Computer-Based Recovery Support for Patients Receiving Residential Treatment for Alcohol/Drug Dependence: Relationship Between Program Use and Outcomes

Abstract

Objective: This analysis of administrative data examined whether use of a Web-based recovery support program was related to self-reported post-treatment alcohol use among patients attending residential treatment for a substance use disorder. Materials and Methods: The program delivered individually tailored clinical content in a multimedia format over the initial 18 months after discharge from treatment. Post-treatment logins to the program and access of clinical content were measured, as was post-treatment alcohol use. Results: Use of the program was frequent in the first 30 days following treatment but steadily decreased over time. Regression analyses revealed a significant relationship between the number of program logins and self-reported alcohol use in the first 6 months following treatment when controlling for other covariates related to alcohol use. Conclusion: These results replicate a previous study of the My Ongoing Recovery Experience (MORE ^® ) program (Hazelden, Center City, MN) and suggest that computerized support programs hold therapeutic potential for patients with alcohol/drug dependence.

Introduction

The use of Web-based disease management programs to manage chronic health conditions has increased over the last several years, and these programs show promise in managing diseases such as cancer, diabetes, obesity, and hypertension.^1
–3 Substance use disorders are chronic disorders characterized by high rates of relapse,^4
–6 and studies indicate that Web-based programs are also effective in preventing relapse in patients with these disorders,^7,8 particularly when used to supplement therapist-delivered interventions.^9,10 However, patient engagement with these programs is often low.^7,8,11

Many computerized interventions for alcohol/drug use have focused on either formal treatment (such as computer-based cognitive behavioral therapy) or preventative methods such as screening and brief intervention.^8,12 The efficacy of Web-based recovery support programs in preventing relapse among individuals recently treated for substance dependence has yet to be extensively examined. Several years ago, Hazelden (Center City, MN), a private nonprofit addiction treatment center, launched an innovative, novel computerized recovery support program called My Ongoing Recovery Experience (MORE^®). MORE was developed by a team of licensed alcohol and drug counselors and mental health professionals for use with adult patients completing residential treatment for alcohol or drug dependence at Hazelden sites. The program incorporates concepts from a variety of evidence-based treatment approaches, such as Twelve Step facilitation, motivational interviewing, and cognitive behavioral therapy. Designed specifically as a post-treatment recovery support/continuing care program, the program provides patients with a wide variety of recovery-related information and services for a period of 18 months following discharge from treatment. MORE is an iteratively tailored, Web-based, mixed-media program that provides interactive recovery-related activities, videos, an extensive electronic library of clinical content, opportunities for contact and fellowship with other recovering individuals, and additional sources of support. The videos cover a wide range of content, including patient testimonials and information on topics such as how to manage a co-occurring disorder, recognizing factors that may put one at risk for relapse, how to protect oneself from developing an addiction to prescription medications, and how to effectively manage stress and adverse life events. Other computerized resources include a patient journal, workbook activities for practicing and applying recently learned information, and an electronic library of articles pertaining to specific issues such as relapse prevention, dealing with cravings, maintaining emotional health, and forming healthy relationships. Participants also have access to a large network of Hazelden alumni and a number of resources for staying in contact with other individuals in recovery, such as recovery blogs, forums, podcasts, and online Twelve Step meetings.

The program also includes telephone and e-mail contact between patients and recovery coaches, who are licensed alcohol and drug counselors employed by Hazelden. The majority of clinical content is delivered over the computer through seven distinct program modules, each of which begins with an electronically administered assessment where patients provide detailed information about substance use and other aspects of life functioning. Based on the responses, the module then delivers content individually tailored to the needs of the patient. The modules are administered in a sequential manner over an 18-month period, and modules need to be completed in order. In other words, content from later modules is only available if the patient has accessed prior modules. In addition, the availability of the modules is timed across the 18-month period. The assessment for Module 1 is completed while patients are still in residential treatment just prior to discharge. The remaining modules (and assessments) are available at distinct time periods after treatment discharge. Each module focuses on specific issues that are likely to be experienced at that particular time in recovery. Table 1 provides a summary of each module and its key topics and activities.

