The Design of a Data Management System for a Multicenter Palliative Care Cohort Study

Abstract

Context: Prospective cohort studies of individuals with serious illness and their family members, such as children receiving palliative care and their parents, pose challenges regarding data management.

Objective: To describe the design and lessons learned regarding the data management system for the Pediatric Palliative Care Research Network's Shared Data and Research (SHARE) project, a multicenter prospective cohort study of children receiving pediatric palliative care (PPC) and their parents, and to describe important attributes of this system, with specific considerations for the design of future studies.

Methods: The SHARE study consists of 643 PPC patients and up to two of their parents who enrolled from April 2017 to December 2020 at seven children's hospitals across the United States. Data regarding demographics, patient symptoms, goals of care, and other characteristics were collected directly from parents or patients at 6 timepoints over a 24-month follow-up period and stored electronically in a centralized location. Using medical record numbers, primary collected data was linked to administrative hospitalization data containing diagnostic and procedure codes and other data elements. Important attributes of the data infrastructure include linkage of primary and administrative data; centralized availability of multilingual questionnaires; electronic data collection and storage system; time-stamping of instrument completion; and a separate but connected study administrative database used to track enrollment.

Conclusions: Investigators planning future multicenter prospective cohort studies can consider attributes of the data infrastructure we describe when designing their data management system.

Keywords: Cohort studies; Data management; Medical record linkage; Multicenter study; Palliative care.