Apple ResearchKit
Apple ResearchKit was release in April 2015 to make it easier for researchers and developers to create mobile apps for medical research as an open-source framework. Through our integration, all functions of ResearchKit are called through configurations in REDCap or the Investigator Dashboard, dispensing with the need for computer programming resources and the development of custom code to implement research protocols as a mobile app. Below are brief descriptions of the three main ResearchKit components.
a) Surveys
Surveys are the primary mechanism for interacting with the user in the app. The platform downloads the study definition from REDCap using the provided API and renders the Data Collection Instruments as ResearchKit Surveys. Our integration maps most REDCap functions to the appropriate ResearchKit object to provide a user-interface specifically designed for iOS devices, enabling efficient and valid data input by users. Branching logic, data types, data validation, and calculated fields are REDCap features implemented in our platform. The platform also provides additional functions related to surveys that are not part of the REDCap feature set and are called by providing commands in REDCap field annotations. The commands are passed into the app using the REDCap API and issued at a specific date defined by REDCap Events, and specific times defined in the Investigator Dashboard. An example field-annotation command that directs survey flow to a specific field follows:
statusPost.survey.goto[“variable_name”:”destination_field”]
The following command instructs Siri to read the field title, text, question, and responses to the user:
statusPost.survey.speak_interaction[“auto_advance":"false","tags":"title;text;question;responses"]
b) Active Tasks
Active Tasks are activities that study participants perform under partially controlled conditions while iPhone sensors collect data. As an example, the active task for analyzing gait and balance asks the user to walk a short distance, while collecting accelerometer data on the device.
ResearchKit currently includes 19 Active Tasks, all of which are implemented by our platform. The tasks fall into seven categories: motor activities, fitness, cognition, speech, hearing, hand dexterity, and vision, and are listed below. ResearchKit is an open-source software framework, so developers are able to produce their own Active Tasks to meet the needs of their study. We have produced an Active Task for cue reactivity. Many of the Active Tasks generate multiple result files, which are zipped together by the app and uploaded to REDCap.
For studies employing real-time monitoring using Gene Doe Advocate, results of Active Tasks are retrieved from REDCap by Apache Spark shortly after completion by the user. A specific function for each Active Task type computes the most helpful metrics for visualization and longitudinal review.
Investigators can call for the delivery of an Active Task to the study participant at any time by issuing commands in REDCap field annotations:
researchkit.activetask.motor.gait_and_balance["intendedUseDescription":"Evaluating your gait”,"duration":20]
c) Consent
Research Studies involving human subjects typically require ethics review by an institutional review board (IRB) or an ethics committee (EC). In some cases, these reviews may lead to the requirement to obtain informed consent from study participants to ensure that each is fully informed about the nature of the study.
The ResearchKit framework provides a number of predefined consent pages that are commonly included in consent documents: Overview, Data gathering, Privacy, Data use, Time commitment, Surveys, Tasks, and Withdrawal.
The predefined sections include appropriate images and localized titles to which appropriate content from the ethics review document is added. Custom consent pages can also be defined to address additional concepts and concerns identified by the ethics review. A consent review step allows final review for the participant, which can repeat the information provided through the prior screens or provide entirely separate review content.
If required, the user may provide a signature to finalize the process, which is uploaded to REDCap. A PDF of the signed consent form can be generated and provided to the user.