Table 1.

My Ongoing Recovery Experience Module Content

MODULE	MODULE AREA OF FOCUS	SAMPLE OF KEY ACTIVITIES AND TOPICS
Module 1	Mind, Body, and Spirit	Identifying people, places, and things that are triggers to relapse and developing coping strategies to deal with these phenomena. Creating a safe home environment free of potential relapse triggers and cues. Engaging in AA or NA, finding groups, obtaining a home group, finding a sponsor, and using the Twelve Steps
Module 2	Getting Connected	Developing a supportive network of healthy people, finding friends in recovery, reconnecting with family, and asking for help
Module 3	Balancing Work and Play	Addressing workplace issues and stressors and establishing ways to have sober fun. Achieving lifestyle balance
Module 4	Healthy and Emotional Living	Coping with normal emotions in recovery including anger, anxiety, sadness, and loneliness. Addressing unhealthy thinking processes or compulsive behaviors that may emerge
Module 5	Healthy Connections	Developing healthy relationships, including friendships and intimate relationships, communicating effectively with others, establishing trust
Module 6	Sober Identity	Establishing an identity as a sober, recovering individual, identifying and abandoning old thought patterns and beliefs, and exploring ways to continuously improve emotional well-being and physical health
Module 7	Life Goals	Re-examining career, education, and life goals as they relate to long-term recovery

AA, Alcoholics Anonymous; NA, Narcotic Anonymous.

In an earlier study of the MORE program, Klein et al.¹³ found that patients who accessed a large number of MORE modules reported less frequent post-treatment alcohol use than patients who accessed few or no modules. The number of modules accessed was also a significant predictor of post-treatment alcohol use in regression models. It is important to note that at the time of the study of Klein et al.,¹³ the only available measure of MORE program usage was the number of program modules accessed at least once. Since that time additional measures of usage have become available. The primary goal of the present study was to replicate the results of Klein et al.¹³ with a new population of residential patients using two more quantitative measures of MORE program usage: the number of logins to the program and the total number of module pages accessed. This latter measure is a more detailed and complete measure of module access than the previous measure because it measures exactly how much content within each module was accessed. The present study examined whether the number of program logins and the total number of module pages accessed were associated with self-reported post-treatment alcohol use.

Materials and Methods

Participants

All adult patients attending residential treatment at Hazelden are given the MORE program, and clinical staff strongly encourage patients to enroll in the program just prior to discharge. In addition to enrolling, patients are encouraged to access Module 1 and complete the assessment within the module. The sample utilized in this study was comprised of residential patients who enrolled in MORE and completed the Module 1 assessment shortly before discharge; patients were discharged in the period January 1–December 31, 2009 (n=1,682). Only those participants who met dependence criteria for at least one substance were included in the sample; those classified as having abuse but not dependence were excluded. No other exclusion criteria were used. Table 2 outlines the demographic and baseline clinical characteristics of the sample. Patients stayed in the program an average of 27.16 days (standard deviation [SD]=3.50), and 97% (n=1,634) successfully completed treatment. Treatment completion was defined as attending the program for the period of time recommended by clinical staff.

Table 2.

Demographic and Clinical Characteristics of the Sample

VARIABLE	% OR MEAN (SD)
Gender
Male	57%
Female	43
Age (years)	41.81 (11.64)
Education
Didn't complete HS	2%
HS diploma/GED	16%
Associate degree	7%
College graduate	39%
Some college	18%
Graduate degree	17%
Marital status
Married	47%
Single, never married	29%
Divorced	14%
Separated	6%
Widowed	1%
Cohabitating	2%
Employment status
Full-time	43%
Part-time	6%
Unemployed	19%
Self-employed	11%
Retired	5%
Student	4%
Homemaker	5%
Disabled	3%
Military	4%
Dependence diagnoses
Alcohol	87%
Marijuana	30%
Cocaine	25%
Amphetamines	16%
Opiates	24%
Co-occurring disorder	75%

GED, graduate equivalent degree; HS, high school; SD, standard deviation.

