bs_bs_banner
R E V I E W
Drug and Alcohol Review (November 2014), 33, 625–636 DOI: 10.1111/dar.12194
COMPREHENSIVE REVIEW
Solving the problem of non-attendance in substance abuse services JOANNA MILWARD, MICHAEL LYNSKEY & JOHN STRANG Institute of Psychiatry, King’s College London, London, UK
Abstract Issues. Rates of non-attendance are among the highest in substance misuse services. Non-attendance is costly and results in the inefficient use of limited resources. Patients who frequently miss their appointments have worse outcomes including treatment dropout and decreased likelihood of achieving long-term abstinence. Approach. This narrative review evaluates interventions targeting non-attendance in addiction services and draws upon the wider health-care literature to identify interventions that could be adapted for substance-abusing populations. Key Findings. Both fixed value and intermittent reinforcement contingency management demonstrate potential for improving attendance. However, small sample sizes and heterogeneous populations make it difficult to draw firm conclusions.Appointment reminders by letter or telephone have demonstrated moderate evidence for improving attendance in substance-abusing populations. Text message appointment reminders are extensively utilised in general health-care settings and consistently improve attendance; however, there is a paucity of research examining the feasibility and effectiveness of text message reminders in addiction services. Implications. A lack of evidence for methods to improve attendance is reflected in the continuing challenge faced by addiction services attempting to manage high rates of non-attendance. Conclusions. Non-attendance remains a persistent issue for addiction services.While there is limited evidence that contingency management improves attendance, more rigorous research is needed to determine the optimal intervention components and effectiveness in different populations, particularly those receiving maintenance treatments. Multicomponent text message interventions incorporating different delivery and content strategies demonstrate a promise for improving nonattendance and poor engagement. [Milward J, Lynskey M, Strang J. Solving the problem of non-attendance in substance abuse services. Drug Alcohol Rev 2014;33:625–36] Key words: substance misuse, contingency management, appointment reminder, non-attendance, SMS messaging.
Introduction Non-attendance at outpatient appointments in drug and alcohol services is among the highest in medical specialities.Twenty-five to 37% of new patient appointments are specified as ‘did not attend’ each year in addiction services in the UK [1]. Missed outpatient appointments are an inefficient use of limited resources and also impact upon the quality of care a patient receives [2]. Individuals who miss their appointments in substance abuse settings demonstrate elevated rates of adverse treatment outcomes; non-attendance is associated with medication non-adherence [3], an increased risk of being rehospitalised [4] and treatment dropout
[5]. Non-attenders are also less likely to achieve abstinence at long-term follow-up than attenders [6]. Understanding the reasons for non-attendance at appointments is a complex and multifaceted issue for treatment services. It is useful to distinguish between non-attendance at initial (after referral) and follow-up appointments, as well as by attendance behaviour [1]; however, studies do not consistently report on these differences. Rates of initial non-attendance are typically higher than for follow-up appointments [1] with patients who are drug dependent being less likely to attend their first treatment session compared with those who are alcohol dependent [5]. Patients exhibit different patterns of attendance; some will miss
Joanna Milward BSc, MSc, Research Assistant, Michael Lynskey BSc, MSc, PhD, Professor of Addictions, John Strang MBBS, FRCPsych, MD, Head of Department. Correspondence to Ms Joanna Milward, National Addiction Centre, Addictions Department, Institute of Psychiatry, King’s College London, Addictions Sciences Building, 4 Windsor Walk, Denmark Hill, London SE5 8BB, UK. Tel: +44 (0)207 848 0811; E-mail: [email protected] Received 29 January 2014; accepted for publication 15 July 2014. © 2014 Australasian Professional Society on Alcohol and other Drugs
