Informatics technology is in its infancy in the intensive care unit. The wide variability in clinical practice is matched by an even wider variability in the availability of informatics support amongst ICU's.

Some units still access individual, non-integrated, non-networked data from individual bedsides and record these records by hand on nursing flow sheets. This immense data base is contained within patient charts, archived in medical records and never seen again. Data collection is at best willy nilly and inadequate and is prohibitively time consuming.

In those units where automated data collection systems exist, databases are frequently inadequately constructed. This is largely due to the lack of a standard national format for data base construction in critical care.

There also are wide variations in proprietary on-line real-time data acquisitions systems and database storage platforms among institutions. Inter-institutional linking of these idiosyncratic data collection modalities is a major challenge. These realities are a significant challenge for the sharing of data.

Data sharing and access to multi-point critical care data bases has never occurred. Recently attempts have been made to develop research data bases and to form centralized data repositories. They have all failed, primarily because of poor informatics support.

The advent of advanced information technologies, the availability of the Internet information super highway, the coming promise of Internet II, combined with point of care real-time patient data acquisition technologies provides a unique opportunity.

For the first time, it is possible to capture bedside data, provide real quality control on a national basis, perform prospective randomized clinical trials and integrate these within a framework designed to educate critical care providers and push information to the bedside of children requiring critical care, regardless of the their location. Every critically ill child can have a bed in this virtual ICU.

In response to this, and many other informatics needs in the pediatric critical care world, In the spring of 1998, several of us got together in St Louis for a two day symposium to explore ways to develop informatics support to meet the needs of pediatric critical care medicine into the next century. We recognize that much excellent work already has been and is being done by many pediatric intensivists around the country. We think that it would be beneficial to provide support, develop standards, encourage organization and coordination and facilitate communication among the many people interested in this area so that we don't have to all 'go it alone'.

The symposium attendees were: Dana Braner at Portland Children's Hospital. Joe DiCarlo at Stanford. Barry Markovitz at Children's Hospital St. Louis. Carl Weigle at Milwaukee Children's. Judy Verger from Children's Hospital of Philadelphia. Jim Fackler from Johns Hopkins Hospital. Steve Pon from Cornell. Adrienne Randolph from Boston Children's. Randall Wetzel at CHLA. Doug Willson from U.Va. (in absentia). Tom Green from Northwestern has since joined the group.

This group is intent on organizing a 'Virtual PICU' consistent with our goals.