Beyond total treatment effects in randomised controlled trials: Baseline measurement of intermediate outcomes needed to reduce confounding in mediation investigations

Telehealth coaching in Project ImPACT indirectly affects children's expressive language ability through parent intervention strategy use and child intentional communication: An RCT

Parent-mediated, naturalistic developmental behavioral interventions (NDBIs) are a promising approach for supporting social communication development in young autistic children. This study examined the effect of telehealth delivery of a parent-mediated NDBI, Project ImPACT, on children's expressive language ability using a randomized control trial with intent-to-treat analysis. Sixty-four young autistic children and their primary caregiver were matched on age and developmental quotient and randomly assigned to receive 6 months of therapist-assisted Project ImPACT (i.e., telehealth coaching), self-directed Project ImPACT, or an active control. Parent-child interactions were recorded at intake and immediately post-treatment, and the children's expressive language skills were assessed at intake and a 9-month follow-up using standardized measures. Although there was no total effect of treatment group assignment on child outcomes, a serial mediation analysis revealed that therapist-assisted ImPACT had an indirect effect on children's expressive language ability at follow-up through their parents' use of the intervention strategies and their intentional communication immediately post-treatment. Findings support Project ImPACT's program theory and highlight the importance of coaching in achieving positive outcomes when delivered via telehealth.

Self-efficacy with Pelvic floor muscle training mediates the effect of an App-based intervention on improving postpartum urinary incontinence severity among pregnant women: A causal mediation analysis from a randomised controlled trial

BACKGROUND

A pragmatic randomised controlled trial has confirmed the effectiveness of Urinary Incontinence for Women (UIW) app-based intervention in improving postpartum urinary incontinence (UI) severity among pregnant women. However, the causal mechanisms underlying this intervention effect remain unclear.

OBJECTIVE

To examine the mediating role of self-efficacy with pelvic floor muscle training (PFMT) on the effect of the UIW app-based intervention in improving postpartum UI severity.

METHODS

This was a secondary causal mediation analysis of a single-center, 2-arm, unblinded pragmatic randomised controlled trial. Singleton pregnant women without UI before pregnancy aged ≥18 years and between 24 and 28 weeks of gestation were recruited from a tertiary public hospital in China and randomised to receive the UIW app intervention plus oral PFMT instructions (n = 63) or oral PFMT instructions alone (n = 63). The primary outcome was postpartum changes in UI severity at 6 weeks. Changes in self-efficacy with PFMT 2 months after randomisation were a hypothesised mediator. Causal mediation analysis was used to estimate the average causal mediation effect (ACME), average direct effect (ADE), average total effect (ATE), and proportion mediated. A sensitivity analysis was conducted to examine the robustness of the ACME in relation to potential unmeasured confounding.

RESULTS

Data from 103 participants were analyzed. The ATE of UIW app-based intervention on postpartum UI severity was 2.91 points (95 % confidence intervals [CI] 1.69 to 4.12), with ADE of 1.97 points (95 % CI 0.63 to 3.41) and the ACME 0.94 points (95 % CI 0.27 to 1.72). The proportion of ATE mediated by self-efficacy with PFMT was 0.32 (95 % CI 0.08 to 0.67). Sensitivity analysis revealed the robust ACME with respect to the potential effects of unmeasured confounding.

CONCLUSION

An increase in self-efficacy with PFMT partially mediated the effect of the UIW app intervention on improvements in postpartum UI severity.

TRIAL REGISTRATION

The original trial was prospectively registered in the Chinese Clinical Trial Registry under the reference number ChiCTR1800016171 on 16/05/2018. Further details can be accessed at: http://www.chictr.org.cn/showproj.aspx?proj=27455.

Feeling connected but dissimilar to one's future self reduces the intention-behavior gap

The intention-behavior gap is a common phenomenon where people fail to follow through on their intentions to change their behavior and pursue their future goals. Previous research has shown that people are more likely to act in favor of their future selves when they feel similar/connected to their future self and can vividly describe them. This study compared an imagination exercise with an integrated imagination and exposure exercise using virtual reality (VR) to embody age-morphed future selves to an imagination only exercise. We expected that strengthening the similarity/connectedness and the vividness of the future self would reduce the intention-behavior gap, and exposure to the future self would have the greatest effect. Surprisingly, the results showed that strengthening connectedness reduced the intention-behavior gap, but strengthening similarity increased the gap. Additionally, the exercises were equally effective in reducing the intention-behavior gap. These findings suggest that both feeling connected to and recognizing dissimilarity to one's future self play different roles in future-oriented behavior change.

