Randomized Preference Trial Research Articles

Effect of Treatment Preference in Randomized Controlled Trials: Systematic Review of the Literature and Meta-Analysis.

A significant limitation of the traditional randomized controlled trials is that strong preferences for (or against) one treatment may influence outcomes and/or willingness to receive treatment. Several trial designs incorporating patient preference have been introduced to examine the effect of treatment preference separately from the effects of individual interventions. In the current study, we summarized results from studies using doubly randomized preference trial (DRPT) or fully randomized preference trial (FRPT) designs and examined the effect of treatment preference on clinical outcomes. The current systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Studies using DRPT or FRPT design were identified using electronic databases, including PubMed, Cochrane Library, EMBASE, and Google Scholar between January 1989 and November 2018. All studies included in this meta-analysis were examined to determine the extent to which giving patients their preferred treatment option influenced clinical outcomes. The following data were extracted from included studies: study characteristics, sample size, study duration, follow-up, patient characteristics, and clinical outcomes. We further appraised risk of bias for the included studies using the Cochrane Collaboration's risk of bias tool. The search identified 374 potentially relevant articles, of which 27 clinical trials utilized a DRPT or FRPT design and were included in the final analysis. Overall, patients who were allocated to their preferred treatment intervention were more likely to achieve better clinical outcomes [effect size (ES) = 0.18, 95% confidence interval (CI) 0.10-0.26]. Subgroup analysis also found that mental health as well as pain and functional disorders moderated the preference effect (ES = 0.23, 95% CI 0.11-0.36, and ES = 0.09, 95% CI 0.03-0.15, respectively). Matching patients to preferred interventions has previously been shown to promote outcomes such as satisfaction and treatment adherence. Our analysis of current evidence showed that allowing patients to choose their preferred treatment resulted in better clinical outcomes in mental health and pain than giving them a treatment that is not preferred. These results underline the importance of incorporating patient preference when making treatment decisions.

Comparing the Effects of Single- and Multiple-Component Therapies for Insomnia on Sleep Outcomes.

Single- and multiple-component therapies are recommended in professional guidelines for managing chronic insomnia. Systematic reviews point to insufficient evidence of the comparative effectiveness of these therapies, which is required for treatment decision making. To compare the effectiveness of three single-component and one multiple-component therapies on short-term sleep outcomes. The data were obtained from 517 persons with chronic insomnia, enrolled in a partially randomized preference trial. They were allocated to the single-component therapies: sleep education and hygiene (SEH), stimulus control therapy (SCT), and sleep restriction therapy (SRT), or the multiple-component therapy (MCT). The outcomes, perceived insomnia severity and sleep parameters, were assessed with established measures at pre and posttest. Repeated measure analysis of variance was used to compare the outcomes across therapy groups over time. The clinical relevance of the therapies' effects was evaluated by examining the effect size and remission rate. The four therapies differed in their effectiveness in reducing perceived insomnia severity and improving sleep outcomes. SEH was least effective. SCT, SRT, and MCT were moderately effective. SCT and SRT demonstrated slightly higher remission rates than MCT for perceived insomnia severity and some sleep parameters. SCT and SRT are viable single-component therapies that produce clinical benefits. Single-component insomnia treatment may be more convenient to implement in the primary care setting due to the reduced number of treatment recommendations compared to MCT.

The contribution of participant, treatment, and outcome factors to treatment satisfaction.

Treatment satisfaction, which refers to the positive appraisal of process and outcome attributes of a treatment, is a prominent indicator of quality care. Although it is known that participant, treatment, and outcome factors influence treatment satisfaction, it remains unclear which factors contribute to satisfaction with each process and outcome attribute. In this study, we examined the extent to which participant (age, gender, education, race, employment), treatment (type of therapy, method of assignment to therapy), and outcome (self-reported improvement in outcome) factors contribute to satisfaction with the process and outcome attributes of therapies for insomnia. This study consists of a secondary analysis of data obtained from a partially randomized preference trial in which persons with chronic insomnia (N = 517) were assigned to treatment randomly or by preference. Four types of behavioral therapies were included: sleep hygiene, stimulus control therapy, sleep restriction therapy, and multi-component therapy. Self-reported improvement in insomnia and satisfaction were assessed with validated measures at post-test. Multiple regression analysis was used to examine which factors influenced satisfaction with each treatment attribute. The findings showed that treatment and outcome, more so than participant, factors influenced satisfaction with the process and outcome attributes of the behavioral therapies for insomnia. Future research on satisfaction should explore the contribution of treatment (type and preference-matching) and outcome factors on satisfaction to build a better understanding of treatment attributes viewed favorably. Such understanding has the potential to inform modifying or tailoring treatments to improve their acceptance to participants and optimize their effectiveness.

