Drug and Alcohol Review (November 2009), 28, 608–615 DOI: 10.1111/j.1465-3362.2009.00064.x
Persistence of heroin use despite methadone treatment: Poor coping self-efficacy predicts continued heroin use RICHARD SENBANJO1,2, KIM WOLFF2, E. JANE MARSHALL2 & JOHN STRANG2 1
KCA Community Drug Services, Ashford, Kent, and 2National Addiction Centre, Institute of Psychiatry, King’s College London, London, UK
Abstract Aim. To evaluate the association between coping self-efficacy and persistent use of heroin by patients enrolled in a methadone treatment program. Design and Methods. Cross-sectional survey. One hundred and ninety-one patients attending outpatient methadone clinics in South-East England, United Kingdom.Validated questionnaires were used to assess drug use (Maudsley Addiction Profile), alcohol use (Alcohol Use Disorders Identification Test), mental health (Hospital Anxiety and Depression Scale) and coping self-efficacy (brief 8-item Drug Taking Confidence Questionnaire). Results. Half of the participants (95/191) reported heroin use in the preceding 14-day period. Heroin use during methadone treatment was associated with financial problems (P = 0.008), spending time with other drug users (P < 0.001), cocaine use (P = 0.002), low mood (P = 0.002) and dissatisfaction with the daily methadone dose (P = 0.014). Compared with ‘Heroin-abstinent’ patients, the ‘Heroin’group reported significantly lower mean coping self-efficacy scores (t = 9.8,d.f. = 182, P < 0.001,effect size 1.17).After correcting for the effects of co-variants in a logistic regression model, the main determinants of persistent heroin use were ‘coping self-efficacy’ [B -0.05; standard error (SE) 0.008;Wald 36.6; odds ratio (OR) 0.95, 95% confidence interval (CI) 0.94, 0.97; P < 0.001] and ‘dissatisfaction with methadone dose’ (B 0.93; SE 0.46;Wald 4.1; OR 2.5, 95% CI 1.03, 6.25; P = 0.042). Satisfaction with methadone dose showed no association with self-efficacy. Discussion and Conclusions.While heroin use during methadone treatment can partly be explained by inadequate dosing, our data suggest a more complex picture with significant contribution from poor coping self-efficacy. Efforts aimed at enhancing and maintaining the patients’ self-efficacy and social skills are likely to improve heroin and other drug use outcomes with added benefits for treatment completion rates and the throughput of methadone programs. [Senbanjo R, Wolff K, Marshall EJ, Strang J. Persistence of heroin use despite methadone treatment: Poor coping self-efficacy predicts continued heroin use. Drug Alcohol Rev 2009;28:608–615] Key words: methadone treatment, coping self-efficacy, heroin use, cocaine use, low mood.