How does cognitive behavior therapy for dissociative seizures work? A mediation analysis of the CODES trial

BACKGROUND

We compared dissociative seizure specific cognitive behavior therapy (DS-CBT) plus standardized medical care (SMC) to SMC alone in a randomized controlled trial. DS-CBT resulted in better outcomes on several secondary trial outcome measures at the 12-month follow-up point. The purpose of this paper is to evaluate putative treatment mechanisms.

METHODS

We carried out a secondary mediation analysis of the CODES trial. 368 participants were recruited from the National Health Service in secondary / tertiary care in England, Scotland, and Wales. Sixteen mediation hypotheses corresponding to combinations of important trial outcomes and putative mediators were assessed. Twelve-month trial outcomes considered were final-month seizure frequency, Work and Social Adjustment Scale (WSAS), and the SF-12v2, a quality-of-life measure providing physical (PCS) and mental component summary (MCS) scores. Mediators chosen for analysis at six months (broadly corresponding to completion of DS-CBT) included: (a) beliefs about emotions, (b) a measure of avoidance behavior, (c) anxiety and (d) depression.

RESULTS

All putative mediator variables except beliefs about emotions were found to be improved by DS-CBT. We found evidence for DS-CBT effect mediation for the outcome variables dissociative seizures (DS), WSAS and SF-12v2 MCS scores by improvements in target variables avoidance behavior, anxiety, and depression. The only variable to mediate the DS-CBT effect on the SF-12v2 PCS score was avoidance behavior.

CONCLUSIONS

Our findings largely confirmed the logic model underlying the development of CBT for patients with DS. Interventions could be additionally developed to specifically address beliefs about emotions to assess whether it improves outcomes.

Understanding the Mechanisms of Cognitive Remediation on Recovery in People With Early Psychosis: A Mediation and Moderation Analysis

BACKGROUND

To provide precision cognitive remediation therapy (CR) for schizophrenia, we need to understand whether the mechanism for improved functioning is via cognition improvements. This mechanism has not been rigorously tested for potential moderator effects.

STUDY DESIGN

We used data (n = 377) from a randomized controlled trial using CIRCuiTS, a therapist-supported CR, with participants from first-episode psychosis services. We applied structured equation modeling to test whether: (1) CR hours explain the goal attainment functional outcome (GAS) at posttreatment, (2) global cognitive improvement mediates GAS, and if (3) total symptoms moderate the CR hours to cognitive improvement pathway, and/or negative symptoms moderate the cognition to functioning pathway, testing moderator effects via the mediator or directly on CR hours to functioning path.

STUDY RESULTS

CR produced significant functioning benefit for each therapy hour (Coeff = 0.203, 95% CI 0.101-0.304, P < .001). The mediated path from CR hours to cognition and cognition to functioning was small and nonsignificant (Coeff = 0.014, 95% CI = -0.010, 0.037, P = .256). Total symptoms did not moderate the path to cognition (P = .211) or the direct path to outcome (P = .896). However, negative symptoms significantly moderated the effect of cognitive improvements on functioning (P = .015) with high negative symptoms reducing the functional gains of improved cognition.

CONCLUSIONS

Although cognitive improvements were correlated with functioning benefit, they did not fully explain the positive effect of increased therapy hours on functioning, suggesting additional CR factors also contribute to therapy benefit. Negative symptoms interfere with the translation of cognitive improvements into functional gains so need consideration.

CLEAR - clozapine in early psychosis: study protocol for a multi-centre, randomised controlled trial of clozapine vs other antipsychotics for young people with treatment resistant schizophrenia in real world settings

BACKGROUND

Clozapine is an antipsychotic drug with unique efficacy, and it is the only recommended treatment for treatment-resistant schizophrenia (TRS: failure to respond to at least two different antipsychotics). However, clozapine is also associated with a range of adverse effects which restrict its use, including blood dyscrasias, for which haematological monitoring is required. As treatment resistance is recognised earlier in the illness, the question of whether clozapine should be prescribed in children and young people is increasingly important. However, most research to date has been in older, chronic patients, and evidence regarding the efficacy and safety of clozapine in people under age 25 is lacking. The CLEAR (CLozapine in EARly psychosis) trial will assess whether clozapine is more effective than treatment as usual (TAU), at the level of clinical symptoms, patient rated outcomes, quality of life and cost-effectiveness in people below 25 years of age. Additionally, a nested biomarker study will investigate the mechanisms of action of clozapine compared to TAU.

METHODS AND DESIGN

This is the protocol of a multi-centre, open label, blind-rated, randomised controlled effectiveness trial of clozapine vs TAU (any other oral antipsychotic monotherapy licenced in the British National Formulary) for 12 weeks in 260 children and young people with TRS (12-24 years old).

AIM AND OBJECTIVES

The primary outcome is the change in blind-rated Positive and Negative Syndrome Scale scores at 12 weeks from baseline. Secondary outcomes include blind-rated Clinical Global Impression, patient-rated outcomes, quality of life, adverse effects, and treatment adherence. Patients will be followed up for 12 months and will be invited to give consent for longer term follow-up using clinical records and potential re-contact for further research. For mechanism of action, change in brain magnetic resonance imaging (MRI) biomarkers and peripheral inflammatory markers will be measured over 12 weeks.

DISCUSSION

The CLEAR trial will contribute knowledge on clozapine effectiveness, safety and cost-effectiveness compared to standard antipsychotics in young people with TRS, and the results may guide future clinical treatment recommendation for early psychosis.

TRIAL REGISTRATION

ISRCTN Number: 37176025, IRAS Number: 1004947.

TRIAL STATUS

In set-up. Protocol version 4.0 01/08/23. Current up to date protocol available here: https://fundingawards.nihr.ac.uk/award/NIHR131175# /.

Mediation of 6-year mid-childhood follow-up outcomes after pre-school social communication (PACT) therapy for autistic children: randomised controlled trial

BACKGROUND

There are very few mechanistic studies of the long-term impact of psychosocial interventions in childhood. The parent-mediated Paediatric Autism Communication Therapy (PACT) RCT showed sustained effects on autistic child outcomes from pre-school to mid-childhood. We investigated the mechanism by which the PACT intervention achieved these effects.

METHODS

Of 152 children randomised to receive PACT or treatment as usual between 2 and 5 years of age, 121 (79.6%) were followed 5-6 years after the endpoint at a mean age of 10.5 years. Assessors, blind to the intervention group, measured Autism Diagnostic Observation Scale Calibrated Severity Score (ADOS CSS) for child autistic behaviours and Teacher Vineland (TVABS) for adaptive behaviour in school. Hypothesised mediators were child communication initiations with caregivers in a standard play observation (Dyadic Communication Measure for Autism, DCMA). Hypothesised moderators of mediation were baseline child non-verbal age equivalent scores (AE), communication and symbolic development (CSBS) and 'insistence on sameness' (IS). Structural equation modelling was used in a repeated measures mediation design.

RESULTS

Good model fits were obtained. The treatment effect on child dyadic initiation with the caregiver was sustained through the follow-up period. Increased child initiation at treatment midpoint mediated the majority (73%) of the treatment effect on follow-up ADOS CSS. A combination of partial mediation from midpoint child initiations and the direct effect of treatment also contributed to a near-significant total effect on follow-up TVABS. No moderation of this mediation was found for AE, CSBS or IS.

CONCLUSIONS

Early sustained increase in an autistic child's communication initiation with their caregiver is largely responsible for the long-term effects from PACT therapy on autistic and adaptive behaviour outcomes. This supports the theoretical logic model of PACT therapy but also illuminates fundamental causal processes of social and adaptive development in autism over time: early social engagement in autism can be improved and this can have long-term generalised outcome effects.

Improving the mixed model for repeated measures to robustly increase precision in randomized trials

In randomized trials, repeated measures of the outcome are routinely collected. The mixed model for repeated measures (MMRM) leverages the information from these repeated outcome measures, and is often used for the primary analysis to estimate the average treatment effect at the primary endpoint. MMRM, however, can suffer from bias and precision loss when it models intermediate outcomes incorrectly, and hence fails to use the post-randomization information harmlessly. This paper proposes an extension of the commonly used MMRM, called IMMRM, that improves the robustness and optimizes the precision gain from covariate adjustment, stratified randomization, and adjustment for intermediate outcome measures. Under regularity conditions and missing completely at random, we prove that the IMMRM estimator for the average treatment effect is robust to arbitrary model misspecification and is asymptotically equal or more precise than the analysis of covariance (ANCOVA) estimator and the MMRM estimator. Under missing at random, IMMRM is less likely to be misspecified than MMRM, and we demonstrate via simulation studies that IMMRM continues to have less bias and smaller variance. Our results are further supported by a re-analysis of a randomized trial for the treatment of diabetes.

Overcoming blinding confounds in psychedelic randomized controlled trials using biomarker driven causal mediation analysis

INTRODUCTION

There is great interest in the use of psychedelic-assisted therapies to treat a range of mental health conditions and initial randomized controlled trials (RCTs) have generated positive results. However, the effect sizes reported in psychedelic RCTs are likely inflated due to expectancy effects due to the de-blinding of both participants and study personnel to treatment allocation caused by the distinctive psychoactive effects of psychedelic drugs.

