Estimating treatment effects from randomized clinical trials with noncompliance and loss to follow-up: the role of instrumental variable methods

Influence of therapist competence and quantity of cognitive behavioural therapy on suicidal behaviour and inpatient hospitalisation in a randomised controlled trial in borderline personality disorder: further analyses of treatment effects in the BOSCOT study

OBJECTIVES

We investigated the treatment effects reported from a high-quality randomized controlled trial of cognitive behavioural therapy (CBT) for 106 people with borderline personality disorder attending community-based clinics in the UK National Health Service - the BOSCOT trial. Specifically, we examined whether the amount of therapy and therapist competence had an impact on our primary outcome, the number of suicidal acts, using instrumental variables regression modelling.

DESIGN

Randomized controlled trial. Participants from across three sites (London, Glasgow, and Ayrshire/Arran) were randomized equally to CBT for personality disorders (CBTpd) plus Treatment as Usual or to Treatment as Usual. Treatment as Usual varied between sites and individuals, but was consistent with routine treatment in the UK National Health Service at the time. CBTpd comprised an average 16 sessions (range 0-35) over 12 months.

METHOD

We used instrumental variable regression modelling to estimate the impact of quantity and quality of therapy received (recording activities and behaviours that took place after randomization) on number of suicidal acts and inpatient psychiatric hospitalization.

RESULTS

A total of 101 participants provided full outcome data at 2 years post randomization. The previously reported intention-to-treat (ITT) results showed on average a reduction of 0.91 (95% confidence interval 0.15-1.67) suicidal acts over 2 years for those randomized to CBT. By incorporating the influence of quantity of therapy and therapist competence, we show that this estimate of the effect of CBTpd could be approximately two to three times greater for those receiving the right amount of therapy from a competent therapist.

CONCLUSIONS

Trials should routinely control for and collect data on both quantity of therapy and therapist competence, which can be used, via instrumental variable regression modelling, to estimate treatment effects for optimal delivery of therapy. Such estimates complement rather than replace the ITT results, which are properly the principal analysis results from such trials.

The Scleroderma Patient-centered Intervention Network (SPIN) Cohort: protocol for a cohort multiple randomised controlled trial (cmRCT) design to support trials of psychosocial and rehabilitation interventions in a rare disease context

INTRODUCTION

Psychosocial and rehabilitation interventions are increasingly used to attenuate disability and improve health-related quality of life (HRQL) in chronic diseases, but are typically not available for patients with rare diseases. Conducting rigorous, adequately powered trials of these interventions for patients with rare diseases is difficult. The Scleroderma Patient-centered Intervention Network (SPIN) is an international collaboration of patient organisations, clinicians and researchers. The aim of SPIN is to develop a research infrastructure to test accessible, low-cost self-guided online interventions to reduce disability and improve HRQL for people living with the rare disease systemic sclerosis (SSc or scleroderma). Once tested, effective interventions will be made accessible through patient organisations partnering with SPIN.

METHODS AND ANALYSIS

SPIN will employ the cohort multiple randomised controlled trial (cmRCT) design, in which patients consent to participate in a cohort for ongoing data collection. The aim is to recruit 1500-2000 patients from centres across the world within a period of 5 years (2013-2018). Eligible participants are persons ≥18 years of age with a diagnosis of SSc. In addition to baseline medical data, participants will complete patient-reported outcome measures every 3 months. Upon enrolment in the cohort, patients will consent to be contacted in the future to participate in intervention research and to allow their data to be used for comparison purposes for interventions tested with other cohort participants. Once interventions are developed, patients from the cohort will be randomly selected and offered interventions as part of pragmatic RCTs. Outcomes from patients offered interventions will be compared with outcomes from trial-eligible patients who are not offered the interventions.

ETHICS AND DISSEMINATION

The use of the cmRCT design, the development of self-guided online interventions and partnerships with patient organisations will allow SPIN to develop, rigourously test and effectively disseminate psychosocial and rehabilitation interventions for people with SSc.

Pragmatic trials of complex psychosocial interventions: methodological challenges

Having first introduced the pragmatic health care trial, the discussion then focuses on a selected list of technical problems that are important for the design, analysis and inference from such trials. The first is lack of independence of participants' outcomes do to clustering either arising from a cluster randomized design or to the way treatment is delivered (therapist and group effects). The second and third concern the implications of non-adherence to treatment and subsequent loss to follow-up, particularly, when non-adherence is associated with missing outcome data. Finally, it is argued that pragmatism and a desire for a scientific explanation should not be regarded as mutually exclusive.

Intermediate outcomes in randomized clinical trials: an introduction

BACKGROUND

Intermediate outcomes are common and typically on the causal pathway to the final outcome. Some examples include noncompliance, missing data, and truncation by death like pregnancy (e.g. when the trial intervention is given to non-pregnant women and the final outcome is preeclampsia, defined only on pregnant women). The intention-to-treat approach does not account properly for them, and more appropriate alternative approaches like principal stratification are not yet widely known. The purposes of this study are to inform researchers that the intention-to-treat approach unfortunately does not fit all problems we face in experimental research, to introduce the principal stratification approach for dealing with intermediate outcomes, and to illustrate its application to a trial of long term calcium supplementation in women at high risk of preeclampsia.

METHODS

Principal stratification and related concepts are introduced. Two ways for estimating causal effects are discussed and their application is illustrated using the calcium trial, where noncompliance and pregnancy are considered as intermediate outcomes, and preeclampsia is the main final outcome.

RESULTS

The limitations of traditional approaches and methods for dealing with intermediate outcomes are demonstrated. The steps, assumptions and required calculations involved in the application of the principal stratification approach are discussed in detail in the case of our calcium trial.

CONCLUSIONS

The intention-to-treat approach is a very sound one but unfortunately it does not fit all problems we find in randomized clinical trials; this is particularly the case for intermediate outcomes, where alternative approaches like principal stratification should be considered.

Cognitive behavioural therapy as an adjunct to pharmacotherapy for primary care based patients with treatment resistant depression: results of the CoBalT randomised controlled trial

BACKGROUND

Only a third of patients with depression respond fully to antidepressant medication but little evidence exists regarding the best next-step treatment for those whose symptoms are treatment resistant. The CoBalT trial aimed to examine the effectiveness of cognitive behavioural therapy (CBT) as an adjunct to usual care (including pharmacotherapy) for primary care patients with treatment resistant depression compared with usual care alone.