Introduction It is now well established that methadone treatment is effective in helping patients reduce injecting and noninjecting heroin use [1]. Program characteristics play a major role in determining the outcome of methadone treatment with the best programs reporting up to 80% reduction in heroin use after several months in treatment [2]. The greatest reductions in illicit opiate use have been among patients maintained on higher methadone doses and those who remain in treatment for more than a year [3–5]. However, there is substantial overlap between effective and ineffective methadone dosages
and for many patients heroin use does not come to a complete end in spite of retention in treatment and the provision of adequate methadone doses [4,6,7]. This suggests that provision of higher methadone doses may only be beneficial for a subsample of heroin-dependent patients [8]. Methadone dose reductions carry significant risk of relapse to illicit opiate use [2,9]. Psychiatric comorbidity (particularly depression), cocaine use, benzodiazepine use and excessive alcohol consumption are all associated with poor methadone treatment outcomes [8,10,11]. Additional drug use by methadone patients may be motivated by psychological processes (cravings,
Richard Senbanjo MB BS, FMCS, MSc, Clinical Director, Honorary Research Fellow, Kim Wolff PhD, Senior Lecturer in the Addictions, E. Jane Marshall MRCP(I), FRCPsych, Consultant Psychiatrist and Senior Lecturer in the Addictions, John Strang MD, FRCPsych, Director. Correspondence to Dr Richard Senbanjo, Community Drug Service, KCA (UK), 171, Beaver Road, Ashford, Kent, TN23 7SG, UK. Tel: +01233 640040; Fax: +01233 640046; E-mail:
[email protected] Received 16 December 2007; accepted for publication 3 November 2008. © 2009 Australasian Professional Society on Alcohol and other Drugs
Self-efficacy and persistent heroin use
urges, dysfunctional thoughts), negative emotional states (stress, insomnia, conflict with others), positive emotional states (pleasant times, celebrating) or exposure to ‘temptation situations’ (drug availability, spending time with drug users, social pressure to use, testing self-control) [12,13]. The extent to which persistent heroin use by methadone patients blocks access for new, often younger, dependent heroin users who would like to enter treatment remains a source of concern for clinicians and commissioners of drug treatment services. Psychosocial interventions, when tailored to the needs of the individual, can enhance methadone treatment outcomes [1,14]. However, little is known about the personal and social characteristics of patients who continue to use heroin during methadone treatment or the type and amount of psychosocial support needed to give maximum clinical effect [15]. Self-efficacy (confidence in ability to remain abstinent) has been found to be related to alcohol and tobacco treatment outcomes, and client-centred interventions for alcohol problems have sought to strengthen patients’ coping self-efficacy [16,17]. Among opioid-dependent patients who have recently completed residential rehabilitation, adequate repertoire of skills and the use of cognitive, avoidance and distraction strategies have been found to protect against relapse to heroin use [18]. Perceived coping self-efficacy has not received sufficient attention in evaluations of methadone treatment outcomes [19,20]. We hypothesised that low coping self-efficacy would predict persistent heroin use among methadone patients.The objectives of the present study were: (i) to evaluate the coping self-efficacy of methadone patients in high-risk situations for heroin use; (ii) to identify potential moderators of heroin use outcomes during methadone treatment; and (iii) to identify the main determinants of persistent heroin use among these patients. Better understanding of the complex interactions between individual, social and treatment variables might help in the design of effective psychosocial interventions for methadone patients.
Methods Participants Consecutive attendees at a supervised methadone program (maintenance or slow-reduction) were recruited to the study between March and July 2003. Participants had to be more than 18 years old, stabilised on methadone treatment for at least 1 month, not intoxicated by alcohol or drugs, not experiencing withdrawal symptoms and able to give informed consent. Ethical approval was obtained from the local Ethics Committee in Kent and from the Joint Research Ethics
609
Committee at the Institute of Psychiatry/South London & Maudsley NHS Foundation Trust. Written informed consent was obtained and the study questionnaire was completed by participants assisted, when needed, by an independent advisor. Treatment setting Data collection took place within community-based methadone clinics across East Kent, a largely rural/ suburban part of South-East England, United Kingdom. The clinics serve a population of approximately 600 000 and the main provider of drug treatment services in the area is KCA (UK), a non-statutory organisation. Each patient had a named key worker and medical care, including prescribing, is provided by Specialist Medical Practitioners in Addiction. The service operates a flexible, individualised methadone dosing policy and most of the patients were on daily selfadministered doses of methadone, supervised by a community pharmacist. Incentives were not provided other than unsupervised takeout of methadone doses for patients who consistently provide drug-free samples. Measurements The research tool consisted of sociodemographic data and validated instruments. Drug use was assessed using the substance use section of the Maudsley Addiction Profile (MAP) [21]. In an attempt to minimise recall bias, the MAP was modified to record drug use over the preceding 14-day period, instead of the standard 28-day reference period. Alcohol consumption was assessed using the 10-item Alcohol Use Disorders Identification Test (AUDIT) questionnaire [22]. The amount of time spent socialising with other drug users was rated on a 4-point scale (none, little, some, most), while satisfaction with methadone dose was rated on a 5-point scale (very satisfied, satisfied, neither satisfied nor dissatisfied, dissatisfied, very dissatisfied). Mental health was assessed using the Hospital Anxiety and Depression Scale (HADS) [23]. The HADS is a selfassessment instrument consisting of 14 items (scored from 0 to 3). Seven items concerned anxiety and seven relate to depression with total scores for each subscale ranging from 0 to 21. The HADS is a clinically useful tool for screening for mood disorders and a score of ⱖ11 indicates probable presence of anxiety disorder or depression [23,24]. Coping self-efficacy in high-risk situations for heroin use was assessed using the 8-item Drug Taking Confidence Questionnaire (DTCQ-8), a brief, validated instrument that accounts for 95% of the variance in the full, 50-item DTCQ questionnaire [25]. The questionnaire measured situation-specific self© 2009 Australasian Professional Society on Alcohol and other Drugs