626
J. Milward et al.
appointments occasionally (partial attendance), some frequently (low attendance) and others will disengage entirely (dropout), and these patterns may have different prognostic significance and may require different treatment interventions. Reasons for non-attendance in addiction settings have not been well described in the research literature. However, they are likely to vary depending on patient and service-level factors. Patients who miss an appointment are more likely to be male, unmarried, living at home and be past or current poly-substance users [7]. At treatment commencement, patient-level predictors of non-attendance include higher drug use, lower levels of motivation and being unemployed [5]. Regarding follow-up appointments, a single study in an alcohol clinic in Reykjavik identified younger age and higher trait anxiety as predictors of non-attendance [8]. Evidence from the wider health literature suggests that additional factors associated with non-attendance may include low social functioning, more severe mental health problems [9], disillusionment with treatment [10,11] and memory-related problems, including forgetfulness [12,13] and lack of familial support [14]. Clinic-level factors associated with non-attendance include lengthy waiting times between referral and assessment [15] and between assessment and treatment [6], poor therapeutic alliance [16], the clinic environment, staff attitudes, and the level of administrative support, such as the availability of appointment reminders [7]. Evidence from general practice settings suggests that practice error [17] inconvenient appointment times and not being able to cancel may also affect attendance [7,13]; however, research is lacking with regard to substance abuse services. Given evidence that both non-attendance at substance abuse treatment is common and that nonattendance predicts adverse treatment outcomes, further research is necessary to understand the factors associated with non-attendance and to develop interventions to improve attendance rates. We conducted a narrative literature review, searching all available online databases for research papers describing interventions for improving attendance rates in substance abuse services, and draw upon promising interventions from other health-care settings which may be transferable to addiction services. It is acknowledged that there are broader issues of non-attendance which are outside the scope of the present review. Specifically, clinic-level barriers, as just discussed, would be unaffected by psychological and behavioural interventions for improving engagement and require additional research to support service delivery. Concomitantly, the majority of research attention examining key predictors of non-attendance has focused upon psychiatric and general health-care set© 2014 Australasian Professional Society on Alcohol and other Drugs
tings; further elucidation of factors associated with nonattendance in substance-dependent patients is a necessary and timely avenue of research.
Interventions for improving treatment attendance in substance misuse services Strategies to improve attendance in substance abuse services can be broadly divided into clinical and nonclinical interventions. Clinical interventions include motivational interviewing (MI) and contingency management (CM), while non-clinical interventions are service delivery focused and target-accelerated intake and appointment reminders. Trials examining accelerated intake report it to be an effective strategy for improving initial session attendance, and it is an intervention for methadone maintenance recommended by the National Treatment Agency for Substance Misuse in the UK [18]. Accelerated intake (same day, next day and within two days) significantly improves attendance at the initial treatment session [19–23]. A single study has examined the effect of rapid intake on attendance for subsequent appointments and reported no mean difference in later attendance rates, suggesting it is only effective in supporting clients to attend their first appointment [23]. A small number of clinical trials have examined attendance as an outcome for MI interventions [24– 30]. These studies report contradictory findings for the effectiveness of MI; for example, studies have reported MI to be more effective for longer-term treatment retention compared with short-term treatment attendance [26,28], while other studies have reported opposite findings [24,25]. No clear conclusions can be drawn from the evidence on MI for session attendance (for detailed reviews, please see Pulford et al. [31] and Dunn et al. [32]). The two interventions for improving attendance that have been the most extensively studied are CM and appointment reminders. In the remainder of this paper, we focus on the literature exploring the implementation and effectiveness of these interventions.
Contingency Management CM is a behavioural intervention based on the principles of reinforcement to target abstinence, medication adherence, goal performance or treatment attendance. The majority of research has focused on abstinence with a consistent body of research reporting efficacy and effectiveness, although effect sizes vary [33,34]. Compared with other psycho-social treatments, metaanalyses reveal that CM is one of the most effective interventions for abstinence and retention [35].
