Sequential Multiple Assignment Randomized Trial Research Articles

A novel experimental vignette methodology: SMART vignettes

We motivate and present the methodology of vignette studies. The primary contribution of this paper is our proposal of a novel vignette study design: “SMART vignettes.” Our design has two notable features: the first is its use of sequential randomization, which conceptually originates from the sequential multiple assignment randomization trial (SMART) design developed by Murphy (2004). The second feature is adaptive allocation. These new features in vignette studies offer unique advantages not offered by traditional vignettes: (1) valid causal inferences on the conditional distributions of the primary outcome of interest, given other factors, (2) balanced allocations across groups, and (3) a greater degree of interactivity for the survey respondent. We illustrate the utility of our method using a case example of a vignette study used to probe physicians’ attitudes toward an AI-embedded clinical system. In this example, a SMART vignette was used to randomize hypothetical scenarios to gain a better understanding of the causal impact of physician attitudes, given emerging evidence that a range of factors including previous decisions, play a role in influencing clinical decisions. We simulated hypothetical vignette studies under both SMART and conventional (i.e. single randomization at baseline) designs. We varied the number of factors for each study and fixed each factor to have two levels. Relative loss was used to compare the degree of imbalance between groups. Both designs had smaller relative losses with larger sample sizes. The SMART study design had lower loss than its conventional counterpart for all values of [Formula: see text] for all studies, indicating better balance. As demonstrated by the relative loss in our simulations, our proposed SMART vignette design has an advantage over the conventional design. This method holds promise in generating new knowledge in decision making scenarios occurring over multiple and discrete time points.

Understanding Physician's Perspectives on AI in Health Care: Protocol for a Sequential Multiple Assignment Randomized Vignette Study.

As the availability and performance of artificial intelligence (AI)-based clinical decision support (CDS) systems improve, physicians and other care providers poised to be on the front lines will be increasingly tasked with using these tools in patient care and incorporating their outputs into clinical decision-making processes. Vignette studies provide a means to explore emerging hypotheses regarding how context-specific factors, such as clinical risk, the amount of information provided about the AI, and the AI result, may impact physician acceptance and use of AI-based CDS tools. To best anticipate how such factors influence the decision-making of frontline physicians in clinical scenarios involving AI decision-support tools, hypothesis-driven research is needed that enables scenario testing before the implementation and deployment of these tools. This study's objectives are to (1) design an original, web-based vignette-based survey that features hypothetical scenarios based on emerging or real-world applications of AI-based CDS systems that will vary systematically by features related to clinical risk, the amount of information provided about the AI, and the AI result; and (2) test and determine causal effects of specific factors on the judgments and perceptions salient to physicians' clinical decision-making. US-based physicians with specialties in family or internal medicine will be recruited through email and mail (target n=420). Through a web-based survey, participants will be randomized to a 3-part "sequential multiple assignment randomization trial (SMART) vignette" detailing a hypothetical clinical scenario involving an AI decision support tool. The SMART vignette design is similar to the SMART design but adapted to a survey design. Each respondent will be randomly assigned to 1 of the possible vignette variations of the factors we are testing at each stage, which include the level of clinical risk, the amount of information provided about the AI, and the certainty of the AI output. Respondents will be given questions regarding their hypothetical decision-making in response to the hypothetical scenarios. The study is currently in progress and data collection is anticipated to be completed in 2024. The web-based vignette study will provide information on how contextual factors such as clinical risk, the amount of information provided about an AI tool, and the AI result influence physicians' reactions to hypothetical scenarios that are based on emerging applications of AI in frontline health care settings. Our newly proposed "SMART vignette" design offers several benefits not afforded by the extensively used traditional vignette design, due to the 2 aforementioned features. These advantages are (1) increased validity of analyses targeted at understanding the impact of a factor on the decision outcome, given previous outcomes and other contextual factors; and (2) balanced sample sizes across groups. This study will generate a better understanding of physician decision-making within this context. DERR1-10.2196/54787.

Incorporating participants' welfare into sequential multiple assignment randomized trials.