Extensions to the two-stage randomized trial design for testing treatment, self-selection, and treatment preference effects to binary outcomes.

While traditional clinical trials seek to determine treatment efficacy within a specified population, they often ignore the role of a patient's treatment preference on his or her treatment response. The two-stage (doubly) randomized preference trial design provides one approach for researchers seeking to disentangle preference effects from treatment effects. Currently, this two-stage design is limited to the design and analysis of continuous outcome variables; in this presentation, we extend this current design to include binary variables. We present test statistics for testing preference, selection, and treatment effects in a two-stage randomized design with a binary outcome measure, with and without stratification. We also derive closed-form sample size formulas to indicate the number of patients needed to detect each effect. A series of simulation studies explore the properties and efficiency of both the unstratified and stratified two-stage randomized trial designs. Finally, we demonstrate the applicability of these methods using an example of a trial of Hepatitis C treatment.

A randomized preference trial of cognitive-behavioral therapy and yoga for the treatment of worry in anxious older adults

BackgroundWorry is a common problem among older adults. Cognitive-behavioral therapy is the most studied nonpharmacological intervention and it has demonstrated efficacy in reducing late-life worry and anxiety. Although the evidence-base is smaller, yoga has been shown to reduce anxiety and stress. However, little is known about the relative effectiveness of these two nonpharmacological interventions. Further, the impact of patient preference on outcomes is unknown.Purpose: The purpose to this study is to compare the effectiveness of cognitive-behavioral therapy (CBT) with yoga for improving late-life worry, anxiety, and sleep. We will also examine the effects of preference and selection on outcomes, adherence, and attrition.MethodsWe are conducting a two-stage randomized preference trial comparing CBT and yoga for the reduction of worry in a sample of anxious older adults. Five hundred participants will be randomized to either the preference trial (participants choose the intervention; N = 250) or to the randomized trial (participants are randomized to one of the two interventions; N = 250) with equal probability. CBT consists of 10 telephone-based sessions with an accompanying workbook. Yoga consists of 10 weeks of group yoga classes (twice a week) that is modified for use with older adults.ConclusionsThe study design is based on feedback from anxious older adults who wanted more nonpharmacological options for intervention as well as more input into the intervention they receive. It is the first head-to-head comparison of CBT and yoga for reducing late-life worry and anxiety. It will also provide information about how intervention preference affects outcomes.Trial registrationClinicalTrials.gov NCT02968238.

Effects of treatment, choice, and preference on health-related quality-of-life outcomes in patients with posttraumatic stress disorder (PTSD).

Health outcomes may depend on which treatment is received, whether choice of treatment is given, and whether a received treatment is the preferred therapy. We examined the effects of these key factors on the EuroQol-5D (EQ-5D-3L) in patients with PTSD. Two hundred patients aged 18-65years with PTSD diagnosis enrolled in a doubly randomized preference trial (DRPT) examining treatment, choice of treatment, and treatment-preference effects of prolonged exposure therapy (PE) and pharmacotherapy with sertraline (SER) (clinicaltrials.gov Identifier: NCT00127673). We performed difference-in-difference analysis to estimate the treatment effects of prolonged exposure therapy (PE) as compared to pharmacotherapy with sertraline (SER), receipt of choice versus no-choice of treatment, and receipt of preferred versus non-preferred treatment on health-related quality-of-life (HRQOL) outcome using the EQ-5D-3L completed at baseline and 10-week post-treatment. The treatment effects of PE on the EQ-5D scores in overall patients and subgroup of patients who preferred PE were 0.150 (p = 0.025) and 0.223 (p < 0.001), respectively. The effects of treatment choice were 0.088 (p = 0.050) and 0.156 (p = 0.043) in overall patients and subgroup of patients received SER, respectively. The effects of treatment preference were 0.101 (p = 0.038) and 0.249 (p = 0.004) in overall patients and subgroup of patients SER, respectively. Overall, PE is associated with better improved HRQOL, especially in patients who prefer it. Independently, allowing patients to choose their preferred treatment resulted in better HRQOL than either assigning them a treatment or giving them a treatment that is not preferred.

