Do levels of evidence affect breadth of service? An audit of MATRIX therapies in a service for adults with intellectual disabilities.

Pateraki Eleni, Kenneth Macmahon

Research output: Contribution to conferencePosterpeer-review


Objectives: The MATRIX is a Scottish Government/NHS Education Scotland document that aims to provide Health Boards with information on strategic planning for psychological therapies, governance of delivery of these therapies and a summary of empirically validated psychological therapies. It is aimed at aiding clinicians and service managers focus on the timely delivery of evidence-based psychological therapies to service users, including those with intellectual disabilities (ID). Although considered a positive development, concern has been raised with the evidence level required for the MATRIX. Randomised controlled trials were considered the ‘gold’ standard, with less well controlled studies comprising a lower level of evidence. Single-case studies were excluded. Therapies (for a population with ID) were only considered to meet an appropriate standard for five psychological issues (anxiety, depression, challenging behaviour, anger and psychosis). The reliance upon a single diagnostic category may conflict with assessment and formulation and may not reflect the range of interventions (such as indirect work with carers) used in psychological services for adults with ID. This may lead to training and resources being invested only on psychological therapies included in the MATRIX, potentially limiting the range of psychological interventions available. This audit aimed to investigate the effect of adoption of a rigid interpretation of a MATRIX based approach to a psychology service for adults with ID. The following questions were posed: 1. What proportion of clients will have an existing diagnosed disorder? 2. What proportion of clients will have a referring issue that has a psychological intervention included in the MATRIX? 3. What proportion of clients will have more than one referring issue? 4. Are MATRIX therapies sufficient for the assessed needs of clients?
Design: The audit utilised a retrospective review of clinical case notes.
Method: Fifty per cent (N=73) of cases within an adult intellectual disability service, opened since October 2011 (the publication date of the MATRIX) were reviewed. Cases were selected by random number. Assessment-only cases were excluded. Eight files (10.9 per cent) were second-rated by a clinician who was not involved in the design or data-gathering process.
1. Diagnosis: 31.5 per cent of cases had a formal diagnosis (excluding a formal diagnosis of ID).
2. Referring issue: 90.4 per cent of cases had at least one referring problem included within the MATRIX.
3. Multiple referring issues: 39.1 per cent of cases had a single referring issue, 42.5 per cent had two, 17.8 per cent had three or more.
4. Sufficiency of MATRIX therapies: 81.8 per cent of cases, even if they were offered a MATRIX-approved therapy, also required an additional intervention.
Conclusions: Only a third of cases had an existing diagnosis, the majority of cases had more than a single presenting issue and the vast majority of cases were considered to require therapies not included in the MATRIX. This suggests that current clinical practice contains significant additional interventions. There may be significant challenges in applying such rigorous standards of research evidence when working a client group with complex needs. Applying rigid therapeutic recommendations may limit opportunities for eclectic practice.
Original languageEnglish
Publication statusPublished - 2014
Event2014 DCP Faculty for People with Intellectual Disabilities - St George's Hotel, Llandudno, United Kingdom
Duration: 26 Mar 201428 Mar 2014


Conference2014 DCP Faculty for People with Intellectual Disabilities
CountryUnited Kingdom


Dive into the research topics of 'Do levels of evidence affect breadth of service? An audit of MATRIX therapies in a service for adults with intellectual disabilities.'. Together they form a unique fingerprint.

Cite this