Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.
Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.
Author ORCID Identifier
https://orcid.org/0000-0003-2109-5769
AccessType
Open Access Dissertation
Document Type
dissertation
Degree Name
Doctor of Philosophy (PhD)
Degree Program
Industrial Engineering & Operations Research
Year Degree Awarded
2019
Month Degree Awarded
September
First Advisor
Jenna L Marquard
Second Advisor
Hari Balasubramanian
Third Advisor
Benjamin M Marlin
Fourth Advisor
Aaron Z Hettinger
Subject Categories
Health Information Technology | Industrial Engineering
Abstract
Medication errors are a leading cause of death in the United States. Electronic Health Records (EHR) along with Computerized Provider Order Entry (CPOE) are considered promising ways to reduce these errors. However, EHR systems have not eliminated medication errors. Moreover, in some cases they have facilitated errors due to issues such as poor usability and negative effects on clinical workflows. The use of unexpected free text within a CPOE system can serve as a marker that the system does not adequately support clinical workflow. Prior studies have looked at the use of free text within medication orders, but the inclusion of medication related information in communication for non-medication orders (CNMOs), a type of free text order, has not been adequately studied. This mixed-methods study identified the prevalence, nature and reasons for the inclusion of medication related information in CNMOs using a large sample of CNMOs placed at a mid-Atlantic hospital system in 2017, and via interviews with physicians. The study found that more than 42% of CNMOs contain medication related information. Moreover, the use of CNMOs varied significantly across provider types, hospital locations, patient settings and other factors. The study found 10 themes that might cause providers to adopt such workarounds, including missing functionality and poor usability. The viii study also identified several general challenges in communicating medication information in the EHR, and potential solutions to mitigate these challenges. This dissertation also demonstrates how natural language processing could be used to identify medication related CNMOs.
DOI
https://doi.org/10.7275/14986704
Recommended Citation
Kandaswamy, Swaminathan, "COMBINING HUMAN FACTORS AND DATA SCIENCE METHODS TO EVALUATE THE USE OF FREE TEXT COMMUNICATION ORDERS IN ELECTRONIC HEALTH RECORDS" (2019). Doctoral Dissertations. 1690.
https://doi.org/10.7275/14986704
https://scholarworks.umass.edu/dissertations_2/1690