Methods
Data collection occurred during the treatment phase of a randomised controlled, multi-centre trial of the clinical and cost-effectiveness of MICE between 2019 and 2022.4 7
Patient and public involvement
Active involvement of CYPE and their caregivers was integral throughout the MICE research programme including this current study, from the funding application stage, through to project design, implementation, evaluation and dissemination. We recruited a diverse, committed and highly active research advisory group of 5 CYPE and 12 caregivers, and employed a patient and public involvement lead who is a coauthor on this paper (ED).
Therapist participant sample
The MICE intervention was delivered by 21 health professionals from a range of backgrounds (consultant paediatricians n=2, paediatric nurses and epilepsy nurse specialists n=6, assistant psychologists n=9, trainee clinical psychologists n=2 and a junior doctor n=1). 19 of the 21 were female. The mean age was 30.94 years (SD=8.91; range=23–51), average years since qualification in core profession (n=9) was 12.44 years (SD=6.73; range=2–24) and mean years of experience of working with CYPE was 2.1 years (SD=4.24; range=0–13). The assistant psychologists (n=9) all had master level degrees in psychology so may have had theoretical knowledge of CBT but did not have prior practical experience in delivering CBT.
Patient participant sample
CYPE aged 3–18 years were recruited from 13 epilepsy services across England and Northern Ireland. Out of the 334 participants in the study, 166 (85 male; M age=10.5 years; 122 White British) were randomly chosen to receive the MICE intervention. 93 (56%) had a primary diagnosis of disruptive behaviour on a standardised clinical interview (Developmental and Well-being Assessment),12 66 (39.8%) had anxiety and 7 (4.2%) depression. Additionally, 40 (24.1%) had been diagnosed with autistic spectrum conditions, and 67 (40.4%) with intellectual disability.
MICE intervention
The MICE intervention comprised up to 20 sessions delivered via telephone over 6 months, with an additional two booster sessions. Treatment included psychoeducation about mental health and epilepsy, CBT techniques for anxiety, depression and behavioural difficulties, and optional sessions on stigma, parental mental health and transition to adult services. The treatment was flexible, personalised and included epilepsy specific examples throughout.
To assess treatment integrity all sessions were recorded. Participants typically completed 16 sessions (IQR=12–19), resulting in a total of 2269 therapy sessions for analysis. All completed sessions were included in the assessments of integrity, regardless of the total number of sessions completed by the participant. All completed sessions were included in the assessments of integrity, regardless of the total number of sessions completed by the participant. There was no minimum number of sessions a participant had to attend to be classed as having received an adequate amount of the intervention. All participants had an assessment and at least one therapy session, with 158/166 (95%) having at least three sessions.
Therapists completed between 16 and 557 sessions of treatment (M=126.68 and SD=158.80). A single therapist delivered all the sessions for the patient, except when there were specific circumstances where treatment was transferred to a new therapist if the therapist’s job role changed prior to the patient completing treatment (eg, if a therapist embarked on a clinical psychology training programme). This occurred for 25/166 patients.
Therapist training and supervision
Therapists underwent a rigorous 6-month training including 5.5 days of practical workshops, regular supervision and completion of a minimum of one training case. Training included CBT techniques, goal setting, therapeutic relationship and risk management. Supervision was provided by a clinical psychologist (AEC or SB) and focused on skill development and monitoring patient progress. With the exception of the assistant psychologists who received weekly hourly supervision, therapists received flexible supervision of variable frequency and length based on therapist and patient needs. For example, more regular supervision was triggered by a lack of patient progress. Full details of therapist training and supervision have been published elsewhere.4 9
Therapist adherence
Therapist adherence to the MICE protocol was evaluated across all 2269 treatment sessions following previously established guidelines.13 14 The MICE protocol incorporates decision flowcharts to provide flexible treatment plans for each problem area (anxiety, low mood, behavioural problems). These flowcharts outline a standard order of treatment elements and include suggested flexible adaptations tailored to each patient’s unique presentation and needs. Previous research has developed adherence pairings to cover all possible pairings between the session of interest (‘index session’) and the previous session, allowing for sessions to be repeated and revisited if needed.13 14 We assessed therapist adherence to the MICE flowcharts using these existing adherence pairings which we expanded to include any additional content and sequencing from the MICE protocol.
Therapists recorded the specific treatment elements they implemented in each session and these were reviewed to compare the session pairings (index session-previous session) against the MICE flowcharts by an expert in the MICE protocol (AEC). The percentage of sessions where therapists adhered to both the content and sequencing of the MICE protocol was calculated. Given that the index session content was based on therapist self-report, the flowchart adherence pairing was also applied to the 10% of randomly selected therapy session audio recordings selected for competence rating to verify the data provided by therapists.
Each session was categorised into ‘all expected content’, ‘some expected content’ or ‘no expected content’. A session was classified as ‘all expected content’ if any part of the session topic aligned with the flowchart.13 14 For sessions categorised as ‘some expected content’, adaptations were identified as: (1) ‘sequencing’ that is, the session was either ahead or behind in the protocol; (2) ‘unexpected change in focus’ that is, content from the MICE protocol but not aligned with the current clinical focus; and (3) ‘expected change in focus’ that is, due to interference or comorbidity, as indicated in the flowcharts. Sessions were labelled as ‘no expected content’ when essential practice elements outlined in the flowcharts were omitted, or when sessions consisted solely of additional content not included in MICE.
Therapist competence
The Cognitive Therapy Rating Scale Revised (CTS-R)15 is a gold standard tool for measuring therapeutic tactical competence in delivering CBT within a single treatment session with high internal consistency and variable inter-rater reliability.16–20 It was used in this study to measure the extent to which the therapist implemented the chosen session with skill and thoroughness. The CTS-R consists of 12 domains (five generic therapeutic skills and seven areas specific to CBT), each rated on a 6-point scale. We used the national standard for competence on this measure that is, minimum pass mark of 50% and all domains scoring minimum of 2.
A random sample of 251 out of the available 2269 sessions was chosen by an independent member of the clinical trials unit. These sessions were rated by a CBT accredited clinician trained in the use of the CTS-R (AEC and PJ). Each therapist had at least one session rated for competence and to verify the adherence rating (range 1–60 sessions; M=12.3; SD=17.4). 30 of the 251 recordings were independently rated by an expert not associated with the MICE research team (PM-H).