Solving the problem of non-attendance
There are two main methods of CM reinforcement. The first is voucher-based reinforcement therapy (VBRT) whereby vouchers exchangeable for goods are provided in either escalating values or at a fixed value. The second method is the intermittent reinforcement or ‘prize draw’ technique [36], in which, instead of reinforcing behaviours through a fixed voucher of cash value, behaviour is reinforced by the chance to draw a slip of paper from a bowl. Each draw is associated with the chance of winning a prize of varying value (typically between $1 and $100). The behavioural reinforcer is different for the respective techniques; while the reinforcer in VBRT methods is the value of the voucher itself, the reinforcer in intermittent reinforcement techniques is the random chance or ‘lottery’ of receiving a high value reinforcer, not the prize itself. A comparatively small body of research has evaluated treatment attendance as an outcome for CM for adults and adolescents in substance abuse and methadone maintenance treatment programs (Table 1). Two pilot studies have examined low-cost prize draw reinforcement for attendance in adolescent substanceabusing populations [48,50]. Prizes were either social reinforcers (e.g. ‘great job’) or small ($1), medium ($5), or large ($15–20) value prizes. One study reported a significant rise in attendance in the incentive condition [50], while the other study reported a significant rise in treatment retention but not session attendance [48]. Sinha et al. [41] examined the effect of escalating vouchers on attendance on a motivational enhancement therapy course in probationreferred adolescents and reported those who received the vouchers were more likely to complete the course; however, there was not a significance effect on mean days of attendance. Studies evaluating CM for attendance in adult substance-abusing populations have examined both VBRT methods and prize draw techniques with five out of seven studies included in this review using the latter. Five studies reported significant effects upon attendance rates; three used prize draws ranging between $1 and $100 [36,46,49], and the others provided fixed vouchers of $2.50 or $5 per session [40,47]. Alessi et al. [45] reported a slightly different result, with significant effects for abstinence but not for attendance using the $1–100 prize draw method. A possible explanation for this result may be that the prizes for abstinence were of a total higher magnitude than those for attendance, suggesting that the lower value reinforcer was not enough of an incentive for patients to attend when compared with the higher value reinforcer. Finally Businelle et al. [42] compared three monetary values of voucher ($60, $95 and $210) provided over three months and reported that only the two higher value vouchers produced significant results.
627
Firm conclusions are difficult to draw due to the range of designs, small sample sizes and the heterogeneous nature of the drug-using population samples. Notwithstanding this, the initial evidence suggests that prize draw CM is at least an equally efficacious alternative to VBRT in adult populations, although two studies did show contradictory results. In VRBT methods, the monetary value of the reinforcer may be an important factor in successful implementation. Conclusions about prize draw techniques in adolescent populations cannot be drawn due to the low number of trials. CM for attendance has been minimally evaluated in methadone-maintained populations.The majority (four out of five studies included) evaluated VBRT in drugdependent pregnant women with a randomised controlled trial (RCT) design. These studies either examined attendance at outpatient day treatment or attendance at counselling or psychiatric services. Two studies reported significant results; Kidorf et al. [43] and Jones et al. [39] provided vouchers of $25 per week and escalating vouchers of up to $70 per day, respectively. Conversely, Jones et al. [38] offered $5 per day for seven days and reported no improvements for attendance. Svikis et al. [37] offered methadonemaintained patients either $0, $1, $5 or $10 per day for seven days and also reported no significant increase in attendance although non-methadone-maintained patients who received the higher magnitude vouchers ($5/$10) did attend more sessions. The single study using the prize draw technique used reinforcers ranging from $2.50 to $100 and reported that patients who exhibited poor attendance at baseline attended significantly more counselling sessions; however, there were no differences in attendance in patients who attended regularly [44]. Overall, the evidence for CM for attendance in