Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study

Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study - NCBI Bookshelf An official website of the United States government Here's how you know The .gov means it's official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site. The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely. Log in Show account info Close Account Logged in as: username Dashboard Publications Account settings Log out Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation Bookshelf Search database Books All Databases Assembly Biocollections BioProject BioSample Books ClinVar Conserved Domains dbVar Gene Genome GEO DataSets GEO Profiles GTR Identical Protein Groups MedGen MeSH NLM Catalog Nucleotide OMIM PMC Protein Protein Clusters Protein Family Models PubChem BioAssay PubChem Compound PubChem Substance PubMed SNP SRA Structure Taxonomy ToolKit ToolKitAll ToolKitBookgh Search term Search Browse Titles Advanced Help Disclaimer NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health. Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study Lori E. Crosby , PsyD, Constance A. Mara , PhD, Yolanda Johnson , MLS, Rogelle Hackworth , BS, and Charles Quinn , MD. Author Information and Affiliations Authors Lori E. Crosby , PsyD, 1,2 Constance A. Mara , PhD, 1,2 Yolanda Johnson , MLS, 1,2 Rogelle Hackworth , BS, 1,2 and Charles Quinn , MD 1,3 . Affiliations 1 Division of Behavioral Medicine & Clinical Psychology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio 2 Department of Pediatrics, University of Cincinnati College of Medicine, Cincinnati, Ohio 3 Division of Hematology, Cancer and Blood Diseases Institute, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio Washington (DC): Patient-Centered Outcomes Research Institute (PCORI) ; 2024 Feb . Copyright and Permissions Copyright © 2024. Cincinnati Children's Hospital Medical Center. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Structured Abstract Background: Sickle cell disease (SCD) is a genetic blood disorder placing children at risk for serious medical complications, early morbidity and mortality, and high health care utilization. In the United States, SCD affects primarily Black and Hispanic or Latine/x children. Hydroxyurea is 1 of only 4 disease-modifying treatments available for this devastating and life-threatening disease. National evidence-based SCD guidelines recommend the use of a shared decision-making approach to offer hydroxyurea to all children with sickle cell anemia as early as 9 months of age and children with SCD in the presence of complications. Hydroxyurea uptake remains low in this population, however, because parents or caregivers (hereafter referred to as parents) lack information about it and have concerns about its safety and potential long-term side effects (eg, cancer, infertility, birth defects). Also, clinicians may not have the training or tools to facilitate a shared discussion with parents that provides medical evidence and considers parent preferences and values. Objectives: The specific aims of this study ( NCT03442114 ) were to evaluate the effectiveness of the usual care dissemination method and the Hydroxyurea Shared Decision-Making (H-SDM) toolkit dissemination method on parent report of decisional uncertainty (primary outcome chosen by parents of children with SCD), parent perception of experiencing shared decision-making (primary outcome), parent knowledge of hydroxyurea, and secondary outcomes: the number of children offered hydroxyurea, hydroxyurea uptake (those with active prescriptions), and child health outcomes (neurocognitive functioning, health care utilization). Methods: We began recruiting in 2018 for a comparative effectiveness cluster randomized trial of the 2 dissemination methods by using a stepped wedge design at 8 participating SCD clinic sites across the United States; however, because of low accrual, staff turnover, and COVID-19 pandemic complications, we added sites in 2019 and changed to a unidirectional quasiexperimental design (ie, no cluster randomization; August 2020) in which 9 sites started with the usual care dissemination method, and then crossed over to the H-SDM toolkit method. Eligible children were between birth and 5 years of age and had to meet criteria for hydroxyurea treatment. Their parents had 2 research visits during the usual care method (baseline and follow-up 3-7 months later) and 2 or 3 visits during the H-SDM toolkit method (baseline; clinic visit with decision aids, which may have taken up to 2 visits; and follow-up visit 3-7 months later). Stakeholders (parents, clinicians, community-based organizations, community engagement experts, insurance providers) were involved in study design, implementation (eg, recruitment), and dissemination. The co-primary outcomes of parent decisional uncertainty and parent report of shared decision-making were the dependent variables in regression models that accounted for site, when the site crossed over from the usual care method to the toolkit method, COVID-19 timing (before the pandemic or during the pandemic), and COVID-19 impact (impact score on a psychosocial COVID-19 measure). We also calculated effect sizes for all outcomes (primary and secondary) to determine the size of the difference between the outcome means for the dissemination methods and explored clinically important differences on our primary outcome measures (ie, whether the mean scores for parents in both groups were above or below empirically established clinical cutoffs). Results: A total of 176 parents of young children with SCD receiving care at 1 of the 9 sites were assigned to either the usual care or H-SDM toolkit method based on when they were enrolled (most usual care participants were enrolled before March 2020, n = 95; most H-SDM toolkit participants were enrolled after March 2020, n = 81). The COVID-19 pandemic substantially affected both study enrollment and implementation because clinic volumes never rebounded to pre–COVID-19 rates, sites experienced high study staff turnover rates, and some sites had challenges with study implementation and delivery. The mean (SD) Decisional Conflict Scale (DCS) score was 19.8 (17.7) in the toolkit condition and 22.9 (20.3) in the usual care condition ( P = .17, Cohen d = 0.16). There were also no statistically significant differences on any of the DCS subscales; standardized effect size differences ranged from d = 0.10 for Values Clarity to d = 0.22 for Informed (ie, small effect sizes). Although not statistically significant, the difference in the DCS (3.1) (indicating more decisional uncertainty in the usual care group) is above the threshold for a clinically important difference. With respect to parent report of shared decision-making, there were no statistically significant differences between the toolkit method and the usual care method on parent report of shared decision-making, as measured by the dyadic OPTION 1 total score ( b = 2.35 [95% CI, −0.72 to 5.42]; P = .13; Cohen d = 0.09). Dyadic OPTION score observed mean difference between the groups did not surpass the threshold for clinical importance. With respect to secondary outcomes, parents enrolled during the toolkit method reported statistically significantly fewer problems communicating about SCD with important others. Clinicians using both dissemination methods were offering hydroxyurea through shared decision-making (≥58.7%). The odds of being offered hydroxyurea (odds ratio [OR], 0.11 [95% CI, 0.03-0.40]; P < .01) or receiving a hydroxyurea prescription (OR, 0.26 [95% CI, 0.10‑0.70]; P = .01), however, were significantly higher for parents enrolled during the usual care method. Conclusions: This study sought to identify the better way to facilitate shared decision-making about hydroxyurea for parents of young children with SCD. Although parents enrolled during the toolkit vs usual care method showed a clinically important difference in their decisional uncertainty, the use of a quasiexperimental design limited our ability to examine true effectiveness in the absence of statistical significance. The clinician training component of the toolkit method was effective, however, in increasing clinician knowledge and self-confidence and should be evaluated with a larger sample of clinicians. We found that participants enrolled when the toolkit method was used were less likely to be offered and prescribed hydroxyurea; however, the consequences of the COVID-19 pandemic on these findings is undetermined. Additional research is therefore needed to determine the impact of the toolkit method. Limitations: The study experienced accrual challenges, site team turnover, and the COVID-19 pandemic, which resulted in a change in study design, limiting the ability to make definitive conclusions about the differential effectiveness of the 2 dissemination methods. Additionally, all parent participants were required to read or understand English, limiting the generalizability of study results. Background Sickle cell disease (SCD) is a genetic blood disorder that affects approximately 100 000 people in the United States. 2 Sickle cell anemia is a subtype of SCD, which is most common among Black and Latine/x people but is also found in people with Mediterranean, Indian, or Saudi Arabian ancestry. 3 In SCD, the body produces an abnormal form of hemoglobin (hemoglobin S) that causes red blood cells to change shape (crescent or sickle shaped) and clump together, blocking oxygen and blood flow in the blood vessels. These cells also die more quickly than other red blood cells, leading to anemia, jaundice, gallstones, and severe fatigue. The blockage caused by the sickled cells results in severe pain and permanent damage to organs (eg, brain, lungs, eyes, heart, bones, kidneys, liver, spleen). Sickle cell pain often necessitates treatment with opioid medications (eg, oxycodone) and results in emergency department (ED) visits and hospitalizations. Early damage to the spleen makes children more susceptible to life-threatening infections. Hydroxyurea, originally a medication used to treat adult serious illness, is a proven safe and effective treatment for sickle cell anemia. Hydroxyurea increases the production of fetal hemoglobin, which prevents hemoglobin S from sickling and improves the overall functioning of the red blood cells. A daily dose of oral hydroxyurea has been shown to significantly reduce the frequency of pain (2-fold decrease) and serious medical complications: dactylitis, severe pain and swelling in the bones of the hands or feet (5-fold decrease), and acute chest syndrome, a life-threatening condition where the lungs do not get the oxygen they need (3-fold decrease). 4 Children taking hydroxyurea need fewer transfusions and spend less time in the hospital for pain (100 fewer days). 4 This medication also protects organs from the effects of chronic sickling. Children taking hydroxyurea maintain their growth percentile or improve it. 5 For these reasons, children taking hydroxyurea report high levels of health-related quality of life (HRQOL). 6 Because the benefits of hydroxyurea outweigh the risks, in 2014 the National Heart, Lung, and Blood Institute (NHLBI) published Evidence-Based Guidelines for the Management of Sickle Cell Disease , which recommended offering hydroxyurea therapy to children with SCD as early as 9 months of age as part of a shared decision-making process. 7 The NHLBI recommended shared decision-making for hydroxyurea in part because the evidence for hydroxyurea as an effective treatment in children is strong, but additional research on the long-term side effects is needed. Shared decision-making is a collaborative process during which clinicians and patients work together to make a decision about a treatment based on clinical evidence, expected outcomes, and patient values and preferences. 8 The goal is for patients, or in this case parents, to take an active role in a decision about their child's health where there are tradeoffs between likely benefits and potential harms when the evidence is not comprehensive. 9 Data suggest that large numbers of eligible patients are not receiving the recommended therapy. Studies show that only 20% to 40% of eligible patients are receiving hydroxyurea. 10 , 11 Furthermore, many clinicians have not prescribed hydroxyurea to children under 4 years of age. 12 A study found that 46% of clinicians had not prescribed it to this age group. 12 It is unclear whether the number is low because clinicians are failing to offer hydroxyurea or parents are delaying or refusing the treatment because of uncertainty, lack of knowledge about hydroxyurea, or feeling unprepared. Currently, the only tool to assist clinicians with implementing hydroxyurea guidelines for SCD is a clinician pocket guide developed by the American Society of Hematology (ASH). The ASH Hydroxyurea and Transfusion Therapy for the Treatment of Sickle Cell Disease clinician pocket guide provides a summary of the guidelines and organizes them into the following categories: who is eligible; what you should do before starting; when hydroxyurea is started; and how to monitor it over time, including contraindications and other considerations. Pocket guides are widely used and are usually rated as useful at the point of care by clinicians, 13 , 14 but they do not provide (1) training to increase clinician motivation, (2) guided practice in communication and cultural humility skills in the elicitation of preferences to build clinician self-efficacy in shared decision-making, (3) audit and feedback to reinforce behavior change, or (4) decision support tools to help clinicians engage parents and support parents in decision-making. A systematic review of the implementation of shared decision-making in clinical practice using the structure, process, and outcome model identified key barriers and facilitators for implementation. 15 For clinicians, key factors include physician understanding of shared decision-making, motivation, clinical uncertainty, communication skills (more than knowledge or technical skills), time constraints, and work pressure. 15 Research suggests that clinician motivation for shared decision-making and clinical uncertainty may be targeted through printed materials (eg, clinical pocket guide) that justify the need for a treatment choice. 16 Information alone may not be sufficient, however, because lack of training in shared decision-making may lead to ineffective and unauthentic implementation. 17 Clinician training in communication skills and cultural humility 18 may increase clinician self-efficacy, a necessary component for behavior change, 19 while improving clinicians' ability to build rapport and trust with parents. Research suggests that Black patients experience less clinician responsiveness and listening during medical visits. 20 Clinicians who are calm and empathetic make patients feel more comfortable, valued, and open. 21 , 22 This training may therefore be particularly important when working with parents of youth with SCD because research suggests that communication disparities may be an important contributor to poor health outcomes for Black patients with chronic diseases. Other important factors such as clinician time constraints and work pressure have been successfully addressed through application of implementation science strategies such as having site champions, embedding shared decision-making into clinic flow processes, and audit and feedback monitoring. 23 Patient decision aids improve patient knowledge and may prepare patients to be active participants in decision-making. If the patient is prepared, this reinforces clinician continued use of shared decision-making. 16 A Cochrane review found that dissemination strategies targeting patients and clinicians are more effective than those that target clinicians alone; however, additional research is needed to elucidate how these dissemination methods compare with one another and their impact on patient outcomes. 16 Our team used Agency for Healthcare Research and Quality funding to develop, test, and begin to disseminate a parent-centered, technology-enhanced decision support toolkit, the Hydroxyurea Shared Decision-Making (H-SDM) toolkit, to assist clinicians in implementing shared decision-making for hydroxyurea and parents in feeling more confident about their decision. This toolkit targets factors noted in systematic reviews 10 , 15 , 24 and the literature on behavior change 14 (clinician knowledge, motivation, cultural humility, and self-efficacy; system factors; and patient readiness). We hypothesized that this dissemination method would facilitate shared decision-making, reduce parental uncertainty about the decision, increase clinician offering of hydroxyurea, and ultimately increase hydroxyurea uptake. Data from our quality improvement pilot study showed that use of the toolkit was effective in increasing the proportion of eligible patients who were offered hydroxyurea (from 33% to 100%); also, parents in the toolkit condition showed less parental decisional uncertainty and greater parent knowledge about hydroxyurea than parents in a control condition. 25 These improvements led to an increase in the proportion of patients with an active prescription for hydroxyurea (uptake). The current study proposed to address a critical gap in the literature regarding how shared decision-making between clinicians and parents of children with SCD can be facilitated and national guidelines can be implemented, with the hopes of increasing offers and uptake of hydroxyurea, a life-saving treatment. Two dissemination methods were compared (the ASH clinician pocket guide and the H-SDM toolkit) for improving adoption of the NHLBI guideline–recommended practice of shared decision-making about hydroxyurea for children with SCD. We also proposed to examine the impact of these dissemination methods on patient- and parent-centered outcomes. Specifically, our study initially sought to compare the effectiveness of the ASH clinician pocket guide dissemination method (the usual care condition) and the H-SDM toolkit dissemination method (the toolkit condition) on clinician implementation of guidelines (use of shared decision-making and offering hydroxyurea), parent decisional uncertainty, hydroxyurea uptake, and patient-centered outcomes (ie, hydroxyurea adherence, neurocognitive functioning, sickle cell–related QOL, and health care utilization). Accrual challenges, staff turnover, and the impact of the COVID-19 pandemic presented significant barriers to the implementation of our original study design as a cluster randomized trial; however, given the potential benefits of continuing (including findings from our pilot study, closing the guideline to practice gap) we pivoted the study design to a multisite quasiexperimental design (ie, sites were not cluster randomized) to move forward with this important research. Significance This project was intended to inform decisions about how to best support shared decision-making for parents and clinicians, which may improve the quality of care; informed decision-making by parents may include a decision to initiate hydroxyurea, thereby improving health outcomes in young children with SCD. The proposed project is significant because it is designed to answer questions that affect outcomes important to parents, patients, and clinicians. Not only could it help to close the gap between guidelines and practice, but it could also address important factors in this process (ie, the “how to”). Despite being recommended by NHLBI guidelines, hydroxyurea uptake remains low for children of all ages, especially young children (birth to 5 years of age). One reason may be that clinicians do not have the time or resources needed to use a shared decision-making approach and therefore may not be offering hydroxyurea consistently. Thus, there may be a gap between guideline recommendations and clinical practice. Another contributing factor for low uptake may be that parents do not consider hydroxyurea as a possible treatment choice because their discussions with clinicians do not adequately address their knowledge and support needs. If clinicians had the tools necessary to involve parents in a decision-making process that supports their needs and considers their values and preferences, parents might discover that hydroxyurea treatment is consistent with their goals and preferences related to their child's health. Consequently, more parents might agree to try hydroxyurea, resulting in more children experiencing the health benefits of this treatment. By examining different methods for facilitating shared decision-making between parents and clinicians, we hope to increase the number of children offered this life-saving treatment and develop a reliable process for offering this treatment through shared decision-making, an equitable approach to care that offers parents an active role in decision-making about their child's treatment. There is significant health policy support for the advancement of shared decision-making in medicine. The Health and Medicine Division of the National Academies of Science, Engineering, and Medicine identified shared decision-making as a priority for comparative effectiveness research (ie, research that compares the benefits and harms of alternative treatments), 26 and in 2017 the National Quality Forum convened an action team on shared decision-making 27 and has endorsed several measures of shared decision-making including the CollaboRATE decision-making score (endorsed in 2019), which is derived from the measure used in this project. Moreover, we reject the notion that suboptimal care for patients with SCD is tolerable or inevitable because of systemic disparities reflecting patients' race or ethnicity and income level. 28 We posit that suboptimal care for SCD is preventable and correctable and hypothesize that care equity can be achieved through the use of multicomponent dissemination methods, if developed with key stakeholders and designed to address barriers at multiple levels (patient, clinician, health care system, and community or societal). Aims Studies suggest that the most effective dissemination strategies incorporate a model such as the RE-AIM framework in their evaluation. The RE-AIM framework is designed to improve the quality, speed, and impact of dissemination methods by attending to 5 factors: Reach: Does the intervention reach the intended population? Efficacy or effectiveness: Does the intervention affect important outcomes? Adoption: Will the intervention be supported by staff, settings, or institutions? Implementation: Will the intervention be delivered consistently? Maintenance: What is in place to ensure that the intervention will continue over time? We developed our evaluation plan using RE-AIM. In the initial PCORI application, we proposed a comparative effectiveness research trial with 2 study aims: Aim 1. Evaluate the effectiveness of the usual care dissemination method (clinician pocket guide) and the H-SDM toolkit dissemination method (toolkit) on (1) parent report of decisional uncertainty for hydroxyurea (primary outcome chosen by parents of children with SCD is effectiveness) and (2) parent report of experiencing shared decision-making when talking with their clinician about hydroxyurea (effectiveness) in a sample of children (birth to 5 years of age). – Hypothesis. Compared with usual care (clinician pocket guide), the H-SDM toolkit dissemination method will result in an increase in the use of a shared decision-making process for hydroxyurea and parents feeling less uncertain in their decision about children taking hydroxyurea. Aim 2. Evaluate the effectiveness of the usual care dissemination method (clinician pocket guide) and the H-SDM toolkit dissemination method (toolkit) on (1) parent knowledge of hydroxyurea (effectiveness), (2) children offered hydroxyurea (reach), (3) children with an active hydroxyurea prescription (uptake: effectiveness), and (4) child health outcomes: pain, neurocognitive functioning, sickle cell–related QOL and health care utilization (effectiveness). – Hypothesis. Compared with usual care (clinician pocket guide), the H-SDM toolkit dissemination method will result in parents of children with SCD birth to 5 years of age knowing more about hydroxyurea, more children offered and receiving hydroxyurea, and these children experiencing positive health outcomes. Given the change to a quasiexperimental study with a unidirectional crossover design, we were able to provide a robust description of the study population characteristics, clinician outcomes, stakeholder engagement involvement, and lessons learned to guide future research. Because we were not able to randomize clusters, however, we lost the ability to evaluate differences within and between clusters that would have answered important questions about dissemination method effectiveness, including their impact on clinical outcomes (ie, parent decisional uncertainty, clinician adoption of shared decision-making). Participation of Patients and Other Stakeholders Investigators rarely engage communities of color in study design and conduct in meaningful and sustained ways. 29 Because the Engaging Parents of Children With Sickle Cell Anemia and Their Providers in Shared Decision-Making for Hydroxyurea (Engage-HU) Trial sought to develop meaningful connections with caregivers of patients with SCD, we used a community participatory research approach 30 that emphasized cultural humility—that is, an ongoing process of self-reflection of one's beliefs and cultural identities to work toward building honest and trustworthy relationships to support power-sharing. This stakeholder-engaged approach strives to give racialized communities a voice in the research process. 19 , 24 Furthermore, we applied antiracism strategies. 31 To ensure that the Engage-HU Trial was well aligned with the priorities and centered to the needs of caregivers of children with SCD, we involved a diverse group of patients with SCD, caregivers, clinicians, and community-based organization leaders as partners in every step of the research process. Perspectives of racialized caregivers of patients with SCD were the central axis on which study decisions were made. This included having patient voices at all stages of the study through engagement of a stakeholder advisory council (see Table 1 for information about the composition of the council), which met 2 to 4 times per year. Throughout the trial, this council assisted with recruitment and retention strategies and research procedures consistent with the participatory research approach and tenets 32 , 33 of community-based participatory research. 30 Parents also contributed significantly to the development of toolkit components (eg, co-designing pamphlet materials and decision aids), primary outcomes (eg, focusing on caregiver uncertainty), and dissemination methods (eg, co-designing the website), and parents of children with SCD were included as primary research team members and on the Data Safety Monitoring Board. 34 Parents also assisted with developing the virtual reality simulation materials, and Black families were modeled as exemplars in the immersive virtual reality environment. 