Substance use diagnoses were made through a comprehensive assessment administered by treatment center staff in the course of routine healthcare operations. All diagnoses were based on DSM-IV criteria.¹⁴ Eighty-seven percent (n=1,465) of the sample received a diagnosis of alcohol dependence. The percentage diagnosed as dependent on both alcohol and at least one other drug was 32% (n=547).

Patients also completed follow up (i.e., post-treatment) surveys administered roughly 1 and 6 months after discharge. Eighty-one percent of the sample (n=1,355) completed the 1-month survey, and 70% (n=1,167) completed the 6-month survey. Sixty-one percent (n=1,026) completed both surveys.

Measures

Measures collected within the MORE program

The Module 1 assessment of MORE consisted of questions measuring motivation and self-efficacy. The question for self-efficacy asked, “How confident do you feel in your ability to work an active recovery program at this time?” (response scale: 1=not confident at all to 7=extremely confident). The question for motivation asked, “How motivated do you feel in your ability to work an active recovery program at this time?” (response scale: 1=not motivated at all to 7=extremely motivated). Because all patients in the sample completed the Module 1 assessment just prior to discharge, these questions represented motivation and self-efficacy levels at the end of treatment. We included these measures in the analyses of self-reported drinking days (reported below) because prior research has indicated that motivation and self-efficacy are related to substance use outcomes after treatment.^15,16

Measures collected after treatment

In the course of routine healthcare operations, research department staff administered follow-up surveys containing questions about behaviors and events that occurred since discharge; only those questions pertinent to this study will be discussed here. Patients completed the follow-up survey at two different times post-discharge: 1 month (25–44 days) and 6 months (173–210 days). Questions referenced the time period since treatment discharge or since the previous follow-up survey date if applicable. We did not analyze 1-month survey data because these data showed extremely low variability, with over 85% of patients reporting no substance use during the period. However, any substance use data reported at the 1-month follow-up were accounted for in the 6-month follow-up.

Post-treatment alcohol and drug use

Questions from the modified Form 90¹⁷ assessed substance use for the time period since treatment discharge. Alcohol items included the number of drinking days, the average number of drinks consumed on drinking days (drinks per drinking day), and the number of days the patient used each type of recreational drug (e.g., marijuana, cocaine, heroin, etc.). The Form 90 has solid validity and test–retest reliability in both adolescents and adults.¹⁸ It is important to note that data on post-treatment drinking days were self-report only and not confirmed via drug screens or corroborating reports. However, self-report measures of substance use are commonly used in treatment outcomes studies, and several studies have suggested a high (although not perfect) degree of correspondence between self-reported use and more objective measures such as drug screens.^19
–21

MORE usage outcomes measures. Number of program logins

The number of times each participant logged onto MORE was recorded for each month post-discharge. The number of logins served as the primary measure of program use in this study and was also used as a predictor in subsequent regression analyses for substance use outcomes.

Number of module pages accessed

The second measure of MORE usage was the total number of module pages (summed across modules) that participants accessed during each month post-discharge. Module content was deemed an important metric apart from program logins because most of the clinical/therapeutic content of the program is delivered through the modules. The total number of module pages accessed was entered into regression models for substance use outcomes.

Number of post-treatment drinking days

At each follow-up, patients self-reported the total number of days they consumed alcohol during the time period referenced by the assessment.

Analysis Strategy

The first set of analyses was purely descriptive and examined the two types of MORE program usage during the 6 months following treatment discharge (the total number of programs logins and the total number of module pages accessed).

