Impact of information quality on the use and effectiveness of computerized clinical reminders
Abstract
A computerized clinical reminder (CCR) system is a type of decision support system that constantly reminds clinicians of recommended health services. It is designed to enhance clinical practice guideline adherence and, ultimately, quality of care; however, it has not been used as routinely as desirable. In order to identify the paths to improve the effective use of a CCR system, we analyzed a cross-sectional survey of 261 VA CCR users with a regression tree algorithm, and found that “easy to use” and “helpful in delivering care” were most important in predicting the use of a CCR system. We further conducted a pilot study and found a negative linear correlation between estimated CCR resolution time and adherence rate (R-square= 0.876, 0.997, and 0.670 for pessimistic, expected, and optimistic times respectively). These findings demonstrated the criticality for a CCR system to be easy to use and able to facilitate decision making. These findings also unveiled the issues of providers’ degrading situation awareness, and the VA CCR system’s incapability to assist them in retrieving desired clinical information. This study aimed to improve the information quality of VA’s CCR system to facilitate decision making. A web-based CPRS system was prototyped as a mockup of the current VA system. We further redesigned the CCR system by incorporating a knowledge-based risk factor repository, a role-based filter, a prioritization mechanism, and a new documentation interface powered by content organization heuristics. The performances of the original and modified designs were tested by 16 physicians in a controlled lab in the Indianapolis VAMC. The subjects individually simulated a scenario-based patient encounter through an interactive simulator. These subjects then addressed and resolved clinical reminders in a two-stage experiment. A semi-structured interview and survey was conducted to elicit subjects’ CCR prioritization heuristics and satisfaction. This study laid out the methodology to improve CCR information quality by aligning information flow with clinicians’ mental model in decision making. The results concluded the modified CCR features were perceived useful and better. Moreover, this proposed information quality framework impacted 80% of subjects’ decision making, and improved 44% of overall CCR prioritization decisions.
Degree
Ph.D.
Advisors
Yih, Purdue University.
Subject Area
Industrial engineering|Information science|Health care management
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.