Tracking Re-hospitalization in Public Health Data

Author: Richard Gilder, RN

Primary Advisor: Kim Dunn, MD, PhD (co-author)

Committee Members: Susan McBride, RN, PhD (co-author); Mari Tietze, RN, PhD (co-author)

Masters thesis, The University of Texas School of Health Information Sciences at Houston.


Analysis of re-hospitalization from aggregated hospital inpatient administrative data across multiple hospitals requires the ability to identify the same patient in the dataset at each encounter regardless of hospital. This was accomplished through a unique patient identifier codification (UPIC) key field generated by proprietary probabilistic linking software in a vast dataset containing over six million inpatient admission records. Description of a novel adaptation of common business spreadsheet software is provided for resolving scale measure values of lag day intervals between admissions and zip code distances in miles between patient and hospital at each admission. The resulting hospital to patient distance can be associated with geo-location distance and direction as a variable, enabling the mapping, calculation, and analysis of geographical central place areas of high health care need density. Clinically significant informatics patterns and signals present within the migration phenomena of multiple and serial chronic re-hospitalization events are rendered tangible and visible in a vast dataset. Based upon regional community inpatient data that is recent and refreshed quarterly, automated and timely public reporting with high priority targeting of specific disease populations is achieved without violation of patient privacy.