AREAS COVERED

An introduction to causal inference for RCTs, the underlying assumptions, and potential confounders along with graphical illustrations is provided. It is proposed that causal mediation analysis using objectively measured mediating biomarkers could be used to identify causal pathways between treatment and outcome in psychedelic RCTs, even with de-blinding of participants and give greater confidence as to the mechanistic basis and efficacy of psychedelic therapies.

EXPERT OPINION

It is argued that psychedelic therapies should not be approved as licensed medicines until causal pathways are clearly established between treatment and outcome. Potential downsides of doing so include, future indication expansion based on low quality clinical trial evidence, the approval of other therapies based on similarly low-quality evidence, and the potential for efficacy to change over time after approvals has been granted.

Study protocol: BRInging the Diabetes prevention program to GEriatric Populations

In the Diabetes Prevention Program (DPP) randomized, controlled clinical trial, participants who were  ≥ 60 years of age in the intensive lifestyle (diet and physical activity) intervention had a 71% reduction in incident diabetes over the 3-year trial. However, few of the 26.4 million American adults age ≥65 years with prediabetes are participating in the National DPP. The BRInging the Diabetes prevention program to GEriatric Populations (BRIDGE) randomized trial compares an in-person DPP program Tailored for Older AdulTs (DPP-TOAT) to a DPP-TOAT delivered via group virtual sessions (V-DPP-TOAT) in a randomized, controlled trial design (N = 230). Eligible patients are recruited through electronic health records (EHRs) and randomized to the DPP-TOAT or V-DPP-TOAT arm. The primary effectiveness outcome is 6-month weight loss and the primary implementation outcome is intervention session attendance with a non-inferiority design. Findings will inform best practices in the delivery of an evidence-based intervention.

The association of amygdala-insula functional connectivity and adolescent e-cigarette use via sleep problems and depressive symptoms

BACKGROUND

Adolescent electronic cigarette (e-cigarette) use remains high. Elucidating contributing factors may enhance prevention strategies. Neurobiologically, amygdala-insula resting-state functional connectivity (rsFC) has been linked with aspects of sleep, affect, and substance use (SU). As such, we hypothesized that amygdala's rsFC with the insula would be associated with e-cigarette use via sleep problems and/or depression levels.

METHODS

An adolescent sample (N = 146) completed a rs-fMRI scan at time 1 and self-reports at time 2 (∼15 months later). Given consistent associations between mental health outcomes and the rsFC of the laterobasal amygdala (lbAMY) with the anterior insula, we utilized a seed region (lbAMY) to region of interest (ROI) analysis approach to characterize brain-behavior relationships. Two serial mediation models tested the interrelations between amygdala's rsFC with distinct anterior insula subregions (i.e., ventral insula [vI], dorsal insula [dI]), sleep problems, depression levels, and days of e-cigarette use.

RESULTS

An indirect effect was observed when considering the lbAMY's rsFC with the vI. Greater rsFC predicted more sleep problems, more sleep problems were linked with greater depressive symptoms, and greater depressive symptoms were associated with more e-cigarette use (indirect effect = 0.08, CI [0.01,0.21]). Indicative of a neurobiological dissociation, a similar indirect effect linking these variables was not observed when considering the lbAMY's rsFC with the dI (indirect effect = 0.03, CI [-0.001,0.10]).

CONCLUSIONS

These outcomes highlight functional interactions between the amygdala and insula as a neurobiological contributor to sleep problems, depressive symptoms, and ultimately SU thereby suggesting potential intervention points to reduce teen e-cigarette use.

Transdiagnostic therapy for persistent physical symptoms: A mediation analysis of the PRINCE secondary trial

The PRINCE secondary trial did not find any evidence that transdiagnostic cognitive behavioural therapy (TDT-CBT) plus standard medical care (SMC) was more efficacious than SMC for patients with Persistent Physical Symptoms (PPS) for the primary outcome Work and Social Adjustment Scale (WSAS) at final follow-up (52 weeks). There was a significant treatment effect for TDT-CBT plus CBT compared with SMC for two secondary outcomes: WSAS at the end of active treatment (20 weeks) and symptom severity (Patient Health Questionnaire, PHQ-15) at 52 weeks. To understand mechanisms that lead to effects of TDT-CBT plus SMC versus SMC we performed a planned secondary mediation analysis. We investigated whether TDT-CBT treatment effects on these two secondary outcomes at the end of the treatment could be explained by effects on variables that were targeted by TDT-CBT during the initial phase of treatment. We pre-specified mediator variables measured at mid-treatment (9 weeks). Reductions in catastrophising and symptom focusing were the strongest mediators of TDT-CBT treatment effects on WSAS at the end of treatment. Improvements in symptom focusing also mediated the effect of TDT-CBT on PHQ-15. Future developments of the TDT-CBT intervention could benefit from targeting these mediators.

