Comparative effectiveness research (CER) bridges the gap between knowledge gained from clinical research and knowledge needed to make informed decisions on disease or care management. Large-scale CER can involve multiple organizations, broad populations, and a variety of care delivery systems, and often depends upon the sharing of patient-level data between multiple data partners. However, privacy and confidentiality considerations place limits on this data sharing. Distributed research networks (DRNs) have been proposed as a mechanism for supporting multi-site CER, while minimizing the need to transfer patient-level data.

The objectives of SPAN are to develop a distributed research network that 1) is interoperable across a range of health-care systems; 2) permits menu-driven querying of data; and 3) utilizes patient-level data for analyses.

The architecture for Scalable PArtnering Network for Comparative Effectiveness Research, or SPAN Network, is based on an existing distributed research network (the DEcIDE DRN2 network). The SPAN Network includes 11 data partners: 9 from an existing research network; and 2 community partners with different delivery systems, data structures, and patient populations. The design specifics of the SPAN Network include:

• a central web-based querying interface with supporting network software
• a user-friendly, menu-driven query tool
• local datasets, which adhere to a common data model and remain at each site
• secure, encrypted transfer of all queries and query results

Participating sites provide data about various treatment strategies used for patients with two Agency for Healthcare Research and Quality (AHRQ) priority conditions: obesity in adults and attention deficit hyperactivity disorder (ADHD) in children. Individual, identifiable patient data reside at each organization, protected behind their firewalls, while information necessary to conduct research is extracted in a limited or de-identified format for analyses. The SPAN Network has the potential to improve the efficiency of conducting multi-site research, and could facilitate the study of a broad spectrum of priority health care conditions.