1. Findings from the Pathways to Recovery and Recovery Management Checkups (RMC) Experiments
Michael L. Dennis, Ph.D.
& Christy K Scott, Ph.D.
Chestnut Health Systems
Bloomington and Chicago, IL
Presentation at the State Systems Development Program VIII Conference: Partnering to Support Recovery-Oriented Systems of Care, Washington, DC, August 20-22, This presentation was supported by funds from and data from NIDA grants no. R01 DA15523, R37-DA11323 and CSAT contract no It is available electronically at . The opinions are those of the authors do not reflect official positions of the government. Please address comments or questions to the author at or .
Thank you…
I would like to start by acknowledging that this presentation builds on a series of studies and papers the Christy Scott, Mark Godley, Susan Godley and I have done… as well as data from the several large community and clinical epidemiological data bases made available to the public by the U.S. Government.
Also note that i have posted an electronic copy of this presentation at

2. Problem and Purpose
Over the past several decades there has been a growing recognition that a subset of substance users suffers from a chronic condition that requires multiple episodes of care over several years.
This presentation will present
Epidemiological data to illustrate the chronic nature of substance disorders and how it relates to a broader understanding of recovery
The results of two experiments designed to improve the ways in which recovery is managed across time and multiple episodes of care.

3. Nine Year Pathways to Recovery Study (Scott & Dennis)
Recruitment: to 1997
Sample: 1,326 participants from sequential admissions to
a stratified sample of 22 treatment units in 12
facilities, administered by 10 agencies on
Chicago's west side.
Substance: Cocaine (33%), heroin (31%), alcohol (27%), marijuana (7%).
Levels of Care: Adult OP, IOP, MTP, HH, STR, LTR
Instrument: Augmented version of the Addiction Severity
Index (A-ASI)
Follow-up: Of those alive and due, follow-up interviews were
completed with 94 to 98% in annual interviews out
to 9 years; over 80% completed within +/- 1 week of target date.
Funding: CSAT grant # T100664, contract #
NIDA grant 1R01 DA15523
Purpose: Substance use disorders are chronic relapsing conditions

4. Substance Use Careers Last for Decades
100%
90%
80%
Percent in Recovery
70%
Median duration of 27 years
(IQR: 18 to 30+)
60%
Years from first use to 1+ years abstinence
50%
40%
30%
Using data from a clinical sample of 1,271 people in Dr. Scott’s 9 year pathways to recovery study, this figure presents a survival analysis estimating the time from first use to 1 or more years of sobriety. The line shows the percent with 1 or more years of abstinence.
1- The median duration was 27 years – with an inter-quartile range (i.e., middle 50%) of 18 to over 30 years.
For Reference only…
We stopped at that point due to censoring in the data due limiting us to 100 or less people and due to 2.6% who had died. We are currently working to extend this another 5 to 10 years and also look at the impact of recovery on the likelihood of death.
20%
10% 0%
Source: Dennis , Scott, Funk & Foss ( 2005) (n=1,271) 5 10 15 20 25 30

5. Source: Dennis et al 2005 (n=1,271)
Substance Use Careers are Longer,
the Younger the Age of First Use
100%
90%
21+
80%
Percent in Recovery
15-20*
Age of 1st Use Groups
70%
60%
Years from first use to 1+ years abstinence
under 15*
50%
40%
30%
Substance use careers are longer the earlier the age of first use ---
20%
* p<.05 (different from 21+)
10% 0% 5 10 15 20 25 30
Source: Dennis et al 2005 (n=1,271)

6. Source: Dennis et al 2005 (n=1,271)
Substance Use Careers are Shorter
the Sooner People get to Treatment
100%
0-9*
90%
80%
10-19*
Percent in Recovery
Years to 1st Tx Groups
70%
60%
Years from first use to 1+ years abstinence
50%
40%
20+
30%
1- However, entering treatment in the first decade of use was associated with reducing the lifetime substance use career by over half.
It does not take long to realize that the potential saving in societal costs justifies more research on to improve participation in and the effectiveness of adolescent substance abuse treatment.
20%
10%
* p<.05 (different from 20+) 0% 5 10 15 20 25 30
Source: Dennis et al 2005 (n=1,271)