35 During the study, regular meetings with our stakeholder advisory council facilitated participant recruitment. Specifically, stakeholders brainstormed ways to improve study procedures by leveraging key strengths and resilience of the sickle cell community (eg, sickle cell community–based organizations sending study information directly to families), which increased the recruitment rate. We provided flexible options for involvement based on stakeholder preference and availability (eg, from consultation to co-production of research), used a variety of accessible communication methods (eg, email, phone, virtual meetings), and fostered partnerships based on aligned purpose and mutual benefits (stakeholders valued access to information about hydroxyurea and decision aids). Table 1 Composition of the Stakeholder Advisory Council. The stakeholder advisory council also played a significant role in dissemination planning by identifying target groups, including community partners and policy-focused organizations, that might have an interest in study findings. After the pause in study operations in response to the COVID-19 pandemic, stakeholders led the creation of a patient- and family-facing website to house the decision aids used in the H-SDM toolkit. The plan is for stakeholders to co-present with members of the research team at scientific and community meetings. Stakeholders, including patients and families, were compensated for their time, effort, and expertise. For detailed information about how stakeholders shared power, made decisions, influenced the study direction, and contributed to the overall research process, please see Williford et al. 31 Methods Study Overview This study was originally designed as a stepped wedge cluster randomized controlled trial in which each site is randomly assigned to a cluster, where each cluster recruits during a baseline and control period without intervention, and then clusters cross over 1 by 1 to the intervention on specific timelines. Consistent with a stepped wedge design, each clinic was to begin enrolling patients using the usual care dissemination method. Then, each cluster of clinics, 1 by 1, was to cross over to using the H-SDM toolkit dissemination method (1 cluster will cross over every 6 months). Training for the H-SDM toolkit was scheduled to begin during the last month of the usual care period for each cluster. Each cluster was set to enroll approximately 7 to 9 participants per time period. Enrollment was scheduled to end 36 months after study initiation. Because of low accrual followed by the COVID-19 pandemic (which decreased the feasibility of addressing previous accrual barriers and introduced new and unknown challenges to enrollment and implementation), however, we changed the study design and analyses such that sites were added ad hoc to increase recruitment and no longer cluster randomized (see the section “ Changes to the Original Study Protocol ”). Because randomization of clinics to clusters was not possible, the last 3 clinics were added later in the study and thus could not be randomly assigned to a specific crossover ordering. In other words, we could not standardize the time periods when clusters crossed over (because of enrollment challenges and COVID-19), so we could not implement the stepped wedge design as intended. We did not randomize by cluster, and the time recruiting in each dissemination method (usual care vs toolkit) was not standardized. Therefore, the design was changed to a quasiexperimental design. Yet our design maintained elements of a stepped wedge design including a key feature: unidirectional crossover. Other key benefits maintained were that (1) site motivation was high because all sites received the intervention and (2) each site contributed participants to each condition (usual care dissemination method and toolkit dissemination method), which minimized the impact of site biases on study outcomes. We were able to continue with 9 of the original 12 sites from across the United States. To ensure representation and generalizability of the sample, we included clinics of varying sizes. In this design, each clinic began to enroll patients using the usual care dissemination method, and then each clinic, 1 by 1, crossed over to using the toolkit dissemination method. In essence, our design retained critical elements of the original, which could not be perfectly implemented because of a delay in study start, recruitment challenges, and COVID-19–related interruptions. The revised design ensured that because individual sites spent time recruiting in both conditions (usual care and H-SDM toolkit), we were still able to test our hypotheses and address our primary aims. Our outcomes, and therefore the analyses, occurred at the individual patient level (not at the site level) and were measured only at a single time point (immediately after the discussion with the clinician). Although sites crossed over and recruited for both usual care and H-SDM toolkit methods (as is usual for most trial designs), the patients in the 2 treatment groups were independent and not affected by crossover between interventions or length of recruitment time specific to a site or condition. The crossover was a design choice that ensured no contamination between the usual care and toolkit conditions at the site level but had no impact on the patient-level analyses. Study Setting Twelve sites participated in the study; however, 4 sites were terminated before the end of the study for 1 of 2 reasons: (1) site principal investigators (PIs) and staff left and were not replaced or (2) accrual was consistently low. We selected sites that varied in SCD population size (<270 patients, small; 270-500 patients, medium; >500 patients, large) and geographic location (South, Northeast, Midwest). Sites were also selected because their patient populations mirrored the larger US SCD population with respect to income status and racial and ethnic diversity (eg, Latine/x, Black, Asian, more than 1 race). In addition, these SCD clinics work with patient populations from urban, suburban, and rural communities. Table 2 displays relevant site characteristics. Table 2 Site Characteristics. Participants A total of 176 participants (parents of young children with SCD, birth to 5 years of age) were enrolled in the study. Most completed measures at baseline and 3 to 7 months after a shared decision-making discussion with their clinician. All participants sought care for their children at 1 of the 12 sites. Inclusion Criteria We used the following inclusion criteria: Diagnosis: SCD Age: birth to 5.99 years, inclusive Eligible for hydroxyurea (genotype SS, sickle cell beta thalassemia, or other genotype and clinical complications) Child's parent, legal guardian, or designated decision-maker (caregiver) must participate in both study visits Child's parent, legal guardian, or designated decision-maker (caregiver) must be able to read, understand, and speak English Exclusion Criteria We used the following exclusion criteria: Parent or legal guardian has previously been approached or made a decision about whether to initiate hydroxyurea Any and all other diagnoses or conditions that, in the opinion of the site investigator or hematologist, would prevent the patient from being a suitable candidate for the study Sibling of participant actively enrolled in the study or enrolled in the past Interventions and Comparators or Controls Usual Care In this condition, sites received current guidelines for offering hydroxyurea and used the ASH pocket guide as a reference. The ASH developed The Hydroxyurea and Transfusion Therapy for the Treatment of Sickle Cell Disease clinician pocket guide based on the NHLBI Evidence Based Management of Sickle Cell Disease: Expert Panel Report, 2014 . The pocket guide recommends the use of shared decision-making. Toolkit Hydroxyurea Shared Decision-Making toolkit The H-SDM toolkit method includes (1) support tools for clinicians (templates for identifying eligible patients and monitoring progress) and optional quality improvement tools; (2) training for clinicians in cultural humility (eg, incorporating reflection on clinician nonverbal messages and parent values valuing identities, awareness of intersectionality, and exploration of preferences), shared decision-making, and communication skills; (3) a virtual reality simulation for deliberate practice of shared decision-making by clinicians; and (4) parent visit decision aids and video narratives. In this condition, sites developed a reliable method for identifying eligible patients and monitoring their progress; received clinician training in shared decision-making, cultural humility, and communication skills; and practiced using the visit decision aids in a virtual reality environment (eg, simulation of clinic room with parent avatars). 35 Additional details about the clinician training can be found in Crosby et al (2019). 35 The H-SDM toolkit includes 4 visit decision aids to support parents in their decision about hydroxyurea: previsit brochure, in-visit issue card (ie, card displaying issues parents identified as key to decision-making), after-visit booklet, and video narratives (videos of parents telling their stories about how they made their decisions about hydroxyurea). 25 Study Outcomes Primary Outcome Measures The primary outcome measures were as follows: Parent-reported decisional uncertainty as measured by the Decisional Conflict Scale (DCS). 36 , 37 The DCS measures uncertainty experienced when feeling uninformed about options, unclear about personal values, or unsupported in making a choice. Parents respond to 16 items on a 5-point Likert scale, from 0 (strongly agree) to 4 (strongly disagree). Items are summed, divided by 16, and multiplied by 25. Scores range from 0 (no decisional uncertainty) to 100 (extremely high decisional uncertainty), with a Cronbach α = .96. Parent perception of shared decision-making as measured by the dyadic OPTION. 38 This measure describes clinician behaviors that involve a patient or parent in decision-making (eg, “My doctor and I made the decision together”). A total score is calculated, ranging from 0 (no involvement) to 100 (maximal involvement). Both primary outcome measures were completed by the parent immediately after the discussion with the clinician about hydroxyurea. Secondary Outcome Measures The secondary outcome measures were as follows: Hydroxyurea offered. The site research coordinator reviewed the electronic health record (EHR) or talked with the clinician to determine, based on review of EHR data, whether hydroxyurea was offered, not offered, or previously prescribed. If hydroxyurea was not offered, coordinators chose a reason (ie, not eligible because patient is on transfusions, not eligible because patient has a comorbid condition, no time to offer, clinician forgot, ill visit, or an open field to enter another reason). For analysis, this variable was coded as 1 (hydroxyurea was offered during the study period) or 0 (hydroxyurea was not offered during the study period). Hydroxyurea prescribed (uptake). The site research coordinator reviewed the EHR and reported whether patients had an active prescription for hydroxyurea (prescription in the 6 months since the discussion visit). For analysis, this variable was coded as 1 (hydroxyurea prescribed) or 0 (hydroxyurea not prescribed). Parent report of neurocognitive functioning, as measured by the Ages & Stages Questionnaire. 39 This questionnaire is a reliable, accurate developmental and social-emotional screener for children between birth and 6 years of age, with a Cronbach α = .60 to .85. Scores range from 0 to 60, with higher scores indicating that the child's development is on schedule. Health care utilization. Electronic health record data on the number of hospitalizations, ill visits, and ED visits in the 12 months before enrollment, if available (some participants were 9 months of age or younger), and the months after enrollment. For analyses, the health care utilization variables were coded as 0 (no) or 1 (yes) for each of the 3 health care use types. Health-related quality of life. Parent report of SCD-specific QOL and pain, as measured by the Pediatric Quality of Life (PedsQL) SCD Module, 40 which assesses several domains of HRQOL, including pain impact, fatigue, pain management, emotions, communication, and treatment adherence. Scores range from 0 to 100, with higher scores indicating higher HRQOL or higher functioning. Hydroxyurea knowledge. Eight-item survey developed based on the existing literature, the Ottawa Knowledge User Manual, 41 and parent and clinician stakeholders and used in our pilot work. 25 Items are summed to obtain a total score ranging from 0 to 9, with higher scores indicating more knowledge. Satisfaction with decision-making. Eight-item survey adapted from the Satisfaction With Decision scale 42 (4 items) and the Agency for Healthcare Research and Quality's Consumer Assessment of Healthcare Providers and Systems survey related to patient experience of care (4 items). 43 Items are summed to obtain a total score ranging from 0 to 28, with higher scores indicating higher satisfaction. Secondary outcomes were measured after discussion with the clinician or at follow-up. Covariates The covariates were as follows: Demographic survey consisting of 10 items assessing family demographics, including patient and parent age in years, sex assigned at birth, race and ethnicity, socioeconomic status, insurance (public vs private), and parent highest level of education completed Health literacy, as measured by the Newest Vital Sign, which tests literacy skills for both numbers and words. 44 Cronbach α > .76, with scores ranging from 0 to 6, with higher scores indicating higher health literacy COVID-19 Exposure and Family Impact Survey (CEFIS). 