610
R. Senbanjo et al.
efficacy in resisting heroin use on a 6-point scale ranging from ‘not at all confident’ (scored zero) to ‘very confident’ (scored 100). Overall self-efficacy score was obtained for each patient by calculating the mean of the eight responses. Sample characteristics The 191 participants were predominantly male (70.3%), white European (97.9%) and unemployed (73.0%) with a mean age of 33.0 years. Nearly twothirds of the sample (60.7%) reported cannabis use in the preceding 14-day period and approximately onequarter reported benzodiazepine (28.6%) or cocaine (21.5%) use. Heroin use data are reported in the Results section. Approximately one-third of the sample (56/191) scored positively (ⱖ8) on the AUDIT questionnaire and the majority (96.2%) smoked cigarettes or hand-rolled tobacco. Methadone doses ranged from 2 to 105 mg/day with a mean ⫾ standard deviation (SD) of 38 mg/day ⫾ 16. A third of the sample (35.3%) scored ⱖ11 on the HADS ‘anxiety’ scale (mean ⫾ SD 9.0 ⫾ 4.9) while one-fifth (20.7%) scored ⱖ11 on the HADS ‘depression’ scale (mean ⫾ SD 7.3 ⫾ 4.6). Classification of the participants The participants were classified into two groups—those who reported no heroin use in the preceding 14-day period (the ‘Heroin-abstinent’ group) and those who reported heroin use on at least one occasion over the same 14-day period (the ‘Heroin’ group). Statistical analyses Satisfaction with methadone dose was recoded as ‘satisfied’ (very satisfied and satisfied) or ‘not satisfied’ (neither satisfied nor dissatisfied, dissatisfied and very dissatisfied). Normally distributed variables were described with means and SD. Groups were compared using odds ratios (OR), mean differences, confidence intervals (CI), c2 or independent t-test. Logistic regression was performed [spss version 15 (SPSS Inc., Chicago, IL, USA)] to assess the impact of a number of factors on persistent heroin use. Effect size was calculated as mean difference in group scores divided by the SD for the whole sample. Significance was designated at 95% (P < 0.05). Results Only one of the 192 patients who completed the study questionnaire failed to provide information on heroin use. Half of the remaining patients (95/191) reported heroin use on at least one occasion in the preceding © 2009 Australasian Professional Society on Alcohol and other Drugs
14-day period (the ‘Heroin’ group). The remaining 96 patients (50.3%) formed the ‘Heroin-abstinent’ group. Sociodemographic data There was no significant difference in the mean age ⫾ SD of the ‘Heroin-abstinent’ and ‘Heroin’ groups: 33.1 years ⫾ 9.4 and 32.9 years ⫾ 7.8, respectively.The groups also showed no significant differences in the following variables (mean ⫾ SD): ‘age left school’ (15.4 years ⫾ 1.3 and 15.6 years ⫾ 1.4, respectively); ‘age at first heroin use’ (20.9 years ⫾ 5.2 and 21.7 years ⫾ 6.9, respectively); and ‘age at regular heroin use’ (22.2 years ⫾ 5.4 and 23.3 years ⫾ 6.8, respectively). Table 