methadone-maintained patients is mixed; methadone is a powerful reinforcer in itself [37], and additional incentives that are below a certain magnitude may not improve attendance further. The two studies that reported significant results provided incentives for a longer period of time (12 week and 14 day trial periods), and thus, the total ‘earnable’ amount was higher than in the studies providing incentives for a seven-day period. Conclusions about the effectiveness of prize draw CM for attendance in methadone-maintained patients cannot be drawn as it has only been evaluated in a single trial and not subjected to an RCT design. Research into CM for attendance is at an early stage and a number of questions remain unanswered. Regarding VBRT methods, it is unclear as to what the minimal value of an effective reinforcer is, whether there is a ceiling effect of prize value and whether the total sum of the reinforcer is more important than the © 2014 Australasian Professional Society on Alcohol and other Drugs
VBRT
Reinforcer
142
25
80
20
65
190
125
Jones et al. [38]
Jones et al. [39]
Helmus et al. [40]
Sinha et al. [41]
Businelle et al. [42]
Kidorf et al. [43]
n
Svikis et al. [37]
Study
© 2014 Australasian Professional Society on Alcohol and other Drugs MM
SA
MA
DD (substance use + mental health diagnosis)
MM
MM
MM + non-MM
Substance group
Adult outpatient psychiatric service
Veteran outpatient clinic
Probation-referred adolescents outpatient
Adult outpatient
Pregnant women outpatient
Pregnant women outpatient
Pregnant women outpatient
Clinical population
RCT: Contingent vouchers versus no voucher control, parallel groups
Retrospective contingent vouchers (three conditions) versus TAU
Contingent vouchers versus no voucher control, parallel groups
A B A within subjects
Contingent vouchers versus no voucher control, parallel groups
Contingent vouchers versus no voucher control, parallel groups
Contingent vouchers versus no voucher control, parallel groups
Design
12
12
28 days
12
2
1
1
Intervention duration (weeks)
Total full days outpatient treatment attended
Total full days outpatient treatment attended
% on time counselling attendance
Total MET sessions attended
Total aftercare sessions attended (max two per month)
Total psychiatric sessions attended per month
$5 per day
Escalating; $5 day 1, increasing $5 per day
$2.50 per session attended (max 2 per week)
Escalating; $25 session 1, increasing $10 per session Contract 1: $5/$10/$15 max $60 Contract 2: $5/$10/$15/ $20 max $95 Contract 3: $25/$30/$50 max $210
$25 per week
Outcome
Total full days outpatient treatment attended
Incentive (value)
$0/$1/$5/$10 per day
Table 1. Contingency management for attendance studies
MM group: No effect of incentives Non-MM group: $5/$10 incentive groups attended 1 more day of treatment per week (M = 3.3) compared with $0/$1 groups (M = 2.3)* MM group: Attended 1.5 more treatment days per week (M = 5.2) compared with AT (M = 3.7)* Incentive condition attended 1.5 more treatment days per week (M = 12.1) than control condition (M = 10.6)* Attendance increased by 20% during CM phase compared with baseline** Higher (NS) attendance rate for CM group compared with controls (M = 2.8 vs. 2.3) Participants in the two CM conditions with the highest voucher values (median = 6 for contract 3, median = 5 for contract 2) attended more sessions than TAU (median = 2)*** CM condition attended 3.6 more sessions (mean) than TAU in month 1, 3.5 days (mean) in month 2, 3.3 days (mean) in month 3***
Results
628 J. Milward et al.
50
52
77
103
75
54
336
239
52
Rhodes et al. [44] Study 1 (variable reinforcer)
Rhodes et al. [44] Study 2 (variable vs. fixed)
Petry et al. [36]
Alessi et al. [45]
Ledgerwood et al. [46]
Bride and Humble [47]
Lott and Jencius [48]
Petry et al. [49]
Branson et al. [50]
SA
SA
SA
SA
SA
SA
Cocaine-dependent MM
MM
MM
Adolescent outpatient
Adult community outpatient
Adolescent outpatient
Adult African American women
Adult community outpatient
Adult community outpatient
Adult community outpatient
Inner city adult outpatient
Inner city adult outpatient
Historical control (TAU) versus CM + TAU
TAU and CM versus TAU control, parallel groups
Observational: Pre-/ post-introduction of CM program
Historical control (TAU) versus CM + TAU
Open group design, TAU followed by TAU + CM (and vice versa)
Crossover design: TAU versus CM
TAU and CM versus TAU control, parallel groups
A B A within subjects
A B A within subjects
52
12
52
4
16
12
12
8
8
% of sessions attended (weekly)
Total days in treatment
Prize value of $1/$20/ $80–$100
Gift certificates $5 for perfect weekly attendance $0/$1/$5/$15
Social reinforcer/$1/$5/$20
$1/$20/$100
% treatment days attended
Prize value of $1/$20/$100
% sessions attended
Total group counselling sessions attended
(i) % days attended (ii) Mean length of stay
Weekly group attendance
% on time counselling attendance
$2.50/$5/$10/$25/$100 versus fixed rate $3.25
Small ($1), medium (e.g. Walkman), jumbo (e.g. television)
% on time counselling attendance
$2.50/$5/$10/$25/$100
23% mean increase of attendance for poor attenders in CM condition (attended
R E V I E W
Drug and Alcohol Review (November 2014), 33, 625–636 DOI: 10.1111/dar.12194
COMPREHENSIVE REVIEW
Solving the problem of non-attendance in substance abuse services JOANNA MILWARD, MICHAEL LYNSKEY & JOHN STRANG Institute of Psychiatry, King’s College London, London, UK
Abstract Issues. Rates of non-attendance are among the highest in substance misuse services. Non-attendance is costly and results in the inefficient use of limited resources. Patients who frequently miss their appointments have worse outcomes including treatment dropout and decreased likelihood of achieving long-term abstinence. Approach. This narrative review evaluates interventions targeting non-attendance in addiction services and draws upon the wider health-care literature to identify interventions that could be adapted for substance-abusing populations. Key Findings. Both fixed value and intermittent reinforcement contingency management demonstrate potential for improving attendance. However, small sample sizes and heterogeneous populations make it difficult to draw firm conclusions.Appointment reminders by letter or telephone have demonstrated moderate evidence for improving attendance in substance-abusing populations. Text message appointment reminders are extensively utilised in general health-care settings and consistently improve attendance; however, there is a paucity of research examining the feasibility and effectiveness of text message reminders in addiction services. Implications. A lack of evidence for methods to improve attendance is reflected in the continuing challenge faced by addiction services attempting to manage high rates of non-attendance. Conclusions. Non-attendance remains a persistent issue for addiction services.While there is limited evidence that contingency management improves attendance, more rigorous research is needed to determine the optimal intervention components and effectiveness in different populations, particularly those receiving maintenance treatments. Multicomponent text message interventions incorporating different delivery and content strategies demonstrate a promise for improving nonattendance and poor engagement. [Milward J, Lynskey M, Strang J. Solving the problem of non-attendance in substance abuse services. Drug Alcohol Rev 2014;33:625–36] Key words: substance misuse, contingency management, appointment reminder, non-attendance, SMS messaging.
Introduction Non-attendance at outpatient appointments in drug and alcohol services is among the highest in medical specialities.Twenty-five to 37% of new patient appointments are specified as ‘did not attend’ each year in addiction services in the UK [1]. Missed outpatient appointments are an inefficient use of limited resources and also impact upon the quality of care a patient receives [2]. Individuals who miss their appointments in substance abuse settings demonstrate elevated rates of adverse treatment outcomes; non-attendance is associated with medication non-adherence [3], an increased risk of being rehospitalised [4] and treatment dropout
[5]. Non-attenders are also less likely to achieve abstinence at long-term follow-up than attenders [6]. Understanding the reasons for non-attendance at appointments is a complex and multifaceted issue for treatment services. It is useful to distinguish between non-attendance at initial (after referral) and follow-up appointments, as well as by attendance behaviour [1]; however, studies do not consistently report on these differences. Rates of initial non-attendance are typically higher than for follow-up appointments [1] with patients who are drug dependent being less likely to attend their first treatment session compared with those who are alcohol dependent [5]. Patients exhibit different patterns of attendance; some will miss
Joanna Milward BSc, MSc, Research Assistant, Michael Lynskey BSc, MSc, PhD, Professor of Addictions, John Strang MBBS, FRCPsych, MD, Head of Department. Correspondence to Ms Joanna Milward, National Addiction Centre, Addictions Department, Institute of Psychiatry, King’s College London, Addictions Sciences Building, 4 Windsor Walk, Denmark Hill, London SE5 8BB, UK. Tel: +44 (0)207 848 0811; E-mail: [email protected] Received 29 January 2014; accepted for publication 15 July 2014. © 2014 Australasian Professional Society on Alcohol and other Drugs