7. Multiple Episodes of Treatment
It Takes Decades and
Multiple Episodes of Treatment
100%
90%
80%
Percent in Recovery
70%
Median duration of 9 years
(IQR: 3 to 23) and 3 to 4 episodes of care
Years from first Tx to 1+ years abstinence
60%
50%
40%
30%
Note that one episode of treatment is rarely sufficient once someone has a chronic substance use disorder with multiple co-occurring problems. This figure shows the time from 1st treatment to 1+ years of abstinence, the median was 9 years with an inter-quartile range of 3 to 23 years. Moreover the average person did not have 1+ years of abstinence until then had been in treatment 3 to 4 times.
The traditional acute care model focusing on 1-12 sessions or even 90 days of treatment and 6 month outcomes are simply not sufficient for treating or studying the treatment of this kind of chronic substance use disorders.
20%
10% 0% 5 10 15 20 25
Source: Dennis et al 2005 (n=1,271)

8. Duration of Treatment Career By Level of Internal Distress at Index TX
100%
Internal Distress
90%
High*
Moderate
80%
Percent in Recovery
Low
70%
Years from first Tx to 1+ years abstinence
60%
50%
40%
Note that it takes several years to see the difference
30%
Note that one episode of treatment is rarely sufficient once someone has a chronic substance use disorder with multiple co-occurring problems. This figure shows the time from 1st treatment to 1+ years of abstinence, the median was 9 years with an inter-quartile range of 3 to 23 years. Moreover the average person did not have 1+ years of abstinence until then had been in treatment 3 to 4 times.
The traditional acute care model focusing on 1-12 sessions or even 90 days of treatment and 6 month outcomes are simply not sufficient for treating or studying the treatment of this kind of chronic substance use disorders.
20%
10%
p<.05
(High v low) 0% 5 10 15 20 25
Source: Dennis et al 2005 (n=1,271)

9. Over half change status annually
The Cyclical Course of Relapse, Incarceration,
Treatment and Recovery
Over half change status annually
P not the same in both directions
Incarcerated
(37% stable) 6%
28%
13% 7%
29%
30% 8%
25%
Treatment is the most likely path to recovery
31% 4%
44%
In the
Community
In Recovery
Using
(58% stable)
(53% stable)
To shift to a chronic condition management paradigm it is necessary to start by recognizing that most people are typically cycling through multiple periods of being in the community using, incarceration, treatment and being in the community in recovery for a month or more. The circles in this figure show the average percent staying in each condition between two annual interviews for this same large sample % stayed in the same status for 12 months and 55% changed.
1,2,3, these arrows show the percent moving along each pathway from one state to another.
5- Notice that the percentages are not the same in both directions.
6- Also notice that the probability of moving from the three states on the left to recovery on the right are not the same --- with treatment being the most likely pathway to recovery over a 12 month period.
In Treatment
(21% stable)
Source: Scott, Dennis, & Foss (2005)

10. Predictors of Change Also Vary by Direction
Probability of Transitioning from Using to Abstinence
mental distress (0.88) + older at first use (1.12)
ASI legal composite (0.84) + homelessness (1.27)
+ # of sober friend (1.23)
+ per 8 weeks in treatment (1.14)
In the
28%
Community
In Recovery
Using
(58% stable)
29%
(53% stable)
The predictors of change also vary by the direction of movement.
1- The probability of transitioning from using to abstinence was inversely related to mental distress and ASI legal composite score and directly related to being older at first use, being homeless, having more sober friends, and the amount of treatment received during the year.
2- The probability of relapsing from abstinence was directly related to the lifetime number of treatment episodes and homelessness and inversely related to being female, the number of arrests, ASI legal composite score, # of sober friends, and participation in self help sessions.
Notice that some things like legal involvement work to hold a person in either condition while others like homeless work to change the condition.
In contrast, the amount of treatment is one of the better predictors of who will transition to recovery while self help participation is the better predictor of who stays there.
The question arises whether increasing early treatment participation would hasten the transition to recovery.
Probability of Relapsing from Abstinence
+ times in treatment (1.21) - Female (0.58)
+ homelessness (1.64) - ASI legal composite (0.84)
+ number of arrests (1.12) - # of sober friend (0.82)
- per 77 self help sessions (0.55)
Source: Scott, Dennis, & Foss (2005)

11. Percent Sustaining Abstinence Through Year 8 by Duration of Abstinence at Year 7
100%
Even after 3 to 7 years of abstinence about 14% relapse .
86%
86%
90%
80%
It takes a year of abstinence before less than half relapse
70%
66%
60%
% Sustaining Abstinent through Year 8
50%
36%
40%
30%
20%
A key question that is often asked is how long someone is at risk of relapse. This figure shows the odds of sustaining relapsed from year 7 to year 8 based on the duration of abstinence.
10% 0%
1 to 12 months
1 to 3 years
3 to 5 years
5+ years
(n=157; OR=1.0)
(n=138; OR=3.4)
(n=59; OR=11.2)
(n=96; OR=11.2)
Duration of Abstinence at Year 7
Source: Dennis, Foss & Scott (2007)

12. Other Aspects of Recovery by Duration of Abstinence of 8 Years
1-12 Months:
Immediate increase in clean and sober friend
1-3 Years:
Decrease in Illegal Activity; Increase in Psych Problems
3-5 Years:
Improved Vocational and Financial Status
5-8 Years:
Improved Psychological Status
100%
% of Clean and
Sober Friens
90%
80%
70%
60%
% Days Worked
For Pay (of 22)
50%
% Above
Poverty Line
40%
30%
This figure classifies people at year 8 of the study based on how long they have been abstinent (if at all) to examine how the duration of abstinence is related to other aspects of recovery
20%
% Days of Illegal
Activity (of 30 days)
% Days of Psych
Prob (of 30 days)
10% 0%
Using
1 to 12 ms
1 to 3 yrs
3 to 5 yrs
5 to 8 yrs
(N=661)
(N=232)
(N=127)
(N=65)
(N=77)
Source: Dennis, Foss & Scott (2007)

13. Death Rate by Years of Abstinence
The Risk of Death goes down with
years of sustained abstinence
4.5%
11.9%
7.1%
3.8% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
10%
11%
12%
13%
14%
15%
Household
(OR=1.00)
Less than 1
(OR=2.87) 1 -
3 Years
(OR=1.61) 4
8 Years
(OR=0.84)
Users/ Early Abstainers 2.87 times more likely to die in the next year
It takes 4 or more years of abstinence for risk to get down to community levels
A key question that is often asked is how long someone is at risk of relapse. This figure shows the odds of sustaining relapsed from year 7 to year 8 based on the duration of abstinence.
Source: Scott, Dennis, & Funk (2008)

14. Other Predictors of Death
+ Pre-existing chronic illness (RR=1.87)
+ Age (RR=1.45)
+ Living on less than 50% of poverty line (RR=1.71)
+ Hospitalization during the 6 months prior to intake (RR=1.26 per week)
+ Illegal activity for money during the 6 months prior to intake (RR=1.14 per 30 days)
- Self Help Sessions attended first 6 months (RR=0.88 per 30 days)
- Years of abstinence (RR=0.83 per year)
Of all the variables in Table 1, the only two significant predictors of the duration of abstinence at the last wave are the days of self-help at 6 months (b=.007, p=.000) and being in treatment at month 6 (b=.429, p=.000).
The days of self help attendance in the first six months were significantly lower for people with a history of chronic illness, living below 50% of the poverty line, and with more years of regular heroin use. They were significantly higher for people with more years of regular use, those having unprotected sex with multiple partners, and those still in treatment at month 6.
Weeks of treatment predicted weeks of self help and reduced risk of death univariately
Being in treatment at month six predicted self help, years of abstinence, and reduced risk of death univariately

15. The Early Re-Intervention (ERI) Experiments (Dennis & Scott)
Recruitment
Recruited 446 from Community Based Treatment in Chicago in 2004 (93% of eligible recruited)
Design
Random assignment to Recovery Management Checkups (RMC) or control
Follow-Up
Quarterly for 4 years (95 to 97% per wave)
Data Sources
GAIN, CEST, CAI, Neo, CRI, Urine, Staff logs
Publication
Dennis, Scott & Funk (2003); Scott, Dennis, & Foss (2005); Dennis & Scott (2007); Scott & Dennis, (under review); Riley, Scott, & Dennis (2008)
In the final part of this presentation I want to tell you about two experiments in which we are trying to alter SUD trajectories through early intervention.
In the first ERI experiment we recruited 448 adults from the largest community based treatment provider in Chicago in We only excluded people who were from outside of Chicago, headed to prison for the study period, did not speak English or Spanish, or who were too cognitively impaired to give informed consent. Of those who were eligible, 84% agreed to participant and were randomly assigned to either a Recovery Management Checkups or a control group. Both groups were re-interviewed quarterly for 2 years – with 95 to 97% follow-up per wave. The Data sources included the Global Appraisal of Individual Needs (GAIN), motivational scales from TCU Client Evaluation of Self and Treatment (CEST), Urine & Salvia tests and staff logs.
In the second ERI experiment we recruited 446 adults from the same community based program 4 years later, used the same exclusion rules, and achieved 93% participation. Again we randomly assigned them to RMC or a control group, but this time for four years. Here I will present findings from the first 12 months – with 95 to 97% follow-up per wave. In addition to the earlier measures, this time we also collected additional scales on treatment process from NDRI’s Client Assessment Interview, personality measures from the NEO and the Moos Coping Response Inventory.
Funding Source NIDA grant R37-DA11323