Methods of analysis for survival outcomes with time-updated mediators, with application to longitudinal disease registry data

Mediation analysis is a useful tool to illuminate the mechanisms through which an exposure affects an outcome but statistical challenges exist with time-to-event outcomes and longitudinal observational data. Natural direct and indirect effects cannot be identified when there are exposure-induced confounders of the mediator-outcome relationship. Previous measurements of a repeatedly-measured mediator may themselves confound the relationship between the mediator and the outcome. To overcome these obstacles, two recent methods have been proposed, one based on path-specific effects and one based on an additive hazards model and the concept of exposure splitting. We investigate these techniques, focusing on their application to observational datasets. We apply both methods to an analysis of the UK Cystic Fibrosis Registry dataset to identify how much of the relationship between onset of cystic fibrosis-related diabetes and subsequent survival acts through pulmonary function. Statistical properties of the methods are investigated using simulation. Both methods produce unbiased estimates of indirect and direct effects in scenarios consistent with their stated assumptions but, if the data are measured infrequently, estimates may be biased. Findings are used to highlight considerations in the interpretation of the observational data analysis.

The Effect of Perceived Threats and Response Efficacy on Adaptation to Smog: An Instrumental Variables Design

Threats and response efficacyperceptions are core conceptsof the protection motivationtheory, and recent years have witnessed a considerable growth of research on the effect of thesefactors on adaptation to air pollution. However, few studies use appropriate designs to deal with endogeneity issues, a situation that raises serious questions on the validity of their findings. To overcome this problem, this study uses the instrumental variables method to test the effect of perceived threats and response efficacy on adaptation to smog episodes. The results of this study show that the conjunction of a moderate to high perception of threats with a high perception of response efficacy is positively associated with the adoption of the recommended behavior. The increase of perceived threats does not seem to have an effect on the behavior of individuals with low response efficacy perception. Moreover, change in perceived response efficacy does not lead to any change in the behavior of individuals with low threat perceptions. Concerning policy implications, this study suggests that smog warnings and health communication campaigns could be more effective if they provide accurate information simultaneously on air pollution level, its adverse effects, and advice on how to mitigate these effects.

Combined social communication therapy at home and in education for young autistic children in England (PACT-G): a parallel, single-blind, randomised controlled trial

BACKGROUND

Autistic children can have difficulty generalising treatment effects beyond the immediate treatment context. Paediatric Autism Communication Therapy (PACT) has been successful when delivered in the clinic. Here we tested the Paediatric Autism Communication Therapy-Generalised (PACT-G) intervention combined between home and education settings for its overall effect and mechanistic transmission of effect across contexts.

METHODS

In this parallel, single-blind, randomised, controlled trial, we recruited autistic children aged 2-11 years in urban or semi-urban areas in Manchester, Newcastle, and London, England. Children needed to meet core autism criteria on Autism Diagnostic Observation Schedule-second edition (ADOS-2) and parent-rated Social Communication Questionnaire (SCQ-lifetime), and children older than 5 years were included if they had intentional communication but expressive language equivalent of age 4 years or younger. Eligible children were randomly assigned (1:1), using block randomisation (random block sizes of 2 and 4) and stratified for site, age (2-4 years vs 5-11 years), and gender, to either PACT-G plus treatment as usual or treatment as usual alone. Research assessors were masked to treatment allocation. The PACT-G intervention was delivered by a therapist in parallel to the child's parents at home and to learning-support assistants (LSA) at their place of education, using both in-person and remote sessions over a 6 month period, to optimise adult-child social interaction. Treatment as usual included any health support or intervention from education or local community services. The primary outcome was autism symptom severity using the ADOS-2, as measured by researchers, at 12 months versus baseline. Secondary outcomes were Brief Observation of Social Communication Change (BOSCC) and dyadic social interaction between child and adult across contexts, both at 12 months. Other secondary outcome measures were assessed using the following composites: language, anxiety, repetitive behaviour, adaptive behaviour, parental wellbeing, child health-related quality of life, and disruptive behaviour. Assessments were done at baseline, 7 months, and 12 months. We used an intention-to-treat (ITT) analysis of covariance for the efficacy outcome measures. Adverse events were assessed by researchers for all trial families at each contact and by therapists in the PACT-G group at each visit. This study is registered with the ISRCTN Registry, ISRCTN 25378536.

FINDINGS

Between Jan 18, 2017, and April 19, 2018, 555 children were referred and 249 were eligible, agreed to participate, and were randomly assigned to either PACT-G (n=122) or treatment as usual (n=127). One child in the PACT-G group withdrew and requested their data be removed from the study, giving an ITT population of 248 children. 51 (21%) of 248 children were female, 197 (79%) were male, 149 (60%) were White, and the mean age was 4·0 years (SD 0·6). The groups were well balanced for demographic and clinical characteristics. In the PACT-G group, parents of children received a median of 10 (IQR 8-12) home sessions and LSAs received a median of 8 (IQR 5-10) education sessions over 6 months. We found no treatment effect on the ADOS-2 primary outcome compared with treatment as usual (effect size 0·04 [95% CI -0·19 to 0·26]; p=0·74), or researcher-assessed BOSCC (0·03 [-0·25 to 0·31]), language composite (-0·03 [-0·15 to 0·10]), repetitive behaviour composite (0·00 [-0·35 to 0·35]), adaptive behaviour composite (0·01 [-0·15 to 0·18]), or child wellbeing (0·09 [-0·15 to 0·34]). PACT-G treatment improved synchronous response in both parent (0·50 [0·36 to 0·65]) and LSA (0·33 [0·16 to 0·50]), mediating increased child communication with parent (0·26 [0·12 to 0·40]) and LSA (0·20 [0·06 to 0·34]). Child dyadic communication change mediated outcome symptom alteration on BOSCC at home (indirect effect -0·78 [SE 0·34; 95% CI -1·44 to -0·11]; p=0·022) although not in education (indirect effect -0·67 [SE 0·37; 95% CI -1·40 to 0·06]; p=0·073); such an effect was not seen on ADOS-2. Treatment with PACT-G also improved the parental wellbeing composite (0·44 [0·08 to 0·79]) and the child disruptive behaviour composite in home and education (0·29 [0·01 to 0·57]). Adverse events on child behaviour and wellbeing were recorded in 13 (10%) of 127 children in the treatment as usual group (of whom four [31%] were girls) and 11 (9%) of 122 in the PACT-G group (of whom three [33%] were girls). One serious adverse event on parental mental health was recorded in the PACT-G group and was possibly study related.

INTERPRETATION

Although we found no effect on the primary outcome compared with treatment as usual, adaptation of the 12-month PACT intervention into briefer multicomponent delivery across home and education preserved the positive proximal outcomes, although smaller in effect size, and the original pattern of treatment mediation seen in clinic-delivered therapy, as well as improving parental wellbeing and child disruptive behaviours across home and school. Reasons for this reduced efficacy might be the reduced dose of each component, the effect of remote delivery, and the challenges of the delivery contexts. Caution is needed in assuming that changing delivery methods and context will preserve an original intervention efficacy for autistic children.

FUNDING

National Institute for Health Research and Medical Research Council Efficacy and Mechanism Evaluation Award.

Producing Clinically Meaningful Reductions in Disability: A Causal Mediation Analysis of a Patient Education Intervention

Patient education is recommended as first-line care for low back pain (LBP), although its efficacy for providing clinically meaningful reductions in disability has been questioned. One way to improve treatment effects is to identify and improve targeting of treatment mechanisms. We conducted a pre-planned causal mediation analysis of a randomized, placebo-controlled trial investigating the effectiveness of patient education for patients with acute LBP. 202 patients who had fewer than six-weeks' duration of LBP and were at high-risk of developing chronic LBP completed two, one-hour treatment sessions of either intensive patient education, or placebo patient education. 189 participants provided data for the outcome self-reported disability at three-months and the mediators, pain self-efficacy, pain catastrophizing, and back beliefs at one-week post treatment. This causal mediation analysis found that pain catastrophizing (mediated effect, -0.64; 95% Confidence Interval [CI], -1.31 to -0.15) and back beliefs (mediated effect, -0.51; 95% CI, -1.15 to -0.02) partly explained the effect of patient education on disability but pain self-efficacy did not (mediated effect, -0.40; 95% CI -1.13 to 0.28). Considering the mediator-outcome relationship, patient education would need to induce an 8 point difference on the pain self-efficacy questionnaire (0-60); an 11 point difference on the back beliefs questionnaire (9-45); and a 21 point difference on the pain catastrophizing scale (0-52) to achieve a minimally clinically important difference of 2 points on the Roland Morris Disability Questionnaire (0-24). PERSPECTIVE: Understanding the mechanisms of patient education can inform how this treatment can be adapted to provide clinically meaningful reductions in disability. Our findings suggest that adapting patient education to better target back beliefs and pain self-efficacy could result in clinically meaningful reductions in disability whereas the role of pain catastrophizing in acute LBP is less clear.

The mediating effect of pain catastrophizing on pain intensity: The influence of the timing of assessments

BACKGROUND

Pain catastrophizing underpins several psychosocial theories of pain, but there is limited evidence to support the proposal that changes in pain catastrophizing cause changes in pain. Results from mediation analyses have conflicting results, and one reason for these might be the timing of the assessment of pain catastrophizing. This study aimed to test the effect of the timing of the assessment of pain catastrophizing on its mediating role on pain intensity.

METHODS

Causal mediation analysis using data from a randomized controlled trial which included 100 participants with chronic low back pain. The trial found that clinical hypnosis, compared to pain education, reduced worst pain intensity and pain catastrophizing. In model 1, we used data from 2-week follow-up for pain catastrophizing and 3-month follow-up for pain. In model 2, we used data from 3-month follow-up for both pain catastrophizing and pain.

RESULTS

The intervention had a significant average total effect on pain (-1.34 points, 95% CI -2.50 to -0.13). The average causal mediation effect was significant when pain catastrophizing, and pain were measured at the same time (-0.62 points, 95% CI -1.30 to -0.11) but not significant when pain catastrophizing and pain intensity were measured at different times (-0.10 points, 95% CI -0.42 to 0.09).

CONCLUSIONS

The timing of the assessment influenced the mediating role of pain catastrophizing on pain intensity. These results raise questions on the casual role that pain catastrophizing has on pain intensity. Psychosocial interventions such as clinical hypnosis can reduce pain intensity even when there has been no change in pain catastrophizing.

Optimising AVATAR therapy for people who hear distressing voices: study protocol for the AVATAR2 multi-centre randomised controlled trial

BACKGROUND

AVATAR therapy is a novel intervention targeting distressing auditory verbal hallucinations (henceforth 'voices'). A digital simulation (avatar) of the voice is created and used in a three-way dialogue between participant, avatar and therapist. To date, therapy has been delivered over 6 sessions, comprising an initial phase, focusing on standing up to a hostile avatar, and a second phase in which the avatar concedes and focus shifts to individualised treatment targets, including beliefs about voices. The first fully powered randomised trial found AVATAR therapy resulted in a rapid and substantial fall in voice frequency and associated distress that was superior to supportive counselling at 12 weeks. The main objective of this AVATAR2 trial is to test the efficacy of two forms of AVATAR therapy in reducing voice-related distress: AVATAR-brief (standardised focus on exposure, assertiveness and self-esteem) and AVATAR-extended (phase 1 mirroring AVATAR-brief augmented by a formulation-driven phase 2). Secondary objectives include the examination of additional voice, wellbeing and mood outcomes, the exploration of mediators and moderators of therapy response, and examining cost-effectiveness of both forms of therapy compared with usual treatment (TAU).

METHODS

This multi-site parallel group randomised controlled trial will independently randomise 345 individuals to receive AVATAR-brief (6 sessions) plus TAU or AVATAR-extended (12 sessions) plus TAU or TAU alone (1:1:1 allocation). Participants will be people with a diagnosis of schizophrenia spectrum and other psychotic disorders who have heard distressing voices for more than 6 months. The primary outcome is the PSYRATS Auditory Hallucinations Distress dimension score at 16 and 28 weeks, conducted by blinded assessors. Statistical analysis will follow the intention-to-treat principle and data will be analysed using linear mixed models. Mediation and moderation analyses using contemporary causal inference methods will be conducted as secondary analyses. Service costs will be calculated, and cost-effectiveness assessed in terms of quality-adjusted life years accrued.

DISCUSSION

This study will clarify optimal therapy delivery, test efficacy in a multi-site study and enable the testing of the AVATAR software platform, therapy training and provision in NHS settings.

TRIAL REGISTRATION

ISRCTN registry ISRCTN55682735 . Registered on 22 January 2020. The trial is funded by the Wellcome Trust (WT).