METHODS

This two parallel-group multicentre randomised controlled trial recruited 469 patients aged 18-75 years with treatment resistant depression (on antidepressants for ≥6 weeks, Beck depression inventory [BDI] score ≥14 and international classification of diseases [ICD]-10 criteria for depression) from 73 UK general practices. Participants were randomised, with a computer generated code (stratified by centre and minimised according to baseline BDI score, whether the general practice had a counsellor, previous treatment with antidepressants, and duration of present episode of depression) to one of two groups: usual care or CBT in addition to usual care, and were followed up for 12 months. Because of the nature of the intervention it was not possible to mask participants, general practitioners, CBT therapists, or researchers to the treatment allocation. Analyses were by intention to treat. The primary outcome was response, defined as at least 50% reduction in depressive symptoms (BDI score) at 6 months compared with baseline. This trial is registered, ISRCTN38231611.

FINDINGS

Between Nov 4, 2008, and Sept 30, 2010, we assigned 235 patients to usual care, and 234 to CBT plus usual care. 422 participants (90%) were followed up at 6 months and 396 (84%) at 12 months, finishing on Oct 31, 2011. 95 participants (46%) in the intervention group met criteria for response at 6 months compared with 46 (22%) in the usual care group (odds ratio 3·26, 95% CI 2·10-5·06, p<0·001).

INTERPRETATION

Before this study, no evidence from large-scale randomised controlled trials was available for the effectiveness of augmentation of antidepressant medication with CBT as a next-step for patients whose depression has not responded to pharmacotherapy. Our study has provided robust evidence that CBT as an adjunct to usual care that includes antidepressants is an effective treatment, reducing depressive symptoms in this population.

FUNDING

National Institute for Health Research Health Technology Assessment.

Randomised controlled trials and clinical maternity care: moving on from intention-to-treat and other simplistic analyses of efficacy

BACKGROUND

The obstetrical literature is dominated by Randomised Controlled Trials (RCTs), with the vast majority being analysed using an intention-to-treat (ITT) approach. Whilst this approach may reflect well the consequence of assignment to therapy and hence the 'trialists'perspective', it may fail to address the consequence of actually receiving therapy (the patient's perspective).

DISCUSSION

This review questions the ubiquitous adherence to the ITT approach, and gives examples of where this may have misled the maternity care professions. It gives an overview of techniques to overcome potential deficiencies in result presentation, using method effectiveness models such as 'Per Protocol' (PP) or 'As-Treated' (AT) that may give more accurate clinical meaning to the presentation of obstetrical results. It then proceeds to cover the added benefits, considerations and potential pitfalls of the use of Instrumental Variable (IV) models in order to better reflect the clinical context.

SUMMARY

While ITT may achieve statistical purity, it frequently fails to address the true clinical or patient's perspective. Though more complex and potentially beset by problems of their own, alternative methods of result presentation may better serve the latter aim. Each of the other methods may rely on untestable assumptions and therefore it is wisest that study results are presented in multiple formats to allow for informed reader evaluation.

Strengthening and stretching for rheumatoid arthritis of the hand (SARAH): design of a randomised controlled trial of a hand and upper limb exercise intervention--ISRCTN89936343

BACKGROUND

Rheumatoid Arthritis (RA) commonly affects the hands and wrists with inflammation, deformity, pain, weakness and restricted mobility leading to reduced function. The effectiveness of exercise for RA hands is uncertain, although evidence from small scale studies is promising. The Strengthening And Stretching for Rheumatoid Arthritis of the Hand (SARAH) trial is a pragmatic, multi-centre randomised controlled trial evaluating the clinical and cost effectiveness of adding an optimised exercise programme for hands and upper limbs to best practice usual care for patients with RA.

METHODS/DESIGN

480 participants with problematic RA hands will be recruited through 17 NHS trusts. Treatments will be provided by physiotherapists and occupational therapists. Participants will be individually randomised to receive either best practice usual care (joint protection advice, general exercise advice, functional splinting and assistive devices) or best practice usual care supplemented with an individualised exercise programme of strengthening and stretching exercises. The study assessors will be blinded to treatment allocation and will follow participants up at four and 12 months. The primary outcome measure is the Hand function subscale of the Michigan Hand Outcome Questionnaire, and secondary outcomes include hand and wrist impairment measures, quality of life, and resource use. Economic and qualitative studies will also be carried out in parallel.

DISCUSSION

This paper describes the design and development of a trial protocol of a complex intervention study based in therapy out-patient departments. The findings will provide evidence to support or refute the use of an optimised exercise programme for RA of the hand in addition to best practice usual care.

The effects of reducing worry in patients with persecutory delusions: study protocol for a randomized controlled trial

Background

Our approach to advancing the treatment of psychosis is to focus on key single symptoms and develop interventions that target the mechanisms that maintain them. In our theoretical research we have found worry to be an important factor in the development and maintenance of persecutory delusions. Worry brings implausible ideas to mind, keeps them there, and makes the experience distressing. Therefore the aim of the trial is to test the clinical efficacy of a cognitive-behavioral intervention for worry for patients with persecutory delusions and determine how the worry treatment might reduce delusions.

Methods/Design

An explanatory randomized controlled trial - called the Worry Intervention Trial (WIT) - with 150 patients with persecutory delusions will be carried out. Patients will be randomized to the worry intervention in addition to standard care or to standard care. Randomization will be carried out independently, assessments carried out single-blind, and therapy competence and adherence monitored. The study population will be individuals with persecutory delusions and worry in the context of a schizophrenia spectrum diagnosis. They will not have responded adequately to previous treatment. The intervention is a six-session cognitive-behavioral treatment provided over eight weeks. The control condition will be treatment as usual, which is typically antipsychotic medication and regular appointments. The principal hypotheses are that a worry intervention will reduce levels of worry and that it will also reduce the persecutory delusions. Assessments will be carried out at 0 weeks (baseline), 8 weeks (post treatment) and 24 weeks (follow-up). The statistical analysis strategy will follow the intention-to-treat principle and involve the use of linear mixed models to evaluate and estimate the relevant between- and within-subjects effects (allowing for the possibility of missing data). Both traditional regression and newer instrumental variables analyses will examine mediation. The trial is funded by the UK Medical Research Council (MRC)/NHS National Institute of Health Research (NIHR) Efficacy and Mechanism Evaluation (EME) Programme.

Discussion

This will be the first large randomized controlled trial specifically focused upon persecutory delusions. The project will produce a brief, easily administered intervention that can be readily used in mental health services.

Trial registration

Current Controlled Trials ISRCTN23197625

Classroom based cognitive behavioural therapy in reducing symptoms of depression in high risk adolescents: pragmatic cluster randomised controlled trial

OBJECTIVE

To compare the effectiveness of classroom based cognitive behavioural therapy with attention control and usual school provision for adolescents at high risk of depression.

DESIGN

Three arm parallel cluster randomised controlled trial.

