Evidence
Commentary on selected new evidence
Commentary on selected new evidence
With advice from topic experts we selected 1 study for further commentary.
Secondary care mental health services – Psychological and psychosocial interventions
We selected the randomised controlled trial by Graham et al. (2016) for a full commentary because this study may provide an insight into an area of psychological treatment with a current paucity of evidence.
What the guideline recommends
NICE guideline CG120 recommends (1.4.18) that evidence-based treatments are offered to treat both psychosis and coexisting substance misuse in adults and young people. The guideline also refers (1.4.19–1.4.20) to the related guidelines for the treatment of psychosis and substance misuse which recommend the use of brief motivational interventions for this population.
Methods
Graham et al. (2016) conducted a pilot randomised controlled trial of Brief Integrated Motivational Intervention (BIMI) in people aged at least 18 years and diagnosed with schizophrenia-related, bipolar or recurrent depressive disorders. Patients newly admitted to inpatient mental health units were recruited if they were users of community mental health services and misused substances over the past month. Substance misuse was classified as a minimum score of 3 over the previous 3 months on the Clinicians Alcohol Use Scale (CAUS) or the Clinicians Drug Use Scale (CDUS).
Eligible participants completed outcome assessments at baseline prior to randomisation and at 3 month follow-up. Outcomes were assessed at 3 months by independent assessors blinded to treatment allocation. Participants were randomly allocated to receive either BIMI with treatment as usual or treatment as usual alone and were analysed by intention to treat. Treatment as usual was defined as consisting of assessment, monitoring and stabilisation of mental state, and medication.
For the primary outcome, the Substance Abuse Treatment scale (SATs) measured changes in treatment engagement. Secondary outcomes consisted of readiness to change substance use and the cost-effectiveness of the BIMI. The Stages of Change Readiness And Treatment Eagerness Scale and the Importance-Confidence Ruler were used to measure readiness and motivation to change. Cost-effectiveness was evaluated from total service costs and quality-adjusted life years ascertained with the EQ-5D measure. Substance use was measured using the Maudsley Addiction Profile, the Severity of Dependence Scale and the Alcohol Use Disorders Identification Test. Psychological functioning was measured using the Recovery Style Questionnaire, Insight Scale and the Hospital Anxiety and Depression Scale. Qualitative interviews at 3 months with participants and therapists were conducted to ascertain satisfaction and feasibility of the intervention.
Results
A total of 59 participants were randomised to receive BIMI with treatment as usual (n=30) or treatment as usual alone (n=29). However, 9 participants in the BIMI group did not receive the intervention and 1 participant in the treatment as usual group was excluded from the analysis.
For the primary outcome as measured with the SAT scale, the BIMI group was associated with a statistically significant increase in treatment engagement compared with treatment as usual (Odds Ratio [OR] =1.63, 95% Confidence Interval [CI] 1.01 to 2.65, p=0.047).
For the secondary outcome, statistical analysis of readiness to change was not conducted due to missing data and the inability to combine substance use data. However, the study notes that at follow-up both the BIMI and treatment as usual groups continued to report low readiness to change.
Motivation to change measured by importance and confidence scores yielded similar results at baseline and follow-up for both groups. Although, it is not specified in the study whether higher or lower scores indicate better outcomes. Importance scores at baseline (BIMI=6.77, Standard Deviation [SD] =3.23; treatment as usual=7.19, SD=3.58) and scores at follow-up (BIMI=7.08, SD=3.74; treatment as usual=6.89, SD=3.30). Confidence scores at baseline (BIMI=8.12, SD=2.30; treatment as usual=7.50, SD=2.94) and scores at follow-up (BIMI=8.15, SD=2.19; treatment as usual=8.02, SD=2.83).
Substance use in the previous 30 days was found to reduce at follow-up compared to baseline for both groups. However, the difference in reduction between groups was not statistically significant (Relative risk=1.02, 95% CI 0.82 to 1.26, p=0.85). Full data was available for 50 participants to be included in this analysis.
Psychological functioning was not found to be statistically significantly different between groups with HADS anxiety scores (−0.80, 95% CI 3.93 to 2.34, p=0.611) or HADS depression scores (−1.89, 95% CI −4.51 to 0.74, p=0.156). However, the 95% CI reported for HADS anxiety scores does not appear to be accurate and is not explained further in the study. Although the style of recovery of participants was measured, missing data prevented any detailed analysis of differences between groups. Also, no significant differences in participants' insight into mental health problems were found between groups.
The cost-effectiveness analysis found that at follow-up the BIMI group (£18,651) was associated with an increased total mean cost of services compared with the treatment as usual group (£15,698). This equated to an average of £3,279 increased cost for the BIMI group compared to the treatment as usual group (95% CI −£3,933 to £10,876). Quality-adjusted life years for both groups were found to be similar however no statistical data is provided.
Qualitative data from interviews found the intervention to be feasible and acceptable to BIMI therapists and 21 participants in the BIMI group.
Strengths and limitations
Strengths
The population in this study is largely relevant to NICE guideline CG120 with the inclusion of schizophrenia and bipolar disorder patients. The inpatient setting for the study is also relevant and an under-researched subgroup considering the paucity of evidence for this population.
The study methodology has strengths with the use of randomisation, allocation concealment and blinding. These methods go some way to reduce the risk of bias. The outcomes measured are relevant to NICE guideline CG120 and provide interest to clinical practice for this population.
Limitations
Although the population in the study is largely relevant, there is some uncertainty regarding the number of included participants with a diagnosis of psychosis. The sample size, as this was a pilot study, is relatively small with 59 included participants of which only 21 received the BIMI treatment and 50 retained in the final follow-up. This sample size is also below the power calculation figure of 68 to find a clinically important difference.
The study used blinding at baseline however the authors note that bias may have been introduced with outcome assessors being aware of treatment allocation at follow-up. As primary clinicians provided some of the outcome ratings, further bias in reporting may have been introduced through the influence of a therapeutic alliance.
Although all outcomes are addressed in the results, the extent of missing data prevented many statistical analyses from being conducted. There is also a lack of detail in the results reported for cost-effectiveness, particularly regarding quality-adjusted life years. The outcomes are reported at a 3-month follow-up period which may be considered short for this complex population and presentation.
There are a number of confounding factors unaccounted for within the study. The BIMI treatment is not relatively brief which may impact on both costs and the primary outcome of treatment engagement. Inherently, participants in the BIMI group are receiving increased contact with healthcare professionals and services which may influence outcome reporting at follow-up. The training and potentially increased availability of specialist staff on the inpatient ward may have played a role in increased engagement beyond the effect of the intervention. There is also a possibility that the increased communication between the intervention therapists and community care co-ordinators promoted engagement either alone or at least in conjunction with the intervention.
Impact on guideline
The results of this study are unlikely to impact on NICE guideline CG120 given the limitations described. Primarily, as this is a pilot feasibility study the sample size and methodology were likely to have limitations as described above. It is also clear that the outcomes measured, although relevant, are lacking in sufficient detail to provide conclusive evidence for the intervention. For further relevance to NICE guideline CG120, the study could have included outcomes related to symptom severity. The study offers some support, albeit very limited given the methodological problems, for the recommendations on providing psychological and psychosocial support to this population. However, taken alone this study provides an interesting and potentially feasible point from which to conduct more research in this population.
This page was last updated: