Using a Cascading Approach to Improve Multisite Study Procedures

Abstract

Background: Multisite studies offer larger, more diverse samples to successfully capture populations and clinical practices of interest at the point of care. However, investigators face challenges with site recruitment and sampling, differences in clinical practices across sites, and data integrity. Addressing these issues a priori can improve the rigor and reproducibility of the research.Objective: This article aims to describe a cascading approach to multisite research. An exemplar is provided of a study using this approach, which aimed to evaluate the prevalence of pain and the pain management practices provided to critically ill children in pediatric intensive care units in the United States.Methods: The cascading approach includes two or more pilot study procedures with a progressively increasing number of sites prior to a full-scale study. Following each pilot, study procedures are evaluated; feedback was obtained from site personnel and content experts; procedures were revised accordingly; approvals were obtained; sites were trained; and the revised procedures are repeated with a larger, more diverse number of sites.Results: In the exemplar provided, improvements in the efficiency and integrity of data collection were noted for the full-scale study following the pilots. All sites that completed the agreements and approvals for study participation were retained for the duration of the two pilots and full-scale study.Discussion: Borrowing from principles of process improvement, the cascading approach allows knowledge to be gained regarding site differences and informs the revision of study procedures while potentially maximizing efficiency and data integrity, minimizing site burden, and maintaining site engagement for multisite studies.

To Error Is Human, to Control Is Divine!

“…Profession-wide efforts to overcome barriers to the conduct of these studies (eg, costs, communication, training, credit, authorship, etc.) are needed 15,16…”

“…are needed. 15,16 Rehabilitation intervention research would benefit greatly from a shift away from individual investigators at individual institutions operating in a local orbit with limited support to produce small sample studies with inconclusive findings. Such studies may accommodate entry level student research project schedules and, if published, aid faculty in achieving tenure and promotion, but their evidentiary value is limited.…”