SETTING

Eight UK secondary schools.

PARTICIPANTS

Adolescents (n=5030) aged 12-16 years in school year groups 8-11. Year groups were randomly assigned on a 1:1:1 ratio to cognitive behavioural therapy, attention control, or usual school provision. Allocation was balanced by school, year, number of students and classes, frequency of lessons, and timetabling. Participants were not blinded to treatment allocation.

INTERVENTIONS

Cognitive behavioural therapy, attention control, and usual school provision provided in classes to all eligible participants.

MAIN OUTCOME MEASURES

Outcomes were collected by self completed questionnaire administered by researchers. The primary outcome was symptoms of depression assessed at 12 months by the short mood and feelings questionnaire among those identified at baseline as being at high risk of depression. Secondary outcomes included negative thinking, self worth, and anxiety. Analyses were undertaken on an intention to treat basis and accounted for the clustered nature of the design.

RESULTS

1064 (21.2%) adolescents were identified at high risk of depression: 392 in the classroom based cognitive behavioural therapy arm, 374 in the attention control arm, and 298 in the usual school provision arm. At 12 months adjusted mean scores on the short mood and feelings questionnaire did not differ for cognitive behavioural therapy versus attention control (-0.63, 95% confidence interval -1.85 to 0.58, P=0.41) or for cognitive behavioural therapy versus usual school provision (0.97, -0.20 to 2.15, P=0.12).

CONCLUSION

In adolescents with depressive symptoms, outcomes were similar for attention control, usual school provision, and cognitive behavioural therapy. Classroom based cognitive behavioural therapy programmes may result in increased self awareness and reporting of depressive symptoms but should not be undertaken without further evaluation and research.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN19083628.

Effects of drop-out on efficacy estimates in five Cochrane reviews of popular antipsychotics for schizophrenia

OBJECTIVE

Our aim was to find out how Cochrane reviews of five popular or frequently prescribed second-generation antipsychotics in the UK (olanzapine, risperidone, quetiapine, amisulpride and aripiprazole) approached the problem of high drop-out in placebo-controlled trials.

METHOD

We examined the following: (i) whether reviews included data from studies with a level of drop-out exceeding their stated exclusion criterion; (ii) the level of missing data each efficacy outcome in each review relied upon; and (iii) impact of excluding studies with high drop-out.

RESULTS

All reviews included data they stated they would exclude because of unacceptable levels of attrition, four (risperidone, olanzapine, amisulpride, aripiprazole) without clear acknowledgement or justification. Several reviews also excluded data from a number of relatively low-attrition studies because of missing standard deviations.

CONCLUSION

Cochrane reviews of five popular antipsychotics for schizophrenia misrepresented the available evidence on their efficacy. The impact of including high-attrition studies was difficult to quantify because of the exclusion of relevant low-attrition studies. Further analysis of the efficacy of these drugs in studies with acceptable rates of attrition is required. To reduce the problem of high attrition, trialists should gather follow-up data from people who leave the double-blind process early.

Facilitated physical activity as a treatment for depressed adults: randomised controlled trial

OBJECTIVE

To investigate the effectiveness of facilitated physical activity as an adjunctive treatment for adults with depression presenting in primary care.

DESIGN

Pragmatic, multicentre, two arm parallel randomised controlled trial.

SETTING

General practices in Bristol and Exeter.

PARTICIPANTS

361 adults aged 18-69 who had recently consulted their general practitioner with symptoms of depression. All those randomised had a diagnosis of an episode of depression as assessed by the clinical interview schedule-revised and a Beck depression inventory score of 14 or more.

INTERVENTIONS

In addition to usual care, intervention participants were offered up to three face to face sessions and 10 telephone calls with a trained physical activity facilitator over eight months. The intervention was based on theory and aimed to provide individually tailored support and encouragement to engage in physical activity.

MAIN OUTCOME MEASURES

The primary outcome was self reported symptoms of depression, assessed with the Beck depression inventory at four months post-randomisation. Secondary outcomes included use of antidepressants and physical activity at the four, eight, and 12 month follow-up points, and symptoms of depression at eight and 12 month follow-up.

RESULTS

There was no evidence that participants offered the physical activity intervention reported improvement in mood by the four month follow-up point compared with those in the usual care group; adjusted between group difference in mean Beck depression inventory score -0.54 (95% confidence interval -3.06 to 1.99; P=0.68). Similarly, there was no evidence that the intervention group reported a change in mood by the eight and 12 month follow-up points. Nor was there evidence that the intervention reduced antidepressant use compared with usual care (adjusted odds ratio 0.63, 95% confidence interval 0.19 to 2.06; P=0.44) over the duration of the trial. However, participants allocated to the intervention group reported more physical activity during the follow-up period than those allocated to the usual care group (adjusted odds ratio 2.27, 95% confidence interval 1.32 to 3.89; P=0.003).

CONCLUSIONS

The addition of a facilitated physical activity intervention to usual care did not improve depression outcome or reduce use of antidepressants compared with usual care alone.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN16900744.

Effective elements of cognitive behaviour therapy for psychosis: results of a novel type of subgroup analysis based on principal stratification

BACKGROUND

Meta-analyses show that cognitive behaviour therapy for psychosis (CBT-P) improves distressing positive symptoms. However, it is a complex intervention involving a range of techniques. No previous study has assessed the delivery of the different elements of treatment and their effect on outcome. Our aim was to assess the differential effect of type of treatment delivered on the effectiveness of CBT-P, using novel statistical methodology.

METHOD

The Psychological Prevention of Relapse in Psychosis (PRP) trial was a multi-centre randomized controlled trial (RCT) that compared CBT-P with treatment as usual (TAU). Therapy was manualized, and detailed evaluations of therapy delivery and client engagement were made. Follow-up assessments were made at 12 and 24 months. In a planned analysis, we applied principal stratification (involving structural equation modelling with finite mixtures) to estimate intention-to-treat (ITT) effects for subgroups of participants, defined by qualitative and quantitative differences in receipt of therapy, while maintaining the constraints of randomization.

RESULTS

Consistent delivery of full therapy, including specific cognitive and behavioural techniques, was associated with clinically and statistically significant increases in months in remission, and decreases in psychotic and affective symptoms. Delivery of partial therapy involving engagement and assessment was not effective.

CONCLUSIONS

Our analyses suggest that CBT-P is of significant benefit on multiple outcomes to patients able to engage in the full range of therapy procedures. The novel statistical methods illustrated in this report have general application to the evaluation of heterogeneity in the effects of treatment.

