HSOA Journal of Addiction & Addictive Disorders Improving Retention in Opioid Treatment

Problematic substance use significantly contributes to morbid ity and mortality and affects individuals and communities system -atically. Indeed, responses to such issues require a multidimen sional, evidence informed approach. Medical, psychological and social responses have been identified as those with most effica cies within the literature. Illicit opiate use brings with it perhaps a more severe symptomatology when we consider the physiological phenomenon. Therefore, treatment needs to be multi-dimensional with both psychosocial and medical interventions. However, treat ment attrition rates are very high with this cohort of people. Ap proaches to identify those at risk of dropout and poor outcomes have been established within the wider psychiatric and psycholog ical domains. Client feedback entails soliciting data on outcome and process of care and utilising the information to adapt and tai lor service deliver on a session to session (unit of care) basis. The present review paper seeks to build on previous research by extending these methods to the treatment of opioids in an Irish context.


Evidence Informed Treatments
The literature on addiction interventions provides us with robust research evidence of the efficacy of several approaches [8,9]. Moreover, evidence suggests [10], that psychosocial interventions have efficacy across the spectrum; including, brief and harm reduction interventions, motivational strategies and abstinence/rehabilitative based approaches. However, the gap between research and practice in naturalistic settings remains [11,12]. While psycho-social approaches work in various ways, medical interventions are also utilised for the physiological aspects of opioid addiction.
Mayet et al. [13], conducted a Cochrane Review and suggested that there was inadequate evidence during 2004 to support the efficacy of psychosocial interventions as a standalone treatment for opioid use. This is not surprising when we consider the wider evidence base exploring the effectiveness of medically assisted treatment as reported in several Cochrane reviews [14][15][16][17][18][19]]. Yet, in her Irish data set research, Carnew [20], suggests that 83% of her study cohort attended for psychosocial interventions alone. Why this was the case has not been identified, however, stigmatisation around methadone use and how the protocols have historically been delivered could be one issue.
While methadone has shown to help retain people in treatment and reduce criminality two Irish studies provide for worrying research. Maycock et al. [21], suggest that the quality of life of many people in their study on opiate replacement treatment may not have improved. Indeed, in another Irish study found that clients voices were lost and feedback not listened to. 'Service users described negative aspects centring on the patients lack of choices, humiliating experiences in consuming methadone in a public space, difficulties complying with punitive contracts and urine screening and engaging with uncaring service providers. This sentiment was echoed by Maycock et al.
[21], study "a prominent feature of the treatment experience was a perception that, as clients or patients, they had no say in their treatment. By and large, participants felt controlled rather than in control with little evidence of them feeling able or entitled to share their experiences or to articulate any aspirations or needs related to their treatment".
While the preceding quotes are more specific to the medical model mode of treatment, practitioners of psychosocial interventions also impact upon attrition rates through their relational interactions. Much has been discussed regarding the characteristics that can impact upon the treatment endeavour, especially related to the drop out phenomena.
However, other correlational factors can impede retention in treatment; for example, as many as 30% of clients with poly use drop out of treatment, alcohol-substance. Mode and modality can impact these statistics differentially, with detox drop out ranging from 21.5%-43% , conducted a systematic review that included 122 studies with 199,000 participants exploring drop out from treatment; they concluded that young age, mental health difficulties and the therapeutic alliance were key variables of drop out.
Research on the Therapeutic Alliance (TA) is well established, in fact, it is probably one of the most studied variables in therapy and one of the strongest predictors of therapy outcome [27-33].
In a key study of the therapeutic alliance, Baldwin et al. [34], elucidates the power of this construct by suggesting that 97% of the difference in client outcomes between therapists can be attributed to the alliance. Moreover, it was the clients rating of the alliance that was the important factor and notably, the clients' contribution was not a variable for outcomes. Said another way, the difference in therapist outcomes is mainly due to their ability to build an alliance with different clients, who rate that alliance strongly, while clients' contribution has little in the way of impact on alliance contribution.