Dynamic treatment regimes (DTRs) are sequences of decision rules that recommend treatments based on patients' time-varying clinical conditions. The sequential, multiple assignment, randomized trial (SMART) is an experimental design that can provide high-quality evidence for constructing optimal DTRs. In a conventional SMART, participants are randomized to available treatments at multiple stages with balanced randomization probabilities. Despite its relative simplicity of implementation and desirable performance in comparing embedded DTRs, the conventional SMART faces inevitable ethical issues, including assigning many participants to the empirically inferior treatment or the treatment they dislike, which might slow down the recruitment procedure and lead to higher attrition rates, ultimately leading to poor internal and external validities of the trial results. In this context, we propose a SMART under the Experiment-as-Market framework (SMART-EXAM), a novel SMART design that holds the potential to improve participants' welfare by incorporating their preferences and predicted treatment effects into the randomization procedure. We describe the steps of conducting a SMART-EXAM and evaluate its performance compared to the conventional SMART. The results indicate that the SMART-EXAM can improve the welfare of the participants enrolled in the trial, while also achieving a desirable ability to construct an optimal DTR when the experimental parameters are suitably specified. We finally illustrate the practical potential of the SMART-EXAM design using data from a SMART for children with attention-deficit/hyperactivity disorder.

The sequential, multiple assignment, randomized trial (SMART) design

The sequential, multiple assignment, randomized trial (SMART) design

SMART use of medications for the treatment of adolescent severe obesity: A sequential multiple assignment randomized trial protocol

BackgroundSevere obesity is a complex, chronic disease affecting nearly 9% of adolescents in the U.S. Although the current mainstay of treatment is lifestyle therapy, pediatric clinical practice guidelines recommend the addition of adjunct anti-obesity medication (AOM), such as phentermine and topiramate. However, guidance regarding when adjunct AOM should be started and how AOM should be used is unclear. Furthermore, an inherent limitation of current treatment guidelines is their “one-size-fits-all” approach, which does not account for the heterogeneous nature of obesity and high degree of patient variability in response to all interventions. MethodsThis paper describes the study design and methods of a sequential multiple assignment randomized trial (SMART), “SMART Use of Medications for the Treatment of Adolescent Severe Obesity.” The trial will examine 1) when to start AOM (specifically phentermine) in adolescents who are not responding to lifestyle therapy and 2) how to modify AOM when there is a sub-optimal response to the initial pharmacological intervention (specifically, for phentermine non-responders, is it better to add topiramate to phentermine or switch to topiramate monotherapy). Critically, participant characteristics that may differentially affect response to treatment will be assessed and evaluated as potential moderators of intervention efficacy. ConclusionData from this study will be used to inform the development of an adaptive intervention for the treatment of adolescent severe obesity that includes empirically-derived decision rules regarding when and how to use AOM. Future research will test this adaptive intervention against standard “one-size-fits-all” treatments.

Contingency management plus acceptance and commitment therapy for initial cocaine abstinence: Results of a sequential multiple assignment randomized trial (SMART)

BackgroundThis study tested an adaptive intervention for optimizing abstinence outcomes over phases of treatment for cocaine use disorder using a SMART design. Phase 1 assessed whether 4 weeks of contingency management (CM) improved response with the addition of Acceptance and Commitment Therapy (ACT). Phase 2 assessed pharmacological augmentation with modafinil (MOD) vs. placebo (PLA) for individuals not achieving abstinence during Phase 1. MethodFor Phase 1 of treatment, participants (N=118) were randomly allocated to ACT+CM or Drug Counseling (DC+CM), the comparison condition. At week 4, treatment response was defined as the submission of six consecutive cocaine-negative urine drug screens (UDS). Phase 1 non-responders were re-randomized to MOD or PLA as adjunct to their initial treatment. Phase 1 responders continued receiving their initial treatment. Primary outcomes included response rate and proportion of cocaine-negative UDS for Phase 1 and 2. Analyses used Bayesian inference with 80% pre-specified as the posterior probability (PP) threshold constituting moderate evidence that an effect exists. ResultsPhase 1 response was higher in the ACT+CM group (24.5%) compared to the DC+CM group (17.5%; PP = 84.5%). In Phase 2, the proportion of cocaine-negative UDS among Phase 1 responders did not differ by initial treatment (PP = 61.8%) but remained higher overall compared to Phase 1 non-responders (PPs > 99%). No evidence of an effect favoring augmentation with MOD was observed. DiscussionAdding ACT to CM increased abstinence initiation. Initial responders were more likely to remain abstinent compared to initial non-responders, for whom modafinil was not an effective pharmacotherapy augmentation strategy.

Effect of Psychological and Medication Therapies for Insomnia on Daytime Functions

Daytime functional impairments are the primary reasons for patients with insomnia to seek treatment, yet little is known about what the optimal treatment is for improving daytime functions and how best to proceed with treatment for patients whose insomnia has not remitted. To compare the efficacy of behavioral therapy (BT) and zolpidem as initial therapies for improving daytime functions among patients with insomnia and evaluate the added value of a second treatment for patients whose insomnia has not remitted. In this sequential multiple-assignment randomized clinical trial conducted at institutions in Canada and the US, 211 adults with chronic insomnia disorder were enrolled between May 1, 2012, and December 31, 2015, and followed up for 12 months. Statistical analyses were performed on an intention-to-treat basis in April and October 2023. Participants were randomly assigned to either BT or zolpidem as first-stage therapy, and those whose insomnia had not remitted received a second-stage psychological therapy (BT or cognitive therapy) or medication therapy (zolpidem or trazodone). Study outcomes were daytime symptoms of insomnia, including mood disturbances, fatigue, functional impairments of insomnia, and scores on the 36-item Short-Form Health Survey (SF-36) physical and mental health components. Among 211 adults with insomnia (132 women [63%]; mean [SD] age, 45.6 [14.9] years), 104 were allocated to BT and 107 to zolpidem at the first stage. First-stage treatment with BT or zolpidem yielded significant and equivalent benefits for most of the daytime outcomes, including depressive symptoms (Beck Depression Inventory-II mean score change, -3.5 [95% CI, -4.7 to -2.3] vs -4.3 [95% CI, -5.7 to -2.9]), fatigue (Multidimensional Fatigue Inventory mean score change, -4.7 [95% CI, -7.3 to -2.2] vs -5.2 [95% CI, -7.9 to -2.5]), functional impairments (Work and Social Adjustment Scale mean score change, -5.0 [95% CI, -6.7 to -3.3] vs -5.1 [95% CI, -7.2 to -2.9]), and mental health (SF-36 mental health subscale mean score change, 3.5 [95% CI, 1.9-5.1] vs 2.5 [95% CI, 0.4-4.5]), while BT produced larger improvements for anxiety symptoms relative to zolpidem (State-Trait Anxiety Inventory mean score change, -4.1 [95% CI, -5.8 to -2.4] vs -1.2 [95% CI, -3.0 to 0.5]; P = .02; Cohen d = 0.55). Second-stage therapy produced additional improvements for the 2 conditions starting with zolpidem at posttreatment in fatigue (Multidimensional Fatigue Inventory mean score change: zolpidem plus BT, -3.8 [95% CI, -7.1 to -0.4]; zolpidem plus trazodone, -3.7 [95% CI, -6.3 to -1.1]), functional impairments (Work and Social Adjustment Scale mean score change: zolpidem plus BT, -3.7 [95% CI, -6.4 to -1.0]; zolpidem plus trazodone, -3.3 [95% CI, -5.9 to -0.7]) and mental health (SF-36 mental health subscale mean score change: zolpidem plus BT, 5.3 [95% CI, 2.7-7.9]; zolpidem plus trazodone, 2.0 [95% CI, 0.1-4.0]). Treatment benefits achieved at posttreatment were well maintained throughout the 12-month follow-up, and additional improvements were noted for patients receiving the BT treatment sequences. In this randomized clinical trial of adults with insomnia disorder, BT and zolpidem produced improvements for various daytime symptoms of insomnia that were no different between treatments. Adding a second treatment offered an added value with further improvements of daytime functions. ClinicalTrials.gov Identifier: NCT01651442.

Chatbot-interfaced and cognitive-affective barrier-driven messages to improve colposcopy adherence after abnormal Pap test results in underserved urban women: A feasibility pilot study.