Allowing for missing outcome data and incomplete uptake of randomised interventions, with application to an Internet-based alcohol trial

Missing outcome data and incomplete uptake of randomised interventions are common problems, which complicate the analysis and interpretation of randomised controlled trials, and are rarely addressed well in practice. To promote the implementation of recent methodological developments, we describe sequences of randomisation-based analyses that can be used to explore both issues. We illustrate these in an Internet-based trial evaluating the use of a new interactive website for those seeking help to reduce their alcohol consumption, in which the primary outcome was available for less than half of the participants and uptake of the intervention was limited.

For missing outcome data, we first employ data on intermediate outcomes and intervention use to make a missing at random assumption more plausible, with analyses based on general estimating equations, mixed models and multiple imputation. We then use data on the ease of obtaining outcome data and sensitivity analyses to explore departures from the missing at random assumption. For incomplete uptake of randomised interventions, we estimate structural mean models by using instrumental variable methods.

In the alcohol trial, there is no evidence of benefit unless rather extreme assumptions are made about the missing data nor an important benefit in more extensive users of the intervention. These findings considerably aid the interpretation of the trial's results. More generally, the analyses proposed are applicable to many trials with missing outcome data or incomplete intervention uptake. To facilitate use by others, Stata code is provided for all methods. Copyright © 2011 John Wiley & Sons, Ltd.

On-line randomized controlled trial of an internet based psychologically enhanced intervention for people with hazardous alcohol consumption

Background

Interventions delivered via the Internet have the potential to address the problem of hazardous alcohol consumption at minimal incremental cost, with potentially major public health implications. It was hypothesised that providing access to a psychologically enhanced website would result in greater reductions in drinking and related problems than giving access to a typical alcohol website simply providing information on potential harms of alcohol. DYD-RCT Trial registration: ISRCTN 31070347.

Methodology/Principal Findings

A two-arm randomised controlled trial was conducted entirely on-line through the Down Your Drink (DYD) website. A total of 7935 individuals who screened positive for hazardous alcohol consumption were recruited and randomized. At entry to the trial, the geometric mean reported past week alcohol consumption was 46.0 (SD 31.2) units. Consumption levels reduced substantially in both groups at the principal 3 month assessment point to an average of 26.0 (SD 22.3) units. Similar changes were reported at 1 month and 12 months. There were no significant differences between the groups for either alcohol consumption at 3 months (intervention: control ratio of geometric means 1.03, 95% CI 0.97 to 1.10) or for this outcome and the main secondary outcomes at any of the assessments. The results were not materially changed following imputation of missing values, nor was there any evidence that the impact of the intervention varied with baseline measures or level of exposure to the intervention.

Conclusions/Significance

Findings did not provide support for the hypothesis that access to a psychologically enhanced website confers additional benefit over standard practice and indicate the need for further research to optimise the effectiveness of Internet-based behavioural interventions. The trial demonstrates a widespread and potentially sustainable demand for Internet based interventions for people with hazardous alcohol consumption, which could be delivered internationally.

Trial Registration

Controlled-Trials.com ISRCTN31070347

The statistical pitfalls of the partially randomized preference design in non-blinded trials of psychological interventions

In a partially randomized preference trial (PRPT) patients with no treatment preference are allocated to groups at random, but those who express a preference receive the treatment of their choice. It has been suggested that the design can improve the external and internal validity of trials. We used computer simulation to illustrate the impact that an unmeasured confounder could have on the results and conclusions drawn from a PRPT. We generated 4000 observations ("patients") that reflected the distribution of the Beck Depression Index (DBI) in trials of depression. Half were randomly assigned to a randomized controlled trial (RCT) design and half were assigned to a PRPT design. In the RCT, "patients" were evenly split between treatment and control groups; whereas in the preference arm, to reflect patient choice, 87.5% of patients were allocated to the experimental treatment and 12.5% to the control. Unadjusted analyses of the PRPT data consistently overestimated the treatment effect and its standard error. This lead to Type I errors when the true treatment effect was small and Type II errors when the confounder effect was large. The PRPT design is not recommended as a method of establishing an unbiased estimate of treatment effect due to the potential influence of unmeasured confounders.

Using the instrumental variables estimator to analyze noninferiority trials with noncompliance

The standard intent-to-treat (ITT) approach can lead to erroneous conclusions about treatment efficacy in noninferiority trials with noncompliance. Per-protocol and as-treated analyses are also known to result in biased comparisons of treatment effects. Alternative statistical methods are therefore needed to better address the effects of noncompliance in noninferiority trials. In this paper, we consider the use of the instrumental variables (IV) estimator in a noninferiority trial with a binary outcome and evaluate the performance of this approach in comparison to other conventional analytic methods. Unlike the ITT, per-protocol, and as-treated approaches, the IV method provides an unbiased estimate of the average causal effect of treatment among the subgroup of compliers and maintains the nominal type I error rate, but does increase the sample size requirements of the study as the expected proportion of noncompliers increases. Further development of the IV estimator for more general patterns of noncompliance would be useful and would encourage broader application of this method in noninferiority trials.

A method to estimate treatment efficacy among latent subgroups of a randomized clinical trial

Subgroup analysis arises in clinical trials research when we wish to estimate a treatment effect on a specific subgroup of the population distinguished by baseline characteristics. Many trial designs induce latent subgroups such that subgroup membership is observable in one arm of the trial and unidentified in the other. This occurs, for example, in oncology trials when a biopsy or dissection is performed only on subjects randomized to active treatment. We discuss a general framework to estimate a biological treatment effect on the latent subgroup of interest when the survival outcome is right-censored and can be appropriately modelled as a parametric function of covariate effects. Our framework builds on the application of instrumental variables methods to all-or-none treatment noncompliance. We derive a computational method to estimate model parameters via the EM algorithm and provide guidance on its implementation in standard software packages. The research is illustrated through an analysis of a seminal melanoma trial that proposed a new standard of care for the disease and involved a biopsy that is available only on patients in the treatment arm.

Protocol for a randomised controlled trial of a school based cognitive behaviour therapy (CBT) intervention to prevent depression in high risk adolescents (PROMISE)

BACKGROUND

Depression in adolescents is a significant problem that impairs everyday functioning and increases the risk of severe mental health disorders in adulthood. Relatively few adolescents with depression are identified and referred for treatment indicating the need to investigate alternative preventive approaches.

STUDY DESIGN

A pragmatic cluster randomised controlled trial evaluating the effectiveness of a school based prevention programme on symptoms of depression in "high risk" adolescents (aged 12-16). The unit of allocation is year groups (n = 28) which are assigned to one of three conditions: an active intervention based upon cognitive behaviour therapy, attention control or treatment as usual. Assessments will be undertaken at screening, baseline, 6 months and 12 months. The primary outcome measure is change on the Short Mood and Feeling Questionnaire at 12 months. Secondary outcome measures will assess changes in negative thoughts, self esteem, anxiety, school connectedness, peer attachment, alcohol and substance misuse, bullying and self harm.

