Mining NIS data for trends of fundoplication for GERD in children

Data mining has found a growing number of medical applications, where it is also called informatics. For instance, gastroesophageal reflux disease (GERD) is common in children with chronic conditions and diseases. Utilization of laparoscopy among several other surgical procedures in anti-reflux surgery has been reported to improve recovery, reduce length of stay and decrease the risk of complications. Very little knowledge is available regarding the surgical trends and hospital outcomes for GERD in children in the U.S. Investigating clinical outcomes on a regional or national scale has become possible recently. The Nationwide Inpatient Sample (NIS) contains more than 7 million hospital discharges per year from 1988 to 2000. This database represents 20 percent of all hospital discharges in the U.S. during that time period. This huge database has enabled clinical researchers to investigate clinical outcomes and perform national estimates for both common and rare diseases. Through close teamwork of Bradley University and University of Illinois, we have applied data mining to discover some surgical trends and hospital outcomes for fundoplication applied to GERD in children under age 19 in the U.S. and other clinical inquiries between 1998-2000. This paper presents methods, techniques, and software packages used in handling this huge database and reports some preliminary results.