16. Recovery Management Checkups (RMC)
Quarterly Screening to determining “Eligibility” and “Need”
Linkage meeting/motivational interviewing to:
provide personalized feedback to participants about their substance use and related problems,
help the participant recognize the problem and consider returning to treatment,
address existing barriers to treatment, and
schedule an assessment.
Linkage assistance
reminder calls and rescheduling
Transportation and being escorted as needed
Treatment Engagement Specialist

17. ERI-2 Time to Treatment Re-Entry
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
The size of the effect is growing every quarter
= -384 days
Percent Readmitted 1+ Times
55% ERI-2 RMC*
(n=221)
37% ERI-2 OM
(n=224)
*Cohen's d=+0.41
Wilcoxon-Gehen 0% 90
180
270
360
450
540
630
Statistic (df=1)
=16.56, p <.0001
Days to Re-Admission (from 3 month interview)
Source: Scott & Dennis (under review)

18. ERI-2: Impact on Outcomes at 45 Months
74%
More days of abstinent
100%
71%
61%
RMC Increased Treatment Participation OM
RMC
90%
47%
Less likely to be in
Need at 45m
80%
38%
Fewer Seq.
Quarters
in Need
70%
67%
56%
60%
55%
50%
Percentage
50%
41%
40%
30%
20%
What this figures shows is the total number of days --- green RMC, red control….. Keep in mind that the control group did not get any help. The right….literature
10% 0%
Re-entered
of 180 Days
of 1260 Days
of 14 Subsequent
Still in need of
Tx at Mon 45
Treatment
of Treatment
Abstinent
Quarters in Need
(d=0.22)*
(d= 0.26) *
(d= 0.26)*
(d= -0.32)*
(d= -0.22) *
Source: Scott & Dennis (under review)
* p<.05

19. ERI: Impact Treatment Re-entry by Comorbidity and Condition
RMC’s Impact on Treatment Participation was robust across
levels of Comorbidity
Returning to treatment varied by Comorbidity
100% OM
RMC
90%
80%
70%
62%
63%
49%
60%
53%
Percentage
47%
50%
40%
33%
30%
20%
What this figures shows is the total number of days --- green RMC, red control….. Keep in mind that the control group did not get any help. The right….literature
10% 0%
Substance Use
Disorder (SUD)
(d=0.23)*
Substance Use +
Internalizing
Disorders
(d=0.38)*
Substance Use +
Internalizing +
Externalizing
Disorders
(d=0.31)*
Source: Rush, Dennis, Scott, Castel, & Funk (2008)
* p<.05

20. 34% Changed Status in an Average Quarter
ERI 2: Average Quarterly Transitions over 3 years
34% Changed Status in an Average Quarter
Incarcerated
(56% stable) 4%
10% 6%
24% 3%
23% 8%
13%
Again the Probability of Entering Recovery is Higher from Treatment
35% 7%
25%
In the
Community
In Recovery
Using
(58% stable)
(75% stable)
To shift to a chronic condition management paradigm it is necessary to start by recognizing that most people are typically cycling through multiple periods of being in the community using, incarceration, treatment and being in the community in recovery for a month or more. The circles in this figure show the average percent staying in each condition between two annual interviews for this same large sample % stayed in the same status for 12 months and 55% changed.
1,2,3, these arrows show the percent moving along each pathway from one state to another.
5- Notice that the percentages are not the same in both directions.
6- Also notice that the probability of moving from the three states on the left to recovery on the right are not the same --- with treatment being the most likely pathway to recovery over a 12 month period.
In Treatment
(32% stable)
Source: Riley, Scott & Dennis, 2008