DISCUSSION

As of August 2010, all 28 year groups (n = 5023) had been recruited and the assigned interventions delivered. Final 12 month assessments are scheduled to be completed by March 2011.

TRIAL REGISTRATION

ISRCTN19083628.

Physical activity as a treatment for depression: the TREAD randomised trial protocol

BACKGROUND

Depression is one of the most common reasons for consulting a General Practitioner (GP) within the UK. Whilst antidepressants have been shown to be clinically effective, many patients and healthcare professionals would like to access other forms of treatment as an alternative or adjunct to drug therapy for depression. A recent systematic review presented some evidence that physical activity could offer one such option, although further investigation is needed to test its effectiveness within the context of the National Health Service.The aim of this paper is to describe the protocol for a randomised, controlled trial (RCT) designed to evaluate an intervention developed to increase physical activity as a treatment for depression within primary care.

METHODS/DESIGN

The TREAD study is a pragmatic, multi-centre, two-arm RCT which targets patients presenting with a new episode of depression. Patients were approached if they were aged 18-69, had recently consulted their GP for depression and, where appropriate, had been taking antidepressants for less than one month. Only those patients with a confirmed diagnosis of a depressive episode as assessed by the Clinical Interview Schedule-Revised (CIS-R), a Beck Depression Inventory (BDI) score of at least 14 and informed written consent were included in the study. Eligible patients were individually randomised to one of two treatment groups; usual GP care or usual GP care plus facilitated physical activity. The primary outcome of the trial is clinical symptoms of depression assessed using the BDI four months after randomisation. A number of secondary outcomes are also measured at the 4-, 8- and 12-month follow-up points including quality of life, attitude to and involvement in physical activity and antidepressant use/adherence. Outcomes will be analysed on an intention-to-treat (ITT) basis and will use linear and logistic regression models to compare treatments.

DISCUSSION

The results of the trial will provide information about the effectiveness of physical activity as a treatment for depression. Given the current prevalence of depression and its associated economic burden, it is hoped that TREAD will provide a timely contribution to the evidence on treatment options for patients, clinicians and policy-makers.

TRIAL REGISTRATION

ISRCTN 16900744.

CRIMSON [CRisis plan IMpact: Subjective and Objective coercion and eNgagement] protocol: a randomised controlled trial of joint crisis plans to reduce compulsory treatment of people with psychosis

Background

The use of compulsory treatment under the Mental Health Act (MHA) has continued to rise in the UK and in other countries. The Joint Crisis Plan (JCP) is a statement of service users' wishes for treatment in the event of a future mental health crisis. It is developed with the clinical team and an independent facilitator. A recent pilot RCT showed a reduction in the use of the MHA amongst service users with a JCP. The JCP is the only intervention that has been shown to reduce compulsory treatment in this way. The CRIMSON trial aims to determine if JCPs, compared with treatment as usual, are effective in reducing the use of the MHA in a range of treatment settings across the UK.

Methods/Design

This is a 3 centre, individual-level, single-blind, randomised controlled trial of the JCP compared with treatment as usual for people with a history of relapsing psychotic illness in Birmingham, London and Lancashire/Manchester. 540 service users will be recruited across the three sites. Eligible service users will be adults with a diagnosis of a psychotic disorder (including bipolar disorder), treated in the community under the Care Programme Approach with at least one admission to a psychiatric inpatient ward in the previous two years. Current inpatients and those subject to a community treatment order will be excluded to avoid any potential perceived pressure to participate. Research assessments will be conducted at baseline and 18 months. Following the baseline assessment, eligible service users will be randomly allocated to either develop a Joint Crisis Plan or continue with treatment as usual. Outcome will be assessed at 18 months with assessors blind to treatment allocation. The primary outcome is the proportion of service users treated or otherwise detained under an order of the Mental Health Act (MHA) during the follow-up period, compared across randomisation groups. Secondary outcomes include overall costs, service user engagement, perceived coercion and therapeutic relationships. Sub-analyses will explore the effectiveness of the JCP in reducing use of the MHA specifically for Black Caribbean and Black African service users (combined). Qualitative investigations with staff and service users will explore the acceptability of the JCPs.

Discussion

JCPs offer a potential solution to the rise of compulsory treatment for individuals with psychotic disorders and, if shown to be effective in this trial, they are likely to be of interest to mental health service providers worldwide.

Trial registration

Current Controlled Trials ISRCTN11501328

An IV for the RCT: using instrumental variables to adjust for treatment contamination in randomised controlled trials

Although the randomised controlled trial is the "gold standard" for studying the efficacy and safety of medical treatments, it is not necessarily free from bias. When patients do not follow the protocol for their assigned treatment, the resultant "treatment contamination" can produce misleading findings. The methods used historically to deal with this problem, the "as treated" and "per protocol" analysis techniques, are flawed and inaccurate. Intention to treat analysis is the solution most often used to analyse randomised controlled trials, but this approach ignores this issue of treatment contamination. Intention to treat analysis estimates the effect of recommending a treatment to study participants, not the effect of the treatment on those study participants who actually received it. In this article, we describe a simple yet rarely used analytical technique, the "contamination adjusted intention to treat analysis," which complements the intention to treat approach by producing a better estimate of the benefits and harms of receiving a treatment. This method uses the statistical technique of instrumental variable analysis to address contamination. We discuss the strengths and limitations of the current methods of addressing treatment contamination and the contamination adjusted intention to treat technique, provide examples of effective uses, and discuss how using estimates generated by contamination adjusted intention to treat analysis can improve clinical decision making and patient care.

Noncompliance in lifestyle intervention studies: the instrumental variable method provides insight into the bias

OBJECTIVE

In lifestyle intervention trials, participants of the control group often change their behavior despite the request to maintain their usual lifestyle pattern. These changes in the control group and changes in addition to the intended in the intervention group can lead to undesirable confounding effects.

STUDY DESIGN AND SETTING

We address several considerations for study design to prevent noncompliance or minimize its effects. Furthermore, we demonstrate how the instrumental variable method can give insight into the extent of bias introduced by noncompliance in randomized trials, within the context of the Sex Hormones and Physical Exercise study.

RESULTS

Noncompliance can be prevented by measures taken in the design phase of a study, for example, limited duration of the study, clear recommendations, power calculation, intensity of the intervention, involvement of the control group, waiting-list control group, and single-consent design nested within an observational study. When nevertheless noncompliance does occur, the instrumental variable method estimates the intervention effect of treatment among the compliers.

