Automated respiratory care charting: artifact rejection and data reduction.

Update Item Information
Title Automated respiratory care charting: artifact rejection and data reduction.
Publication Type thesis
School or College School of Medicine
Department Biomedical Informatics
Author Young, Wang Hsueh-fen
Date 1993-06
Description The objective of this study was to develop an automated charting algorithm (for artifact rejection and data reduction) for parameters from ventilators used for routine respiratory care (RC) charting. Six artifact rejections (moving average, moving medium, moving exponential weighted using two different weighting factors, Lowell and moving LOWESS) and a data reduction method using two different critical time were tried. These six artifact rejection and two data reduction methods mad a combination of 12 automated charting algorithms. Real time data collected by Medical Information Bus (MIB) from 10 different patients (total 617.1 hours) were processed by these 12 automated charging algorithms and the results were compared to the respiratory therapists’ manual charted data. These ventilator data could be categorized into groups: measured ventilator data and machine settings. Only measured ventilator data were processed by artifact rejection algorithm, machine setting were not processed by artifact rejection algorithms since machine settings data had almost no artifact. The differences between manual and automated charting of data for machine settings were not statistically significant except respiratory rate setting, sensitivity setting and plateau pause time setting, but these difference were caused by therapists’ manual entry errors and in any event were not found to be clinically important. Automated chartings could easily detect setting changes automatically and accurately. The differences between manual and automated charting data for measured ventilator data were statistically significant for machine respiratory rate, spontaneous respiratory rate, corrected tidal volume and spontaneous tidal volume. In most cases the difference were less than 10% of the normal range of each parameter and were not considered clinically important. It improved that automated charting could also easily detect measured ventilator changes automatically and accurately. The algorithms developed eliminated not only artifact from the ventilatory data by selecting representative dta, but also reported only clinically important events to the host computer to minimize the amount of data stored in patient’s record. All the 12 automated charting algorithms could also record clinically important data into patient’s record more timely than manual charting, eliminate manual data entry errors, save time for medical practitioner’s on paperwork. Among the 12 automated charging algorithms tired in this study, the moving median with critical time of 3 minutes provided the bet results for artifact rejection and data reduction. Two user questionnaires survey were completed. The first determined RC therapists’ general opinions abut automated ventilator charting using computers. The results showed that therapists were neutral about automated charting system. However, some of the therapists agreed that computer system would be helpful, and some worried about the “ownership†of the data and the inconvenience of using computerized system. The second questionnaire was designed to determine how medical practitioners selected data in their routine charting. Seven medical practitioners participated in the second survey. The results showed that they each had different opinions about what were “important events†to chart. Automated charting using computers can be used to provide a more efficient health care. It is essential for the health care. It is essential for health providers to establish a set of rules for automated charting that emphasize how often and what kind of data should be charted.
Type Text
Publisher University of Utah
Subject Data Processing; Medicine
Subject MESH Respiration; Intensive Care; Medical Informatics
Dissertation Institution University of Utah
Dissertation Name MS
Language eng
Relation is Version of Digital reproduction of "Automated respiratory care charting: artifact rejection and data reduction." Spencer S. Eccles Health Sciences Library. Print version of "Automated respiratory care charting: artifact rejection and data reduction." available at J. Willard Marriott Library Special Collection. RC39.5 1993 .Y68.
Rights Management © Wang Hsueh-fen Young.
Format application/pdf
Format Medium application/pdf
Identifier us-etd2,5158
Source Original: University of Utah Spencer S. Eccles Health Sciences Library (no longer available).
ARK ark:/87278/s64j0vmp
Setname ir_etd
ID 192316
Reference URL https://collections.lib.utah.edu/ark:/87278/s64j0vmp
Back to Search Results