Do Parent Preferences for Child Conduct Problem Interventions Impact Parenting Outcomes? A Pilot Study in Community Children's Mental Health Settings.

A pilot, doubly randomized preference trial was conducted to investigate the impact of providing parents preferences on parenting outcomes. Families with children having conduct problems were randomly assigned to a choice group in which they received their preferred treatment among the four intervention options or a no-choice group in which they were randomized assigned to one of the four options. Results of mixed-effects models showed that parents in the choice group who selected Parent Management Training-Oregon Model (PMTO) had better parenting outcomes over time compared to parents in the choice group who selected child therapy. It highlights the importance of incorporating parent preferences in the delivery of evidence-based treatments.

On Measuring and Reducing Selection Bias With a Quasi‐Doubly Randomized Preference Trial

AbstractRandomized experiments provide unbiased estimates of treatment effects, but are costly and time consuming. We demonstrate how a randomized experiment can be leveraged to measure selection bias by conducting a subsequent observational study that is identical in every way except that subjects choose their treatment—a quasi‐doubly randomized preference trial (quasi‐DRPT). Researchers first strive to think of and measure all possible confounders and then determine how well these confounders as controls can reduce or eliminate selection bias. We use a quasi‐DRPT to study the effect of class time on student performance in an undergraduate introductory microeconomics course at a large public university, illustrating its required design elements: experimental and choice arms conducted in the same setting with identical interventions and measurements, and all confounders measured prospectively to treatment assignment or choice. Quasi‐DRPTs augment randomized experiments in real‐world settings where participants choose their treatments.

Sample size and power for a stratified doubly randomized preference design.

The two-stage (or doubly) randomized preference trial design is an important tool for researchers seeking to disentangle the role of patient treatment preference on treatment response through estimation of selection and preference effects. Up until now, these designs have been limited by their assumption of equal preference rates and effect sizes across the entire study population. We propose a stratified two-stage randomized trial design that addresses this limitation. We begin by deriving stratified test statistics for the treatment, preference, and selection effects. Next, we develop a sample size formula for the number of patients required to detect each effect. The properties of the model and the efficiency of the design are established using a series of simulation studies. We demonstrate the applicability of the design using a study of Hepatitis C treatment modality, specialty clinic versus mobile medical clinic. In this example, a stratified preference design (stratified by alcohol/drug use) may more closely capture the true distribution of patient preferences and allow for a more efficient design than a design which ignores these differences (unstratified version).

What Can We Learn From A Doubly Randomized Preference Trial?—An Instrumental Variables Perspective

AbstractThe doubly randomized preference trial (DRPT) is a randomized experimental design with three arms: a treatment arm, a control arm, and a preference arm. The design has useful properties that have gone unnoticed in the applied and methodological literatures. This paper shows how to interpret the DRPT design using an instrumental variables (IV) framework. The IV framework reveals that the DRPT separately identifies three different treatment effect parameters: the Average Treatment Effect (ATE), the Average Treatment Effect on the Treated (ATT), and the Average Treatment Effect on the Untreated (ATU). The ATE, ATT, and ATU parameters are important for program evaluation research because in realistic settings many social programs are optional rather than mandatory and some people who are eligible for a program choose not to participate. Most of the paper is concerned with the interpretation of the research design. To make the ideas concrete, the final section provides an empirical example using data from an existing DRPT study.

A randomized preference trial to inform personalization of a parent training program implemented in community mental health clinics.

Incorporating participant preferences into intervention decision-making may optimize health outcomes by improving participant engagement. We describe the rationale for a preference-based approach to the personalization of community-based interventions. Compensating for the limitations of traditional randomized controlled trials (RCTs) and partially randomized preference trials (PRPTs), we employed a doubly randomized preference trial in the present study. Families (N = 129) presenting to community mental health clinics for child conduct problems were randomized to choice or no-choice conditions. Within each condition, parents were again randomized, or offered choices between home- and clinic-based, individual and group versions of a parent training program or services-as-usual. Participants were assessed at baseline, and treatment retention data were gathered. Families assigned to the choice condition were significantly less likely to drop out of treatment than those in the no-choice condition. In the choice condition, in-home treatment was the preferred modality, and across conditions, families were less likely to be retained in group and clinic modalities. Research on preferences may boost participant engagement and inform shared decision-making.

