The Impact of Community and Patient Engagement Practices on Vaccine Confidence in the United States

Overview

The primary objective of this study is to demonstrate an empirical relationship between community- engagement practices, and between patient-centric clinical trial design, execution and communication practices, and vaccine adoption experience. Our hypothesis is that patient-centric clinical trial activity, and community engagement in late-stage clinical trials and early-stage commercialization, reduces vaccine hesitancy and increases vaccine confidence among health care providers overall and within diverse patient communities and ultimately drives faster vaccine adoption.

We will use a mixed methods approach: Quantitative Assessment The purpose of the quantitative assessment is to map diffusion curves and to determine an empirical relationship between community-engagement activity, patient-centric clinical trial design, execution and communication and vaccine adoption experience. The research team will build and analyze a dataset of relevant vaccines - those developed for infectious diseases impacting the general adult population -- approved by the Food and Drug Administration (FDA) since 2005. We will next map the diffusion of each of these vaccines at the national and regional level and identify 5 outliers that experienced significantly faster adoption. Data on the pre-approval pivotal trials supporting these vaccines will be compiled from a proprietary database developed by the Tufts Center for the Study of Drug Development (Tufts CSDD). In 2020, Tufts CSDD compiled data on pivotal trials of all new drugs and biologics approved since 2007. Data was drawn and compiled from a number of sources in the public domain including the FDA, clinicaltrials.gov, and published articles in the literature. Detailed information about each pivotal trial supporting an approved drug or biologic - including drug information, pivotal trial scope and patient demographic data - was gathered. Since 2020, Tufts CSDD has continued to maintain and update this dataset. For this study, we will supplement the dataset with additional data on post-approval research activity and on patient-centric practices supporting pre- and post-approval clinical trials. Data will be gathered on national and regional community collaborations and partnerships, and on public and patient involvement and partnership. This data can be manually gathered from listings on clinicaltrials.gov and from articles in peer-reviewed literature and the trade press. Next, using a commercially available database, Tufts CSDD will draw data (organized by NCT#) to map the detailed diffusion rates of each of the 39 vaccines. Data will include launch date, monthly units dispensed, demographics of patients who received the vaccine including gender, race and ethnicity. All patient data provided is anonymized and not patient will be contacted by the study team. Using geocodes, we will map national and regional diffusion rates. Independent variables for the 39 vaccines include NCT#, trade and generic name, approval date, launch date, race and ethnicity of patients enrolled, number and location of investigative sites, number of participants, patient advocacy group involvement, community-engagement practices, patient and site input into protocol design, and clinical trial execution strategies supporting patient participation access and convenience. Dependent variables include monthly units dispensed nationally and regionally, mean month-over- month growth rate of units dispensed, demographics of patients who received the vaccine including gender, race, and ethnicity nationally and regionally, and cycle times associated with launch, uptake, and peak levels of diffusion. We will gather univariate descriptive statistics including mean, median, standard deviation and coefficients of variation. Bivariate analysis will also be conducted to understand the relationship between vaccine diffusion rates and the independent variables, including patient-centric approaches used, location, other pivotal trial characteristics and patient participant demographics. Hypothesis testing, specifically t-tests and analysis of variance where appropriate, will test for statistically significant differences between independent variable subgroups by vaccine. Regression modeling will be used to determine whether independent variables are predictive of vaccine diffusion rates overall, by region and demographic community subgroup. We will assess binary variables to understand interactions between community engagement and patient-centric approaches and to account for common combinations of approaches used. The demographic makeup of the pivotal trial for the vaccine will be used to determine if it is predictive of patient demographic uptake (for example, whether the proportion of Asian participants in the pivotal trial influences vaccine uptake in Asian communities). Correlation analysis between vaccine diffusion rates and cycle times will also be conducted. All analyses will be completed in R (statistical programming software). Choropleth mappings will be built to display national and regional diffusion rates and location of pivotal trials. Analysis of density of diffusion rates and proximity to pivotal trial site locations will be conducted. All mapping and related analysis will be completed in ArcGIS (mapping and analytics software). The five vaccines that experienced the fastest national adoption rates will be the focus of the Qualitative Assessment. Qualitative Assessment Qualitative research methods provide rich, contextualized data about complex social and organizational phenomena, such as vaccine diffusion, that are difficult or impossible to obtain through traditional quantitative methods. We will use qualitative, individual interviews to gather information about factors perceived to have aided in successful dissemination of each of the 5 vaccines identified during the quantitative phase of this research. We elected to utilize individual interviews, as opposed to other qualitative data gathering techniques (e.g., focus groups) because they are less expensive, can be conducted faster, and permit us to probe more deeply among a diverse group of individuals. We will interview healthcare providers, subject matter experts, and community representatives/ influencers (e.g., patient and disease advocates; clergy, community leaders) to gather critical insights into vaccine diffusion rates. We plan to conduct 10 interviews in total, among each of these three groups - two interviews per group for each of the five vaccines for a total of 30 interviews. We will continue until the point of data saturation. Interview guides will be based on findings from existing literature, investigator expertise, and data gathered during the quantitative phase of research. Interview guides will be finalized based on the quantitative analysis. Interview questions will be provided in advance to help interviewees prepare. Interviewees will be identified from the literature and through referral by subject matter experts. Some community representatives/influencers will be individuals who have been profiled and outspoken in the media about the importance of community access to health care. All interviewees will sign an informed consent form. Interviews, which are expected to take approximately 30-60 minutes, will be conducted virtually on Zoom by trained research staff. Interviews will be recorded for professional (verbatim) transcription and interview texts will subsequently be analyzed. We will analyze qualitative data using standard methods of thematic content analysis which we have used in many prior studies. Thematic analysis includes a hybrid of inductive and deductive approaches. Before analysis commences, interviewers will review transcribed interviews for accuracy. Next, members of the interview team will independently review each transcript and identify initial themes. In a series of meetings, team members will compare independently identified themes and through an iterative group process of consensus, super-ordinate and subordinate categories. Following discussion and consensus regarding the superordinate themes, team members will next independently conduct line-by-line coding by compiling themes and descriptive quotes into Excel spreadsheets. These documents will be reviewed and compared. If disagreement regarding the meaning of a specific quote arises, team members will review the interview recording and transcript and come to consensus regarding meaning. To further understand the development strategies of pharmaceutical companies sponsoring the five fastest vaccines adopted, we will also speak with 1 to 2 representatives from the clinical and medical affairs teams associated with each of the five vaccines to fill in any gaps in the vaccine diffusion dataset and to gather additional insights explaining the outlier vaccines with the fastest adoption rates. These interviews will be conducted virtually and we anticipate that each will be 45- to 60-minutes in duration. Tufts CSDD has extensive experience engaging with clinical research professionals in empirical, pre-competitive studies of this nature.

Conditions

Interventions

Eligibility

Outcomes