Objective To set up a new grading system of intraventricular hemorrhage (IVH) and determine the value of predicting the probability of post-hemorrhagic hydrocephalus (PHH) in IVH. Methods We first modified the Graeb c...Objective To set up a new grading system of intraventricular hemorrhage (IVH) and determine the value of predicting the probability of post-hemorrhagic hydrocephalus (PHH) in IVH. Methods We first modified the Graeb criteria, then compared the value of prediction for PHH assessed by the Graeb criteria with the modified Graeb criteria. One hundred and thirty one IVH patients were divided into two groups: the upper group (n=67) and the lower group (n=64). Gold standard of PHH was assessed by CT scan or by out-drainage. The diagno-stic parameters such as sensitivity (SE), specificity (SP) were analyzed. In the cutoff point of SE and SP curves, diagnostic efficiency (DE), and Kappa value (K) were analyzed. The probability of PHH was estimated by binary logistic regressions. Results In all ventricular group, to Graeb criteria in the cutoff point, SE, SP, and K was 0.78, 0.84, and 0.60; and to modified Graeb criteria SE, SP, and K was 0.90, 0.84, and 0.74 respectively. The probability of PHH from point of 3-12 was 0.011, 0.032, 0.085, 0.212, 0.435, 0.689, 0.865, 0.949, 0.981, and 0.994 respectively according to modified Graeb criteria. Conclusion The modified Graeb criteria combined with logistic regression were useful methods to assess the severity of IVH and to predict the probability of PHH in IVH.展开更多
文摘Objective To set up a new grading system of intraventricular hemorrhage (IVH) and determine the value of predicting the probability of post-hemorrhagic hydrocephalus (PHH) in IVH. Methods We first modified the Graeb criteria, then compared the value of prediction for PHH assessed by the Graeb criteria with the modified Graeb criteria. One hundred and thirty one IVH patients were divided into two groups: the upper group (n=67) and the lower group (n=64). Gold standard of PHH was assessed by CT scan or by out-drainage. The diagno-stic parameters such as sensitivity (SE), specificity (SP) were analyzed. In the cutoff point of SE and SP curves, diagnostic efficiency (DE), and Kappa value (K) were analyzed. The probability of PHH was estimated by binary logistic regressions. Results In all ventricular group, to Graeb criteria in the cutoff point, SE, SP, and K was 0.78, 0.84, and 0.60; and to modified Graeb criteria SE, SP, and K was 0.90, 0.84, and 0.74 respectively. The probability of PHH from point of 3-12 was 0.011, 0.032, 0.085, 0.212, 0.435, 0.689, 0.865, 0.949, 0.981, and 0.994 respectively according to modified Graeb criteria. Conclusion The modified Graeb criteria combined with logistic regression were useful methods to assess the severity of IVH and to predict the probability of PHH in IVH.