Cost-Effectiveness of Prolonged Exposure Therapy Versus Pharmacotherapy and Treatment Choice in Posttraumatic Stress Disorder (the Optimizing PTSD Treatment Trial)

Cost-effectiveness of treatment for posttraumatic stress disorder (PTSD) may depend on type of treatment (eg, pharmacotherapy vs psychotherapy) and patient choice of treatment. We examined the cost-effectiveness of treatment with prolonged exposure therapy versus pharmacotherapy with sertraline, overall treatment preference, preference for choosing prolonged exposure therapy, and preference for choosing pharmacotherapy with sertraline from the US societal perspective. Two hundred patients aged 18 to 65 years with PTSD diagnosis based on DSM-IV criteria enrolled in a doubly randomized preference trial. Patients were randomized to receive their treatment of choice (n = 97) or to be randomly assigned treatment (n = 103). In the choice arm, patients chose either prolonged exposure therapy (n = 61) or pharmacotherapy with sertraline (n = 36). In the no-choice arm, patients were randomized to either prolonged exposure therapy (n = 48) or pharmacotherapy with sertraline (n = 55). The total costs, including direct medical costs, direct nonmedical costs, and indirect costs, were estimated in 2012 US dollars; and total quality-adjusted life-year (QALY) was assessed using the EuroQoL Questionnaire-5 dimensions (EQ-5D) instrument in a 12-month period. This study was conducted from July 2004 to January 2009. Relative to pharmacotherapy with sertraline, prolonged exposure therapy was less costly (-$262; 95% CI, -$5,068 to $4,946) and produced more QALYs (0.056; 95% CI, 0.014 to 0.100) when treatment was assigned, with 93.2% probability of being cost-effective at $100,000/QALY. Independently, giving a choice of treatment also yielded lower cost (-$1,826; 95% CI, -$4,634 to $749) and more QALYs (0.010; 95% CI, -0.019 to 0.044) over no choice of treatment, with 87.0% probability of cost-effectiveness at $100,000/QALY. Giving PTSD patients a choice of treatment appears to be cost-effective. When choice is not possible, prolonged exposure therapy may provide a cost-effective option over pharmacotherapy with sertraline. ClinicalTrials.gov identifier: NCT00127673.

PHS56 - Euroqol (EQ-5D) health utility scores in post-traumatic stress disorder (PTSD) patients: Results from a doubly randomized preference trial (DRPT)

PHS56 - Euroqol (EQ-5D) health utility scores in post-traumatic stress disorder (PTSD) patients: Results from a doubly randomized preference trial (DRPT)

Minimal clinically important differences for the EQ-5D and QWB-SA in Post-traumatic Stress Disorder (PTSD): results from a Doubly Randomized Preference Trial (DRPT)

ObjectiveTo determine the minimal clinically important difference (MCID) for the health-utility measures EuroQol-5 dimensions (EQ-5D) and Quality of Well Being Self-Administered (QWB-SA) Scale in PTSD patients.Research design and methodsTwo hundred patients aged 18 to 65 years with PTSD enrolled in a doubly randomized preference trial (DRPT) examining the treatment and treatment-preference effects between cognitive behavioral therapy and pharmacotherapy with sertraline and completed the EQ-5D and QWB-SA at baseline and 10-week post-treatment. The anchor-based methods utilized a Clinical Global Impression-Improvement (CGI-I) and Clinical Global Impression-Severity. We regressed the changes in EQ-5D and QWB-SA scores on changes in the anchors using ordinary least squares regression. The slopes (beta coefficients) were the rates of change in the anchors as functions of change in EQ-5D and QWB, which represent our estimates of MCID. In addition, we performed receiver operating characteristic (ROC) curve analysis to examine the relationship between the changes in EQ-5D and QWB-SA scores and treatment-response status. The MCIDs were estimated from the ROC curve where they best discriminate between treatment responders and non-responders. The distribution-based methods used small to moderate effect size in terms of 0.2 and 0.5 of standard deviation of the pre-treatment EQ-5D and QWB-SA scores.ResultsThe anchor-based methods estimated the MCID ranges of 0.05 to 0.08 for the EQ-5D and 0.03 to 0.05 for the QWB. The MCID ranges were higher with the distribution-based methods, ranging from 0.04 to 0.10 for the EQ-5D and 0.02 to 0.05 for the QWB-SA.ConclusionsThe established MCID ranges of EQ-5D and QWB-SA can be a useful tool in assessing meaningful changes in patient’s quality of life for researchers and clinicians, and assisting health-policy makers to make informing decision in mental health treatment.Clinical trial registrationClinicaltrials.gov; Identifier: NCT00127673.