1 shows no significant betweengroup differences in the sex, ethnicity, employment and housing status of the two groups of methadone patients. A higher proportion of the ‘Heroin’ group reported living with a drug-using partner (52.2% vs. 32.6%), but the difference was not statistically significant (c2 = 3.6; OR 2.3; 95% CI 0.96, 5.32; P = 0.059). Compared with ‘Heroin-abstinent’ patients, those in the ‘Heroin’ group were significantly more likely to spend ‘some’ or ’most’ of their time with other drug users (93.7% vs. 63.5%; c2 = 25.7; OR 2.1; 95% CI 1.66, 2.64; P < 0.001) and to be in debt (c2 = 7.1; OR 2.2; 95% CI 1.23, 3.98; P = 0.008). Drug use, alcohol consumption and tobacco smoking Patients in the ‘Heroin’ group were more likely to use cocaine (crack or powder) (c2 = 9.2; OR 3.1; 95% CI 1.46, 6.49; P = 0.002), but there were no significant between-group differences in the use of cannabis, benzodiazepines, amphetamines, alcohol (as measured by the AUDIT) or cigarette/tobacco smoking (Table 1). Methadone treatment A higher proportion of patients in the ‘Heroinabstinent’ group had been in treatment for more than 6 months (67.7%) compared with the ‘Heroin’ group (54.7%), but the difference was not statistically significant (c2 = 3.4; OR 1.7; 95% CI 0.96, 3.12; P = 0.066). There was no significant between-group difference in the daily methadone dose (mean ⫾ SD: 40 mg ⫾ 13 for the ‘Heroin’ group and 37 mg ⫾ 19 for ‘Heroinabstinent’ patients; t = -1.3; d.f. = 189; P = 0.200). However, patients in the ‘Heroin’ group were significantly more likely than their counterparts to express dissatisfaction with their methadone dose (c2 = 6.1; OR 1.6; 95% CI 1.05, 2.30; P = 0.014). HADS scores There was no significant between-group difference in the mean ⫾ SD of HADS anxiety scores (‘Heroin’
Self-efficacy and persistent heroin use
611
Table 1. Sociodemographic and substance use data among 191 methadone patients according to heroin use in the preceding 14-day period Whole sample ‘Heroin-abstinent’ ‘Heroin’ group (n = 191) group (n = 96) (n = 95) n (%) n (%) n (%) Sex Male Ethnicity White European Employment (n = 188) Unemployed Finance (n = 190) In debt Housing Homeless Living with Parents Partner A drug user (n = 90) Time with drug users ‘Some’ or ‘Most’ Drug use Cannabis Benzodiazepines (n = 189) Cocaine Amphetamines Alcohol use AUDIT positive Smoking (n = 188) Tobacco smoker Time in treatment >6 months Methadone Not satisfied with current dose
Statistics c2
OR
95% CI
P (two-tailed)
135 (70.7)
68 (70.8)
67 (70.5)
0.0 0.99 0.53–1.84
0.963
187 (97.9)
93 (96.9)
94 (98.9)
1.0 3.03 0.31–29.68
0.317
150 (79.8)
75 (80.6)
75 (78.9)
0.1 1.11 0.54–2.27
0.772
109 (57.4)
46 (47.9)
63 (67.0)
7.1 2.21 1.23–3.98
0.008a
26 (13.6)
11 (11.4)
15 (15.8)
0.8 0.69 0.30–1.59
0.383
36 (18.8)
15 (15.6)
21 (22.1)
1.3 1.53 0.74–3.19
0.252
38 (42.2)
15 (32.6)
23 (52.2)
3.6 2.26 0.96–5.32
0.059
150 (78.5)
61 (63.5)
89 (93.7)
25.7 2.10 1.66–2.64
