Abstract
The patient base of the Partners HealthCare System in Boston exceeds 1.8 million. Many of these patients are desirable for participation in research studies. To facilitate their discovery, we developed a data warehouse to contain clinical characteristics of these patients. The data warehouse contains diagnosis and procedures from administrative databases. The patients are indexed across institutions and their demographics provided by an Enterprise Master Patient Indexing service. Characteristics of the diagnoses and procedures such as associated providers, dates of service, inpatient/outpatient status, and other visit-related characteristics are also fed from the administrative systems. The targeted users of this system are research clinician s interested in finding patient cohorts for research studies. Their data requirements were analyzed and have been reported elsewhere. We did not expect the clinicians to become expert users of the system. Tools for querying healthcare data have traditionally been text based, although graphical interfaces have been pursued. In order to support the simple drag and drop visual model, as well as the identification and distribution of the patient data, a three-tier software architecture was developed. The user interface was developed in Visual Basic and distributed as an ActiveX object embedded in an HTML page. The middle layer was developed in Java and Microsoft COM. The queries are represented throughout their lifetime as XML objects, and the Microsoft SQL7 database is queried and managed in standard SQL.
Full text
PDFPage 1174
Selected References
These references are in PubMed. This may not be the complete list of references from this article.
- Banhart F., Klaeren H. A graphical query generator for clinical research databases. Methods Inf Med. 1995 Sep;34(4):328–339. [PubMed] [Google Scholar]
- Murphy S. N., Morgan M. M., Barnett G. O., Chueh H. C. Optimizing healthcare research data warehouse design through past COSTAR query analysis. Proc AMIA Symp. 1999:892–896. [PMC free article] [PubMed] [Google Scholar]