Conversely, clients of therapists who cultivate weaker alliances tend to drop out at higher rates and experience poorer outcomes [35,36]. This research is consistent across general or substance use therapy. Thus, monitoring of the therapeutic alliance would seem necessary. Notably, the interventionists ability to establish a bond, collaborative agreement on the task, goals and methods to be used in treatment, is essential, as outlined by conceptualisation of the therapeutic alliance. Heinonen et al. [37], contend that "the ability to professionally relate to clients with empathy, warmth, positive regard, clear and positive communication and the capacity to take critical feedback predicted better outcomes across all levels of therapist expertise"

Feedback Informed Approaches
Routine Outcome Monitoring (ROM), also referred to as feedback informed or systemic feedback, is gathering support over the last decade in both academic and practice settings. Vast literature on this approach postulates that intentionally and formally soliciting live feedback from clients on a session by session basis can improve therapy outcomes, reduce dropout rates and identify those at risk for deterioration [38][39][40].
This method of assessing clients in an empirical and standardised manner may be needed as research suggests that practitioners do not adequately predict the deterioration of clients, those at risk of dropping out or at risk of null outcomes when assessing these issues informally [41].
Although routine outcome monitoring and feedback informed treatment are widely supported in the research as pan-theoretical practice-based evidence approaches, across broader psychiatric and psychological environments, specificity is very often the criteria sought out by policy makers and commissioning bodies. Therefore, it is incumbent on the review to highlight studies of client feedback approaches within this specific arena.
However, there is a relative dearth of research specifically addressing this question in addiction settings in general and opioid use specifically. McCaul and Svikis [42], showed that providing clinicians with feedback on client attendance subsequently improved retention rates. Schuman et al. [43], conducted research into the use of feedback with soldiers attending group therapy (n=263) for substance use. Compared to the control group which consisted of group therapy treatment as usual, the experimental group showed moderate effect size gains and attended more sessions. Crits-Christoph et a.l [44], showed that for off-track patients, "feedback compared with no feedback led to significant linear reductions in alcohol use throughout treatment and also in OQ-45 total scores and drug use from the point of the second feedback instrument to Session 12", Brorson et al. [45], failed to predict those at risk of dropout within a small sample utilising the OQ -Analyst, however, notably, measures of alliance were not used in conjunction with this outcome measurement tool.
Likewise, Crits-Christoph et al. [46], found no effects of feedback, however, it was identified that providing aggregated scores at the group level as opposed to individual clients left practitioners unmotivated to use this feedback. This would seem to reflect the issues in the Delargy et al. study where informal feedback was not listened to or acted on. Raes et al. found that the group that received assessment and feedback were significantly more likely to remain in treatment at and beyond 8 and 12 weeks. This is an important finding considering the attrition rates that we currently exhibit in services. While these studies show some good and promising results, they equally demonstrate limited methodologies, such as non-randomisations and lack of appropriate outcome measures.
However, a strong aspect of these studies is the diversity of ethnicity, age, substance use and socioeconomic status, in addition to the environments the research were carried out in (drug courts, community settings, inpatient).Thus, the heterogeneous composition lends credence to the findings and it is not overly ambitious to postulate that the use of client feedback in substance use will likely mirror that of the wider behavioural health care arena, where it has been given evidence based recognition by American Psychological Association (APA).
In a study exploring the acceptability of feedback approaches in substance use treatment, both clients and practitioners highlighted its value and usefulness for treatment planning [ suggest patient-therapist goal consensus and collaboration enhance psychotherapy outcome" At the same time, two meta-analyses Swift et al. [49] and Lindhiem et al. [50], demonstrated that preference accommodation within treatment produces better client outcomes. Hence, the call by Carlier and Van Eeden [51], who suggest that training should be provided to clinicians in administration, interpretation and using feedback to discuss treatment goals and to ignite a culture of feedback within service provision. Baldwin et al.
