OUP user menu

Drug misuse treatment services in Scotland: predicting outcomes

Zoe Slote Morris, Maria Gannon
DOI: http://dx.doi.org/10.1093/intqhc/mzn019 271-276 First published online: 20 May 2008


Objective To investigate which aspects of treatment satisfaction are the best predictors of improved health, improved mental health and achievement of abstinence in drug misuse treatment services.

Design Data were collected as part of the Drug Outcome Research in Scotland study, a prospective cohort study designed to evaluate drug misuse treatment provided in Scotland. Data were collected using a structured interview. Participants were recruited between 1 October 2001 and 30 June 2002. Follow-up interviews were carried out ∼8 months later. Logistic regression analysis is used to explore client satisfaction with treatment on outcomes, using the Treatment Perceptions Questionnaire (discussed in Marsden et al., Assessing client satisfaction with treatment for substance use problems and the development of the Treatment Perceptions Questionnaire (TPQ). Addict Res 2000;8:455–70).

Setting Prison, residential and community facilities.

Participants A total of 841 drug users starting a new episode of drug treatment in Scotland in 2000–01.

Interventions Methadone, substitute drugs other than methadone, residential rehabilitation, residential detoxification and non-clinical.

Main outcome measures Reported improvements in physical health, mental health and abstinence.

Results Client satisfaction predicted positive outcomes, independent of treatment setting. Predicting abstinence and improved physical and mental health were the items: ‘I have received the help that I was looking for’ and ‘The staff have helped to motivate me to sort out my problems’.

Conclusions Feeling that treatment is appropriate, finding staff motivating, and having enough time to sort out problems are important aspects of satisfaction with treatment among users of drug treatment services who achieved positive treatment outcomes. Services should seek to provide more individualized services based on understanding of individual client needs. This may require longer treatment periods and greater client involvement.

  • addiction
  • patient-centred
  • patient satisfaction
  • quality


An increased focus on client satisfaction with public services, including health care, is an international issue. Client and patient satisfaction is used to measure service quality [1], in particular patient-centeredness [2]. Satisfaction data allow services to be held accountable [3] and, if the feedback is used constructively, can contribute to the improvement of services [4, 5]. This paper focuses on client satisfaction in an addiction setting. Reflecting practice in UK addiction services, service users are referred to as clients rather than patients throughout the paper.

Client satisfaction is particularly important to quality provision of addiction services, because it has been found to predict retention in treatment [3, 6, 7] and better outcomes for treatment [7, 813]. Aspects of service provision relevant to client satisfaction in addiction settings include differences across treatments modalities and settings [9, 14], programme policies and regulations, the quality of staff and the programme, and morale among staff and patients [15], achieving personal treatment goals and being allowed time to deal with problems [16]. These issues have particular importance to satisfaction where clients have limited choice over treatment options [17].

This paper explores client satisfaction with addiction treatment services in Scotland. Previous analysis of data from a national Scottish sample of drug treatment users [7] found that greater satisfaction was associated with positive treatment outcomes. The purpose of this paper is to investigate which particular aspects of treatment are the best predictors of outcomes in order to understand more precisely what it is about satisfaction that predicts positive outcomes for clients, and to consider the implications for service provision and policy.

The analysis focuses on three health-related outcome variables: physical health, mental health and abstinence. International studies usually show that illicit drug use is associated with poor health [18, 19], including vulnerability to infectious disease, collapsed veins, malnutrition and poor sleep [1, 2], as well as increased prevalence of psychiatric problems [20, 21]. Addiction treatment is typically located within health care services and seeks to improve physical and mental health outcomes for clients. Achievement of abstinence was also included because it was the declared aim of treatment for most clients in the Scottish study on which the analysis is based [22] and therefore salient to clients.



Data were collected as part of the Drug Outcome Research in Scotland (DORIS) study, a prospective cohort study designed to evaluate drug misuse treatment provided in Scotland. The study was conducted with the full approval of the Scottish Multi Centre Research Ethics Committee. Over 1000 clients (n = 1033) were recruited from 33 agencies across Scotland. The agency from which the respondent was recruited to the project is referred to as the ‘index agency’. Entry criteria were: having a primary dependence on illicit drugs, starting a new episode of drug treatment and being able to give contact details for follow-up. Participants were recruited between 1 October 2001 and 30 June 2002. Follow-up interviews were conducted ∼8 months after the initial interview.

A structured one-to-one interview, conducted by trained interviewers employed by the Centre for Drug Misuse Research at Glasgow University, was used to collect baseline personal data, information on treatment modality (methadone, substitute drugs other than methadone, residential rehabilitation, residential detoxification and non-clinical interventions such as one-to-one counselling and group work) and agency setting (prison, residential and community).

