Application for 2012 University of California Larry L. Sautter Award for

Innovation in Information Technology

Date:May 15, 2012

Project Title:Research Volunteer Registry

University of California Davis Health System

Submitted:Kent Anderson, M.S., Research Technology Manager; Associate Director, CTSC Biomedical Informatics

UC Davis Health System IT, Co-Director, Biomedical Informatics

(916) 703-9123

Dan Cotton, Educational Technology& Application Development Manager

UC Davis Health System IT

(916) 734-2019

Michael Minear

Chief Information Officer; Director, CTSC Biomedical Informatics

(916) 734-7131

I. Project Leaders and Team Members

UC Davis strongly promotes a team science approach as a methodology for translational work. One of the key elements leading to the success of this project was the consistent partnering of individual experts throughout UC Davis Health System and other institutions to achieve team goals.

Project Leadership

  • Daniel Cotton, Educational Technology & Application Development Manager
  • Kent Anderson, M.S., Research Technology Manager; Associate Director, CTSC Biomedical Informatics

Information Technology

  • Peter Ruan, Lead Developer
  • Ann Mankser, Analyst
  • Alesia Minyard, Analyst

Clinical Translation Science Center (CTSC)

  • Ayan Patel, Application Programmer, CTSC Biomedical Informatics
  • Sam Morley, M.S., Biomedical Informatics Program Manager

M.I.N.D. Institute

  • Lou Ann Barnett, Ph.D., PMP, Autism Phenome Project Manager
  • Brenda Shelton, Senior Community Health Representative

Executive Sponsors

  • Michael Minear, Chief Information Officer, UC Davis Health System; Director, CTSC Biomedical Informatics
  • Robert Hales, M.D., Joe P. Tupin Endowed Chair, Professor of Clinical Psychiatry and Chair
  • Narriman Shahrokh, MIND Chief Administrative Officer

II. Summary

A Research Volunteer Registry has been created by UC Davis Health System in order to support research subject recruitment across the Health System for all types of research. This application has allowed the Health System to consolidate multiple registry databases from different departments into a single,more compliant, feature rich, and efficient tool for both community volunteers and research staff alike. The application integrates with the Research Electronic Data Capture (REDCap) software system so that once the study participants have confirmed their participation, they can be seamlessly imported into REDCap where a study profile has been defined and study data collection can begin. This new system reduces liability by ensuring access to volunteer data is restricted until the proper consent is received.

III. Project Description

Background

The UC Davis Health System conducts more than 1,000 studies each year in areas of medical, clinical, and behavioral intervention and observation involving human volunteers.Multiple databases following varying recruitment and security methodologies had been used to track research volunteers. In the past, Excel, Access, and custom built software were common applications used to manage data for tracking research volunteers. These legacy systems had propagated throughout the research enterprise, making security and privacy compliance difficult to manage.

The MIND Institute (Medical Investigation of Neurodevelopmental Disorders) was managing volunteer registry data in a legacy fashion within unsupported application technology lacking desired functionality and improvements. UCDHS Information Technology (IT)partnered with the MIND research staff and faculty to build a generic set of software tools and services to incorporate the requested enhancements, while modernizing the Health System’s approach to volunteer management.The success of the RVR is due to the collaboration between IT, MIND executives, faculty and staff.

The Research Volunteer Registry created a starting point for UC Davis Health System researchers to connect with community members interested in volunteering for research. IT was able to address and improve several problematic areas of research subject recruitment experienced in the past. Three specific areas of improvement were focused on:

  1. Research volunteers become “owners” of their information – In legacy systems tracking community research volunteers, information was collected once, and rarely updated with any regularity. In each system volunteer information was static, and rapidly became obsolete as volunteers moved, changed contact information, or in the case of pediatric subjects, grew older and became ineligible for pediatric studies. With the Volunteer Registry, UC Davis Health System IT shifted the data management paradigm to allow research volunteers to directly and easily manage their own information. The system sends reminder emails annually to encourage volunteers to update and validate their information and to affirm interest in research participation. The volunteer is empowered to be selective about their research participation, and can anonymously decline to participate in any study without obligation.
  2. Bridging and protecting the gap between volunteer and research recruitment – This valuable repository of research volunteers with their demographic and associated disease information and study preferences requires an interface to navigate and select cohorts for follow up. In legacy registries, the information obtained about research subjects was often poorly protected from data loss or theft of personal information, or multiple researchers could have open access to the entire data set. The UC Davis Research Volunteer Registry was designed to comply with standard health data security practices (Health Insurance Portability and Accountability Act, HIPAA) to ensure the privacy of volunteer data, including “blinding” investigators to personally identifiable information about the volunteers until explicit consent is obtained. This added layer of sophisticated protection automates even the “honest broker” approach commonly used to identify research cohorts.
  3. Integration with REDCap study data management software – Only once a volunteer has given explicit consent to participate in a given study is their personal information made available to the researcher or their staff. As the researcher receives responses to their recruitment efforts they can click a button to bring the volunteer data over into the REDCap study data management system. This not only saves the research staff a lot of time rekeying data, but this greatly limits the opportunity for a breach of information or loss of sensitive data. The direct integration of the Volunteer Registry to the REDCap software further enhances security, and encourages best practices for study data management. This actually simplifies and accelerates the researcher’s experience in study subject recruitment, which is often the greatest limitation to study progress and research success

Research Volunteer Registry Overview

The Research Volunteer Registry is one application but has two distinctly different user interfaces and authentication/authorization mechanisms. One component of this tool includes functionality designed for volunteer users that will be referred to as the Volunteer Registry (VR). The second module includes an interface for research users that will be referred to as the Volunteer Registry Search (VRS). A third module was built to integrate the registry with REDCap.

Figure 1- Depicts the user interaction with each of the modules in the Research Volunteer Registry.

The innovation of the Volunteer Registry, Volunteer Registry Search, and REDCap integration effort is in streamlining the entire sequence of the events (Figure 1).

  1. Creating a starting point for community members interested in volunteering for research.
  2. Improving accuracy of research volunteer cohort selection to participation in a research study and while protecting volunteer confidentiality
  3. Automating the handoff from cohort selection to research subject tracking, improving data quality and reducing level of effort to initiate a study

Volunteer Registry (VR)

The VR has a public user interface where unauthenticated users can access information on the Registry and such as an overview, privacy details, and information on how to register(Appendix A - Figure 5). Volunteers can use the VR to then register to receive information on studies that they may want to participate in. Joining the registry is easy, private, and leaves the decision to participate in any study in the volunteer’s hands.

A volunteer can create an account to log in and out of the VR. Once an account has been created,account holders can create one or more research profiles. The research profiles are created for the individual as well as family members, and can be completed in three easy steps.

The first step in completing a registry profile is to enter demographic data on the research volunteer. If the volunteer is not the user, the system prompts the user to answer a set of questions on the relationship between himself/herself and the family member.

The second step is to enter any medical conditions that the volunteer has been diagnosed with (Appendix A -Figure 6). It is important for researchers to get accurate medical conditions collected in order to filter out volunteers who are not a match for their study. The interface was designed so that as a user enters the first few letters of the medical condition in a text box, a list of suggested conditions appear that can be selected. The more letters the user types the better the suggestions become.

The list of medical conditions has been imported into the VR database and is coded with Systematized Nomenclature of Medicine- Clinical Terms (SNOMED-CT). If a user selects a condition that is in the SNOMED-CT list the medical condition is coded in the database automatically and can be easily queried by researchers. Volunteers who are unable to find their condition in the list using the terms that they are familiar with can enter the term they use for the condition in the text field and save it.

An administrator from the CTSC reviews a dashboard report from the VR showing the number of medical conditions that were not assigned a SNOMED-CT code. They can log into the system and try to reconcile the data entered with the correct code for the condition which will not overwrite the data that the user entered to describe their condition.

As administrators reconcile inaccurate medical conditions entered by the volunteers with correct SNOMED-CT codes, heuristics built into the system will suggest the correct medical conditions to volunteers based on common data entry errors.

The third step in creating a registry profile is selecting research studies of interest to the subject. The list of studies is populated from a CTSC database of active studies.

Once a researcher determines that a research profile matches his/her study criteria, they will be able to use the system to send the volunteer an email with details about the study and a request to participate. In the body of the email there are two links that can be clicked indicating their preference to participate. If the user clicks that they are interested in a study they are taken to the VR to further confirm that a researcher may contact them (Appendix A -Figure 7).

A research volunteer may keep their data up to date by logging into their account at any time, allowing the user to become the “owner” of their information. They can also set their profile as inactive which will prohibit researchers from contacting them about future studies.

Volunteer Registry Search (VRS)

The second module of the Research Volunteer Registry is the search functionality built for research users. This module includes functionality to search the VR for the number of potential volunteers that meet the specific study criteria, to invite volunteers to participate, and to transfer the data into their study database.

A researcher begins their work in the VRS by entering an Institutional Review Board (IRB) approved email into the system. The email template can be edited later as long as no email has been sent. The content the researcher enters in a template is sandwiched between boiler plate content that will be sent with all emails from the VR to any research volunteers. The template editor is very simple and does not allow for the researcher to format the data or include live links. The reason for this design is to avoid the emails getting caught in the recipient’s spam filters.

During the analysis stage of implementation, the Health System reviewed the UC Davis Cohort Discovery Tool[1] (powered by i2b2) to determine the optimal search mechanism for research subject populations. It was determined that the search functionality created for researchers should be similar to the widely used Cohort Discovery application in order to lower the learning curve and foster early adoption of the new Volunteer Registry Search software.

Before any email is sent using the template, the research staff must query for volunteers to include in their recruitment effort. Queries can be built using three types of data; 1) demographic data fields, 2) medical conditions (by terms entered by volunteer user or by SNOMED), and 3) research studies that the volunteers have identified interest in. Complex queries can be built in the user interface designed similarly to the Cohort Discovery query functionality (i2B2) (Appendix B -Figure 8). The criteria in each of the groups are combined together with ORs and the groups are combined together as the screen shows with ANDs. More groups can be added allowing for large and complicated searches.

When a researcher has run his/her query and a required number of possible participants were returned,they can start Recruitment. In this step the user selects the name of a study and the email template they want to use and the system shows the user a preview of the email. If they are satisfied with the preview they can send the email and await the responses.

Using the Study Screening Summary section of the application researcher users can review the responses and the name of volunteers who have consented to being contacted for this study. By clicking the volunteer’s name a research user can view the details of the volunteer and can contact them as well as indicate if they are qualified or not after being contacted.

Now the cohorts are ready to be moved into REDCap. Users can click a button to send Cohorts to REDCap (Appendix B - Figure 9). As more volunteers respond that they would like to participate, these last few steps can be repeated.

REDCap Integration with Volunteer Registry Data

REDCap is a modernsoftware application for building and managing online clinical research databases. A growing number of UC Davis researchers and their collaborators are using REDCap to securely manage research data. The application allows users to build and manage online surveys and databases quickly and securely, and to link data collected with existing statistical software tools. REDCap is a valuable tool that is expected to replace all Microsoft Excel spreadsheets and Access databases that researchers frequently use to collect study data. The application allows users to access and manage the data from anywhere in the world over a secure web connection with authentication and data logging. REDCap is fully customizable with advanced features that allow researchers to shape their database or survey, and incorporate mid-study modifications, auto-validation, calculated fields and branching logic.

Developed by an informatics core at Vanderbilt University with ongoing support from the National Center for Research Resources and National Institutes of Health grants, REDCap was designed to address the high cost of commercial research software and the limited informatics and information-technology support for research at universities and other organizations -- common problems for academic biomedical researchers hoping to use electronic databases. The software is supported by a consortium of 300+ active academic health institutions worldwide and is currently being used in more than 36,000 studies with over 50,000 end-users across the consortium.

REDCap supports:

  • Rapid creation and design of projects by constructing a “data dictionary” template file in Microsoft Excel, that is then uploaded into REDCap for execution
  • Audit trails for tracking data manipulation and user activity
  • Automated export procedures for seamless data downloads to Excel, PDF, and common statistical packages (Appendix C - Figure 10)
  • Built-in project calendar and scheduling capabilities
  • Ad hoc reporting tools
  • Advanced features, such as branching logic, file uploading, and calculated fields

UC Davis Health System launched REDCap for general use in October 2011. The application has been warmly received by the research community and adopted for a growing number of studies (Figure 2), by a growing number of researchers and their staff (Figure 3).


Figure 2– UC Davis Research studies created in REDCap in the past six months /
Figure 3– UC Davis REDCap Active Users

Project Benefits

There are many benefits to this new application for volunteers, research staff, and the Health System as well. Volunteers can create one account and if their data changes they can easily update their information. This application provides an account holder the ability to create research profiles for themselves as well as family members. Volunteers are able to enter information that a researcher may be looking for such as height and weight, but are also able to indicate the research studies they are interested in and can see what types of studies are currently being done. The user interface is intuitive and makes it easy for a volunteer to respond to participation requests. At any time a volunteer canexclude themselves from receiving new invitations. There is an extensive user guide and FAQ section to assist users in using the registry.