Challenges in ensuring adherence to colposcopy and follow-up recommendations, particularly within underserved communities, hinder the delivery of appropriate care. Informed by our established evidence-based program, we sought to assess the feasibility and acceptability of a novel cognitive-affective intervention delivered through a Chatbot interface, aimed to enhance colposcopy adherence within an urban inner-city population. We developed the evidence-based intervention, CervixChat, to address comprehension of colposcopy's purpose, human papillomavirus (HPV) understanding, cancer-related fatalistic beliefs, procedural concerns, and disease progression, offered in both English and Spanish. Females aged 21-65, with colposcopy appointments at an urban OBGYN clinic, were invited to participate. Enrolled patients experienced real-time counseling messages tailored via a Chatbot-driven barriers assessment, dispatched via text one week before their scheduled colposcopy. Cognitive-affective measures were assessed at baseline and through a 1-month follow-up. Participants also engaged in a brief post-intervention satisfaction survey and interview to capture their acceptance and feedback on the intervention. The primary endpoints encompassed study adherence (CervixChat response rate and follow-up survey rate) and self-evaluated intervention acceptability, with predefined feasibility benchmarks of at least 70% adherence and 80% satisfaction. Among 48 eligible women scheduled for colposcopies, 27 (56.3%) agreed, consented, and completed baseline assessments. Participants had an average age of 34 years, with 14 (52%) identifying as non-Hispanic White. Of these, 21 (77.8%) engaged with the CervixChat intervention via mobile phones. Impressively, 26 participants (96.3%) attended their diagnostic colposcopy within the specified timeframe. Moreover, 22 (81.5%) completed the follow-up survey and a brief interview. Barriers assessment revealed notable encodings in the Affect and Values/Goals domains, highlighting concerns and understanding around HPV, as well as its impact on body image and sexual matters. Persistent and relatively high intrusive thoughts and lowered risk perceptions regarding cervical cancer were reported over time, unaffected by the intervention. Post-intervention evaluations documented high satisfaction and perceived usefulness, with recommendations for incorporating additional practical and educational content. Our findings underscore the robust satisfaction and practicality of the CervixChat intervention among a diverse underserved population. Moving forward, our next step involves evaluating the intervention's efficacy through a Sequential Multiple Assignment Randomized Trial (SMART) design. Enhanced by personalized health coaching, we aim to further bolster women's risk perception, address intrusive thoughts, and streamline resources to effectively improve colposcopy screening attendance.

Abstract 15365: The Effect of Mobile Phone Messaging in the Reduction of Ambulatory Blood Pressure in Adults: A Systematic Review

Introduction: Ample evidence supports a lower risk for cardiovascular disease (CVD) with controlled blood pressure (BP), yet hypertension (HTN) remains a leading cause of CVD. Mobile health such as text messaging is commonly included in BP management programs; however, the association of texting with BP interventions is less well known. We sought to examine the effectiveness of mobile messaging on BP in adults in the ambulatory setting. Methods: In this systematic review, we searched 4 databases from date of inception through February 2023 using search terms (texting, messaging, mobile phone, BP, HTN). After removing duplicate citations, two reviewers independently reviewed the references for eligibility (randomized control trial [RCT], participants > 18 years of age, participants either with elevated BP, and/or a diagnosis of HTN). Results: A total of 1,786 RCTs were identified of which 48 met eligibility criteria, for a combined total of 19,720 participants from 21 countries, including 13 US studies. Twenty-seven studies reported significant reductions in BP (systolic, diastolic or both). All three RCTs using texting and pharmacist counseling (n=3) demonstrated significant changes in BP, while 13 of 16 RCTs with texting and provider involvement observed significant improvement in BP. For the studies reporting between group changes in systolic BP from baseline, 7 studies noted significant improvements in BP (Figure 1). Conclusions: Despite heterogeneity in study components, we found text messaging when used with other intervention components was associated with BP reductions among adults with HTN. Larger studies using novel designs such as SMART (sequential multiple assignment randomized trials) are warranted to understand efficacy of texting with other intervention components.