45 Various aspects of the COVID-19 epidemic are likely to affect families and may influence the findings of research in pediatric health. The CEFIS Impact Score was used for the current project and consists of 12 items. Mean scores are used, with higher scores indicating more negative impact and higher distress. The first CEFIS was administered in May 2020, and 66 participants (those currently enrolled and newly enrolled) completed the CEFIS Sample Size Calculations and Power Original Power Analysis The original sample size calculation was based on minimal effect sizes from studies on the DCS (effect sizes range from 0.4 to 1.2), our primary outcome, and a cluster randomized design with unidirectional crossover (Hussey and Hughes approach 46 ). Specifically, power analyses were calculated using Optimal Design power analysis software, assuming (1) site (Youden J = 11) level variation at an intraclass correlation of 0 after group mean centering of all analysis variables, (2) populations between 154 and 209 individuals (assuming proper missing data handling) available for analysis, (3) effect sizes (δ) between 0.40 and 0.60, (4) the inclusion of control covariates (parent age, parent health literacy, participant sex assigned at birth, participant age, disease severity, and socioeconomic status), and (5) block randomizing by site size to reduce response variable error variance by ( R 2 = 0.40) 40%. Results showed that power will be greater than 0.80, even if sample size (N = 154) and effect size (δ = 0.40) are at their minimal values, if the treatment effect variation (σ 2 δ) is minimal (<0.011; see Table 3 ). Table 3 Minimal Sample Sizes for Treatment Effect Variation. Revised Power Analysis A revised power analysis was conducted using Monte Carlo simulation in Mplus, version 8.5, software with 5000 replications, assuming 83 patients in usual care and 77 patients in the H-SDM toolkit condition and up to 3 additional covariates in the model. The power analysis indicated that we would have more than 80% power to detect an effect size group difference on our continuous outcome measures of d = 0.44 or larger. To clarify the “revised” power analyses, d is a standardized effect size metric developed by Cohen, 47 with the following thresholds typically used to evaluate these effect sizes: d < 0.20 is considered trivial, d = 0.20 is considered small, d = 0.50 is considered a medium effect, and d ≥ 0.80 is considered a large effect. In this case, the effect size d more specifically refers to the standardized difference in the outcome between the toolkit method (treatment) and usual care (control). The Monte Carlo simulations assumed (1) within-site and within-cluster intraclass correlations would be 0.01 (because we expected trivial impacts from the clusters or sites) and (2) the variation in the outcome explained by covariates would be small ( R 2 = 0.10, a conservative estimate). For the revised power, we had completed recruitment in the control group, with a final sample at the time of 83 patients, because all sites had crossed over to the intervention condition at that time and approximately 26 patients had been recruited for the toolkit method by October 2021. We projected how many more we could reasonably recruit from October 2021 to the end of the study, given the existing recruitment challenges, which resulted in a projected sample size of 77 patients for the toolkit method. The revised power analyses therefore focused on the smallest effect size difference (in the Cohen d metric) that we would be able to detect between the toolkit and usual care dissemination methods, with a total sample size of 160 patients (ie, 83 already completed usual care and an anticipated 77 would complete toolkit), with at least 80% power in a Monte Carlo simulation study, and the assumptions listed earlier. Notably, the projected final sample size was within the range of the sample sizes from the original power analyses (see Table 3 ) and the effect sizes (ie, 0.40-0.60; note that the initial power analysis used δ, which referred to the effect size difference between the groups in Cohen d metric). Time Frame for the Study The study was conducted from 2018 to 2022. Recruitment began in July 2018 and finished in November 2021. Participants had 1 to 3 study visits. Some participants completed baseline measures during 1 visit, but others had 2 visits because they did not have a shared discussion until the second visit. We had planned to conduct follow-up visits 6 months after the discussion but shortened the time frame to 3 months with the goal of ending the study closer to the date originally planned (see the section “ Changes to the Original Study Protocol ”). We attempted to collect participant follow-up data for all participants recruited within 6 months of February 2022. Data Collection and Sources Data Collection Process We enrolled eligible young children with SCD who had been identified by clinician referral or EHR review. The research team approached the child's clinician to obtain approval before contacting the potential participant. Eligible parents of patients received a letter or flyer by mail or a phone call. A research coordinator with experience recruiting patients with SCD then either followed up by phone or approached families in clinic. Data were collected on how families are recruited, and there were no demographic differences between those who were recruited by phone or in clinic. Interested families provided consent in clinic or by phone (e-consent), and the baseline assessments were scheduled. Study visits were in person or by telehealth and were scheduled at the family's convenience (before or after clinic visits, evenings, and SCD events). If a parent or caregiver did not complete baseline assessments within 30 days of consent, then the parent or caregiver was rescreened, asked to provide consent again, and asked to complete baseline assessments again. A participant could be asked to provide consent again only 1 time. All parent data were collected in Research Electronic Data Capture (REDCap), a HIPAA-protected and secure research database for online surveys, or printouts of the surveys. All coordinator and site PI data were collected in REDCap, as well. Retention strategies included (1) scheduling visits at times convenient for the family (during clinic or telehealth visits), (2) making reminder phone calls for visits or assessments, (3) allowing questionnaires to be completed online, and (4) making check-in calls to promote retention during the COVID-19 pandemic. In addition, the rationale for 2 visits and the importance of follow-up were reviewed with each family at baseline to engage them as partners in the research process. To prevent loss to follow-up, we collected patient permission to contact 2 close family members or friends if the participant's family was difficult to reach. Data Sources We used the following data sources: EHR data. The site study team conducted a medical chart review to obtain information about the child's hemoglobin type, hydroxyurea status, hospitalizations and ED utilization, clinic visit notes, and clinic processes (eg, use of hydroxyurea guidelines). Questionnaires. Parent surveys are described in the “ Study Outcomes ” section of this report. Clinicians also completed an electronic demographics form and a self-rating of whether they completed the toolkit clinician training. Fidelity. Fidelity was measured using a parent survey and a clinical research coordinator case report form. We had also planned to record at least 3 clinic visits per site to assess fidelity to ensure that the clinician portion of the toolkit was being implemented as instructed; however, sites were unable to submit recordings after the COVID-19 pandemic. Analytical and Statistical Approaches We redesigned the study so that the time that individual sites spent recruiting in a specific condition (usual care or H-SDM toolkit) would not affect our hypotheses or our primary aims. Our outcomes, and therefore the analyses, occur at the individual patient level (not at the site level) and were measured only at a single time point. Although sites recruited for both usual care and H-SDM toolkit methods (as is usual for most trial designs), each site was recruiting for only 1 condition at any given time, and thus the patients in the treatment groups are independent and were not affected by crossover or length of recruitment time specific to a site or condition. The crossover was a design choice that ensured no contamination between the usual care and toolkit conditions at the site level because once sites began the toolkit method they did not revert to the usual care method; thus, this has no impact on the patient-level analyses. Overall, there is little difference between our original analysis plan and revised analysis plan; both involved comparing patient-level outcomes for those in the treatment group with those in the usual care control group. Originally, we planned to use group mean centering to eliminate response variable variance due to site and assumed that cluster variance would be negligible. Instead, we have controlled for cluster explicitly in the regression models detailed below and accounted for site-level variation by specifying that the SEs allow for intragroup correlations, given that the individual observations within sites are likely to be dependent (eg, the vce(cluster site) command in Stata [StataCorp]). We have also accounted for the pandemic that interrupted the study, which was not part of the originally planned analyses. Impacts of COVID-19 were accounted for in 2 ways: (1) We created a variable at the individual patient level that describes when data were collected: 0 (before COVID-19) or 1 (affected by COVID-19). We controlled for this variable in the analyses. (2) We administered a measure that assesses the psychosocial impacts of COVID-19, and we included these scores as a covariates in the regression analyses. 48 To determine the effect of our H-SDM toolkit condition or usual care on parent-reported decisional uncertainty, as measured by DCS total scores (the primary outcome) assessed immediately after implementation of the treatment condition, we used a linear regression model (a method to describe the relationship between an independent and a response variable such as height and weight) in Stata, version 17, software 39 with our main predictor of interest entered as a binary indicator of group (toolkit vs usual care). The model was estimated using full information maximum likelihood estimation on the full analysis sample (N = 146) to account for missing data. To account for site-level variation, we specified that the SEs allow for intragroup correlations and controlled for when the sites crossed from the usual care to the toolkit condition (ie, the cluster). Finally, a variable at the individual patient level that described when data were collected (0 [before COVID-19] and 1 [affected by COVID-19]) was included as a covariate in the model, as were the impact scores collected from the CEFIS. The same modeling strategy and same covariates were used for our other primary outcome, parent report of shared decision-making, as measured by dyadic OPTION total scores, as well as our secondary outcome scores from the Ages & Stages Questionnaire and the PedsQL. Health care variables were all binary, with 0 (no) and 1 (yes) and analyzed using logistic regression in Mplus, version 8.6, software 41 again controlled for site, cluster, and COVID-19 timing and impact using full-information maximum likelihood estimation on the full analysis sample (N = 146; n = 83 usual care and n = 63 toolkit). We also calculated effect sizes for all outcomes to determine how large the difference was between the means for the dissemination methods and examined clinically important differences on our primary outcomes (ie, whether the mean scores for parents in both groups were above or below clinical cutoffs). The trial experienced accrual challenges (eg, delayed recruitment), so data were systematically collected on recruitment barriers with the goal of developing targeted interventions. We conducted analyses to examine clinically important differences on the primary outcome measures for the groups. For the DCS, mean scores below 25 indicate making a decision, and mean scores above 35 are associated with high decisional regret. 37 Point differences between groups can be clinically important because studies have found that with every unit increase, the likelihood of poor decisional outcomes (regret, delay, blame) increased from 3 to 59 times. 