At second interview ∼8 months later, interviewees were asked ‘Has [index agency] helped you to achieve any of the following?’, including ‘better physical health’ and ‘improved mental emotional health’. Respondents were also asked ‘Do you think that coming/going to [index agency] has helped you change your drug use in any of following ways?’, including ‘becoming abstinent/drug free’. Responses to both sets of questions were coded 1 for a positive response, and 0 for no response. Outcome measures used in the analysis are, therefore, reported data.

Assessment of satisfaction

The Treatment Perceptions Questionnaire (TPQ) was also administered at second interview in order to assess client satisfaction with their index agency. The TPQ was developed by Marsden et al. [13] to measure quality in an addictions setting. Scores are derived from 10 items relating to perceptions about staff and programme design. Each item is scored on a five-point scale (disagree strongly–agree strongly; 0–4). Scores on the negative items are recoded to measure positive evaluations on all measures. Higher scores indicate greater satisfaction. For statements that are worded negatively, a higher score indicates greater disagreement with the statement therefore.

Previous analysis of the TPQ found the index to be valid and reliable in the Scottish drug misuse treatment context [7]. Higher scores were associated with positive outcomes including retention in treatment, abstinence, better health and better mental health [7]. The internal consistency of the TPQ was good as measured by Cronbach's Alpha (α = 0.83).


This analysis focuses on those respondents who were interviewed at initial intake and ∼8 months later. Data from the first two time points were selected for analysis, because these data represent the closest time points between treatment and evaluation and are likely to represent the most accurate, and therefore most valid, assessment of clients. There have been few major changes to drug misuse treatment services in Scotland between 2003 and 2007, and the data can be considered relevant to contemporary services.

There was an overall follow-up rate of 85% (n = 878), but this differed by setting. A smaller fraction of respondents in the prison treatment setting were re-interviewed: 81% (n = 361) compared with 89% (n = 203) of those in residential settings and 88% (n = 314) of respondents from the community treatment setting.

Of the total sample (n = 1033), 26 were excluded from the analysis because they had been recruited from needle exchange services, the ethos of which the TPQ is not designed to capture. Of the 859 respondents followed-up at the second time point, 18 cases were excluded because they were missing responses to an item in the TPQ. This left a total of 841 individuals included in the analysis, with no missing values.

Logistic regression was used to investigate the relationship between treatment setting and items of the TPQ with positive outcomes. Standardized coefficients (odds ratios) were used to allow comparisons of the magnitude of the effects of different items.

Agency setting was also included as a control variable in the analysis of the outcome variables. Different settings have different aims and ethos. For example, treatment offered in the prison setting is typically shorter and more limited [16]. In addition, earlier analysis of the DORIS data found significant differences between prison and non-prison clients [7, 23]. Analysis was undertaken using SPSS for Windows Version 14.


Sample characteristics

Of the 841 individuals included in this analysis, age at intake ranged from 16 to 51, with a mean age of 27 years. Five hundred and eighty (69%) of the respondents were male. Of the sample, 835 (99%) of all respondents were ‘white’. Three hundred and fifty-four (42%) respondents were recruited from prison units, 199 (24%) recruited from the residential setting and 288 (34%) from the community setting. In terms of treatment modality, 234 (28%) respondents were starting methadone treatment, 229 (27%) another substitute drug, 107 (13%) were starting a residential rehabilitation treatment, 105 (12%) a residential detoxification, and a further 166 (20%) were in one-to-one counselling or group work.

At second interview, of all individuals followed up 396 (47%) reported that contact with their index agency had helped them achieve reported improved health, 331 (39%) better mental health and 230 (27%) a period of abstinence.

The summary score was computed in earlier analysis [7]. TPQ score was 20.1 (SD = 7.6) on a 0–40 scale. A one-way ANOVA test indicated significant differences between TPQ scores for clients in different settings (F(2838) = 72.7, P ≤ 0.01) [7]. Those in the prison setting reported the lowest levels of satisfaction (mean = 16.8, SD = 6.6), followed by those in residential settings (mean = 21.5, SD = 7.1), and those in community setting the highest levels of satisfaction (mean = 23.2, SD = 7.3). This analysis also showed that higher TPQ scores predicted higher odds of interviewees reporting better health (β = 1.12; P ≤ 0.01), better mental and emotional health (β = 1.12; P ≤ 0.01), and having had a period of abstinence (β = 1.10; P ≤ 0.01) as a result of contact with their index agency.

Table 1 shows the results of logistic regression analysis using the TPQ items as predictors of having achieved better physical health, better mental health and a period of abstinence between interviews. The univariate analysis shows the effect of each predictor variable on the outcome independent of the other variables. The multivariate model controls for treatment setting as this influences the type and length of treatment offered [17] and the opportunity for abstinence.

View this table:
Table 1.