The relationship between MORE program use and self-reported drinking days after treatment was examined via a regression analysis. This analysis was important to conduct because it is one thing to show that a particular variable (i.e., use of MORE) is related to another variable (i.e., number of post-treatment drinking days), but it is another thing to show that the first variable statistically predicts the second variable even when other variables potentially related to the second variable are taken into account. Past research indicates that demographic variables such as gender and age are often predictors of substance use after treatment, as are motivation, self-efficacy, and successful completion of treatment.¹⁵ In the present analysis, self-efficacy and motivation levels at the end of treatment were available because these measures were collected during the Module 1 assessment; demographic variables and treatment completion were available via the patient medical record.

The regression model was created by designating the total number of self-reported drinking days after treatment as the to-be-predicted variable and entering the total number of MORE program logins and the total number of module pages accessed as predictors. In addition, gender, age, marital status, treatment completion, self-efficacy, and motivation were added to the model as covariates. Because the number of self-reported drinking days at 6 months was skewed and overdispersed (i.e., the SD was larger than the mean), negative binomial regression (a type of generalized linear model) was used to create the model for drinking days, as this model is more appropriate than linear regression for this type of data.²²

Results

Total Program Logins and Total Module Pages Accessed

Table 3 shows the mean number of program logins and the mean number of module pages accessed during each month of the 6-month post-discharge period. For both measures, usage was fairly high in the first month and steadily dropped across subsequent months. Collapsed across months, the mean number of program logins during the 6-month period was 10.59 (SD=15.22). The mean number of module pages accessed during the period was 15.41 (SD=24.40). To better understand the pattern of logins within the sample, we also calculated the percentage of patients who logged onto the program a given number of times during the 6 months. Forty-four percent (n=745) of the sample logged onto the program zero to three times, 16% (n=279) logged on four to six times, and 40% (n=673) logged on seven or more times. Regarding module pages accessed, 27% (n=453) accessed 0 pages, 13% (n=215) accessed 1–5 pages, 27% (n=454) accessed 6–10 pages, and 33% (n=543) accessed over 10 pages.

Table 3.

Average Number of Logins and Module Pages Accessed During Each Month Post-Discharge

MONTH	MEAN (SD) LOGINS	MEAN (SD) MODULE PAGES ACCESSED
1	5.03 (5.57)	7.97 (9.11)
2	1.87 (3.73)	2.68 (7.51)
3	1.32 (3.16)	1.90 (5.90)
4	1.00 (2.81)	1.17 (4.37)
5	0.75 (2.20)	1.00 (4.07)
6	0.61 (2.12)	0.68 (2.89)

SD, standard deviation.

Negative Binomial Regression Analyses

A negative binomial regression examined the relationship between MORE usage and self-reported post-treatment drinking days during the 6 months post-treatment. The total number of logins and the total number of module pages accessed during the 6-month period were entered simultaneously as predictors along with the covariates discussed above. The overall fit of the negative binomial model was significant: χ²(14)=599.30, p<0.001. It is important to note that the total number of program logins was a significant predictor (B=−0.02; 95% confidence interval [CI]=−0.037 to −0.004; p=0.015), indicating that the greater the number of logins, the fewer the number of self-reported post-treatment drinking days. The total number of module pages accessed was also a significant predictor (B=−0.01; 95% CI=−0.021 to −0.002; p=0.024); the greater the number of module pages, the fewer the number of drinking days. Other significant predictors were age (B=0.032; 95% CI=0.013–0.05; p=0.001), gender (B=−498; 95% CI=−0.841 to −0.156; p=0.004), marital status (B=−0.797; 95% CI=−1.29 to −0.299; p=0.002), and self-efficacy (B=−0.218; 95% CI=−0.432 to −0.004; p=0.046). Older patients had a greater number of drinking days than younger patients, and females had a greater number of drinking days than males. Single patients had more drinking days than married patients. Patients with low self-efficacy had a greater number of drinking days than patients with high self-efficacy.