37 , 49 For the dyadic OPTION, there are no published clinical cutoffs. Changes to the Original Study Protocol The original study design was a stepped wedge cluster randomized controlled trial in which all clinics (N = 9) would begin enrolling patients using the usual care dissemination method. Then, each cluster (group of 2 sites or clinics), 1 by 1, would cross over to using the H‑SDM toolkit dissemination method (1 cluster will cross over every 5 months) and stop doing usual care. 50 The order in which the clusters crossed over was to be random; the cluster was to be the unit of randomization ( Figure 1 ). We planned to cluster sites together based on characteristics such as the number of clinicians, number of patients, region, and patient racial and ethnic mix in an attempt to balance the clusters. There were to be 4 clusters, with each cluster consisting of 2 sites for a total of 8 sites. Training for the H-SDM toolkit was planned to begin during the last month of the usual care period for each cluster. Participants (parents of young children with SCD) would complete measures at the baseline clinic visit and 6 months later in REDCap. The study experienced many recruitment challenges, however, with delays in the first few months of the study, and had low accrual rates (<50% of projected milestones). We therefore added a site and changed to 3 clusters (3 sites per cluster), but 2 sites were closed early on in the study (2019), 1 due to low accrual and the other due to staff turnover. To increase accrual, we added 3 more sites and lowered the overall enrollment target from 220 to 176 (February 2020; Figure 2 ). After several months, the study was meeting monthly enrollment goals, and we were on track to meet study enrollment goals. Figure 1 Initial Engage Hydroxyurea Study Design and Timeline. Figure 2 Engage Hydroxyurea Revised Study Design and Timeline. In March 2020, the first COVID-19 shutdown occurred, initially halting all recruitment and enrollment. At this time, all sites but 1 had completed H-SDM toolkit training. Approximately the last 64 participants enrolled in the study were enrolled during the COVID-19 pandemic, with most of them being assigned to the H-SDM toolkit group because their sites had crossed over to this dissemination method. Once some sites were able to resume research studies (July-August 2020), most clinic and research visits were switched to telehealth. After the COVID-19 shutdowns, clinic volumes never rebounded to pre–COVID-19 rates, which made meeting enrollment targets impossible. In addition, sites had challenges with study implementation and delivery (see “Lessons Learned”), and staff turnover also resulted in recruitment delays and low accrual. For example, although we collected fidelity data from parent participants, sites were unable to record sessions for fidelity checks as planned after the COVID-19 pandemic. The study team made multiple attempts to obtain recordings, but only half of the sites submitted recordings. Follow-up was impeded by difficulties reaching participants during the COVID-19 pandemic because participants were attending fewer clinic visits and had less communication with medical teams. Although we collected data on friends and family members, many participants did not answer calls or return messages. Despite several retention strategies, many participants could not be reached and did not complete follow-up visits. Efforts made to mitigate these challenges included retraining staff, clarifying recruitment targets, offering incentives for coordinators, extending recruitment, and adding new sites. Recruitment challenges persisted, and a decision was made to lower the overall recruitment goal and change the study design (August-September 2021; see Figure 3 later in this report). The following modifications were made to the study protocol: Figure 3 Engage Hydroxyurea Study Consort Flow Diagram. The study design was changed from a stepped wedge to an observational quasiexperimental design. Two sites closed because of low accrual; 3 sites were added to bolster recruitment. The follow-up period was shortened from 6 months to 3 months, which allowed for the collection of as much follow-up data as possible given that the study enrollment goal was not met until the end of the research period (November 2021). Analyses to examine the timing of data collection and the impact of COVID-19 were added to the analysis plan. The modified protocol made it possible to test the original hypothesis despite the smaller overall sample size. Although we recruited 82 participants for the toolkit condition, data were available for only 63 rather than the 77 participants targeted, limiting our ability to make definitive statements about the effectiveness of the dissemination methods. Results Allocation Overview of Participant Flow A total of 692 parents and caregivers were identified for study eligibility across sites. Of these, 515 parents and caregivers were excluded ( Figure 3 ) for not meeting inclusion criteria (n = 107), declining to participate (n = 37), missing or canceling appointments (n = 100), or other medical or psychosocial reasons (n = 271). A total of 176 parents consented to participate and were assigned to either the usual care or toolkit condition based on when they were enrolled in the study (ie, site in usual care time period or H-SDM toolkit time period). For the primary outcomes, data were collected at or after the shared discussion (before follow-up visits). Secondary outcomes were collected after discussion with the clinician or at follow-up. Twenty-eight participants (11 from usual care and 17 from toolkit) were excluded from the analyses for 1 or more of the following reasons: (1) did not complete primary outcome measures before the study ended, (2) lost to follow-up, or (3) withdrawn from the study (individually or due to site closure). Regardless of the reason, we had no baseline data to use in the analyses for these participants (n = 28). In total, data from 146 participants were available to calculate baseline characteristics (83 in usual care and 63 in toolkit; Table 4 ). All parent participants completed an informed consent process with a member of the study team and provided written consent with their electronic signature (e-signature) on an electronic copy of the consent form housed in an e-consent portal managed by the central study team. There were no differences in demographic characteristics between parent participants who only completed the consent form and those who participated in at least some part of the study (ie, completed at least 1 survey). Figure 3 documents the study flow. Table 4 Study Participant Characteristics Overall and by Condition. Enrollment Study Participant Characteristics A total of 176 parents consented to participate and were assigned to either the usual care or toolkit condition based on when they were enrolled in the study (ie, site in usual care time period or H-SDM toolkit time period). The mean (SD) participant, parent, and caregiver age was 30.13 (7.35) years. Most parent participants reported a female sex assigned at birth (93.1%); with respect to racial identity, participants self-identified as Black (89.0%), mixed race (6.2%), White (1.4%), Asian (0.7%), or other (2.1%). Participants were primarily mothers (90.4%), but some were fathers (5.5%) or grandmothers (2.7%), and there was 1 uncle (0.8%). The mean (SD) age of the children was 1.11 (1.43) years, and most of them had the homozygous sickle cell anemia (94.5%) or sickle cell beta thalassemia (2.7%) genotype. There were no significant differences in child age, parent age, genotype, or parent sex assigned at birth between participants in usual care or the toolkit conditions. We obtained health literacy data on participants at baseline. Parent participants obtained a mean (SD) score of 4.24 (1.52) on the Newest Vital Sign, which indicates adequate literacy. Our participants (n = 42) obtained a mean (SD) score of 2.61 (0.89) on part 2, Impact of the CEFIS, suggesting COVID-19 had a moderate impact on study participants (scores closer to 4 indicate more negative impact). Analyses Aim 1 Results Evaluate the effectiveness of the usual care dissemination method (clinician pocket guide) and the H-SDM toolkit dissemination method on (1) parent report of decisional uncertainty for hydroxyurea (effectiveness) and (2) parent report of experiencing shared decision-making when talking with their clinician about hydroxyurea (effectiveness) . After we controlled for COVID-19 timing (completing the study before or after the pandemic), COVID-19 impact (CEFIS Impact score), cluster (when in the study sites switched to the toolkit condition), and site, there was no statistically significant difference between the toolkit and usual care methods on parent-reported decisional uncertainty, as measured by the DCS total scores ( b = −5.73 [95% CI, 0-100]; P = .16; Cohen d = 0.16). There were no statistically significant differences on any of the DCS subscales (effect size differences ranged from d = 0.10 for Values Clarity to d = 0.22 for Informed; Table 5 ). Analyses of clinically important differences found that the observed (ie, unadjusted) mean score on the DCS total score for parents enrolled during the usual care time period was 3.1 points higher. Table 5 Observed (ie, Unadjusted) Means, CIs, SDs, and Standardized Effect Size Differences by Group on the DCS With Regression Coefficient Results From the Aim 1 Analysis ( b Coefficient, P Value). After we controlled for COVID-19 timing (completing the study before or after the pandemic), COVID-19 impact (CEFIS Impact score), cluster (when the study sites switched to the toolkit condition), and site, there was no statistically significant difference between the toolkit group and the usual care group on parent perception of shared decision-making as measured by the dyadic OPTION measure ( b = 2.35 [95% CI, −12 to 48]; P = .13; Cohen d = 0.09). Analyses found that the observed (unadjusted) mean score on the dyadic OPTION measure for parent participants in the toolkit method was 40.3; the mean for those in usual care method was 39.8. Aim 2 Results Evaluate the effectiveness of the usual care dissemination method (clinician pocket guide) and the H-SDM toolkit dissemination method on (1) parent knowledge of hydroxyurea (effectiveness); (2) children offered hydroxyurea (reach); (3) children with an active hydroxyurea prescription (uptake: effectiveness); and (4) child health outcomes: pain, neurocognitive functioning, sickle cell–related QOL, and health care utilization (effectiveness) . After we controlled for COVID-19 timing, COVID-19 impact, cluster, and site, there were no statistically significant differences for parent-reported SCD HRQOL as measured by the PedsQL Sickle Cell Disease Module (total score), parent report of neurocognitive functioning as measured by the Ages & Stages Questionnaire, or parent satisfaction with decision-making or parent hydroxyurea knowledge between the methods (toolkit or usual care) at follow-up ( Table 6 ). There was a statistically significant difference on 1 subscale of the SCD Health-Related Quality of Life measure, Communication 2 Subscale, suggesting that parents in the toolkit condition reported fewer problems communicating about their child's SCD and pain with important others ( b = 6.45 [95% CI, 0.90-12.00]; P = .02; d = 0.22). Notably, the observed (ie, unadjusted) means for this subscale indicate that the usual care condition had higher Communication 2 scores (M = 84.8 vs M = 78.3). After adjustment for covariates in the model, however, the toolkit condition had higher Communication 2 Subscale scores (marginal, ie, adjusted M = 77.46 vs marginal M = 71.00 for the usual care group). Table 6 Observed (ie, Unadjusted) Means, 95% CIs, SDs, and Standardized Effect Size Differences, by Group, on the Secondary Outcomes, With Regression Coefficient Results From the Aim 2 Analysis ( b Coefficient, P Value). Table 7 displays health care utilization and hydroxyurea outcomes. With respect to health care utilization, there was no significant difference in the odds of health care use between children whose parents participated during the usual care vs toolkit dissemination methods at baseline or follow-up (OR, 1.32 [95% CI, 0.64-2.73]; P = .45). Approximately 28.4% of children with SCD whose parents participated during the usual care method reported their child had at least 1 acute pain episode compared with 26.3% of those enrolled during the toolkit method. Likewise, there were no statistically significant differences in the odds of hospitalization (OR, 0.47 [95% CI, 0.11-2.04]; P = .32) or ED visits (OR, 0.63 [95% CI, 0.22-1.78]; P = .38) for children of parents enrolled during the toolkit method or usual care method. Furthermore, 51.2% of parents enrolled during the usual care method use period reported that their child had a recent hospitalization and 38.1% had a recent ED visit compared with 44.3% of those enrolled during the toolkit method use period reporting hospitalizations and 35.4% reporting ED visits. Table 7 Odds of Health Care Use, Offering Hydroxyurea, and Receiving Hydroxyurea Prescription. Most participants enrolled during the usual care method use period (80.7%) were offered hydroxyurea through shared decision-making compared with only 58.7% enrolled during the toolkit method use period. This difference was statistically significant in that the odds of being offered hydroxyurea during the toolkit method (OR, 0.11 [95% CI, 0.03-0.40]; P = .00) were significantly lower than during the usual care method. There was also a significant difference in the odds of receiving a hydroxyurea prescription between the conditions (OR, 0.26 [95% CI, 0.10-0.70]; P = .01), such that participants enrolled during the toolkit method had lower odds of receiving a hydroxyurea prescription during the study than those enrolled when the usual care method was used (48.2% children of parent participants during usual care receiving a prescription vs 39.7% of children of parent participants when the toolkit method was used). There were no significant differences between the groups with respect to demographics ( Table 8 ). Additional analyses were not possible because of the low numbers in the dropout group. We also examined differences in primary outcomes for older (3-5 years of age) and younger children (birth to 2 years of age; Table 8 ) enrolled in the study. Table 8 Outcomes by Completion Status and Age Group. Discussion Summary of Results This study sought to examine a better way to facilitate shared decision-making about hydroxyurea for parents of young children with SCD. There were no statistically significant differences between the dissemination methods with respect to the primary study outcomes: parent decisional uncertainty and parent report of shared decision-making. Study findings indicated a clinically important difference in that the observed unadjusted means between the groups differed for decisional uncertainty. Specifically, parent participants whose clinicians used the toolkit dissemination method reported less decisional uncertainty (3.1 points lower mean score) than parents whose clinicians used the usual care method, surpassing the empirically established clinically important difference threshold of 1 point. 