Sequenced treatment alternatives to relieve adolescent depression (STAR-AD): a multicentre open-label randomized controlled trial protocol

BackgroundAdolescent major depressive disorder (MDD) is a prevalent mental health problem with low treatment success rates. Whether fluoxetine or fluoxetine combined with cognitive-behavioural therapy (CBT) is the more effective initial treatment for adolescent MDD remains controversial, and few studies have investigated whether treatment switching or augmentation is preferred when the initial treatment is not working well.MethodsWe developed a multicentre open-label Sequential Multiple Assignment Randomized Trial (SMART) design, consisting of two phases lasting 8 weeks each. In phase 1 (at baseline), patients will be recruited and grouped in fluoxetine group or fluoxetine combined with CBT group by patient self-selection. In phase 2 (after 8 weeks of treatment), the nonresponders will be randomly assigned to six groups, in which participants will switch to sertraline, vortioxetine, or duloxetine or added aripiprazole, olanzapine, or lithium carbonate to fluoxetine. After the full 16 weeks of treatment, we will assess the long-term sustainability of the treatment effects by evaluating participants during their subsequent naturalistic treatment. The primary outcome will be the response rate, determined by the Children’s Depression Rating Scale-Revised (CDRS-R). Secondary outcomes include the change in scores on the Beck Depression Inventory (BDI), the Screen for Child Anxiety-Related Emotional Disorders (SCARED) and the Safe Assessment.DiscussionThe results from this study will aid clinicians in making informed treatment selection decisions for adolescents with MDD.Trial registrationThis protocol was registered at ClinicalTrials.gov with Identifier: NCT05814640.

Social Media and Online Dating Safety Practices by Adolescent Sexual and Gender Diverse Men: Mixed-Methods Findings From the SMART Study

PurposeSexual and gender diverse youth (SGDY) develop and employ safety strategies on their own to manage risks while using dating apps. This study aimed to describe the online dating safety practices of SGDY and determine the effectiveness of an eHealth HIV-prevention educational intervention with dating safety content to promote future safety behaviors. MethodsSGDY aged 13–18 from across the United States (N = 1087) were assigned to increasingly intensive HIV-prevention educational programs using a sequential multiple-assignment randomized trial design. Data were collected at three time points, each three months apart. Participants were asked multiple–choice and open-ended questions about the safety practices they used online and in person; mixed-methods described the safety behaviors of SGDY. Logistic regressions were used to determine psychosocial predictors of safety behaviors and the effectiveness of the intervention in promoting future safety practices. Results60% (n = 662) of participants used dating apps, most of whom reported using online (96.4%, n = 638) and in-person (92.9%, n = 615) safety strategies, such as limiting the disclosure of personal information or meeting other users in public. Outness and previous victimization were important psychosocial predictors of engaging in safety behaviors. Additionally, participants who received the online safety education were over 50% more likely to employ certain safety behaviors than SGDY who did not. DiscussionWhile most SGDY reported at least 1 type of safety strategy when dating online, safety practices differed across psychosocial variables, such as outness. This study provides evidence for the effectiveness of an eHealth educational intervention tailored to SGDY to promote additional safety behaviors.

Optimizing remote access to urinary incontinence treatments for women veterans (PRACTICAL): Study protocol for a pragmatic clinical trial comparing two virtual care options

ObjectivesIn this pragmatic clinical trial, the primary objective is to increase access to behavioral treatment of urinary incontinence (UI) for women Veterans by comparing the effectiveness of two virtual care delivery modalities. MethodsVeterans Affairs (VA) clinical sites in AL, GA, NC will virtually randomize 286 women Veterans with UI (ie, stress, urge, or mixed). We will compare the effectiveness of our mHealth UI application (MyHealtheBladder) to a single VA Video Connect (VVC) session delivered by trained UI providers. Women without improvement after 8 weeks will receive an optimization VVC visit using a sequential, multiple assignment, randomized trial (SMART) design. The primary outcome is UI symptom improvement at 12-weeks with or without optimization; secondary outcomes include improvements in lower urinary tract symptoms, adherence, retention rates, perceptions of improvement, and visit-related miles saved. Sample size needed to identify a 2.5-point change (range 0–21) in the International Consultation on Incontinence Questionnaire – Urinary Incontinence Short Form (ICIQ-UI SF) from baseline to 12-weeks post-randomization is 200 participants. Allowing for an attrition rate of 25%, 286 participants are required. Key resultsStudy team initiated remote recruitment on April 2020. Recruitment is on target with a 75% retention rate. We expect completion in fall of 2023 (clinicaltrials.govNCT04237753). Discussion/ConclusionEngaging women Veterans with virtual modalities for initial UI treatment may increase access to UI care while also improving symptoms. After assessing efficacy, adherence, and retention, the next step is to implement the most effective option for remote delivery of evidence-based behavioral UI treatment for women Veterans.Trial registration:ClinicalTrials.gov number NCT04237753