Discussion

In an earlier analysis of the MORE program, Klein et al.¹³ found a positive relationship between the number of program modules accessed and self-reported substance use outcomes after treatment. The present study replicated these preliminary findings using more comprehensive and detailed measures of usage, namely, the number of times patients logged onto the program and the number of module pages accessed during the first 6 months post-treatment. Both usage measures were significant predictors of self-reported post-treatment alcohol use in regression analyses. Patients with a higher number of logins had significantly fewer self-reported drinking days than patients with a lower number of logins, and patients who accessed a greater number of module pages had fewer self-reported drinking days than patients who accessed a lower number of pages.

Although we found a relationship between MORE use and self-reported alcohol use after treatment, we cannot conclude whether the relationship is causal. The findings should be interpreted very cautiously until they are replicated with a stronger experimental design such as a randomized controlled trial. It is also important to note that this analysis used administrative data collected in the course of routine healthcare operations. A randomized controlled trial comparing outcomes of MORE participants with those of participants not using the program is the only way to demonstrate program efficacy.

The present study also found that although patient usage of MORE was high in the month following treatment, both measures of usage steadily declined over time. This finding replicates that of Klein et al.¹³ and many other studies of computerized disease management programs.^7,8,11

Future research should focus on various methods to increase MORE engagement, such as providing patients with incentives, incorporating use of MORE directly into each patient's treatment plan, and increasing the interactions between each patient and his or her recovery coach.

Several limitations should be considered when interpreting the results of this study. Although the follow-up rate at 6 months was fairly high (70%), sampling bias may have been present and may have impacted the results. A related concern is that participants who were successfully contacted after treatment had better substance use outcomes than those not contacted, which may have inflated outcomes measures. Post-treatment alcohol use data were also self-report in nature and may have been inaccurate in some cases. As mentioned earlier, the basis of this study was an analysis of administrative data collected during the course of clinical operations, and we did not have a control group of patients who were not enrolled in MORE. Finally, it is possible that some patients who used MORE did so in an attempt to “game” the system. More specifically, patients may have used the program to specifically give a more favorable impression of their efforts to stay abstinent. This possibility seems fairly unlikely given that patients were not told how their data would be analyzed, nor were they told that analyses regarding MORE use and post-treatment substance use would be conducted.

It is important to note that this study also had several strengths. Because it replicates previous findings regarding MORE usage and outcomes using better, more comprehensive measures of usage, this study provides further evidence that computerized recovery support tools can be effective for individuals with substance use disorders. The usage measures used here provided a fairly detailed picture of patient usage, whereas some studies of computerized interventions have reported more qualitative measures.²³ The results of this study suggest that computerized recovery support programs hold promise for individuals treated for substance use disorders, although long-term engagement with these programs remains a challenge.

Footnotes

Disclosure Statement

A.A.K. and J.J.A. are employees of the Hazelden Foundation. Neither of the authors has any commercial associations that might create a conflict of interest in connection with the present article, nor do any competing financial interests exist.

References

Gustafson

, McTavish

, Stengle

, Ballard

, Hawkins

, Shaw

, Jones

, Julesberg

, McDowell

, Chen

, Volrathongchai

, Landucci

. Use and impact of eHealth system by low-income women with breast cancer. J Health Commun, 2005; 10,Suppl 1:195–218.

Glasgow

, Nutting

, King

, Nelson

, Cutter

, Gaglio

, Rahm

, Whitesides

. Randomized effectiveness trial of a computer-assisted intervention to improve diabetes care. Diabetes Care, 2005; 28:33–39.

Williams

, Lynch

, Glasgow

. Computer-assisted intervention improves patient-centered diabetes care by increasing autonomy support. Health Psychol, 2007; 26:728–734.

Grella

, Hser

, Hsieh

. Predictors of drug treatment re-entry following relapse to cocaine use in DATOS. J Subst Abuse Treat, 2003; 25:145–154.

Moos

, Moos

. Rates and predictors of relapse after natural and treated remission from alcohol use disorders. Addiction, 2006; 101:212–222.

Neto

, Lambaz

, Tavares

. Compliance with aftercare treatment, including disulfiram, and effect on outcome in alcohol-dependent patients. Alcohol Alcohol, 2007; 42:604–609.

Rooke

, Thorsteinsson

, Karpin

, Copeland

, Allsop

. Computer-delivered interventions for alcohol and tobacco use: A meta-analysis. Addiction, 2010; 105:1381–1390.

White

, Kavanagh

, Stallman

, Klein

, Kay-Lambkin

, Proudfoot

, Drennan

, Connor

, Baker

, Hines

, Young

. Online alcohol interventions: A systematic review. J Med Internet Res, 2010; 12:e62.

Bickel

, Buchhalter

, Marsch

, Badger

. Computerized behavior therapy for opioid-dependent outpatients: A randomized clinical trial. Exp Clin Psychopharmacol, 2008; 16:132–143.

10.

Kay-Lambkin

, Baker

, Lewin

, Carr

. Computer-based psychological treatment for comorbid depression and problematic alcohol and/or cannabis use: A randomized controlled trial of clinical efficacy. Addiction, 2009; 104:378–388.

11.

Campbell

, Nunes

, Miele

, Matthews

, Polsky

, Ghitza

, Turrigiano

, Bailey

, VanVeldhuisen

, Chapdelaine

, Froias

, Stitzer

, Carroll

, Winhusen

, Clingerman

, Perez

, McClure

, Goldman

, Crowell

. Design and methodological considerations of an effectiveness trial of a computer-assisted intervention: An example from the NIDA Clinical Trials Network. Contemp Clin Trials, 2012; 33:386–395.

12.

Kypri

, Langley

, Saunders

, Cashell-Smith

, Herbison

. Randomized controlled trial of web-based alcohol screening and brief intervention in primary care. Arch Intern Med, 2008; 168:530–536.

13.

Klein

, Slaymaker

, Dugosh

, McKay

. Computerized continuing care support for alcohol and drug dependence: A preliminary analysis of usage and outcomes. J Subst Abuse Treat, 2012; 42:25–34.

14.

American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4 ^th. Washington, DC: American Psychiatric Association, 2000.

15.

Ciraulo

, Piechniczek-Buczek

, Iscan

. Outcome predictors in substance use disorders. Psychiatr Clin North Am, 2003; 26:381–409.

16.

Kadden

, Litt

. The role of self-efficacy in the treatment of substance use disorders. Addict Behav, 2011; 36:1120–1126.

17.

Project MATCH Research Group. Project MATCH (Matching Alcoholism Treatment to Client Heterogeneity): Rationale and methods for a multisite clinical trial matching patients to alcoholism treatment. Alcohol Clin Exp Res, 1993; 17:1130–1145.

18.

Tonigan

, Miller

, Brown

. The reliability of the Form 90: An instrument for assessing alcohol treatment outcome. J Stud Alcohol, 1997; 58:358–364.

19.

Napper

, Fisher

, Johnson

, Wood

. The reliability and validity of drug users' self reports of amphetamine use among primarily heroin and cocaine users. Addict Behav, 2010; 35:350.

20.

Del Boca

, Darkes

. The validity of self-reports of alcohol consumption: state of the science and challenges for research. Addiction, 2003; 98,Suppl 2:1–12.

21.

Babor

, Steinberg

, Anton

, Del Boca

. Talk is cheap: Measuring drinking outcomes in clinical trials. J Stud Alcohol, 2000; 61:55–63.

22.

Neal

, Simons

. Inference in regression models of heavily skewed alcohol use data: A comparison of ordinary least squares, generalized linear models, and bootstrap resampling. Psychol Addict Behav, 2007; 21:441–452.

23.

Eysenbach

. The law of attrition. J Med Internet Res, 2005; 7:e11.