37 , 49 Future studies are needed to confirm this finding. Findings on the parent report of shared decision-making measure suggest that clinicians who used either dissemination method demonstrated similar levels of shared decision-making behaviors (unadjusted mean scores were equivalent). There were statistically significant differences in secondary outcomes. Parents enrolled during the toolkit method use period reported fewer problems communicating about SCD with important others. Clinicians using both dissemination methods were offering hydroxyurea through shared decision-making (≥58.7%; ie, adhering to national evidence-based guidelines). The odds of being offered hydroxyurea (OR, 0.11 [95% CI, 0.03-0.40]; P < .01) or receiving a hydroxyurea prescription (OR, 0.26 [95% CI, 0.10-0.70]; P = .01), however, were significantly higher for parents enrolled during the usual care method. Subgroup analyses revealed no significant differences in demographics between those who completed the study and those who dropped out. Clinician fidelity with the toolkit dissemination method was high. With respect to toolkit dissemination method components, clinicians showed statistically significant increases in communication and shared decision-making skills and self-confidence after receiving training as part of the toolkit dissemination method. Results in Context Despite the change in study design, this quasiexperimental study found that clinicians were using a shared decision-making process most of the time (≥58.7%; ie, adhering to national evidence-based sickle cell guidelines). This is encouraging because closing the gap in translation of guidelines into practice can be challenging especially in light of the COVID-19 pandemic. The literature has shown that decisional uncertainty is worse when a person feels uniformed, is unclear about their values, and is unsupported in the decisional process 30 and that decision-making interventions can lower decisional uncertainty. Both sets of parents had DCS scores lower than 25, which have been associated with implementing decisions, 37 suggesting that the parents in the sample are highly likely to make a decision about hydroxyurea. Parent participants in the toolkit condition had a mean DCS score of 19.8, whereas those in the usual care condition had a mean of 23.5. Although not statistically significant, this difference is clinically important because every unit increase in the DCS total score is associated with negative decisional outcomes such as decisional regret (5 times more likely), delaying the decision (23 times more likely), and changing their mind (59 times more likely). 37 , 49 Studies have also found that with every point increase in DCS score, people were 19% more likely to blame their clinician for a bad health outcome. 51 These results would need to be replicated in future studies. Shared decision-making involves a dyadic and relationship-centered approach. 38 , 52 We trained clinicians in the toolkit condition by using a didactic approach to shared decision-making and communication and motivational interviewing skills, followed by deliberate practice using virtual reality simulation. 35 Evaluation data showed significant pre-post differences in clinicians' self-reported confidence in key shared decision-making skills: discussing benefits, costs, and the impact of hydroxyurea on daily life. Clinicians also reported increased confidence in motivational interviewing and cultural humility skills: using open-ended questions, asking about specific concerns, valuing differences and perceptions by confirming understanding, and using an elicit-provide-elicit approach. These results suggest high feasibility and acceptability of the toolkit as a potential method for facilitating confidence with shared decision-making for hematology clinicians. The toolkit dissemination method should therefore be further evaluated for its potential to support clinician implementation of shared decision-making; the extent to which the COVID-19 pandemic and the sample being highly likely to make a decision affected study results is unclear. Additional research is therefore needed to determine whether this training reduces decisional uncertainty in families of young children with SCD. Study results were substantially affected by the COVID-19 pandemic as clinical volumes never rebounded, staff turnover was high, and both study recruitment and implementation challenges were experienced. In addition, most parents assigned to the H-SDMT toolkit method were enrolled during the COVID-19 pandemic. Our analyses found that parents enrolled during the toolkit method reported fewer problems communicating about SCD with important others. Perhaps this is because these parents had been talking with daycares, schools, or family members more about their child's condition because of the COVID-19 pandemic and were thus more comfortable with discussions about their child's diagnosis and pain. We also found that children of parent participants were significantly more likely to be offered and prescribed hydroxyurea when the usual care dissemination method was used; however, additional information is needed to understand this finding because it may be an artifact of time of enrollment in the study and other unknown pandemic-related factors. Almost all of the usual care dissemination method data were collected before COVID-19, whereas almost all of the toolkit dissemination method data were collected during the pandemic (from 64 parents, most of them assigned to the toolkit condition). At that time, clinicians' priorities and practices may have changed and been more focused on immunizations and COVID-19 vaccination rather than prescribing new disease-modifying therapies; in addition, clinicians may have been fatigued from the pandemic. We also know that overall enthusiasm for studies decreases over time, and this may also have been a contributor. Six of 9 sites in the study reported that they offered prescriptions 100% of the time and maintained that rate throughout the study. Data collected from the 3 remaining sites showed high rates of offering but differences in hydroxyurea prescription rates. For these sites, during the usual care dissemination method, hydroxyurea was offered approximately twice as often (mean [SD] usual care, 87.5% [33.35%]; mean [SD] toolkit, 48.91% [50.27%]). Children at these sites were prescribed hydroxyurea more often, however, when the toolkit dissemination method was used (mean [SD] toolkit, 37.5% [47.1%]; mean [SD] usual care, 14.63% [27.45%]). These data suggest that there might be differences in site characteristics that affect offering hydroxyurea through shared decision-making. Future research should therefore examine barriers and facilitators to shared decision-making and prescribing at the site level (eg, setting factors). Potential to Affect Health Care Decision-Making Study results have the potential to affect the quality of care for children with SCD because shared decision-making has been identified as a practice to reduce health care inequities. We expect that parents who are more certain about their decision, are clearer about their values, and feel supported through a shared decision-making process will feel more empowered. This study provided data to support further evaluation of the toolkit method to support health care decision-making in this population. Toolkit components were designed to support the decisional needs of parents and clinicians when discussing preventive treatment with hydroxyurea for asymptomatic children with SCD. Toolkit-trained clinicians reported being more confident in their ability to discuss what matters most to families about hydroxyurea (ie, benefits, impact on daily life) in a manner that promotes an open dialogue and encourages parents to take an active role in decision-making (promotes equity). Use of the toolkit method thus may help to meet both parent and clinicians needs. Ultimately, this toolkit may be most beneficial for sites that are not routinely using a shared decision-making process with their patients and families because it may increase the consistent use of shared decision-making at those sites and their adherence with national evidence-based SCD care guidelines. Lessons Learned Our experience implementing this study can provide lessons for future studies of shared decision-making in this population and those with other rare pediatric diseases. In the United States, SCD affects primarily people of African and Hispanic or Latine/x origin. Although our study design incorporated best practice strategies for recruiting and retaining people of color in research, we experienced recruitment challenges that were exacerbated by the COVID-19 pandemic. 34 Recruitment was slower than anticipated and varied across sites. We had to close sites with low enrollment and add additional sites. This was challenging given that SCD is a rare disease with a small number of available participants. We used the Consolidated Framework for Implementation Research (CFIR) 53 to categorize recruitment and enrollment barriers and to guide the development of targeted strategies to improve enrollment. After implementation of the recruitment strategies, the number of caregivers identified for prescreening increased from 54 to 164, and enrollment more than tripled from 14 to 46 caregiver participants. 54 The CFIR constructs guided the development of targeted strategies that were effective in reducing attrition, increasing site participation, and increasing enrollment and could be implemented within a quasiexperimental design. The reflective process of applying CFIR constructs reframes recruitment challenges as the responsibility of the research team rather than characterizing the study population as “difficult” or “hard to reach,” which could happen in studies where the population is primarily Black or Hispanic or Latine/x. Future trials with patients with SCD and minoritized populations may benefit from applying CFIR constructs to study implementation. Stakeholder engagement was an integral part of this study, as mentioned previously. In addition, we applied social justice, antiracism, and community participatory strategies throughout the study (eg, research question, implementation, dissemination). The participatory process between researchers and stakeholders resulted in selection of a research question relevant and valuable to the community. Aligning study goals with community priorities protects against prioritizing the needs or interests of the dominant White group. In this way, it can empower racialized communities by ensuring that stakeholder voices remain centered. Using these strategies also ensures that the research team remains attentive to equity and inclusion throughout the research process. 55 Ongoing engagement of stakeholders helps to redistribute power between the researchers and participants by acknowledging the value of stakeholders. Furthermore, we applied an intersectionality lens to the research. For example, when developing site training, the research team included information on institutional practices and how a patient's intersecting identities reframe and shape patient-clinician interactions. 56 This approach was also used in designing a training to help site coordinators feel comfortable approaching this primarily Black population during the dual pandemics of COVID-19 and social unrest. The training provided an in-depth understanding of the complexity of experiences of Black mothers who may be experiencing higher disease burden and racial trauma. It also focused on the use of trauma-informed skills such as empathy and support. These strategies could be applied in other clinical trials with racialized or marginalized populations to engage them in research. Subgroup Analyses To better understand the study population and the potential effectiveness of the dissemination methods, we conducted several subgroup analyses. With respect to study participants, first we examined the impact of telehealth delivery to ensure that it did not have a negative impact on our outcomes. Data from the Telehealth Usability Questionnaire (n = 42) indicated that parent participants' overall satisfaction with telehealth was good (mean [SD] score, 17.2/21 [10.5]). Participants felt telehealth was easy to use (mean [SD] score, 26.9/28 [15.4]) and effective (high interaction quality; mean [SD], 21.5/28 [13.0]). Ratings, however, were lower for telehealth reliability (mean [SD] score, 10.8 [7.4]) and usefulness (mean [SD] score, 12.17/21 [7.7]). Second, we sought to examine the impact of parent health literacy on outcomes. The high level of parent health literacy in our sample was a strength and precluded our conducting subgroup analyses from examining whether outcomes differed by parent health literacy level. Third, we had planned to examine differences between parents who declined and parents who enrolled; however, the only data collected on parents who declined study participation were the sex assigned at birth of the patient and the site. Once participants declined, we did not collect any additional information. Fourth, to understand differences between parents who completed the study and those who did not (dropouts), we reviewed the demographics of both groups and mean scores on the primary outcomes. There were no significant differences between the groups with respect to demographics (see Table 8 ). Additional analyses were not possible because of the low numbers in the dropout group. Finally, we sought to determine whether there were differences in primary outcomes for older (3-5 years of age) and younger children (birth to 2 years of age; see Table 8 ) enrolled in the study. We also sought to conduct subgroup analyses at the clinician and clinic or site level with the goal of identifying clinician or clinic differences with respect to outcomes. First, to identify differences between clinicians, we collected demographics on them. Most clinicians (n = 44) were physicians (51.8%) or nurses (12.5%), identified as female (64.3%), were between 35 and 44 years of age (35.7%), and had worked with children with SCD for more than 10 years (67.5%). We had hoped to identify differences between clinicians who adopted NHLBI evidence-based SCD guidelines and those who adopted the toolkit vs not. Most clinicians, however, reported adopting NHLBI guidelines for hydroxyurea in young children 9 months of age or older (83.4%). Second, to assess adoption of the toolkit, we used fidelity assessments. Fidelity was very high in that 95.7% of toolkit parent participants reported receiving at least 1 decision aid during their clinic visit (ie, their clinician used the toolkit components as instructed). Because fidelity was so high and all clinicians reported using the toolkit, there were no sites meeting criteria for “less than full adoption.” We were therefore unable to examine differences in sites that adopted the full toolkit vs core components. Third, to assess continued implementation of the toolkit, parents reported on use of decision aids or shared decision-making at follow-up visits. Data indicate that 69.4% of participants reported experiencing some element of shared decision-making at follow-up visits. Given study enrollment challenges, we examined differences in sites with respect to study and intervention implementation. Specifically, we collected data from weekly site calls and used CFIR constructs to categorize barriers and facilitators to implementation. 