21. ERI 2: Average Quarterly Transitions over 3 years
Transition Tx to Recovery (vs. relapse)
Freq. of Use (0.01) Self Help Act. (1.22)
Tx Resistance (0.85) + Wks Self Help (1.20)
In the
Community
In Recovery
Using
(58% stable)
(75% stable)
To shift to a chronic condition management paradigm it is necessary to start by recognizing that most people are typically cycling through multiple periods of being in the community using, incarceration, treatment and being in the community in recovery for a month or more. The circles in this figure show the average percent staying in each condition between two annual interviews for this same large sample % stayed in the same status for 12 months and 55% changed.
1,2,3, these arrows show the percent moving along each pathway from one state to another.
5- Notice that the percentages are not the same in both directions.
6- Also notice that the probability of moving from the three states on the left to recovery on the right are not the same --- with treatment being the most likely pathway to recovery over a 12 month period.
35%
25%
10%
Transition to Tx (vs use)
- Tx Resistance (0.94)
+ Freq. of Use (3.74)
+ Desire for Help (1.15)
+ Wks of Self Help (1.14)
+ Self Help Act. (1.31)
+ Prior Wks of Tx (1.10)
+ RMC (2.00)
In Treatment
(32% stable)
Source: Riley, Scott & Dennis, 2008

22. These studies provide converging evidence demonstrating that
substance use disorders are often chronic in the sense that they last for years and the risk of relapse is high
the majority of people accessing publicly funded substance abuse treatment have been in treatment before, are likely to return, and may need several additional episodes of care before they reach a point of stable recovery.
Yet over half do achieve recovery
Recovery is broader than just abstinence
The odds of getting to and staying in recovery can be improved with proactive checkups and management.

23. Sources and Related Work
Dennis, M.L., Foss, M.A., & Scott, C.K (2007). An eight-year perspective on the relationship between the duration of abstinence and other aspects of recovery. Evaluation Review, 31(6),
Dennis, M. L., Scott, C. K. (2007). Managing Addiction as a Chronic Condition. Addiction Science & Clinical Practice , 4(1),
Dennis, M. L., Scott, C. K., Funk, R., & Foss, M. A. (2005). The duration and correlates of addiction and treatment careers. Journal of Substance Abuse Treatment, 28, S51-S62.
Dennis, M. L., Scott, C. K., & Funk, R. (2003). An experimental evaluation of recovery management checkups (RMC) for people with chronic substance use disorders. Evaluation and Program Planning, 26(3),
Epstein, J. F. (2002). Substance dependence, abuse and treatment: Findings from the 2000 National Household Survey on Drug Abuse (NHSDA Series A-16, DHHS Publication No. SMA ). Rockville, MD: Substance Abuse and Mental Health Services Administration, Office of Applied Studies. Retrieved from
Office Applied Studies (2002). Analysis of the 2002 National Survey on Drug Use and Health (NSDUH) on line at .
Office Applied Studies (2002). Analysis of the 2002 Treatment Episode Data Set (TEDS) on line data at
Riley, B.B.,, Scott, C.K, & Dennis, M.L. (2008). The effect of recovery management checkups on transitions from substance use to substance abuse treatment and from treatment to recovery. Poster presented at the UCLA Center for Advancing Longitudinal Drug Abuse Research Annual Conference, August 13-15, 2008, Los Angless, CA. .
Rush, B., Dennis, M.L., Scott, C.K, Castel, S., & Funk, R.R. (2008). The Interaction of Co-Occurring Mental Disorders and Recovery Management Checkusp on Treatment Participation and Recovery.
Scott, C. K., & Dennis, M. L. (under review). Results from Two Randomized Clinical Trials evaluating the impact of Quarterly Recovery Management Checkups with Adult Chronic Substance Users. Addiction.
Scott, C. K., Dennis, M. L., & Foss, M. A. (2005). Utilizing recovery management checkups to shorten the cycle of relapse, treatment re-entry, and recovery. Drug and Alcohol Dependence, 78,
Scott, C. K., Dennis, M. L., & Funk, R.R. (2008). Predicting the relative risk of death over 9 years based on treatment completion and duration of abstinence . Poster 119 at the College of Problems on Drug Dependence (CPDD) Annual Meeting, San Juan, PR, June 16, Available at .
Scott, C. K., Foss, M. A., & Dennis, M. L. (2005). Pathways in the relapse, treatment, and recovery cycle over three years. Journal of Substance Abuse Treatment, 28, S61-S70.