Document Type

Capstone Experience

Graduation Date

5-2019

Degree Name

Master of Public Health

Department

Biostatistics

First Committee Member

Gleb Haynatzki, PhD

Second Committee Member

Jason Shiffermiller, MD, MPH

Third Committee Member

Kristy Carlson, PhD

Fourth Committee Member

Tricia LeVan, PhD

Abstract

The main goal of this project was to examine the differences in perioperative hyperglycemia treatment received by patients with a diagnosis of diabetes mellitus (DM) and patients without a diagnosis of diabetes (NDM); and how these treatment differences can affect the length of hospital stay. Studies have revealed that, when comparing DM and NDM patients with the same degree of perioperative hyperglycemia, NDM patients suffer worse outcomes. It has been suggested in previous research that this may be because NDM patients receive treatment that does not measure up to the standard of care treatment that DM patients receive. In this project, we study these standard of care discrepancies between DM and NDM patients, in order to examine the difference in postoperative hospital length of stay (LOS).

We conducted a retrospective cohort study, using a subset of data comprised of inpatients who underwent a broad spectrum of surgery types at Nebraska Medicine over a period of twenty months. All patients in the study experienced perioperative hyperglycemia. Roughly half of them were DM patients and roughly half were NDM patients. We performed bivariate regression analysis on all available covariates in the data set in order to find candidates for final linear and logistic regression models. We also performed a time-to-event survival analysis to see if different lengths of stay among the two groups of patients were explained by possible differences in time lags between hyperglycemia trigger and insulin administration. We aimed to find whether aspects of glucose control practices were associated with hyperglycemic NDM patients experiencing longer LOS than DM patients. This information may bolster existing study conclusions and better assist us in understanding barriers to glucose control measures in NDM patients.

Download

Share

COinS