Study protocol for a sequential multiple assignment randomized trial to reduce risky drinking among service members and their partners

BackgroundMilitary spouses and partners in relationships with a heavy drinking service member report high levels of mental health concerns and consequences, which are compounded when both partners drink heavily. Military spouses and partners —termed “concerned partners” (CPs)—may be an important gateway for motivating service members (SMs) to seek care. However, CPs may first need to reduce their own drinking and improve their communication to effectively support and encourage changes for their service member partner. Partners Connect is a web-based intervention aimed at improving communication and relationship quality and increasing SM help-seeking. MethodsThe current study design is a two-stage Sequential Multiple Assignment Randomized Trial (SMART) to develop an adaptive CP intervention to decrease CP drinking and increase SM help-seeking. CPs aged 18 and older (n = 408) will be recruited via social media and followed for six months. In stage one, we will randomize CPs to either a 4-session web-based intervention (Partners Connect) or to receive communication resources from the Gottman Institute website. The goal is to have CPs invite their SM to complete an online personalized normative feedback (PNF) session. If their SM completes the PNF at stage one, CPs will be considered “responders,” if the SM does not complete, CPs who are “non-responders” will be re-randomized during stage two to receive either (1) a CRAFT workbook or (2) phone-based CRAFT if in Partners Connect; or (1) Partners Connect or (2) a CRAFT workbook if in Gottman. DiscussionBy first intervening with the service member's CP, we aim to better equip them to engage their service member partner in treatment services. In doing so, we develop a model that increases treatment accessibility and appeal among a group that may not otherwise seek care.ClinicalTrials.gov Identifier: NCT05619185

Social determinants of health, psychological distress, and caregiver burden among informal cancer caregivers of cancer survivors during treatment.

This study compared three ethnic/racial groups of informal cancer caregivers on social determinants of health and tested social determinants of health as predictors of psychological distress and caregiver burden. This study was a secondary analysis of baseline data of a sequential multiple assignment randomized trial (SMART) testing symptom management interventions with caregiver-survivor dyads. Caregivers completed baseline measures of social determinants of health (SDoH), functional limitations, psychological distress, and caregiver burden. Hispanic, non-Hispanic White, and non-Hispanic other races caregivers were compared on these variables. Multivariate tests of associations between SDoH and caregiver burden and psychological distress were conducted in structural equation modeling with caregiver burden and psychological distress as latent variables. Hispanic caregivers reported significantly higher caregiver burden, specifically for finances, family, and schedules. Caregiver burden was significantly predicted by having income barely or not meeting needs, being female, socially isolated, married, Hispanic, and having poor physical functioning. Significant predictors of caregivers' psychological distress: being female, being socially isolated, and having poor physical functioning. Hispanic caregivers experience significant challenges associated with caregiver burden, especially if they are female, socially isolated, and have poor physical functioning. Assessment of these SDoH is important in caregiver health to provide supportive care during caregiving. NCT03743415 www.clinicaltrials.gov.

Statistical Considerations and Software for Designing Sequential, Multiple Assignment, Randomized Trials (SMART) with a Survival Final Endpoint

ABSTRACT Sequential, multiple assignment, randomized trial (SMART) designs are appropriate for comparing adaptive treatment interventions, in which intermediate outcomes (called tailoring variables) guide subsequent treatment decisions for individual patients. Within a SMART design, patients may be re-randomized to subsequent treatments following the outcomes of their intermediate assessments. In this paper, we provide an overview of statistical considerations necessary to design and implement a two-stage SMART design with a binary tailoring variable and a survival final endpoint. A chronic lymphocytic leukemia trial with a final endpoint of progression-free survival is used as an example for the simulations to assess how design parameters, including, choice of randomization ratios for each stage of randomization, and response rates of the tailoring variable affect the statistical power. We assess the choice of weights from restricted re-randomization on data analyses and appropriate hazard rate assumptions. Specifically, for a given first-stage therapy and prior to the tailoring variable assessment, we assume equal hazard rates for all patients randomized to a treatment arm. After the tailoring variable assessment, individual hazard rates are assumed for each intervention path. Simulation studies demonstrate that the response rate of the binary tailoring variable impacts power as it directly impacts the distribution of patients. We also confirm that when the first stage randomization is 1:1, it is not necessary to consider the first stage randomization ratio when applying the weights. We provide an R-shiny application for obtaining power for a given sample size for SMART designs.