[34], expand on this point from their alliance study by concluding that "clinical implications include therapists monitoring their contribution to the alliance, clinics providing feedback to therapists about their alliances and therapists receiving training to develop and maintain strong alliances" Several psychometrically sound instruments with normative data are available in order to work within a feedback informed approach [52][53][54]. Generally, feedback is solicited based on measures assessing the quality of the therapeutic alliance and outcome questionnaires based on specific or global levels of distress. Hatfield and Ogles conducted a national survey of psychologists and found that uptake of such instruments was limited due to perceived barriers such as; time and money and practicalities of their in-session brevity. This may be true for longer more tiresome instruments such as the Outcome Questionnaire-45 or the Symptom Checklist-90. However, shorter tools such as the Outcome Rating Scale (ORS) or the Clinical Outcomes in Routine Evaluation (CORE10) have utility without losing much validity or reliability; more, their in-session brevity and utility are high. However, pre-defined outcome measures, or often politicalised, especially in the substance use domain where commissioning bodies mandate their use. In a scoping review of the outcome measurement literature in the substance use field Alves, Sales and Ashworth [55], found 42 measure covering 54 domains. While the measures generally covered the important topics, "we found that several topics of relevance for patients were not covered by any of the measures included in our study".
Thus, the need for measures that can capture subjective and important issues for clients are needed. However, it is also integral that such instruments are reliable, valid and have clinical utility and brevity. The Outcome Rating Scale (ORS) is used in conjunction with the Session Rating Scale (SRS) as the main protocols in Feedback Informed Treatment (FIT). Both tools are ultra-brief 4 item measures, capturing the process and outcome of care. Importantly, both measures meet the criteria above, as they have strong psychometric properties, and are brief subjective measures. In relation to the ORS, the 4 items (individual, interpersonal, social and overall) are a reflection of general-well being as identified by the client's subjective experience as opposed to pre-determined domains that providers and commissioning bodies put forward.
Although measurement of outcomes is utilised in some services, these are generally used to evidence outcomes for commissioning bodies as opposed to dialogue tools for therapeutic conversations and are used at pre-post destinations. At the same time, excellent research is carried out in naturalistic settings assessing the needs of clients within services to influence policy and practice, yet some limitations are evident here to. Current processes often consist of longitudinal, retrospective programme/interventions or process evaluations presented thematically that identify service user's needs and experiences [21, [56][57][58][59] and are thus extrapolated to future services users or policy positions. Although providing important and rich research in different contexts, such methodologies do not capture live data which can be used to adapt the treatment approach in real time, based on client preferences and needs as they relate to both process and outcomes of care [60][61][62][63][64][65][66][67][68][69][70][71][72][73][74].

Discussion
The treatment of opioid dependency represents a particular challenge for treatment practitioners, providers and policy makers. Although we have many different interventions across the broader psychosocial and medical domains, keeping this cohort of people engaged in the treatment process long enough to benefit from such interventions is problematic. Much research has been conducted on the client factors that correlate with early termination from the treatment endeavour. However, less has been explicitly carried out on treatment provider factors that impact this issue. Yet, it would seem that the therapeutic alliance plays a key role in mediating client satisfaction and ultimately positive outcomes.
The research literature presents a compelling case for involving those in our care to have a voice in their treatment approach. Establishing client preferences through the use of outcome measures is robustly supported within the extent literature. Within the Irish context it is clear that listening to the voices of clients is necessary, not just because it serves a clinical purpose by improving outcomes. But, also, at a human and social justice level, clients should be able to feel empowered and voice their concerns, preferences, wishes and needs. Feedback Informed Treatment (FIT) is one method well positioned to help those working with clients to articulate their needs in a formalised manner and use the resulting information to address these needs in real time. 14. Clark N, Lintzeris N,