CONCLUSION

Noncompliance can seriously affect validity of lifestyle trial results. Its occurrence should be prevented by taking measures during the design phase of a study. The instrumental variable method can give insight into confounding by noncompliance in randomized trials.

Intention to treat, per protocol, as treated and instrumental variable estimators given non-compliance and effect heterogeneity

We consider the behaviour of three approaches to efficacy estimation--using so-called 'as treated' (AT), 'per protocol' (PP) and 'instrumental variable' (IV) analyses--and of the Intention to Treat estimator, in a two-arm randomized treatment trial with a Normally distributed outcome when there is treatment effect heterogeneity and non-random compliance with assigned treatment. Formulae are derived for the bias of estimators when used either to estimate average treatment effect (ACE) or to estimate complier average treatment effect (CACE) under several models for the relationship between compliance and potential outcomes. These enable the expected values of AT, PP and IV estimators to be ranked in relation to ACE, and show that AT and PP estimators are generally biased for both ACE and CACE even under homogeneity. However, we show that the difference between any pair of (AT, PP, IV) estimates can be used to estimate the correlation between the latent variable determining compliance behaviour and one potential outcome. In the absence of measures that predict compliance, bounds for ACE can only be set given strong assumptions. Regarding the Intention to Treat estimator, while this is 'biased towards the null' if viewed as a measure of CACE, we show that it is not always so in relation to ACE. Finally we discuss the behaviour of the estimators under weak and strong null hypotheses.

Therapist-delivered Internet psychotherapy for depression in primary care: a randomised controlled trial

BACKGROUND

Despite strong evidence for its effectiveness, cognitive-behavioural therapy (CBT) remains difficult to access. Computerised programs have been developed to improve accessibility, but whether these interventions are responsive to individual needs is unknown. We investigated the effectiveness of CBT delivered online in real time by a therapist for patients with depression in primary care.

METHODS

In this multicentre, randomised controlled trial, 297 individuals with a score of 14 or more on the Beck depression inventory (BDI) and a confirmed diagnosis of depression were recruited from 55 general practices in Bristol, London, and Warwickshire, UK. Participants were randomly assigned, by a computer-generated code, to online CBT in addition to usual care (intervention; n=149) or to usual care from their general practitioner while on an 8-month waiting list for online CBT (control; n=148). Participants, researchers involved in recruitment, and therapists were masked in advance to allocation. The primary outcome was recovery from depression (BDI score <10) at 4 months. Analysis was by intention to treat. This trial is registered, number ISRCTN 45444578.

FINDINGS

113 participants in the intervention group and 97 in the control group completed 4-month follow-up. 43 (38%) patients recovered from depression (BDI score <10) in the intervention group versus 23 (24%) in the control group at 4 months (odds ratio 2.39, 95% CI 1.23-4.67; p=0.011), and 46 (42%) versus 26 (26%) at 8 months (2.07, 1.11-3.87; p=0.023).

INTERPRETATION

CBT seems to be effective when delivered online in real time by a therapist, with benefits maintained over 8 months. This method of delivery could broaden access to CBT.

FUNDING

BUPA Foundation.

Mediation and moderation of treatment effects in randomised controlled trials of complex interventions

Complex intervention trials should be able to answer both pragmatic and explanatory questions in order to test the theories motivating the intervention and help understand the underlying nature of the clinical problem being tested. Key to this is the estimation of direct effects of treatment and indirect effects acting through intermediate variables which are measured post-randomisation. Using psychological treatment trials as an example of complex interventions, we review statistical methods which crucially evaluate both direct and indirect effects in the presence of hidden confounding between mediator and outcome. We review the historical literature on mediation and moderation of treatment effects. We introduce two methods from within the existing causal inference literature, principal stratification and structural mean models, and demonstrate how these can be applied in a mediation context before discussing approaches and assumptions necessary for attaining identifiability of key parameters of the basic causal model. Assuming that there is modification by baseline covariates of the effect of treatment (i.e. randomisation) on the mediator (i.e. covariate by treatment interactions), but no direct effect on the outcome of these treatment by covariate interactions leads to the use of instrumental variable methods. We describe how moderation can occur through post-randomisation variables, and extend the principal stratification approach to multiple group methods with explanatory models nested within the principal strata. We illustrate the new methodology with motivating examples of randomised trials from the mental health literature.

Recombinant human erythropoiesis-stimulating agents and mortality in patients with cancer: a meta-analysis of randomised trials

BACKGROUND

Erythropoiesis-stimulating agents reduce anaemia in patients with cancer and could improve their quality of life, but these drugs might increase mortality. We therefore did a meta-analysis of randomised controlled trials in which these drugs plus red blood cell transfusions were compared with transfusion alone for prophylaxis or treatment of anaemia in patients with cancer.

METHODS

Data for patients treated with epoetin alfa, epoetin beta, or darbepoetin alfa were obtained and analysed by independent statisticians using fixed-effects and random-effects meta-analysis. Analyses were by intention to treat. Primary endpoints were mortality during the active study period and overall survival during the longest available follow-up, irrespective of anticancer treatment, and in patients given chemotherapy. Tests for interactions were used to identify differences in effects of erythropoiesis-stimulating agents on mortality across prespecified subgroups.

FINDINGS

Data from a total of 13 933 patients with cancer in 53 trials were analysed. 1530 patients died during the active study period and 4993 overall. Erythropoiesis-stimulating agents increased mortality during the active study period (combined hazard ratio [cHR] 1.17, 95% CI 1.06-1.30) and worsened overall survival (1.06, 1.00-1.12), with little heterogeneity between trials (I(2) 0%, p=0.87 for mortality during the active study period, and I(2) 7.1%, p=0.33 for overall survival). 10 441 patients on chemotherapy were enrolled in 38 trials. The cHR for mortality during the active study period was 1.10 (0.98-1.24), and 1.04 (0.97-1.11) for overall survival. There was little evidence for a difference between trials of patients given different anticancer treatments (p for interaction=0.42).

INTERPRETATION

Treatment with erythropoiesis-stimulating agents in patients with cancer increased mortality during active study periods and worsened overall survival. The increased risk of death associated with treatment with these drugs should be balanced against their benefits.

FUNDING

German Federal Ministry of Education and Research, Medical Faculty of University of Cologne, and Oncosuisse (Switzerland).

Estimating dose-response effects in psychological treatment trials: the role of instrumental variables

We present a relatively non-technical and practically orientated review of statistical methods that can be used to estimate dose-response relationships in randomised controlled psychotherapy trials in which participants fail to attend all of the planned sessions of therapy. Here we are investigating the effects on treatment outcome of the number of sessions attended when the latter is possibly subject to hidden selection effects (hidden confounding). The aim is to estimate the parameters of a structural mean model (SMM) using randomisation, and possibly randomisation by covariate interactions, as instrumental variables. We describe, compare and illustrate the equivalence of the use of a simple G-estimation algorithm and two two-stage least squares procedures that are traditionally used in economics.

