Engaging Patients in Prenatal Genetic Testing Decisions as a Pathway to Improve Obstetric Outcomes

Overview

The goal of this study is to ensure that pregnant patients have the resources and support needed to access Prenatal Screening \& Diagnostic Testing (PS\&D) in an informed and evidence-based fashion by developing an innovative digital tool to support patients' decision-making and contributing fundamental knowledge to advance science in a way that promotes patients' access to new prenatal applications of genomic science and technology. Our central hypothesis is that, by focusing on patient engagement as a key driver to improve patient outcomes, the use of an evidence-based artificial-intelligence (AI) powered patient engagement tool will increase patients' ability to seek information and structure a decision-making process that, in turn, increases informed decisions about PS\&D and decreases decisional conflict associated with those decisions. Using data from NEST (Ensuring Patients Informed Access to NIPT \[non-invasive prenatal testing\]), the investigators designed the next iteration of NEST, a point-of care shared decision-making tool powered by artificial intelligence (AI) to provide a personalized and dynamic decision support tool: Obstetric Prenatal Genetic Testing Engagement Solution (OPUS). OPUS is an AI-enabled healthcare chatbot (a computer program capable of processing and simulating human conversation) that provides patients with personalized information and decision-making support at different stages of the PS\&D pathway. It functions using a series of questions contained in the NEST with a branching logic sequence of questions and answers based on the responses to and from the patient, using a conversational and adaptable interaction. It also contains nested tiers of information, ranging from introductory to detailed information about patient engagement, health literacy, the different PS\&D options, and resources to learn about insurance coverage for PS\&D. OPUS was designed to be accessed by patients with different technological resources and preferences, using a cell phone, a mobile device, or a computer.

Study overview: The investigators will conduct a randomized cluster trial to examine the impact of the Obstetric Prenatal Genetic Testing Engagement Solution (OPUS) on the decision-making process about Prenatal Screening \& Diagnostic Testing (PS\&D) from the perspectives of patients, partners, and providers. In doing so, the investigators will examine important changes that have taken place in genetic technology and healthcare. In-depth, semi-structured interviews will be conducted to gain greater insight into the decision-making process and strategies throughout pregnancy to gain a longitudinal understanding of how decision-making unfolds with the utilization of tests and gestational age. The investigators will employ a method that allows for the study of decision-making across diverse patient and provider populations, including urban and rural communities and academic and community practices, to understand how these factors in addition to social determinants of health influence patients' informed access to PS\&D. The team will collaborate the Advisory Group and in consultation with the Data Safety monitoring Board (DSMB) assembled for this study, to ensure scientific rigor in the experimental design and conduct of the study, analysis and interpretation of the data, and dissemination of the final results and version of the tool. Methods: Study population and recruitment: Participants will include three groups: (1) Obstetric (OB) providers, (2) patients, and (3) patient partners. Subject recruitment will take place at the three study sites: (1) the Cleveland Clinic Health System, (2) the MetroHealth Medical System, and (3) the Neighborhood Family Practice Community Health Centers. Randomization: Providers will be randomized by practice into: (1) the intervention arm utilizing OPUS or (2) the control arm utilizing standard practice patterns with respect to education and counseling (usual care). The rationale for randomization by practice is to prevent contamination between study arms, as a patient may see more than one provider within a practice over the course of her prenatal care. Comparators: Intervention: OPUS is a chatbot that will provide an artificial intelligence (AI) powered conversational and personalized interaction to assist patients in seeking information about their PS\&D options. It was designed by Cleveland Clinic Information Technology (IT) and Power Objects/Microsoft with the goal of meeting patients where they are in their decision-making regarding the nature and detail of information and decision-support needed to navigate PS\&D decisions. OPUS can be accessed from a cell phone, mobile device, or computer using a web-based or text-based approach. It was also developed based on existing studies demonstrating the benefits of healthcare chatbots. The investigators will implement mechanisms that minimize barriers to access OPUS: making tablets available on-site for patients with limited technology and providing broadband Intent as needed to access OPUS (intervention) or other educational resources (intervention and control groups). OPUS will utilize the Microsoft Azure Health Bot and Dynamics 365 technology and build on the existing Cleveland Clinic health IT solutions to support patient engagement in other clinical contexts that meet requirements of Health Information Portability and Accountability Act (HIPAA) and encryption for protected health information (PHI) management. Participants will be exposed to OPUS at three key time points as determined by our NEST R01. Time Point 1 (T1) - Prior to the initial prenatal visit: This represents the time period before patients initiate discussions with their providers about genetic risk and PS\&D. It may include patients at 7-21 weeks estimated gestational age (EGA). Time Point 2 (T2) - Prior to the initial PS\&D decision: This time point represents the patients' initial step in the PS\&D decision-making pathway. It may include patients at 9-23 weeks EGA. Time Point 3 (T3) - Prior to the 2nd trimester anatomical ultrasound (US): This is a time point with different significance for patients based on initial PS\&D choices. This includes patients who present for (a) an initial assessment after declining all PS\&D prior to the 2nd trimester ultrasound, (b) follow-up assessment after receiving an abnormal result on a prior screening, or (c) subsequent routine assessment after receiving a normal result on a prior screen or diagnostic test. It may include patients at 18-23 weeks EGA. Participants may opt-out of engaging with OPUS at any time point in the study. Control: The control group will receive standard counseling per usual care. They will proceed with the appointment with the scheduled provider, who will discuss genetic risk and assessment per his or her usual counseling and practice patterns. Aim 1. Determine the ability of OPUS to structure a decision-making process about PS\&D that aligns with patient needs and preferences compared to usual care. The investigators hypothesize that patients exposed to OPUS will have higher levels of engagement that will enable them to structure a shared or informed decision-making process that aligns with their needs and preferences. Patient Measures: Information about participants' demographics, reproductive, and medical history will be collected from the electronic medical records (EMR). The investigators will then collect data using administration of a series of surveys, and semi-structured interviews to assess the impact of OPUS at the key time points in the decision-making pathway. All instruments will be distributed and completed via Research Electronic Data Capture (REDCap) Survey, with a paper version available at participant request. The investigators will also make tablets available at the clinical visit for participants to complete the survey. Patient Activation: The investigators will utilize a modified version of the Patient Activation Measure (PAM)-13 to measure patient activation regarding PS\&D decisions. The PAM-13 is a widely used, standardized measure of patient activation in healthcare decisions. It is an interval-level scale developed to assess respondents' understanding of their role in the care process and having the knowledge, skill, and confidence to manage healthcare decisions. Decision preferences status/preferences: The investigators will measure participants' preferences regarding decision-making, experiences with shared decision making (SDM). The investigators will measure decision-making decisions/preferences at the following timepoints: T1 Pre-visit, and T1 Post-visit. This will include questions about printed or online materials accessed for self-education. Decision-making: The research assistant will place a small audio-recording device in the exam room to record the content, interaction, and duration of dialogue about genetic risk and risk assessment and remove it immediately after the visit. This approach will allow for evaluation of patient and patient/partner dyads in healthcare discussions with the provider. The investigators have elected to audio record instead of video record to minimize the Hawthorne effect and because of the sensitive nature of the physical exam during this visit. Our prior work has demonstrated that this approach can be successfully used. Interviews: The investigators will conduct semi-structured interviews with a subset of patients at T1 Post-visit, T2 Post-test, (n = 40) to gain additional information about OPUS's usability, feasibility, and acceptability of the intervention. The interviews will be guided by an interview guide developed in conjunction with the Advisory Group and audio recorded for data analysis purposes. Provider measures: Providers will complete a baseline questionnaire (OB1) at the time of enrollment to collect information about demographics, medical education, and training (including any training specific to healthcare communication), and clinical practice. It will also contain a self-assessment of the providers' ability to facilitate a SDM process. Aim 2. Determine the ability of OPUS to improve a patient's ability to make an informed choice about PS\&D compared to usual care. The investigators hypothesize that patients exposed to OPUS will have higher rates of informed choice regarding the utilization or decline of prenatal genetic testing. Data collection: Patients will complete the Informed Choice for Screens and Diagnostic Tests survey (ICSD), an instrument developed and validated as part of NEST at the following intervals: T1-Pre-visit, T1 Post-visit, T2 Post-test, and T3 Pre-ultrasound (US). The investigators will also assess health literacy using the validated self-administered medical term recognition test (METER), an instrument successfully used during NEST. Data analysis: Analysis will be conducted to determine the impact of OPUS on the variables of knowledge, attitudes, and deliberation; the variables will be analyzed together as components of the informed choice and then separately to understand the effect of exposure to OPUS on each variable. The investigators will also examine the change of these variables together and separately from T1 to T2 and T3. Based on the methods described by the author Marteau and adapted for the ICSD, the variables of knowledge and attitudes, and choice will be categorized. These data will be summed to determine whether responses at different time points represented an informed choice (good knowledge, positive attitude, and intention or decision to utilize consistent with behavior) or an uninformed choice (poor knowledge, negative attitude and attitude was inconsistent with behavior) for the utilized screen or diagnostic test. Logistic regression methods will be used to assess differences in exposure to OPUS and reproductive, demographic (e.g., race/ethnicity, technology access, social economic status \[SES\]), and clinical practice variables. Aim 3. Determine the extent to which OPUS reduces patients' uncertainty and decisional conflict associated with PS\&D decisions. The investigators hypothesize that patients exposed to OPUS will have lower levels of decisional uncertainty and conflict throughout the PS\&D pathway compared to patients exposed to usual care. Data collection: The investigators will use the validated Decisional Conflict Scale (DCS) to examine the presence of decisional conflict, understand its impact on patient experience, and identify modifiable factors to reduce it at T2, T3 and T4. This validated instrument was used successfully as part of NEST that uses Likert scale items to measure personal perceptions of uncertainty in the decision-making process and modifiable factors contributing to uncertainty. Our goal in selecting this measure is to assess the process of obtaining information and making a choice, not the outcome of the decision to utilize (or not utilize) PS\&D. Data analysis: Analysis will be conducted to determine the impact of OPUS on decisional conflict independently and as a function of informed choice (including variables of knowledge, attitudes, and deliberation separately and together), patient activation, and SDM/IDM. Analysis will be conducted in a fashion described by the author O'Connor to determine the level of decisional conflict. Logistic regression will be used to assess correlations between informed choice measures, patient demographic measures, and measures of decisional conflict. The data for each time point will be combined, and the analysis re-run using repeated measures logistic regression methods and including test time as an additional independent variable. Analyses will be conducted to examine differences in reduction of decisional conflict in addition to baseline uncertainty and anxiety across subgroups, including by race/ethnicity, knowledge/health literacy, insurance status, reproductive history.