usage.

• As displayed in Figure 1, with the exception of the first year of college, males reported significantly higher opportunities to use than women (41% vs. 31% overall). Males reported an increase in exposure opportunity over time (Year 1 to Year 2:  p <.03; Year 2 to Year 3: p=.005), while females did not. This suggests a gender difference in opportunities to use.

 

• However, there was no significant difference between males and females in the sub-sample (n=584) of those who used cocaine, given the opportunity. Both genders showed an overall rate of 37%. While men were more likely to be presented with opportunities to use cocaine, men and women were equal in terms of use when given the opportunity to use.

• Women tended to use cocaine more frequently than men, both in terms of lifetime use as of Year 4 (27.78 days vs. 12.75 days during Year 4) and during the peak year of use (18.39 days vs. 8.83 days). For females, annual changes in frequency of cocaine use were statistically significant from Year 1 to Year 2 (p=.001) and Year 2 to Year3 (p=.009).

 

• Women were more likely than men to meet DSM-IV criteria for dependence (9.3% vs. 2.5%).

Limitations

• The authors note several limitations, including that their study cannot rule out the possibility of gender differences in reporting dependence symptoms. Males may report fewer symptoms that would meet DSM-IV criteria for dependence.
• They also note that the study’s findings have limited generalizability to other types of colleges and geographic regions.

Discussion

This study’s longitudinal design and high follow-up rate is a strong start for identifying patterns of risky behavior. The results suggest that there are possible gender differences in how college students experience opportunities for use of illicit drugs, in line with previous studies. The results help resolve apparent paradoxes in gender-based patterns of cocaine use. Men reported more opportunities to use cocaine than women. Research that uncovers the settings in which young men typically find opportunities to use cocaine might inform primary prevention efforts. However, among the smaller sample of students who actually used cocaine given the opportunity, there was no significant difference between the two genders. Women who chose to use cocaine used it more frequently and were more likely to meet DSM-IV criteria for dependence. Animal studies have suggested sex differences in endocrine-related cocaine response (Kasperski et al., 2011). Early interventions targeted for women could emphasize these differences by teaching girls about their potential heightened sensitivity to cocaine use. More research remains to be done in specific gender differences in developing substance abuse dependence.

- Kat Belkin

What do you think?  Please use the comment link below to provide feedback on this article.

References

Kasperski, S. J., Vincent, K.B., Caldeira, K.M., Garnier-Dykstra, L.M., O'Grady, K.E., Arria, A.M. (2011). "College students' use of cocaine: Results from a longitudinal study." Addictive Behaviors 36, 408-4011.

Substance Abuse and Mental Health Services Administration (2008). Results from the 2007 National Survey on Drug Use and Health: National findings (DHHS Publication No. SMA 08-4343). Rockville, MD: Office of Applied Studies.

STASH, Vol. 7(7) - Flying “High”: Drug Use Prevalence among Aviation Employees

Since the early 1980s, more employers have required their employees to complete random drug tests, for drugs such as marijuana, cocaine, and more (Li et al., 2011). In particular, the federal government mandates drug testing among employees who hold “safety-sensitive” occupations, including airline pilots. This week, STASH reviews recent research that examined the prevalence rates of positive drug screens among “safety-sensitive” aviation employees who either completed random drugs tests or recently were involved in an aviation accident (Li et al., 2011). The authors examined whether the use of illicit drugs is associated with an increased risk of airline accident involvement.

Method

• The researchers analyzed aviation employer (i.e., all major airlines, commuter air carriers/air taxi, and non-FAA air traffic controllers) records that detailed the results of drug testing for marijuana, cocaine, amphetamines, opiates, and phencyclidine among employees who held “safety-sensitive” positions (e.g., flight crewmembers, flight attendants, flight instructors, aircraft dispatchers). The records covered the years from 1995 to 2005.
• Analyses utilized two sources of data: standard random drug testing and post-accident drug testing.
  • Random testing accounts for about 96% of all drug related records, and involve urine testing a random selection of at least 25% of “safety-sensitive” employees.
  • Post-accident urine testing occurs within 32 hours of an accident for employees whose performance might have contributed to an aviation accident (i.e., an event resulting in fatality or serious injury, or substantial damage to an aircraft).

Results

• From 1995-2005, the prevalence of positive screens among random drug tests was 0.64% (i.e., 7,211 of 1,129,922). The prevalence of positive screens among post-accident tests was 1.82% (i.e., 91 of 4,977).
• The odds ratio of accident involvement for employees who tested positive for drugs was 2.90 (95% CI = 2.35-3.57). In other words, employees who tested positive for drugs were almost three times as likely to be involved in an accident as employees who tested negative for drugs.
• Accounting for the prevalence of drug violations observed from random tests, this odds ratio suggests that 1.2% of all aviation accidents were attributable to drug violations during this period.^[i]
• Figure 1 (adapted from Li et al., 2011) shows the distribution of drug violations among aviation employees who hold safety-sensitive positions. Overall, marijuana accounted for 67.3% of violations, cocaine accounted for 23.9%, amphetamines accounted for 6.1%, opiates accounted for 2.1%, and phencyclidine accounted for 0.6%.

 

Limitations

• The authors noted several limitations, including that their analyses were restricted to a segment of all aviation events; they could not include general (e.g., non-commercial) aviation accidents, which account for 90% of all aviation accidents.
• They also note that they could not assess potential confounding individual-level variables, due to nature of data (i.e., aggregate reports to FAA).

Discussion

Drug testing for employees who have “safety-sensitive” occupations is popular among the public and policy-makers. Research related to this practice is limited, however. These results suggest that drug violations generally are rare among flight crews; about one in every 2,000 random drug tests produces a positive result. Although the use of illicit drugs is rare, it is consequential; it nearly triples the risk of accident involvement. More research is needed to determine whether this observation extends to general aviation accidents.

-Debi LaPlante

What do you think?  Please use the comment link below to provide feedback on this article.

Reference

Li, G., Baker, S. P., Zhao, Q., Brady, J. E., Lang, B. H., Rebok, G. W., & DiMaggio, C. (2011). Drug violations and aviation accidents: Findings from the US mandatory Drug Testing Programs. Addiction, 106, 1287-1292.

--

^[i]Attributable Risk = [p x (OR-1)]/[1+ p x (OR-1)], where p=the prevalence of drug violations detected in random testing.

STASH, Vol. 7(6) – Primary Care Treatment for People with Opiate Addiction: Stronger Support for a Team Effort

Fewer than 25% of individuals who are opioid dependent receive addiction treatment (SAMHSA, 2007). Within general medical practices, buprenorphine is a promising tool for the treatment of opioid dependence; this strategy is commonly referred to as office-based opioid treatment (OBOT). However, many physicians cite insufficient nursing support as a barrier to their use of OBOT (Walley et al., 2008). Boston Medical Center has instituted a collaborative care OBOT program that emphasizes collaboration among nurses, physicians, and pharmacists. In this week’s STASH, we review a new study that describes the effectiveness of this program and explores potential patient-level moderators of treatment success (Alford et al., 2011).

Methods

• Four hundred eight patients with opioid addiction took part in the program (M age = 39 years; SD = 11; 66.5% White, 16.5% Black, and 15.4% Hispanic; 66.0% Male). Researchers excluded 26 patients who left treatment because of preexisting legal or medical conditions or a need to re-locate from the analyses. 
• The collaborative care OBOT program included a full-time nurse program director, nurse care managers (NCMs), a program coordinator, and generalist physicians with part-time clinical practices. NCMs’ clinical responsibilities included assessment, patient education, development treatment plans, overseeing medication management, monitoring treatment adherence, and communicating with other providers. Physicians examined patients, prescribed buprenorphine, and provided follow-up as needed. 
• The treatment model included 3 stages:
  • Assessment:  The nurse program director screened patients and triaged them to intake appointments. At intake, NCMs and physicians jointly administered a thorough medical evaluation and developed a treatment plan. 
  • Induction and Stabilization: Physicians prescribed buprenorphine and NCMs monitored induction, which occurred on-site on Day 1, and dose stabilization, which occurred at home on Days 2-7. NCMs assessed for withdrawal and adverse effects.
  • Maintenance: Staff provided ongoing monitoring at follow-up appointments at least weekly for the first 4 to 6 weeks followed by visits every 2 to 4 weeks if adherent, gradually tapering out to once a month and then once every 3 months. NCMs conducted scheduled and random callback urine drug tests at least once every 3 months, and encouraged patients to attend self-help groups and mandatory weekly addiction counseling.
• The researchers defined treatment adherence as 4 consecutive urine drug tests with negative results for illicit drugs; study participants also needed to evidence positive results for buprenorphine, and attend at least 3 of 4 counseling sessions per month.
• To fulfill the criteria for successful treatment, patients had to meet three conditions:retention at 12 months with a consistent pattern of treatment adherence
  • buprenorphine dosage taper after 6 months of treatment adherence 
  • absence of illicit opiate use for at least 6 months 

Results

• At 1 year, 196 of 382 patients (51.3%) underwent successful treatment, 162 patients (42.4%) showed an unsuccessful outcome, and 24 (6.3%) transferred to methadone maintenance treatment for more structured care or for full opioid agonist therapy.
• Urinalysis indicated a 4.7% opioid positive test rate for patients who were still in the OBOT program at 12 months.
• As Figure 1 illustrates, moderators of treatment success include age, race, employment, and illicit buprenorphine use. Patients who were older, employed, and self-maintained with illicit buprenorphine had significantly higher odds of treatment success (all p’s < 0.05). Those of African American or Hispanic/Latino race had significantly lower odds of treatment success (p’s < 0.05).

Figure 1
Odds Ratios for Factors that Influence Treatment Success at 12 Months for Patients Entering OBOT in Primary Care

 

 
   
Limitations

• This study did not have a control group. This makes it impossible to directly compare the effectiveness of the program to other traditional methods of care.
• Follow-up information was not available after patient departure for the program. A broader scope of outcomes is needed to assess the success of the OBOT program.

Discussion

This study shows some support for OBOT within a collaborative care model. The results compare favorably to those of similar programs that treat opioid addiction. Gerra et al. (2006) found a 43.8% retention rate at 12 months for heroin users accepting buprenorphine treatment in an outpatient setting, and Pinto et al. (2010) report a 50% retention rate at 6 months for a buprenorphine treatment program with less clinical support. Methadone treatment consistently has evidenced a higher retention rate than buprenorphine treatment, but has shown no difference in suppression of illicit opioid use for those who remain in treatment (Mattick et al., 2008; Fischer et al., 1999). The high level of clinical support and role sharing in the collaborative care model might account for some treatment success, but a 51.3% retention rate is difficult to applaud – treatment can do better. 

The finding that self-maintaining with illicit buprenorphine at admission predicted later success is interesting.  Schuman-Olivier et al. (2010) found that individuals who are dependent on opioids often illicitly use buprenorphine to manage pain and withdrawal rather than to get high. Similarly, Alford et al. speculate that those who used buprenorphine illicitly at pre-admission are highly motivated to find a safe and affordable source of buprenorphine for treatment of withdrawal symptoms. Further study is needed to better understand the ways that race, age, and employment status predict success in OBOT.

-Kat Belkin

What do you think?  Please use the comment link below to provide feedback on this article.

References

Alford, D. P., LaBelle, C.T., Kretsch, N., Bergeron, A., Winter, M., Botticelli, M., Samet, J.H.  (2011). "Collaborative Care of Opioid-Addicted Patients in Primary Care Using Buprenorphine: Five-Year Experience." Archives of Internal Medicine 171(5): 425-431.

Fischer, G., Gombas, W., Eder, H., Jagsch, R., Peternell, A., Stühlinger, G., Pezawas, L., Aschauer, H.N., Kasper, S. (1999). "Buprenorphine versus methadone maintenance for the treatment of opioid dependence." Addiction 94(9): 1337-1347.
   
Gerra, G., Leonardi, C., D’Amore, A., Strepparola, G., Fagetti, R., Assi, C., Zaimovic, A., Lucchini, A. (2006). "Buprenorphine treatment outcome in dually diagnosed heroin dependent patients: A retrospective study." Progress in Neuropsychopharmacology and Biological Psychiatry 30(2): 265-272.

Mattick, R. P., Kimber, J., Breen, C., Davoli, M. (2008). “Buprenorphine maintenance versus placebo or methadone maintenance for opioid dependence.” Cochrane Database of Systematic Reviews (2).
   
Pinto, H., Maskrey, V., Swift, L., Rumball, D., Wagle, A., Holland, R. (2010). "The SUMMIT trial: A field comparison of buprenorphine versus methadone maintenance treatment." Journal of Substance Abuse Treatment 39: 340-352.
   
SAMHSA (2007). Results from the 2006 National Survey on Drug Use and Health: National Findings. Substance Abuse and Mental Health Services Administration. Rockville, MD, Office of Applied Studies.

Schuman-Olivier, Z., Albanese, M., Nelson, S. E., Roland, L., Puopolo, F., Klinker, L., & Shaffer, H. J. (2010). Self-treatment: Illicit buprenorphine use by opioid dependent treatment seekers. Journal of Substance Abuse Treatment, 39, 41-50.
   
Walley A.Y., A. J. K., Cheng D.M. (2008). "Office-based management of opioid dependence
with buprenorphine: clinical practices and barriers." Journal of General Internal Medicine 23(9): 1393-1398.