Odds ratios (OR) and 95% confidence intervals (CI) for achievement of improvement to physical, mental health and abstinence by setting and TPQ items (units of OR, one point on a five point scale)

Outcome variablesPhysical healthMental healthAbstinence
Univariate modelsMultivariate modelUnivariate modelsMultivariate modelUnivariate modelsMultivriate model
Independent variables
 Treatment setting (ref = community)
  Prison setting0.7 (0.5–1.0)1.7 (1.2–2.6)0.6 (0.4–0.8)1.0 (0.7–1.5)1.0 (0.7–1.4)2.0 (1.3–3.2)
  Residential setting1.2 (0.8–1.7)1.1 (0.7–1.7)1.4 (0.9–2.0)1.4 (0.9–2.2)4.3 (2.9–6.3)5.4 (3.3–8.8)
 TPQ items
  The staff have not always understood the kind of help I want.1.6 (1.4–1.8)1.2 (1.0–1.4)1.4 (1.3–1.6)1.0 (0.9–1.2)1.5 (1.4–1.8)1.2 (1.0–1.4)
  I have been well informed about decisions made about my treatment (reverse coded).1.6 (1.4–1.8)1.2 (1.0–1.4)1.5 (1.3–1.8)1.1 (0.9–1.3)1.6 (1.4–1.9)1.3 (1.0–1.5)
  The staff and I have different ideas about my treatment objectives.1.3 (1.2–1.5)0.9 (0.8–1.1)1.3 (1.2–1.5)1.0 (0.9–1.2)1.3 (1.1–1.5)1.0 (0.8–1.2)
  There has always been a member of staff available when I have wanted to talk (reverse coded).1.4 (1.3–1.6)0.9 (0.8–1.1)1.4 (1.3–1.6)1.0 (0.9–1.1)1.4 (1.2–1.6).09 (0.8–1.1)
  The staff have helped motivate me to sort out my problems (reverse coded).1.8 (1.6–2.1)1.4 (1.2–1.7)1.7 (1.5–2.0)1.3 (1.1–1.6)1.7 (1.5–2.0)1.3 (1.1–1.6)
  I have not liked all of the treatment sessions I have attended.1.2 (1.0–1.4)1.1 (0.9–1.3)1.2 (1.1–1.4)1.2 (1.0–1.4)1.0 (0.9–1.2)1.1 (0.9–1.4)
  I have not had enough time to sort out my problems.1.4 (1.2–1.6)1.1 (1.0–1.3)1.4 (1.3–1.6)1.2 (1.1–1.4)1.3 (1.1–1.5)1.2 (1.0–1.3)
  I think the staff have been good at their jobs (reverse coded).1.7 (1.5–2.0)1.0 (0.9–1.3)1.6 (1.4–1.8)0.9 (0.7–1.1)1.6 (1.4–1.9)1.0 (0.8–1.3)
  I have received the help that I was looking for (reverse coded).1.9 (1.7–2.2)1.5 (1.3–1.8)1.9 (1.6–2.1)1.5 (1.2–1.8)1.8 (1.5–2.0)1.3 (1.1–1.6)
  I have not liked some of the treatment rules or regulations.1.1 (1.0–1.3)0.9 (0.8–1.0)1.2 (1.0–1.3)0.9 (0.8–1.1)1.0 (0.9–1.2)0.9 (0.8–1.1)

Physical health

Columns two and three of Table 1 show the odds of reporting better physical health based on treatment setting and satisfaction. In the univariate analysis, treatment setting was not associated with reported improvements in health. All items of the TPQ were positive predictors of better health. In the multivariate model, which uses all the TPQ items and controls for setting, those treated in the prison setting report improved health outcomes compared with community clients. Three TPQ items also predicted better health. In order of magnitude, these were the client having received the help they were looking for, finding the staff motivating, and having had time to sort out problems.

Mental health

Columns four and five of Table 1 show the results for mental health. In the univariate analysis, compared with clients being treated in the community setting, clients who received their index treatment in prison units were less likely to report improved mental health. All items of the TPQ were positive predictors of better health. In the multivariate model, setting was not a significant predictor of improved mental health when satisfaction scores were taken into account. Three TPQ items predicted perceived better health. In order of magnitude, these were the client having received the help they were looking for, finding the staff motivating, and having had time to sort out problems.

Achievement of abstinence

Columns six and seven of Table 1 show the results of the logistic regression analysis on abstinence. In the univariate analysis, compared with community setting, residential clients had significantly higher odds of abstinence, whereas prison unit clients did not. Of the other TPQ items, only having liked all the treatment sessions they attended and having liked the programme rules and regulations were not positive predictors of abstinence. The multivariate model accounts for different opportunities to be abstinent in different treatment settings. Agency type was significant, with those in prison and residential settings more likely to report having achieved a period of abstinence than those in the community setting, controlling for the satisfaction measures. Of these, significant predictors were having been kept informed about treatment, feeling that the staff were motivating, and having received the help they sort. All were positive, with higher satisfaction scores predicting higher odds of abstinence.