626
J. Milward et al.
appointments occasionally (partial attendance), some frequently (low attendance) and others will disengage entirely (dropout), and these patterns may have different prognostic significance and may require different treatment interventions. Reasons for non-attendance in addiction settings have not been well described in the research literature. However, they are likely to vary depending on patient and service-level factors. Patients who miss an appointment are more likely to be male, unmarried, living at home and be past or current poly-substance users [7]. At treatment commencement, patient-level predictors of non-attendance include higher drug use, lower levels of motivation and being unemployed [5]. Regarding follow-up appointments, a single study in an alcohol clinic in Reykjavik identified younger age and higher trait anxiety as predictors of non-attendance [8]. Evidence from the wider health literature suggests that additional factors associated with non-attendance may include low social functioning, more severe mental health problems [9], disillusionment with treatment [10,11] and memory-related problems, including forgetfulness [12,13] and lack of familial support [14]. Clinic-level factors associated with non-attendance include lengthy waiting times between referral and assessment [15] and between assessment and treatment [6], poor therapeutic alliance [16], the clinic environment, staff attitudes, and the level of administrative support, such as the availability of appointment reminders [7]. Evidence from general practice settings suggests that practice error [17] inconvenient appointment times and not being able to cancel may also affect attendance [7,13]; however, research is lacking with regard to substance abuse services. Given evidence that both non-attendance at substance abuse treatment is common and that nonattendance predicts adverse treatment outcomes, further research is necessary to understand the factors associated with non-attendance and to develop interventions to improve attendance rates. We conducted a narrative literature review, searching all available online databases for research papers describing interventions for improving attendance rates in substance abuse services, and draw upon promising interventions from other health-care settings which may be transferable to addiction services. It is acknowledged that there are broader issues of non-attendance which are outside the scope of the present review. Specifically, clinic-level barriers, as just discussed, would be unaffected by psychological and behavioural interventions for improving engagement and require additional research to support service delivery. Concomitantly, the majority of research attention examining key predictors of non-attendance has focused upon psychiatric and general health-care set© 2014 Australasian Professional Society on Alcohol and other Drugs
tings; further elucidation of factors associated with nonattendance in substance-dependent patients is a necessary and timely avenue of research.
Interventions for improving treatment attendance in substance misuse services Strategies to improve attendance in substance abuse services can be broadly divided into clinical and nonclinical interventions. Clinical interventions include motivational interviewing (MI) and contingency management (CM), while non-clinical interventions are service delivery focused and target-accelerated intake and appointment reminders. Trials examining accelerated intake report it to be an effective strategy for improving initial session attendance, and it is an intervention for methadone maintenance recommended by the National Treatment Agency for Substance Misuse in the UK [18]. Accelerated intake (same day, next day and within two days) significantly improves attendance at the initial treatment session [19–23]. A single study has examined the effect of rapid intake on attendance for subsequent appointments and reported no mean difference in later attendance rates, suggesting it is only effective in supporting clients to attend their first appointment [23]. A small number of clinical trials have examined attendance as an outcome for MI interventions [24– 30]. These studies report contradictory findings for the effectiveness of MI; for example, studies have reported MI to be more effective for longer-term treatment retention compared with short-term treatment attendance [26,28], while other studies have reported opposite findings [24,25]. No clear conclusions can be drawn from the evidence on MI for session attendance (for detailed reviews, please see Pulford et al. [31] and Dunn et al. [32]). The two interventions for improving attendance that have been the most extensively studied are CM and appointment reminders. In the remainder of this paper, we focus on the literature exploring the implementation and effectiveness of these interventions.
Contingency Management CM is a behavioural intervention based on the principles of reinforcement to target abstinence, medication adherence, goal performance or treatment attendance. The majority of research has focused on abstinence with a consistent body of research reporting efficacy and effectiveness, although effect sizes vary [33,34]. Compared with other psycho-social treatments, metaanalyses reveal that CM is one of the most effective interventions for abstinence and retention [35].
Solving the problem of non-attendance
There are two main methods of CM reinforcement. The first is voucher-based reinforcement therapy (VBRT) whereby vouchers exchangeable for goods are provided in either escalating values or at a fixed value. The second method is the intermittent reinforcement or ‘prize draw’ technique [36], in which, instead of reinforcing behaviours through a fixed voucher of cash value, behaviour is reinforced by the chance to draw a slip of paper from a bowl. Each draw is associated with the chance of winning a prize of varying value (typically between $1 and $100). The behavioural reinforcer is different for the respective techniques; while the reinforcer in VBRT methods is the value of the voucher itself, the reinforcer in intermittent reinforcement techniques is the random chance or ‘lottery’ of receiving a high value reinforcer, not the prize itself. A comparatively small body of research has evaluated treatment attendance as an outcome for CM for adults and adolescents in substance abuse and methadone maintenance treatment programs (Table 1). Two pilot studies have examined low-cost prize draw reinforcement for attendance in adolescent substanceabusing populations [48,50]. Prizes were either social reinforcers (e.g. ‘great job’) or small ($1), medium ($5), or large ($15–20) value prizes. One study reported a significant rise in attendance in the incentive condition [50], while the other study reported a significant rise in treatment retention but not session attendance [48]. Sinha et al. [41] examined the effect of escalating vouchers on attendance on a motivational enhancement therapy course in probationreferred adolescents and reported those who received the vouchers were more likely to complete the course; however, there was not a significance effect on mean days of attendance. Studies evaluating CM for attendance in adult substance-abusing populations have examined both VBRT methods and prize draw techniques with five out of seven studies included in this review using the latter. Five studies reported significant effects upon attendance rates; three used prize draws ranging between $1 and $100 [36,46,49], and the others provided fixed vouchers of $2.50 or $5 per session [40,47]. Alessi et al. [45] reported a slightly different result, with significant effects for abstinence but not for attendance using the $1–100 prize draw method. A possible explanation for this result may be that the prizes for abstinence were of a total higher magnitude than those for attendance, suggesting that the lower value reinforcer was not enough of an incentive for patients to attend when compared with the higher value reinforcer. Finally Businelle et al. [42] compared three monetary values of voucher ($60, $95 and $210) provided over three months and reported that only the two higher value vouchers produced significant results.
627
Firm conclusions are difficult to draw due to the range of designs, small sample sizes and the heterogeneous nature of the drug-using population samples. Notwithstanding this, the initial evidence suggests that prize draw CM is at least an equally efficacious alternative to VBRT in adult populations, although two studies did show contradictory results. In VRBT methods, the monetary value of the reinforcer may be an important factor in successful implementation. Conclusions about prize draw techniques in adolescent populations cannot be drawn due to the low number of trials. CM for attendance has been minimally evaluated in methadone-maintained populations.The majority (four out of five studies included) evaluated VBRT in drugdependent pregnant women with a randomised controlled trial (RCT) design. These studies either examined attendance at outpatient day treatment or attendance at counselling or psychiatric services. Two studies reported significant results; Kidorf et al. [43] and Jones et al. [39] provided vouchers of $25 per week and escalating vouchers of up to $70 per day, respectively. Conversely, Jones et al. [38] offered $5 per day for seven days and reported no improvements for attendance. Svikis et al. [37] offered methadonemaintained patients either $0, $1, $5 or $10 per day for seven days and also reported no significant increase in attendance although non-methadone-maintained patients who received the higher magnitude vouchers ($5/$10) did attend more sessions. The single study using the prize draw technique used reinforcers ranging from $2.50 to $100 and reported that patients who exhibited poor attendance at baseline attended significantly more counselling sessions; however, there were no differences in attendance in patients who attended regularly [44]. Overall, the evidence for CM for attendance in methadone-maintained patients is mixed; methadone is a powerful reinforcer in itself [37], and additional incentives that are below a certain magnitude may not improve attendance further. The two