Estimating intervention effects of prevention programs: accounting for noncompliance

Individuals not fully complying with their assigned treatments is a common problem encountered in randomized evaluations of behavioral interventions. Treatment group members rarely attend all sessions or do all "required" activities; control group members sometimes find ways to participate in aspects of the intervention. As a result, there is often interest in estimating both the effect of being assigned to participate in the intervention, as well as the impact of actually participating and doing all of the required activities. Methods known broadly as "complier average causal effects" (CACE) or "instrumental variables" (IV) methods have been developed to estimate this latter effect, but they are more commonly applied in medical and treatment research. Since the use of these statistical techniques in prevention trials has been less widespread, many prevention scientists may not be familiar with the underlying assumptions and limitations of CACE and IV approaches. This paper provides an introduction to these methods, described in the context of randomized controlled trials of two preventive interventions: one for perinatal depression among at-risk women and the other for aggressive disruptive behavior in children. Through these case studies, the underlying assumptions and limitations of these methods are highlighted.

Estimating effects from randomized trials with discontinuations: the need for intent-to-treat design and G-estimation

BACKGROUND

Randomized trials provide pivotal evidence for evaluation and approval of therapies. Nonetheless, such trials are often plagued by noncompliance, especially in the form of premature discontinuation of treatment. While intent-to-treat (ITT) analysis can provide valid tests of no-effect hypotheses, some trials may make ITT analysis impossible by ceasing follow-up when patients go off assigned treatment. Furthermore, estimates based on ITT, on-treatment, or per-protocol comparisons can seriously understate harm or benefit.

PURPOSE

To show how g-estimation based on randomization status is a natural generalization of ITT null testing to estimating efficacy from trials with important discontinuation or noncompliance.

METHODS

We contrast with an analysis of the effect of a tiotropium inhaler on the occurrence of chronic obstructive pulmonary disease (COPD) events in a six-month double-blind placebo-controlled trial of 1829 patients with good but imperfect compliance.

RESULTS

The covariate-adjusted point estimates, 95% confidence limits (CL), and null P-values comparing expected COPD event times in placebo versus tiotropium patients were: ITT, 1.21, CL = 1.02, 1.43, P = 0.027; on-treatment, 1.27, CL = 1.06, 1.52, P = 0.009; per-protocol, 1.36, CL = 1.13, 1.63, P = 0.001; and g-estimation, 1.31, CL = 1.03,1.72, P = 0.027. Thus g-estimation preserved the ITT test of the null, but exhibited more uncertainty about the size of the tiotropium effect than the other methods. In particular, it allowed for a much larger potential effect than did ITT analysis, but produced a much larger null P than exhibited by per-protocol analysis.

LIMITATIONS

Like ITT analysis, g-estimation requires all patients be followed to the end of the trial protocol, regardless of whether they comply with the protocol. Like on-treatment and per-protocol analyses, it also requires accurate compliance information be recorded.

CONCLUSION

G-estimation should become a standard procedure for the analysis of trials with noncompliance. Software to do so is available in major packages, and the procedure is easily coded for other packages.

Incomplete adherence to an adjunctive systemic antibiotic regimen decreases clinical outcomes in generalized aggressive periodontitis patients: a pilot retrospective study

AIM

The objective of this study was to explore the effect of incomplete adherence to the prescribed antibiotic regimen, amoxicillin and metronidazole, in the non-surgical treatment of generalized aggressive periodontitis (GAP).

METHODS

This retrospective study included 18 GAP subjects who received a conventional course of full-mouth non-surgical periodontal treatment using machine-driven and hand instruments and an adjunctive course of systemic antibiotics (500 mg amoxicillin and 500 mg metronidazole three times a day for 7 days). Clinical parameters were collected at baseline and at 2 months post-treatment. Self-reported adherence to the prescribed medication regimen was recorded at 2 months.

RESULTS

All clinical parameters, except for the mean clinical attachment level (CAL) in sites with initial probing pocket depth (PPD) < or = 3 mm, improved at 2 months in all subjects. PPD reduction was 3.7 mm [95% confidence interval (CI) 3.2, 4.3 mm] in deep pockets (> or = 7 mm) and 2.2 mm (95% CI 1.9, 2.4 mm) in moderate pockets (4-6 mm), while CAL gain was 2.2 mm (95% CI 1.7, 2.6 mm) and 1.2 mm (95% CI 0.8, 1.5 mm), respectively. However, only 11 subjects (61.1%) reported full adherence to the medication. In deep pockets (> or = 7 mm), the difference between an adherent and non-adherent/partially adherent subject was 0.9 mm (95% CI 0.1, 1.7 mm, ancova, p=0.027) in PPD reduction and 0.8 mm (95% CI -0.2, 1.9, p=0.129) in CAL gain at 2 months. In moderate pockets (4-6 mm) this difference was smaller in magnitude: 0.4 mm (95% CI 0.1, 0.9 mm, p=0.036) in PPD reduction and 0.2 mm (95% CI -0.3, 0.9 mm, p=0.332) in CAL gain.

CONCLUSIONS

Within the limits of this design, these data suggest that incomplete adherence to a 7-day adjunctive course of systemic metronidazole and amoxicillin is associated with decreased clinical outcomes in subjects with generalized aggressive periodontitis.

Adherence to pharmacological treatment for juvenile bipolar disorder

OBJECTIVE

The objective of this study was to describe the prevalence and correlates of adherence to divalproex sodium (DVPX) and lithium carbonate (Li) combination treatment during the initial stabilization treatment phase.

METHOD

Adherence to Li/DVPX combination therapy was measured by the presence or absence of minimum serum concentrations of DVPX (50 microg/mL) or Li (0.6 mmol/L). Secondary measures included pill count, patient/parent report, and clinical judgment. Correlates of adherence, including patient characteristics, medication side effects, and family variables, were evaluated.

RESULTS

One hundred seven patients (70 males and 37 females) were studied. The proportion of serum concentrations in the therapeutic range across the study period was 0.84 for DVPX and 0.66 for Li. Maternal (r = -0.31; p<.01) and paternal (r = -0.44; p < .01) hospitalization for a psychiatric disorder and less adaptive family functioning (r=-0.26; p < .05) related to treatment nonadherence for DVPX. Better treatment adherence to DVPX (r = 0.21; p < .05) and Li (r = 0.23; p < .05) was associated with a greater number of side effects, whereas male sex was associated with worse adherence to both DVPX (r= -0.24; p < .05) and Li (r = -0.22; p < .05) pharmacotherapy. Clinical response to treatment correlated with adherence to DVPX treatment (r = 0.33; p < .01).

CONCLUSIONS

Nonadherence may limit the statistical power of treatment efficacy studies and the effectiveness of pharmacotherapy treatment for juvenile BPD and necessitate strategies to evaluate and enhance levels of treatment adherence.

Laparoscopic versus open colposuspension--results of a prospective randomised controlled trial

OBJECTIVE

To compare the effectiveness and cost effectiveness of open and laparoscopic colposuspension in the treatment of stress urinary incontinence.

DESIGN

A randomised controlled trial. Women were randomised between March 1999 and February 2002 and were seen for assessment at 6, 12 and 24 months, postoperatively.

SETTING

Women were recruited from six gynaecology units in the UK.

POPULATION

Women with proven stress urinary incontinence requiring surgery.

METHODS

Open abdominal retropubic colposuspension or laparoscopic colposuspension.

MAIN OUTCOME MEASURES

Primary outcomes were subjective (satisfaction with outcome) and objective (negative 1-hour pad test). Secondary outcomes were operative and postoperative morbidity and quality of life. The study was powered to demonstrate noninferiority, i.e. that the absolute cure rate of laparoscopic colposuspension was no more than 15% below that of open colposuspension.

RESULTS

A total of 291 women were randomised, with 24-month data on subjective and objective outcomes in 88 and 82.5%, respectively. The intention-to-treat analysis indicated no significant difference in cure rates between open and laparoscopic surgery. The objective cure rates for open and laparoscopic were 70.1 and 79.7%, respectively. Subjective cure rates by satisfaction were lower than objective cure; 54.6 and 54.9%, respectively, and there was considerable nonconcordance both ways.

CONCLUSIONS

Laparoscopic colposuspension is not inferior to open colposuspension in terms of curing stress urinary incontinence.

Randomized controlled trial of effect of prescription of clozapine versus other second-generation antipsychotic drugs in resistant schizophrenia

There is good evidence that clozapine is more efficacious than first-generation antipsychotic drugs in resistant schizophrenia. It is less clear if clozapine is more effective than the other second-generation antipsychotic (SGA) drugs. A noncommercially funded, pragmatic, open, multisite, randomized controlled trial was conducted in the United Kingdom National Health Service (NHS). Participants were 136 people aged 18-65 with DSM-IV schizophrenia and related disorders whose medication was being changed because of poor clinical response to 2 or more previous antipsychotic drugs. Participants were randomly allocated to clozapine or to one of the class of other SGA drugs (risperidone, olanzapine, quetiapine, amisulpride) as selected by the managing clinician. Outcomes were assessed blind to treatment allocation. One-year assessments were carried out in 87% of the sample. The intent to treat comparison showed no statistically significant advantage for commencing clozapine in Quality of Life score (3.63 points; CI: 0.46-7.71; p = .08) but did show an advantage in Positive and Negative Syndrome Scale (PANSS) total score that was statistically significant (-4.93 points; CI: -8.82 to -1.05; p = .013) during follow-up. Clozapine showed a trend toward having fewer total extrapyramidal side effects. At 12 weeks participants who were receiving clozapine reported that their mental health was significantly better compared with those receiving other SGA drugs. In conclusion, in people with schizophrenia with poor treatment response to 2 or more antipsychotic drugs, there is an advantage to commencing clozapine rather than other SGA drugs in terms of symptom improvement over 1 year.

Fatigue Intervention by Nurses Evaluation--the FINE Trial. A randomised controlled trial of nurse led self-help treatment for patients in primary care with chronic fatigue syndrome: study protocol. [ISRCTN74156610]

BACKGROUND

Chronic fatigue syndrome, also known as ME (CFS/ME), is a condition characterised primarily by severe, disabling fatigue, of unknown origin, which has a poor prognosis and serious personal and economic consequences. Evidence for the effectiveness of any treatment for CFS/ME in primary care, where most patients are seen, is sparse. Recently, a brief, pragmatic treatment for CFS/ME, based on a physiological dysregulation model of the condition, was shown to be successful in improving fatigue and physical functioning in patients in secondary care. The treatment involves providing patients with a readily understandable explanation of their symptoms, from which flows the rationale for a graded rehabilitative plan, developed collaboratively with the therapist. The present trial will test the effectiveness and cost-effectiveness of pragmatic rehabilitation when delivered by specially trained general nurses in primary care. We selected a client-centred counselling intervention, called supportive listening, as a comparison treatment. Counselling has been shown to be as effective as cognitive behaviour therapy for treating fatigue in primary care, is more readily available, and controls for supportive therapist contact time. Our control condition is treatment as usual by the general practitioner (GP).

METHODS AND DESIGN

This study protocol describes the design of an ongoing, single-blind, pragmatic randomized controlled trial of a brief (18 week) self-help treatment, pragmatic rehabilitation, delivered by specially trained nurse-therapists in patients' homes, compared with nurse-therapist delivered supportive listening and treatment as usual by the GP. An economic evaluation, taking a societal viewpoint, is being carried out alongside the clinical trial. Three adult general nurses were trained over a six month period to deliver the two interventions. Patients aged over 18 and fulfilling the Oxford criteria for CFS are assessed at baseline, after the intervention, and again one year later. Primary outcomes are self-reported physical functioning and fatigue at one year, and will be analysed on an intention-to-treat basis. A qualitative study will examine the interventions' mechanisms of change, and also GPs' drivers and barriers towards referral.

Structural mean models for compliance analysis in randomized clinical trials and the impact of errors on measures of exposure

Partial compliance with assigned treatment regimes is common in drug trials and calls for a causal analysis of the effect of treatment actually received. As such observed exposure is no longer randomized, selection bias must be carefully accounted for. The framework of potential outcomes allows this by defining a subject-specific treatment-free reference outcome, which may be latent and is modelled in relation to the observed (treated) data. Causal parameters enter these structural models explicitly. In this paper we review recent progress in randomization-based inference for structural mean modelling, from the additive linear model to the structural generalized linear models. An arsenal of tools currently available for standard association regression has steadily been developed in the structural setting, providing many parallel features to help randomization-based inference. We argue that measurement error on exposure is an important practical complication that has, however, not yet been addressed. We show how standard additive linear structural mean models are robust against unbiased measurement error and how efficient, asymptotically unbiased inference can be drawn when the degree of measurement error bias is known. The impact of measurement error is illustrated in a blood pressure example and finite sample properties are verified by simulation. We end with a plea for more and careful use of this methodology and point to directions for further development.