A Viable Alternative When Propensity Scores Fail: Evaluation of Inverse Propensity Weighting and Sequential G-Estimation in a Two-Wave Mediation Model

Two methods from the potential outcomes framework - inverse propensity weighting (IPW) and sequential G-estimation - were evaluated and compared to linear regression for estimating the mediated effect in a two-wave design with a randomized intervention and continuous mediator and outcome. Baseline measures of the mediator and outcome can be considered confounders of the follow-up mediator - outcome relation for which adjustment is necessary to eliminate bias. To adjust for baseline measures of the mediator and outcome, IPW uses stabilized inverse propensity weights whereas sequential G-estimation uses regression adjustment. Theoretical differences between the models are described, and Monte Carlo simulations compared the performance of linear regression; IPW without weight truncation; IPW with weights truncated at the 1st/99th, 5th/95th, and 10th/90th percentiles; and sequential G-estimation. Sequential G-estimation performed similarly to linear regression, but IPW provided a biased estimate of the mediated effect, lower power, lower confidence interval coverage, and higher mean squared error. Simulation results show that IPW failed to fully adjust the follow-up mediator - outcome relation for confounding due to the baseline measures. We then compared the mediated effect estimates using data from a randomized experiment evaluating a steroid prevention program for high school athletes. Implications and future directions are discussed.

Paediatric Autism Communication Therapy-Generalised (PACT-G) against treatment as usual for reducing symptom severity in young children with autism spectrum disorder: study protocol for a randomised controlled trial

Background

Prior evidence shows that behaviours closely related to the intervention delivered for autism are amenable to change, but it is more difficult to generalise treatment effects beyond the intervention context. We test an early autism intervention designed to promote generalisation of therapy-acquired skills into home and school contexts to improve adaptive function and reduce symptoms. A detailed mechanism study will address the process of such generalisation. Objective 1 will be to test if the PACT-G intervention improves autism symptom outcome in the home and school context of the intervention as well as in the primary outcome research setting. Objective 2 will use the mechanism analysis to test for evidence of acquired skills from intervention generalizing across contexts and producing additive effects on primary outcome.

Methods/design

This is a three-site, two-parallel-group, randomised controlled trial of the experimental treatment plus treatment as usual (TAU) versus TAU alone. Children aged 2–11 years (n = 244 (122 intervention/122 TAU; ~ 82/site) meeting criteria for core autism will be eligible. The experimental intervention builds on a clinic-based Pre-school Autism Communication Treatment model (PACT), delivered with the primary caregiver, combined with additional theory- and evidence-based strategies designed to enhance the generalisation of effects into naturalistic home and education contexts. The control intervention will be TAU.

Primary outcome: autism symptom outcome, researcher-assessed using a standardised protocol. Secondary outcomes: autism symptoms, child interaction with parent or teacher, language and reported functional outcomes in home and school settings. Outcomes measured at baseline and 12-month endpoint in all settings with interim interaction measurements (7 months) to test treatment effect mechanisms.

Primary analysis will estimate between-group difference in primary outcome using analysis of covariance with test of homogeneity of effect across age group. Mechanism analysis will use regression models to test for mediation on primary outcome by parent-child and teaching staff-child social interaction.

Discussion

This is an efficacy and mechanism trial of generalising evidence-based autism treatment into home and school settings. It will provide data on whether extending treatment across naturalistic contexts enhances overall effect and data on the mechanism in autism development of the generalisation of acquired developmental skills across contexts.

Trial registration

ISRCTN, ID: 25378536. Prospectively registered on 9 March 2016:

Electronic supplementary material

The online version of this article (10.1186/s13063-018-2881-3) contains supplementary material, which is available to authorized users.

Less Medication Use in Inpatients With Severe Mental Illness Receiving a Multidisciplinary Lifestyle Enhancing Treatment. The MULTI Study III

Besides having an unhealthy lifestyle contributing to premature mortality, inpatients with severe mental illness (SMI) use high dosages of medication. Previous research has shown improved health after lifestyle improvements in SMI. In addition, we aimed to retrospectively study whether a multidisciplinary lifestyle enhancing treatment (MULTI) was associated with changes in medication use after 18 months, as compared with patients that continued treatment as usual (TAU) and explored mediation by a change in physical activity. We conducted an observational study within a cohort of inpatients with SMI, who received MULTI (N = 65) or continued TAU (N = 49). Data on their somatic and psychotropic medications were collected, converted into defined daily dose (DDD), and analyzed using linear multilevel regression, correcting for baseline value and differences between groups in age, diagnosis, and illness severity. Compared with TAU, the DDD for psychotropic medication significantly decreased with MULTI (B = -0.55, P = 0.02). Changes in total activity did not mediate this association, suggesting that multiple components of MULTI contributed. Corrected between-group analyses for subgroups of medication were not possible due to lack of power and skewed distributions. Within-group data showed a decreased proportion of users as well as median DDD in both groups for almost all medications. In addition to previously reported health improvements after 18 months of MULTI, we observed a significant decrease in dose of psychotropic medication in MULTI compared to TAU. This first study evaluating a wide range of medications indicates a possible effect of lifestyle improvements on medication use in inpatients with SMI. Findings need to be confirmed in future controlled studies, however.