studies that reported significant results provided incentives for a longer period of time (12 week and 14 day trial periods), and thus, the total ‘earnable’ amount was higher than in the studies providing incentives for a seven-day period. Conclusions about the effectiveness of prize draw CM for attendance in methadone-maintained patients cannot be drawn as it has only been evaluated in a single trial and not subjected to an RCT design. Research into CM for attendance is at an early stage and a number of questions remain unanswered. Regarding VBRT methods, it is unclear as to what the minimal value of an effective reinforcer is, whether there is a ceiling effect of prize value and whether the total sum of the reinforcer is more important than the © 2014 Australasian Professional Society on Alcohol and other Drugs
VBRT
Reinforcer
142
25
80
20
65
190
125
Jones et al. [38]
Jones et al. [39]
Helmus et al. [40]
Sinha et al. [41]
Businelle et al. [42]
Kidorf et al. [43]
n
Svikis et al. [37]
Study
© 2014 Australasian Professional Society on Alcohol and other Drugs MM
SA
MA
DD (substance use + mental health diagnosis)
MM
MM
MM + non-MM
Substance group
Adult outpatient psychiatric service
Veteran outpatient clinic
Probation-referred adolescents outpatient
Adult outpatient
Pregnant women outpatient
Pregnant women outpatient
Pregnant women outpatient
Clinical population
RCT: Contingent vouchers versus no voucher control, parallel groups
Retrospective contingent vouchers (three conditions) versus TAU
Contingent vouchers versus no voucher control, parallel groups
A B A within subjects
Contingent vouchers versus no voucher control, parallel groups
Contingent vouchers versus no voucher control, parallel groups
Contingent vouchers versus no voucher control, parallel groups
Design
12
12
28 days
12
2
1
1
Intervention duration (weeks)
Total full days outpatient treatment attended
Total full days outpatient treatment attended
% on time counselling attendance
Total MET sessions attended
Total aftercare sessions attended (max two per month)
Total psychiatric sessions attended per month
$5 per day
Escalating; $5 day 1, increasing $5 per day
$2.50 per session attended (max 2 per week)
Escalating; $25 session 1, increasing $10 per session Contract 1: $5/$10/$15 max $60 Contract 2: $5/$10/$15/ $20 max $95 Contract 3: $25/$30/$50 max $210
$25 per week
Outcome
Total full days outpatient treatment attended
Incentive (value)
$0/$1/$5/$10 per day
Table 1. Contingency management for attendance studies
MM group: No effect of incentives Non-MM group: $5/$10 incentive groups attended 1 more day of treatment per week (M = 3.3) compared with $0/$1 groups (M = 2.3)* MM group: Attended 1.5 more treatment days per week (M = 5.2) compared with AT (M = 3.7)* Incentive condition attended 1.5 more treatment days per week (M = 12.1) than control condition (M = 10.6)* Attendance increased by 20% during CM phase compared with baseline** Higher (NS) attendance rate for CM group compared with controls (M = 2.8 vs. 2.3) Participants in the two CM conditions with the highest voucher values (median = 6 for contract 3, median = 5 for contract 2) attended more sessions than TAU (median = 2)*** CM condition attended 3.6 more sessions (mean) than TAU in month 1, 3.5 days (mean) in month 2, 3.3 days (mean) in month 3***
Results
628 J. Milward et al.
50
52
77
103
75
54
336
239
52
Rhodes et al. [44] Study 1 (variable reinforcer)
Rhodes et al. [44] Study 2 (variable vs. fixed)
Petry et al. [36]
Alessi et al. [45]
Ledgerwood et al. [46]
Bride and Humble [47]
Lott and Jencius [48]
Petry et al. [49]
Branson et al. [50]
SA
SA
SA
SA
SA
SA
Cocaine-dependent MM
MM
MM
Adolescent outpatient
Adult community outpatient
Adolescent outpatient
Adult African American women
Adult community outpatient
Adult community outpatient
Adult community outpatient
Inner city adult outpatient
Inner city adult outpatient
Historical control (TAU) versus CM + TAU
TAU and CM versus TAU control, parallel groups
Observational: Pre-/ post-introduction of CM program
Historical control (TAU) versus CM + TAU
Open group design, TAU followed by TAU + CM (and vice versa)
Crossover design: TAU versus CM
TAU and CM versus TAU control, parallel groups
A B A within subjects
A B A within subjects
52
12
52
4
16
12
12
8
8
% of sessions attended (weekly)
Total days in treatment
Prize value of $1/$20/ $80–$100
Gift certificates $5 for perfect weekly attendance $0/$1/$5/$15
Social reinforcer/$1/$5/$20
$1/$20/$100
% treatment days attended
Prize value of $1/$20/$100
% sessions attended
Total group counselling sessions attended
(i) % days attended (ii) Mean length of stay
Weekly group attendance
% on time counselling attendance
$2.50/$5/$10/$25/$100 versus fixed rate $3.25
Small ($1), medium (e.g. Walkman), jumbo (e.g. television)
% on time counselling attendance
$2.50/$5/$10/$25/$100
23% mean increase of attendance for poor attenders in CM condition (attended