Estimating the causal effect of randomization versus treatment preference in a doubly randomized preference trial.

Although randomized studies have high internal validity, generalizability of the estimated causal effect from randomized clinical trials to real-world clinical or educational practice may be limited. We consider the implication of randomized assignment to treatment, as compared with choice of preferred treatment as it occurs in real-world conditions. Compliance, engagement, or motivation may be better with a preferred treatment, and this can complicate the generalizability of results from randomized trials. The doubly randomized preference trial (DRPT) is a hybrid randomized and nonrandomized design that allows for estimation of the causal effect of randomization versus treatment preference. In the DRPT, individuals are first randomized to either randomized assignment or choice assignment. Those in the randomized assignment group are then randomized to treatment or control, and those in the choice group receive their preference of treatment versus control. Using the potential outcomes framework, we apply the algebra of conditional independence to show how the DRPT can be used to derive an unbiased estimate of the causal effect of randomization versus preference for each of the treatment and comparison conditions. Also, we show how these results can be implemented using full matching on the propensity score. The methodology is illustrated with a DRPT of introductory psychology students who were randomized to randomized assignment or preference of mathematics versus vocabulary training. We found a small to moderate benefit of preference versus randomization with respect to the mathematics outcome for those who received mathematics training.

Self-help therapy and recovery in psychosis: methodological considerations and service user involvement in a partially randomised preference trial

A new Recovery Guide for people with experience of psychosis has been developed (Barrowclough et al, 2009), drawing on existing cognitive behavioural interventions and founded on collaborations between service users, psychologists, researchers and healthcare professionals. This poster showcases the involvement of ‘experts by experience’ in the design, delivery and evaluation of this Recovery Guide, and the therapeutic structure that supports its use. The project has input from service users by various methods, including service user researchers, a service user reference group and service user consultants. Previous work has highlighted the strong and rational preferences that people experiencing psychosis hold about their potential treatment options (Gerrard et al, 2009); service users were presented with detailed information about the Recovery Guide and about different levels of therapeutic intervention to support it (low support= telephone cognitive behavioural therapy; high support= as low support, plus group sessions facilitated by a cognitive behavioural therapist and service user researcher). Findings revealed that most would be willing to take part in a trial of psychological treatments, whether as participants receiving extra therapeutic support or receiving treatment as usual only and assisting with the progress of the research. Participants also held strong feelings about which option they would prefer to receive. Participant preferences of this kind have not been taken into account in previous psychological treatment trials for psychosis. The poster describes the trial currently being implemented to evaluate the Recovery Guide; STAR-T (Self-help Therapy and Recovery-Trial), is an NIHR-funded preference trial in which participants can elect to join their treatment option of choice, or be randomly allocated to a condition. The trial emphasises the value of choice for people experiencing psychosis, and the poster will also discuss the methodological considerations necessary to support such a preference trial. The trial also benefits from the involvement of two service user researchers, who have personal experience of psychosis, and a panel of service user consultants. The ways in which this input has helped shape the design, implementation and evaluation of the trial will be outlined.

‘Let the computer choose?’: the experience of participants in a randomised preference trial of medical versus surgical termination of pregnancy

The termination of pregnancy trial (Newcastle upon Tyne, UK), is the only randomised trial on termination of pregnancy methods incorporating a qualitative element that aimed to understand the experiences of women participating in the trial. Based on the results of this qualitative work, this article aims to provide insights into two strands of understanding; firstly, women's experience of participating in research about abortion and secondly, their experience of participating in a randomised preference trial. Semi-structured interviews were conducted of up to 90 minutes with 30 participants recruited at a single hospital site. A total of 20 women from the preference arm and 10 from the random arm were interviewed. The analysis and discussion of our findings use reflexive modernisation as a framework for understanding and interpreting some of the actions of social agents, that is, the participants and trial recruiters in the course of a clinical trial as an expert system. We found that the factors that shape women's experiences and decisions include trust in the expert system and reflexivity and agency on the part of both participants and trial recruiters.

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.

