國立高雄大學圖資館 |

Language: English

Back

Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention.
Author:	Kreuter, Jason Michael.
Description:	85 p.
Notes:	Source: Dissertation Abstracts International, Volume: 64-10, Section: B, page: 5069.
Notes:	Supervisor: William E. Hammond.
Contained By:	Dissertation Abstracts International64-10B.
Subject:	Engineering, Biomedical.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3109012
ISBN:	0496565575

Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention.
Kreuter, Jason Michael.

Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention. [electronic resource] - 85 p.

Source: Dissertation Abstracts International, Volume: 64-10, Section: B, page: 5069.

Thesis (Ph.D.)--Duke University, 2003.

Background. The computer based clinical practice reminder system has grown to be an important form of clinical practice guideline distribution. While computer based guidelines have been studied extensively, a common limitation of the studies is their relatively short duration due to the difficulty of data-mining a long-term medical data set. This study presents a method for long-term analysis of the effectiveness of computer based guidelines. Methods. A twenty-five year data set from The Medical Record (TMR) medical information system at the Duke Family Medicine Clinic (DFMC) was ported to a relational database. Analyses were performed on the effects of practice type (change from non-academic to academic in 1991), release of the 1989 U.S. Preventive Services Taskforce guidelines, and the initial installation of clinical practice reminders in 1984 on compliance with Papanicolaou, mammogram, and influenza vaccination guidelines. In order to account for the overlapping time periods of the interventions, the data was analyzed at 2-year and 5-year "eras" around the three time points. The data was normalized against the total number of patients eligible to follow the guideline directive. Results. The total number of patients receiving influenza vaccinations increased significantly from 1981--1998, but not in patients over age 65. A significant difference was found in mammogram guideline compliance in 2-year and 5-year eras around 1989 and 1991 (p < 0.05). Significant difference was found in Papanicolaou guideline compliance in 2-year era 1989 (p < 0.05), and 5-year eras 1989 and 1991 (p < 0.05). For mammograms, 5-year eras 1984 and 1989, and 2-year era 1991 had significant difference in the rate of change in reminder compliance. For Papanicolaou tests, none of the era comparisons had a significant difference in the rate of change in reminder compliance. Discussion. The use of eras effectively allowed for a long-term analysis that accounted for changes in medical practice over twenty years. The 2-year and 5-year eras enabled measurement and comparison of the effect of each year of interest with and without time overlap with another year of interest. The study showed that computer based guidelines improve guideline compliance at even seven years after their introduction.

ISBN: 0496565575Subjects--Topical Terms:

227004
Engineering, Biomedical.

Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention.
LDR:03292nmm _2200253 _450 001 162260
005 20051017073427.5
008 230606s2003 eng d
020 $a 0496565575
035 $a 00148761
035 $a 162260
040 $a UnM $c UnM
100 0 $a Kreuter, Jason Michael. $3 227386
245 1 0 $a Use of a legacy data repository for multi-year longitudinal analysis of a clinical intervention. $h [electronic resource]
300 $a 85 p.
500 $a Source: Dissertation Abstracts International, Volume: 64-10, Section: B, page: 5069.
500 $a Supervisor: William E. Hammond.
502 $a Thesis (Ph.D.)--Duke University, 2003.
520 # $a Background. The computer based clinical practice reminder system has grown to be an important form of clinical practice guideline distribution. While computer based guidelines have been studied extensively, a common limitation of the studies is their relatively short duration due to the difficulty of data-mining a long-term medical data set. This study presents a method for long-term analysis of the effectiveness of computer based guidelines. Methods. A twenty-five year data set from The Medical Record (TMR) medical information system at the Duke Family Medicine Clinic (DFMC) was ported to a relational database. Analyses were performed on the effects of practice type (change from non-academic to academic in 1991), release of the 1989 U.S. Preventive Services Taskforce guidelines, and the initial installation of clinical practice reminders in 1984 on compliance with Papanicolaou, mammogram, and influenza vaccination guidelines. In order to account for the overlapping time periods of the interventions, the data was analyzed at 2-year and 5-year "eras" around the three time points. The data was normalized against the total number of patients eligible to follow the guideline directive. Results. The total number of patients receiving influenza vaccinations increased significantly from 1981--1998, but not in patients over age 65. A significant difference was found in mammogram guideline compliance in 2-year and 5-year eras around 1989 and 1991 (p < 0.05). Significant difference was found in Papanicolaou guideline compliance in 2-year era 1989 (p < 0.05), and 5-year eras 1989 and 1991 (p < 0.05). For mammograms, 5-year eras 1984 and 1989, and 2-year era 1991 had significant difference in the rate of change in reminder compliance. For Papanicolaou tests, none of the era comparisons had a significant difference in the rate of change in reminder compliance. Discussion. The use of eras effectively allowed for a long-term analysis that accounted for changes in medical practice over twenty years. The 2-year and 5-year eras enabled measurement and comparison of the effect of each year of interest with and without time overlap with another year of interest. The study showed that computer based guidelines improve guideline compliance at even seven years after their introduction.
590 $a School code: 0066.
650 # 0 $a Engineering, Biomedical. $3 227004
650 # 0 $a Computer Science. $3 212513
710 0 # $a Duke University. $3 226880
773 0 # $g 64-10B. $t Dissertation Abstracts International
790 $a 0066
790 1 0 $a Hammond, William E., $e advisor
791 $a Ph.D.
792 $a 2003
856 4 0 $u http://libsw.nuk.edu.tw/login?url=http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3109012 $z http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3109012