Fortunately, over the last 20 years computer science and medical informatics have evolved so that real-time data collection is now possible in the clinical practice setting, at the point-of-care. Ironically, the evolution of computer science and medical informatics mirrors the dynamic and complex evolution of the AIDS epidemic and HIV medicine. Advances in medical informatics have resulted in the development of modern, fully functional EMR systems capable of collecting data entered by clinicians at the time care is delivered using specifically defined data terminology with constrained vocabularies, leading to consistent fully analyzable data.
It is now possible to perform sophisticated, detailed outcomes analyses in close to “real-time”. Yet, even among sites with modern EMR systems, there are not large enough numbers of patients, sufficient availability of linked specimens or results of outcome-specific questionnaires, nor sufficient degree of diversity in the populations to answer many of the contemporary research questions highlighted above, hence, the rationale for CNICS. CNICS is built as a secure server application, using a partially normalized, optimized relational database (SQL), Extended Markup Language (XML), and JAVA. The application and database servers are secured with standard techniques including: strong access controls; software firewall; secure file transfer protocols; daily differential and full back-ups as well as off-site secure storage of backup media; network-wide virus protection software, and institutional IT-support.
The CNICS data system is built to allow flexibility in the way sites express specific data values, but in our experience, data quality has been enhanced by procedures which require sites to produce and transmit data formatted according to CNICS Transmissions Standards. While this uniformity is important, the inherent flexibility of the CNICS system allows sites to develop and refine their capability at different rates, as well as to permit incremental migration to enhancements to our standard.
Transformation and validation of data are done both at the sites, and centrally on the CNICS server. We deploy an XML packager that checks the site submissions with the CNICS definitions and terms and provides the site with immediate notice if data are not formatted correctly. Valid CNICS data are transmitted to the central repository using password-protected, encrypted, secure file transfer protocol. There the data are checked again using a system which permits not only checks of syntax, or structure, but also supports the evaluation of complex rules that may use any prior CNICS database content.