Re-conceptualizing SMART designs as a hybrid of randomized and regression discontinuity designs: opportunities, cautions

ABSTRACT Though previously unacknowledged, a SMART (Sequential Multiple Assignment Randomized Trial) design uses both regression discontinuity (RD) and randomized controlled trial (RCT) designs. This combination structure creates a conceptual symbiosis between the two designs that enables both RCT- and previously unrecognized, RD-based inferential claims. In Phase 1, based on a quantitative cut-point, a SMART design yields successful or unsuccessful participants after the initial RCT. In Phase 2, interventions are randomly assigned to successful and unsuccessful, Phase 1 subgroups, thereby yielding both RCT and RD results. Like RCT-based inferences, many but not all newly recognized, RD-based SMART inferences are potentially unbiased. In particular, comparisons between RCT and RD outcomes are potentially problematic. Principles from the within-study comparison literature are used to inform SMART-based research, thereby introducing otherwise ignored inferential and procedural possibilities. Awareness of RD designs embedded within SMARTs encourages previously unrecognized inferential claims, though potential internal validity threats and limitations must be carefully weighed. Finally, new, SMART-like sequences of quasi-experiments and experiments can be tested and embedded logic applied.

Sample size estimation for comparing dynamic treatment regimens in a SMART: A Monte Carlo-based approach and case study with longitudinal overdispersed count outcomes.

Dynamic treatment regimens (DTRs), also known as treatment algorithms or adaptive interventions, play an increasingly important role in many health domains. DTRs are motivated to address the unique and changing needs of individuals by delivering the type of treatment needed, when needed, while minimizing unnecessary treatment. Practically, a DTR is a sequence of decision rules that specify, for each of several points in time, how available information about the individual's status and progress should be used in practice to decide which treatment (e.g. type or intensity) to deliver. The sequential multiple assignment randomized trial (SMART) is an experimental design widely used to empirically inform the development of DTRs. Sample size planning resources for SMARTs have been developed for continuous, binary, and survival outcomes. However, an important gap exists in sample size estimation methodology for SMARTs with longitudinal count outcomes. Furthermore, in many health domains, count data are overdispersed-having variance greater than their mean. We propose a Monte Carlo-based approach to sample size estimation applicable to many types of longitudinal outcomes and provide a case study with longitudinal overdispersed count outcomes. A SMART for engaging alcohol and cocaine-dependent patients in treatment is used as motivation.

Adaptive interventions for alcohol misuse and violent behaviors among adolescents and emerging adults in the emergency department: A sequential multiple assignment randomized controlled trial protocol

Alcohol use and violent behaviors among youth are associated with morbidity and mortality. An emergency department (ED) visit provides an opportunity to initiate prevention efforts. Despite promising findings from our single session SafERteens brief intervention (BI), impact is limited by modest effect sizes, with data lacking on optimal boosters to enhance effects. This paper describes the protocol for a sequential, multiple assignment, randomized trial (SMART). Adolescents and emerging adults (ages 14–20) in the ED screening positive for alcohol use and violent behaviors (physical aggression) were randomly assigned to: 1) SafERteens BI + Text Messaging (TM), or 2) SafERteens BI + remote Health Coach (HC). Participants completed weekly surveys over 8 weeks after the ED visit to tailor intervention content and measure mechanisms of change. At one-month, intervention response/non-response is determined (e.g., binge drinking or violent behaviors). Responders are re-randomized to continued intervention condition (e.g., maintenance) or minimized condition (e.g., stepped down). Non-responders are re-randomized to continued condition (e.g., maintenance), or intensified condition (e.g., stepped up). Outcomes were measured at 4 and 8 months, including primary outcomes of alcohol consumption and violence, with secondary outcomes of alcohol consequences and violence consequences. Although the original goal was to enroll 700 participants, COVID-19 impacts on research diminished recruitment in this trial (enrolled n = 400). Nonetheless, the proposed SMART is highly innovative by blending real-time assessment methodologies with adaptive intervention delivery among teens with comorbid alcohol misuse and violent behaviors. Findings will inform the content and timing booster interventions to alter risk behavior trajectories.Trial Registration:ClinicalTrials.govNCT03344666. University of Michigan # HUM00109156.

