| Effect of the computerized EKG system on the family medicine residents' interpretation of EKGs. |
|
Yong Eun Kim, Yu Sun Kim, Cheung Kee Kim, Chang Yup Kim, In Hong Whang, Tae Woo Yoo, Bong Yul Huh |
|
|
| 컴퓨터 자동 심전도 판독 시스템이 가정의학 전공의의 심전도 판독에 미치는 영향 |
|
김용은, 김유선, 김청기, 김창엽, 황인홍, 유태우, 허봉렬 |
|
서울대학교병원 가정의학과 |
|
|
|
|
| Abstract |
To know whether computerized EKG system improves the accuracy of family practitioner's EKG interpretation and, if so, whether this improvement brings considerable acquirement of clinically significant informations, we collected all EKGs(total 207 cases) processed by computerized EKG machine from April 1,1989 until March 31,1990 in the outpatient department of Family Medicine, Seoul National University Hospital.
Firstly, these 207 EKGs were evenly and randomly distributed to the 10 family medicine residents and we requested them to interpret each EKG unaided by computerized EKG system. Secondly, these 207 EKGs were evenly and randomly reassigned to the previous 10 family medicine residents and we requested them to interpret each EKG with the aid of computerized EKG system.
Cardiologist's interpretation aided with computerized EKG system was used as a standard interpretation and evaluation of accuracy of residents' interpretation was done according to it.
1. Degree of agreement with cardiologist's interpretation increased from 72.9%(precomputer interpretation) to 76.3%(postcomputer interpretation).
2. Sensitivity of the residents' interpretation on the normal EKGs increased from 0.80(precomputer interpretation) to 0.85(postcomputer interpretation).
3. Specificity of the residents' interpretation on the normal EKGs increased from 0.80(precomputer interpretation) to 0.82(precomputer interpretation).
4. When compared with postcomputer interpretation, precomputer interpretation was changed in the 60 cases out of 207 cases.
In the normal EKGs(total 123 cases), 22 cases(17.9%) were changed.
In the abnormal EKGs(total 64 cases), 38 cases(43.0%) were changed.
5. Analysis of postcomputer interpretation was as followed;
more agreements with cardiologist's interpretation were in 38 cases(60.3%), less agreements were in 20 cases(33.3%) and unchanged cases were 2(6.4%)
6. More agreement brought clinically significant informations in 17 cases(44.7%) out of more agreed EKGs (38 cases).
We concluded that postcomputer interpretation made the family medicine residents' interpretation more agreed with cardiologist interpretation and this increased agreement brought clinically significant informations. Thus this result suggested that computerized EKG system serve as a useful teaching tool for educating EKG interpretation and increase the quality of practice by acting as a backup opinion for more accurate interpretation. |
|
PDF Links
Full text via DOI
Download Citation
