Cardiac Index Predicted Factors for Shock Patients

ABSTRACT

The report aims to formulate an appropriate cardiac index (CI) model with 19 explanatory variables for 113 shock patients. Determinants of CI are focused from the fitted model. There is little study for the predicted factors of CI based on probabilistic modeling. CI model has been developed in the report with 19 explanatory variables for 113 shock patients, and the data site is: http://www.umass.edu/statdata/statdata/data/shock.txt Statistical method of joint generalized linear models (JGLMs) is adopted. Mean CI is negatively associated with age (P = 0.0044), shock type (SHOCKT) at level 3  (P = 0.0586),  diastolic blood pressure (DBP) (P = 0.0032), mean circulation time (MCT) (P<0.0001), hemoglobin (HG) (P = 0.0053), while it is positively with mean arterial pressure (MAP) (P<0.0001), heart rate (HR) (P<0.0001), body surface index (BSI) (P<0.0001), appearance time (AT) (P<0.0001), plasma volume index (PVI) (P<0.0001). Variance of CI is negatively associated with  age (P = 0.0020), height (P = 0.0255), sex (P = 0.0582), SHOCKT at level 2 (P = 0.0356), MAP (P = 0.0664), AT (P = 0.0080), hematocrit (HCT) (P = 0.0039), card sequence order (CSO) (P<0.0001), while it is positively associated with systolic blood pressure (SBP) (P = 0.0241) and MCT (P = 0.0039). CI is high if MAP, or HR, or BSI, or AT, or PVI rises, or age, or DBP, or MCT, or HG decreases. Variance of CI depends on many explanatory factors. These findings are completely a new addition in medical literature.