Testing adaptive interventions to improve PTSD treatment outcomes in Federally Qualified Health Centers: Protocol for a randomized clinical trial

BackgroundPosttraumatic stress disorder (PTSD) disproportionately affects low-income individuals and is untreated in 70% of those affected. One third of low-income Americans are treated in Federally Qualified Health Centers (FQHCs), which do not have the capacity to provide all patients with first-line treatments such as Prolonged Exposure (PE). To address this problem, FQHCs could use low-intensity interventions (e.g., Clinician-Supported PTSD Coach: CS PTSD Coach) and medium-intensity interventions (e.g., PE for Primary Care: PE-PC) to treat PTSD with fewer resources. However, some patients will still require high-intensity treatments (e.g., full-length PE) for sustained clinical benefit. Thus, there is a critical need to develop stepped-care models for PTSD in FQHCs. MethodWe are conducting a Sequential, Multiple Assignment, Randomized Trial (SMART) with 430 adults with PTSD in FQHCs. Participants are initially randomized to CS PTSD Coach or PE-PC. After four sessions, early responders step down to lower frequency interaction within their assigned initial treatment strategy. Slow responders are re-randomized to either continue their initial treatment strategy or step up to Full PE for an additional eight weeks. The specific aims are to test the effectiveness of initiating treatment with PE-PC versus CS PTSD Coach in reducing PTSD symptoms and to test the effectiveness of second-stage strategies (continue versus step-up to Full PE) for slow responders. ConclusionsThis project will provide critical evidence to inform the development of an effective stepped-care model for PTSD. Testing scalable, sustainable sequences of PTSD treatments delivered in low-resource community health centers will improve clinical practice for PTSD.

A sequential, multiple assignment randomized trial comparing web-based education to mobile video interpreter access for improving provider interpreter use in primary care clinics: the mVOCAL hybrid type 3 study protocol

BackgroundIndividuals who use a language other than English for medical care are at risk for disparities related to healthcare safety, patient-centered care, and quality. Professional interpreter use decreases these disparities but remains underutilized, despite widespread access and legal mandates. In this study, we compare two discrete implementation strategies for improving interpreter use: (1) enhanced education targeting intrapersonal barriers to use delivered in a scalable format (interactive web-based educational modules) and (2) a strategy targeting system barriers to use in which mobile video interpreting is enabled on providers’ own mobile devices.MethodsWe will conduct a type 3 hybrid implementation-effectiveness study in 3–5 primary care organizations, using a sequential multiple assignment randomized trial (SMART) design. Our primary implementation outcome is interpreter use, calculated by matching clinic visits to interpreter invoices. Our secondary effectiveness outcome is patient comprehension, determined by comparing patient-reported to provider-documented visit diagnosis. Enrolled providers (n = 55) will be randomized to mobile video interpreting or educational modules, plus standard interpreter access. After 9 months, providers with high interpreter use will continue as assigned; those with lower use will be randomized to continue as before or add the alternative strategy. After another 9 months, both strategies will be available to enrolled providers for 9 more months. Providers will complete 2 surveys (beginning and end) and 3 in-depth interviews (beginning, middle, and end) to understand barriers to interpreter use, based on the Theoretical Domains Framework. Patients who use a language other than English will be surveyed (n = 648) and interviewed (n = 75) following visits with enrolled providers to understand their experiences with communication. Visits will be video recorded (n = 100) to assess fidelity to assigned strategies. We will explore strategy mechanism activation to refine causal pathway models using a quantitative plus qualitative approach. We will also determine the incremental cost-effectiveness of each implementation strategy from a healthcare organization perspective, using administrative and provider survey data.DiscussionDetermining how these two scalable strategies, alone and in sequence, perform for improving interpreter use, the mechanisms by which they do so, and at what cost, will provide critical insights for addressing a persistent cause of healthcare disparities.Trial registrationNCT05591586.