Taking account of different treatment settings, this analysis of client evaluations of satisfaction with Scottish drug misuse treatment services suggests that particular aspects of service predict positive outcomes. Responsive treatment seems important: ‘I have received the help that I was looking for’ was the most strongly predictive satisfaction variable of all three outcomes.

Given that clients contact these services to get help, it is perhaps odd that this item did not render other aspects of treatment redundant. It appears, however, that clients consider a range of factors in their assessments, possibly separating service and technical quality [24]. Cheng's and colleagues' finding [24]—that clients recommend services to others based on perceived technical competence rather than declared satisfaction—has important implications for policy, given that a key policy aim is to ‘increase the participation of problem drug users…in drug treatment programmes’ [25]. Personal recommendations may be especially useful when dealing with socially marginalized clients.

The staff's ability to motivate clients was also important to predicting positive outcomes. Brown [26] argues that the main aim of drug treatment should be to develop motivation and engagement, especially among clients who are negative or ambivalent. From these data, it is not possible to say how they can or should do this, but they do suggest a strong need to explore issues of motivation further in future research.

Also important to both health outcomes was that the clients need to be given sufficient time in treatment. We have argued elsewhere [27] that many treatments are time-bound in a way that is not suitable for clients facing myriad issues. Moreover, treatments fall short of durations ‘critical’ to making them effective [28]. For service planners, this suggests a need to rethink the length of programmes and thus avoid discharging clients from treatment before meeting their goals. This has clear resource implications but may help reduce return visits. Further analysis of this issue would seem prudent.

While two aspects of treatment satisfaction were significant predictors of all the outcomes, a wider pattern did not occur. As already discussed, time was important to both health outcomes, but not abstinence. Conversely, being well informed about treatment decisions was a significant predictor of abstinence only. The reason for this difference is not obvious, and perhaps reflects a need for more insight into service use. Neale [29], for example, found that Scottish drug users approached different agencies for different needs, and valued different characteristics in each. Understanding how clients use different agencies would be a fruitful further line of investigation in the design and provision of cost-effective care.

There are some limitations to the analysis which need to be recognized. Outcomes measured were global, short-term, perceived and not actual. This means, for example, that the data do not indicate whether clients actually experienced better health or they just thought they did. The DORIS survey, from which the data are derived, is not linked with actual medical records or to other treatment records for reasons of confidentiality and logistics. While behavioural measures would provide more robust conclusions, a number of other studies rely on reported outcomes measures [8], including analysis of the SF-36 [18, 30]. Client perceptions may also reflect the client experience which may have more meaning to clients than objective outcome measures.

A further potential weakness is that since people who are positive about one thing may also be positive about another. Isolating the ‘effect’ from the client's perception of a particular agency presents a challenge. Asking clients to evaluate their treatment agency retrospectively may introduce bias because clients who achieved positive outcomes may look back on their treatment experience as positive irrespective of whether they found it good at the time, and vice versa. Furthermore, clients' evaluation of their index agency could reflect more general evaluation of treatment services where individuals have received treatment from other agencies. These issues reinforce the need to understand better client use of the system.

Despite these limitations, however, the data are plausible. For example, agency type predicts achievement of abstinence as would be expected. The analysis also supports previous findings, such as that of satisfaction being associated with treatment modality and predicting outcomes. These findings can, therefore, provide input to providers to develop their services. Clients' feeling that treatment is appropriate, finding staff motivating, and having enough time to sort out problems are important aspects of satisfaction with treatment among users of drug treatment services who achieved positive treatment outcomes. This may signal a need for drug misuse treatment services to provide more individualized services based on good understanding of individual client needs, including longer periods of treatment.

More generally, the study suggests scope for greater involvement of users of drug treatment services in Scotland, where current provision is generally provider-driven with limited user choice [17]. This has the potential to help staff understand more precisely how they can motivate clients, why services are effective, and how to support more client-centred care. It could help bring drug services more in line with developments in other areas of health care and the wider trends in health care policy [1, 2], including choice and patient participation within a broader quality agenda. This remains relatively underdeveloped in drug treatment services [29].


The Drug Outcome Research in Scotland study (DORIS) is funded by the Robertson Trust and supported by the Scottish Executive.


The views expressed in this paper are those of the authors and should not be attributed to either body. Professor Neil McKeganey is the grant holder of the study. Thanks are due to the members of the DORIS team, including Professor Joanne Neale, Professor Michael Bloor and all the participants of the study.


View Abstract