 |
The Journal of the Japanese Society for Clinical Microbiology |
Biblioraphy Information
| ArticleTitle |
Multivariate analysis of clinical data for predicting bacteremia in the patients with infectious diseases |
| Language |
J |
| AuthorList |
Yutaka Kouguchi1), Takao Tashiro2), Masayo Tamura1), Masaki Ishimatsu1), Etsuko Murakami1), Rieko Takagi1), Yukinori Kurokawa1), Kaoru Tohyama3) |
| Affiliation |
1) Department of Clinical Laboratory, Kawasaki Medical School Hospital
2) Human Life and Health Sciences, The Open University of Japan
3) Department of Laboratory Medicine, Kawasaki Medical School |
| Publication |
J.J.C.M.: 24 (2), 114-123, 2014 |
| Received |
November 12, 2013 |
| Accepted |
February 10, 2014 |
| Abstract |
We retrospectively analyzed 236 patients with infectious diseases using a multiple logistic regression analysis to determine the factors predicting bacteremia of these patients. Factors analyzed were age and sex of the patients, vital signs involving Glasgow Coma Scale (GCS), blood pressure, heart rate, body temperature, and basic laboratory examination data. The data were collected on the day of blood culture testing ([0]) and before three days ([-3]) and then the difference values of [0] and [-3] were Δ values. An optimum discriminant equation was calculated from all these parameters (E1), vital signs (E2) and only body temperature (E3). The area under the receiver operating characteristics (ROC) curves were compared among any following three prognostic parameters: the scores by E1 and by E2 and by E3. An optimum discriminant equation (E1) was derived as {P=1/(1+e-X), X=1.27191+0.00795 (age)+0.41470 (sex; man=0, woman=1)-0.16070 ([0] Ht)+0.04127 ([0] heart rate)-0.20440 ([0] mono%)-0.46360 (Δ GCS)+1.18603 (Δ T-Bil)-0.07070 (Δ Plt)}. The score calculated by E1 was the best discriminant, while E3 was the worst predictor for septic patients. The combinations of vital signs and basic blood examination proved to be useful for predicting bacteremia of infectious diseases. |
| Keywords |
|
|