54 Barriers mapped onto 3 CFIR constructs: (1) process barriers (ie, no identified “site champion” and poor recruitment planning at several sites), (2) inner setting barriers (ie, limited communication, low relative study priority at several sites), and (3) outer setting barriers (ie, poor patient attendance at clinic appointments). Targeted strategies to improve implementation included (1) PI site visits and retraining on recruitment procedures to address process barriers; (2) increased frequency of communication through all coordinator, site PI, and individual site calls to address inner setting barriers; and (3) development and implementation of no-show procedures for clinic appointments to address outer setting barriers. Future studies should examine implementation barriers and facilitators to implementation of the toolkit dissemination method that could inform the spread to other SCD clinics. Study Limitations Study findings should be understood in the context of several limitations. First, and most significantly, treatment group and the onset of the COVID-19 pandemic are almost entirely confounded, such that those recruited during the usual care dissemination method occurred before the onset of the pandemic and those recruited to the toolkit condition occurred during the pandemic. This limits our ability to make definitive conclusions about the differential effectiveness of the 2 dissemination methods for facilitating shared decision-making about hydroxyurea in this population. Second, because of challenges with study implementation and low accrual, the study design was changed from a stepped wedge to a quasiexperimental study with a unidirectional crossover. We were able to maintain some of the advantages of the stepped wedge design including that all sites crossed over to the toolkit dissemination method. We considered site and cluster effects in our analyses. Staff turnover and practice variation across the sites, however, greatly affected study implementation and delivery and participant accrual. For example, all sites that terminated early were small, and all but 1 closed after losing staff. Third, the sample size was lower than originally projected because of variable rates of recruitment across sites, study implementation barriers (see “Lessons Learned”), and changes in practice (eg, telehealth visits, less frequent visits) in response to the COVID-19 pandemic. The sample size shortfall did have the potential to affect power, but the observed effect size differences between the groups were substantially lower than anticipated, meaning that even if the target sample size had been recruited, there would still not have been enough power to detect the smaller observed effect sizes. Fourth, the study did not include children from all over the United States or those who may experience language barriers, so findings may not apply to all clinicians working with younger children with SCD; however, the sites were selected because their patient populations mirror the larger US SCD population with respect to racial identity and income. Finally, the study population consisted of parents who come to clinic for services. We do not know the impact of the dissemination methods on decisional uncertainty of parents who do not come to clinic regularly because they are experiencing barriers related to the social determinants of health and access to care. Future Research Study results and limitations highlight opportunities for future research. Future studies may want to use a study design where both dissemination methods are implemented simultaneously; this will help to determine how much differences in study outcomes can be attributed to implementation factors (eg, setting, turnover) or the dissemination methods themselves. Data suggest that some sites may have struggled with reliably collecting and using data to monitor their use of shared decision-making over time (ie, rate of 100%). This suggests that these sites might need more structure to implement the study; different implementation methods could be evaluated in a future implementation trial. It is important that future studies also examine the role of other key clinical team members, including physician extenders, nurses, social workers, and psychologists. Moreover, future studies could be designed to examine the relationship between shared decision-making and hydroxyurea adherence. 57 Understanding context-specific factors will be essential to optimizing the spread of the toolkit dissemination method to additional sites. Finally, it is essential to continue to collect data on parent perceptions of shared decision-making in real-world settings. Conclusions This study compared 2 ways to facilitate shared decision-making about hydroxyurea for parents of young children with SCD by using a pocket guide (usual care) or toolkit dissemination method. The study used a multifaceted approach to assessing the shared decision-making process, measuring not only parent decisional uncertainty (impact) but also clinician shared decision-making behaviors and parent satisfaction with decision-making. We did not observe statistically significant differences between the 2 dissemination methods on our primary outcomes. Specifically, there were no statistically significant differences on the mean DCS total scores, DCS subscales, or dyadic OPTION total scores between the groups. There was a 3.1‑point difference in DCS mean scores, with the toolkit DCS mean score being lower. For clinical context, every unit increase in DCS total score is associated with negative decisional outcomes such as decisional regret (5 times more likely), delaying the decision (23 times more likely), and changing their mind (59 times more likely). 37 , 49 With respect to secondary outcomes, parents enrolled during the toolkit method reported statistically significantly fewer problems communicating about SCD with important others. We found that participants enrolled when the toolkit method was used were less likely to be offered and prescribed hydroxyurea. We also found statistically significant pre-post differences between clinicians in knowledge and self-efficacy for shared decision-making when trained with our toolkit method (didactic skill-based training in motivational interviewing, cultural humility, and deliberate practice with virtual reality simulation). Although the COVID-19 pandemic led to recruitment and implementation challenges, these challenges led to opportunities to better understand contextual factors needed for effective implementation of shared decision-making in busy clinics that serve children with complex social and medical challenges. The CFIR constructs guided the development of system-level interventions that did not blame parents or patients. Finally, this study applied community participatory research and social justice and antiracism approaches and resulted in an article 31 detailing how we applied these approaches throughout the study and the impact that may be of use to future researchers. References 1. Elwyn G, Grande SW, Barr P. Observer OPTION-5 Manual. Measuring shared decision making by assessing recordings or transcripts of encounters from clinical settings.: The Dartmouth Institute for Health Policy and Clinical Practice; 2018. 2. Hassell KL. Population estimates of sickle cell disease in the US. Am J Prev Med. 2010;38(4):S512-S521. doi:10.1016/j.amepre.2009.12.022 [ PubMed : 20331952 ] [ CrossRef ] 3. Brawley OW, Cornelius LJ, Edwards LR, et al. National Institutes of Health Consensus Development Conference statement: hydroxyurea treatment for sickle cell disease. Ann Intern Med. 2008;148(12):932-938. doi:10.7326/0003-4819-148-12-200806170-00220 [ PubMed : 18458271 ] [ CrossRef ] 4. Wang WC, Ware RE, Miller ST, et al. Hydroxycarbamide in very young children with sickle-cell anaemia: a multicentre, randomised, controlled trial (BABY HUG). Lancet. 2011;377(9778):1663-1672. doi:10.1016/S0140-6736(11)60355-3 [ PMC free article : PMC3133619 ] [ PubMed : 21571150 ] [ CrossRef ] 5. Thornburg CD, Dixon N, Burgett S, et al. A pilot study of hydroxyurea to prevent chronic organ damage in young children with sickle cell anemia. Pediatr Blood Cancer. 2009;52(5):609-615. doi:10.1002/pbc.21738 [ PMC free article : PMC5600482 ] [ PubMed : 19061213 ] [ CrossRef ] 6. Thornburg CD, Calatroni A, Panepinto JA. Differences in health related quality of life in children with sickle cell disease receiving hydroxyurea. J Pediatr Hematol Oncol. 2011;33(4):251-254. doi:10.1097/MPH.0b013e3182114c54 [ PMC free article : PMC3729442 ] [ PubMed : 21516020 ] [ CrossRef ] 7. Yawn BP, Buchanan GR, Afenyi-Annan AN, et al. Management of sickle cell disease: summary of the 2014 evidence-based report by expert panel members. JAMA. 2014;312(10):1033-1048. doi:10.1001/jama.2014.10517 [ PubMed : 25203083 ] [ CrossRef ] 8. Rexhepi H. Information Systems for the Practice of Evidence-Based Medicine and Shared Decision Making. Academic Conferences International Limited; 2016:258. 9. Crosby LE, Shook LM, Ware RE, Brinkman WB. Shared decision making for hydroxyurea treatment initiation in children with sickle cell anemia. Pediatr Blood Cancer. 2015;62(2):184-185. doi:10.1002/pbc.25124 [ PMC free article : PMC5278638 ] [ PubMed : 25308571 ] [ CrossRef ] 10. Légaré F, Stacey D, Turcotte S, et al. Interventions for improving the adoption of shared decision making by healthcare professionals. Cochrane Database Syst Rev. 2014;15(9):CD006732. doi:10.1002/14651858.CD006732.pub3 [ PubMed : 20464744 ] [ CrossRef ] 11. Crego N, Douglas C, Bonnabeau E, et al. Sickle-cell disease co-management, health care utilization, and hydroxyurea use. J Am Board Fam Med. 2020;33(1):91-105. doi:10.3122/jabfm.2020.01.190143 [ PMC free article : PMC7942752 ] [ PubMed : 31907250 ] [ CrossRef ] 12. Brandow AM, Jirovec DL, Panepinto JA. Hydroxyurea in children with sickle cell disease: practice patterns and barriers to utilization. Am J Hematol. 2010;85(8):611-613. doi:10.1002/ajh.21749 [ PMC free article : PMC2911651 ] [ PubMed : 20658592 ] [ CrossRef ] 13. Siebens H, Tucker J, Leander K. Use of geriatrics at your fingertips, a pocket guide, to educate physiatrists in geriatric care. Arch Phys Med Rehabil. 2004;85(9):1552-1554. doi:10.1016/j.apmr.2003.09.015 [ PubMed : 15375834 ] [ CrossRef ] 14. Korn LM, Reichert S, Simon T, Halm EA. Improving physicians' knowledge of the costs of common medications and willingness to consider costs when prescribing. J Gen Intern Med. 2003;18(1):31-37. doi:10.1046/j.1525-1497.2003.20115.x [ PMC free article : PMC1494810 ] [ PubMed : 12534761 ] [ CrossRef ] 15. Tang C, Wang A, Yan J. Exploring motivations and resistances for implementing shared decision-making in clinical practice: a systematic review based on a structure–process–outcome model. Health Expect. 2022;25(4):1254-1268. doi:10.1111/hex.13541 [ PMC free article : PMC9327808 ] [ PubMed : 35662361 ] [ CrossRef ] 16. Légaré F, Adekpedjou R, Stacey D, et al. Interventions for increasing the use of shared decision making by healthcare professionals. Cochrane Database Syst Rev. 2018;7(7): CD006732. doi:10.1002/14651858.CD006732.pub4 [ PMC free article : PMC6513543 ] [ PubMed : 30025154 ] [ CrossRef ] 17. Schoenfeld EM, Goff SL, Elia TR, et al. Physician-identified barriers to and facilitators of shared decision-making in the emergency department: an exploratory analysis. Emerg Med J. 2019;36(6):346-354. doi:10.1136/emermed-2018-208242 [ PMC free article : PMC6865804 ] [ PubMed : 31097464 ] [ CrossRef ] 18. Hoang K, Halpern-Felsher B, Brooks M, Blankenburg R. Shared decision-making with parents of hospitalized children: a qualitative analysis of parents' and providers' perspectives. Hosp Pediatr. 2020;10(11):977-985. doi:10.1542/hpeds.2020-0075 [ PubMed : 33037030 ] [ CrossRef ] 19. Godin G, Bélanger-Gravel A, Eccles M, Grimshaw J. Healthcare professionals' intentions and behaviours: a systematic review of studies based on social cognitive theories. Implement Sci. 2008;3(1):36. doi:10.1186/1748-5908-3-36 [ PMC free article : PMC2507717 ] [ PubMed : 18631386 ] [ CrossRef ] 20. Peek ME, Wilson SC, Gorawara-Bhat R, Odoms-Young A, Quinn MT, Chin MH. Barriers and facilitators to shared decision-making among African-Americans with diabetes. J Gen Intern Med. 2009, 24:1135-1139. doi:10.1007/s11606-009-1047-0 [ PMC free article : PMC2762499 ] [ PubMed : 19578818 ] [ CrossRef ] 21. Hayes D, Edbrooke-Childs J, Town R, Wolpert M, Midgley N. Barriers and facilitators to shared decision making in child and youth mental health: clinician perspectives using the Theoretical Domains Framework. Eur Child Adolesc Psychiatry. 