Randomised preference trial of medical versus surgical termination of pregnancy less than 14 weeks' gestation (TOPS)

To determine the acceptability, efficacy and costs of medical termination of pregnancy (MTOP) compared with surgical termination of pregnancy (STOP) at less than 14 weeks' gestation, and to understand women's decision-making processes and experiences when accessing the termination service. A partially randomised preference trial and economic evaluation with follow-up at 2 weeks and 3 months. The Royal Victoria Infirmary, Newcastle upon Tyne, UK. Women accepted for termination of pregnancy (TOP) under the relevant Acts of Parliament with pregnancies < 14 weeks' gestation on the day of abortion. A further group of women attending contraception and sexual health clinics participated in a discrete choice experiment (DCE). STOP: all women > or = 6 weeks' and < 14 weeks' gestation were primed with misoprostol 400 micrograms 2 hours before the procedure. STOP was performed under general anaesthesia using vacuum aspiration. MTOP: all women < 14 weeks' gestation were given mifepristone 200 milligrams orally, returning 36-48 hours later for misoprostol. Main outcome measure was acceptability of TOP method. Secondary outcome measures included strength of preference by willingness to pay (WTP); distress, using the Impact of Event Scale (IES); anxiety and depression; satisfaction with care; experience of care; frequency and extent of symptoms including self-assessment of pain; clinical effectiveness; and complications. A DCE was used to identify attributes that shape women's preferences for abortion services. The trial recruited 1877 women, 349 in the randomised arms and 1528 in the preference arms. Of those in the preference arms, 54% chose MTOP. At 2 weeks after the procedure more women having STOP would choose the same method again in the future. Acceptability of MTOP declined with increasing gestational age. The difference in acceptability between STOP and MTOP persisted at 3 months. At 2 weeks after TOP, women in the preference arms were prepared to pay more to have their preferred option. There was no difference in anxiety or depression scores in women having MTOP or STOP. However, women randomised to MTOP had higher scores on subscales of the IES at both 2 weeks and 3 months. There was no difference in IES scores between MTOP and STOP in the preference arm. Women were more likely to be satisfied overall and with technical and interpersonal aspects of care if they had STOP rather than MTOP. Experience of care scores were lower after MTOP in both randomised and preference arms. During admission women undergoing MTOP had more symptoms and reported higher mean pain scores, and after discharge reported more nausea and diarrhoea. There were no differences in time taken to return to work between groups; around 90% had returned to work and normal activity by 2 weeks. Rates of unplanned or emergency admissions were higher after MTOP than after STOP. Overall complication rates were also higher after MTOP, although this only achieved statistical significance in the preference arm. Overall, STOP cost more than MTOP due to higher inpatient standard costs. Even though complication rates were higher with MTOP, it was still more cost-effective. DCE identified three attributes with an almost equal impact on women's preferences: provision of counselling, number of days delay to the procedure, and possibility of an overnight stay. MTOP was associated with more negative experiences of care and lower acceptability. Acceptability of MTOP declined with increasing gestational age. MTOP was less costly but also less effective than STOP. The majority of women choosing MTOP were satisfied with their care and found the procedure acceptable. RECOMMENDATIONS FOR FURTHER RESEARCH: An audit of provision of MTOP and STOP in England and Wales is urgently required. Further studies exploring the barriers to offering women the choice of method of TOP are needed, together with research on the acceptability and effectiveness of (1) MTOP and manual VA in pregnancies below 9 weeks' gestation and (2) MTOP and dilatation and evacuation after 14 weeks' gestation. Current Controlled Trials ISRCTN07823656.

Causal Inference in Hybrid Intervention Trials Involving Treatment Choice

Randomized allocation of treatments is a cornerstone of experimental design but has drawbacks when a limited set of individuals are willing to be randomized, or the act of randomization undermines the success of the treatment. Choice-based experimental designs allow a subset of the participants to choose their treatments. We discuss here causal inferences for hybrid experimental designs in which some participants are randomly allocated to treatments and others receive their treatment preference. This work was motivated by the Women Take Pride (WTP) study, a doubly randomized preference trial (DRPT) that assessed behavioral interventions for women with heart disease. We propose a model for estimating the causal effects in the subpopulations defined by treatment preferences, and thus preference effects. We describe an EM algorithm for computing maximum likelihood estimates of the model parameters. We illustrate the method by analyzing sickness impact profile (SIP) scores and treatment adherence in the WTP data. Our results show some evidence that SIP scores were improved when women received their preferred treatment and strong preference effects on program adherence; that is, women assigned to their preferred treatment were more likely to adhere to the program. We also provide a framework for assessing the DRPT and other hybrid trial designs and discuss some alternative designs from the perspective of the strength of assumptions required to make causal inferences.Please see the supplemental material for a correction to this article.