2019, 28:655-666. doi:10.1007/s00787-018-1230-0 [ PMC free article : PMC6514112 ] [ PubMed : 30229306 ] [ CrossRef ] 22. Lin C-Y, Renwick L, Lovell K. Patients' perspectives on shared decision making in secondary mental healthcare in Taiwan: a qualitative study. Patient Educ Couns. 2020;103(12):2565-2570. doi:10.1016/j.pec.2020.05.030 [ PubMed : 32487469 ] [ CrossRef ] 23. Scholl I, LaRussa A, Hahlweg P, Kobrin S, Elwyn G. Organizational-and system-level characteristics that influence implementation of shared decision-making and strategies to address them—a scoping review. Implement Sci. 2018;13(1):40. doi:10.1186/s13012-018-0731-z [ PMC free article : PMC5845212 ] [ PubMed : 29523167 ] [ CrossRef ] 24. Légaré F, Ratté S, Stacey Det al., Interventions for improving the adoption of shared decision making by healthcare professionals. Cochrane Database Syst Rev. 2010;12(5):CD006732. doi:10.1002/14651858.CD006732.pub2 [ PubMed : 20464744 ] [ CrossRef ] 25. Crosby LE, Walton A, Shook LM, et al. Development of a hydroxyurea decision aid for parents of children with sickle cell anemia. J Pediatr Hematol Oncol. 2019;41(1):56-63. doi:10.1097/MPH.0000000000001257 [ PMC free article : PMC7359001 ] [ PubMed : 30044352 ] [ CrossRef ] 26. Sox HC, Greenfield S. Comparative effectiveness research: a report from the Institute of Medicine. Ann Intern Med. 2009;151(3):203-205. doi:10.7326/0003-4819-151-3-200908040-00125 [ PubMed : 19567618 ] [ CrossRef ] 27. Elwyn G, Burstin H, Barry MJ, et al. A proposal for the development of national certification standards for patient decision aids in the US. Health Pol. 2018;122(7):703-706. doi:10.1016/j.healthpol.2018.04.010 [ PubMed : 29728288 ] [ CrossRef ] 28. Berry JG, Bloom S, Foley S, Palfrey JS. Health inequity in children and youth with chronic health conditions. Pediatrics. 2010, 126(Supplement_3):S111-S119. doi:10.1542/peds.2010-1466D [ PubMed : 21123473 ] [ CrossRef ] 29. Key KD, Furr-Holden D, Lewis EY, et al. The continuum of community engagement in research: a roadmap for understanding and assessing progress. Prog Community Health Partnersh. 2019;13(4):427-434. doi:10.1353/cpr.2019.0064 [ PubMed : 31866597 ] [ CrossRef ] 30. Lucero J, Wallerstein N, Duran B, et al. Development of a mixed methods investigation of process and outcomes of community-based participatory research. J Mix Methods Res. 2018;12(1):55-74. doi:10.1177/1558689816633309 [ PMC free article : PMC5722275 ] [ PubMed : 29230152 ] [ CrossRef ] 31. Williford DN, McTate EA, Hood AM, et al. Psychologists as leaders in equitable science: applications of anti-racism and community participatory strategies in a pediatric behavioral medicine clinical trial. Am Psychol. 2023;78(2):107-118. doi:10.1037/amp0001086 [ PMC free article : PMC10474572 ] [ PubMed : 37011163 ] [ CrossRef ] 32. Sacristán JA, Aguarón A, Avendaño-Solá Cet al., Patient involvement in clinical research: why, when, and how. Patient Prefer Adherence. 2016, 10:631-640. doi:10.2147/PPA.S104259 [ PMC free article : PMC4854260 ] [ PubMed : 27175063 ] [ CrossRef ] 33. Vaughn LM, Jacquez F. Participatory research methods—choice points in the research process. J Particip Res Methods. 2020;1(1):13244. doi:10.35844/001c.13244 [ CrossRef ] 34. Hood AM, Strong H, Nwankwo C, et al. Engaging caregivers and providers of children with sickle cell anemia in shared decision making for hydroxyurea: protocol for a multicenter randomized controlled trial. JMIR Res Protoc. 2021;10(5):e27650. doi:10.2196/27650 [ PMC free article : PMC8178738 ] [ PubMed : 34018965 ] [ CrossRef ] 35. Crosby LE, Real FJ, Cruse B, et al. An immersive virtual reality curriculum for pediatric providers on shared decision making for hydroxyurea. Blood. 2019, 134:3402. doi:10.1182/blood-2019-128661 [ PMC free article : PMC8943226 ] [ PubMed : 35319512 ] [ CrossRef ] 36. O'Connor A, Cranney A. User Manual. Ottawa Hospital Research Institute; 2002. 37. O'Connor A. User Manual—Decisional Conflict Scale. 1993 (updated 2010). Ottawa Hospital Research Institute; 2014. 38. Melbourne E, Sinclair K, Durand MA, Légaré F, Elwyn G. Developing a dyadic OPTION scale to measure perceptions of shared decision making. Patient Educ Couns. 2010;78(2):177-183. doi:10.1016/j.pec.2009.07.009 [ PubMed : 19647970 ] [ CrossRef ] 39. Squires J, Bricker D, Potter L. Revision of a parent-completed developmental screening tool: Ages and Stages Questionnaires. J Pediatr Psychol. 1997;22(3):313-328. doi:10.1093/jpepsy/22.3.313 [ PubMed : 9212550 ] [ CrossRef ] 40. Panepinto JA, Bonner M. Health-related quality of life in sickle cell disease: past, present, and future. Pediatr Blood Cancer. 2012;59(2):377-385. doi:10.1002/pbc.24176 [ PubMed : 22522407 ] [ CrossRef ] 41. O'Connor A, Cranney A. User Manual—Knowledge. Ottawa Hospital Research Institute; 2004. 42. Wills CE, Holmes-Rovner M. Preliminary validation of the Satisfaction With Decision scale with depressed primary care patients. Health Expect. 2003;6(2):149-159. doi:10.1046/j.1369-6513.2003.00220.x [ PMC free article : PMC5060180 ] [ PubMed : 12752743 ] [ CrossRef ] 43. Weech-Maldonado R, Carle A, Weidmer B, Hurtado M, Ngo-Metzger Q, Hays RD. The Consumer Assessment of Healthcare Providers and Systems (CAHPS) cultural competence (CC) item set. Med Care. 2012, 50(9 suppl 2):S22-31. doi:10.1097/MLR.0b013e318263134b [ PMC free article : PMC3748811 ] [ PubMed : 22895226 ] [ CrossRef ] 44. Osborn CY, Mayberry LS, Mulvaney SA, Hess R. Patient web portals to improve diabetes outcomes: a systematic review. Curr Diab Rep. 2010;10(6):422-435. doi:10.1007/s11892-010-0151-1 [ PMC free article : PMC3086814 ] [ PubMed : 20890688 ] [ CrossRef ] 45. Kazak AE, Alderfer M, Enlow PT, et al. COVID-19 Exposure and Family Impact Scales: factor structure and initial psychometrics. J Pediatr Psychol. 2021;46(5):504-513. doi:10.1093/jpepsy/jsab026 [ PMC free article : PMC8083683 ] [ PubMed : 33749794 ] [ CrossRef ] 46. Hussey MA, Hughes JP. Design and analysis of stepped wedge cluster randomized trials. Contemp Clin Trials. 2007;28(2):182-191. doi:10.1016/j.cct.2006.05.007 [ PubMed : 16829207 ] [ CrossRef ] 47. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Routledge; 2013. 48. Mara CA, Peugh JL. Validity of data collected from randomized behavioral clinical trials during the COVID-19 pandemic. J Pediatr Psychol. 2020;45(9):971-976. doi:10.1093/jpepsy/jsaa078 [ PMC free article : PMC7797740 ] [ PubMed : 32968774 ] [ CrossRef ] 49. Sun Q. Predicting Downstream Effects of High Decisional Conflict: Meta-Analyses of the Decisional Conflict Scale [thesis]. University of Ottawa; 2006. doi:10.20381/ruor-18514 [ CrossRef ] 50. Hemming K, Haines T, Chilton P, Girling A, Lilford R. The stepped wedge cluster randomised trial: rationale, design, analysis, and reporting. BMJ. 2015, 350:h391. doi:10.1136/bmj.h391 [ PubMed : 25662947 ] [ CrossRef ] 51. Gattellari M, Ward JE. Men's reactions to disclosed and undisclosed opportunistic PSA screening for prostate cancer. Med J Aust. 2005;182(8):386-389. doi:10.5694/j.1326-5377.2005.tb06756.x [ PubMed : 15850434 ] [ CrossRef ] 52. Légaré F, Moher D, Elwyn G, LeBlanc A, Gravel K. Instruments to assess the perception of physicians in the decision-making process of specific clinical encounters: a systematic review. BMC Med Inform Decis Mak. 2007;7(1):30. doi:10.1186/1472-6947-7-30 [ PMC free article : PMC2151936 ] [ PubMed : 17937801 ] [ CrossRef ] 53. Costa LL, Bobay K, Hughes R, et al. Using the Consolidated Framework for Implementation Research to evaluate clinical trials: an example from multisite nursing research. Nurs Outlook. 2020;68(6):769-783. doi:10.1016/j.outlook.2020.07.005 [ PubMed : 32859426 ] [ CrossRef ] 54. Strong H, Hood AM, Johnson Y, et al. Using the Consolidated Framework for Implementation Research to identify recruitment barriers and targeted strategies for a shared decision-making randomized clinical trial in pediatric sickle cell disease. Clin Trials. 20(3):211-222. doi:10.1177/17407745231154199 [ PMC free article : PMC10330034 ] [ PubMed : 36794731 ] [ CrossRef ] 55. Buchanan NT, Perez M, Prinstein MJ, Thurston IB. Upending racism in psychological science: strategies to change how science is conducted, reported, reviewed, and disseminated. Am Psychol. 2021, 76:1097-1112. doi:10.1037/amp0000905 [ PubMed : 34990171 ] [ CrossRef ] 56. Herbst R, Corley AMS, McTate E. Clinical framework for dismantling antiblack racism in the clinic room. Clin Pediatr. 2023;62(10):1129-1136. doi:10.1177/00099228231156009 [ PMC free article : PMC10478307 ] [ PubMed : 36852825 ] [ CrossRef ] 57. Ben-Zacharia A, Adamson M, Boyd A, et al. Impact of shared decision making on disease-modifying drug adherence in multiple sclerosis. Int J MS Care. 2018;20(6):287-297. doi:10.7224/1537-2073.2017-070 [ PMC free article : PMC6295876 ] [ PubMed : 30568566 ] [ CrossRef ] Related Publications •. Hood AM, Strong H, Nwankwo C, et al. Engaging caregivers and providers of children with sickle cell anemia in shared decision making for hydroxyurea: protocol for a multicenter randomized controlled trial. JMIR Res Protoc. 2021;10(5):e27650. doi:10.2196/27650 [ PMC free article : PMC8178738 ] [ PubMed : 34018965 ] [ CrossRef ] •. Crosby LE, Real FJ, Cruse B, et al. An immersive virtual reality curriculum for pediatric providers on shared decision making for hydroxyurea. Blood. 2019, 134:3402. doi:10.1182/blood-2019-128661 [ CrossRef ] •. Crosby LE, Walton A, Shook LM, et al. Development of a hydroxyurea decision aid for parents of children with sickle cell anemia. J Pediatr Hematol Oncol. 2019;41(1):56-63. doi:10.1097/MPH.0000000000001257 [ PMC free article : PMC7359001 ] [ PubMed : 30044352 ] [ CrossRef ] •. Hood AM, Johnson Y, Nwankwo C, et al. Clinical practice patterns for hydroxyurea initiation in young children with sickle cell disease. Blood. 2019, 134:4713. doi:10.1182/blood-2019-124835 [ CrossRef ] •. Williford DN, McTate EA, Hood AM, et al. Psychologists as leaders in equitable science: applications of anti-racism and community participatory strategies in a pediatric behavioral medicine clinical trial. Am Psychol. 2023;78(2):107-118. doi:10.1037/amp0001086 [ PMC free article : PMC10474572 ] [ PubMed : 37011163 ] [ CrossRef ] •. Strong H, Hood AM, Johnson Y, et al. Using the Consolidated Framework for Implementation Research to identify recruitment barriers and targeted strategies for a shared decision-making randomized clinical trial in pediatric sickle cell disease. Clin Trials. 2023;20(3):211-222. doi:10.1177/17407745231154199 [ PMC free article : PMC10330034 ] [ PubMed : 36794731 ] [ CrossRef ] Acknowledgments The research team acknowledges others who contributed to this study, including the stakeholder advisory council; medical and graduate students, particularly Joanna Rebitski and Emmanuel Gray; and clinical research coordinators and data analysts Yolanda Johnson, Jennifer Allen, Catharine Whitacre, and Stacey Gomes. Research reported in this report was funded through a Patient-Centered Outcomes Research Institute® (PCORI®) Award (CDR-1609-36055). Further information available at: https://www.pcori.org/research-results/2017/comparing-ways-help-parents-children-sickle-cell-disease-decide-treatment-engage-hu-study Original Project Title: Engaging parents of children with sickle cell anemia and their providers in shared decision making for hydroxyurea PCORI ID: CDR-1609-36055 ClinicalTrials.gov ID: NCT03442114 Suggested citation: Crosby LE, Mara CA, Johnson Y, Hackworth R, Quinn C. (2024). Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study . Patient-Centered Outcomes Research Institute (PCORI). https://doi.org/10.25302/02.2024.CDR.160936055 Disclaimer The [views, statements, opinions] presented in this report are solely the responsibility of the author(s) and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute® (PCORI®), its Board of Governors or Methodology Committee. Copyright © 2024. Cincinnati Children's Hospital Medical Center. All Rights Reserved. This book is distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License which permits noncommercial use and distribution provided the original author(s) and source are credited. (See https://creativecommons.org/licenses/by-nc-nd/4.0/ Bookshelf ID: NBK621485 PMID: 41915777 DOI: 10.25302/02.2024.CDR.160936055 Share Views PubReader Print View Cite this Page Crosby LE, Mara CA, Johnson Y, et al. Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study [Internet]. Washington (DC): Patient-Centered Outcomes Research Institute (PCORI); 2024 Feb. doi: 10.25302/02.2024.CDR.160936055 PDF version of this title (966K) In this Page Background Participation of Patients and Other Stakeholders Methods Results Discussion Conclusions References Related Publications Acknowledgments Other titles in this collection PCORI Final Research Reports Related information NLM Catalog Related NLM Catalog Entries PMC PubMed Central citations PubMed Links to PubMed Recent Activity Clear Turn Off Turn On Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Tr... Comparing Ways to Help Parents of Children with Sickle Cell Disease Decide on Treatment—The ENGAGE HU Study Your browsing activity is empty. Activity recording is turned off. Turn recording back on See more... Follow NCBI Twitter Facebook LinkedIn GitHub NCBI Insights Blog Connect with NLM Twitter Facebook Youtube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov