A novel scoring system for predicting the neurologic prognosis prior to the initiation of induced hypothermia in cases of post-cardiac arrest syndrome: the CAST score

Abstract

BackgroundThe aim of this study was to develop a scoring system for identifying the post-cardiac arrest syndrome (PCAS) patients with a good potential for recovery prior to the initiation of induced therapeutic hypothermia.MethodsA multi-center, retrospective, observational study was performed. Data of a total of 151 consecutive adults who underwent induced hypothermia after cardiac arrest (77 learning cases from two hospitals and 74 validation cases from two other hospitals) were analyzed.ResultsIn the learning set, 8 factors (initial rhythm, witnessed status and time until return of spontaneous circulation, pH, serum lactate, motor score according to the Glasgow Coma Scale (GCS), gray matter attenuation to white matter attenuation ratio (GWR), serum albumin, and hemoglobin) were found to be strongly correlated with the neurological outcomes. A tentative scoring system was created from the learning data using these factors, and the predictive accuracy (sensitivity and specificity) was evaluated in terms of both internal validation (0.85 and 0.84) and external validation (cutoff 50%: 0.95 and 0.90, 30%: 0.87 and 0.98, 15%: 0.67 and 1.00). Finally, using all the data, we established a post-Cardiac Arrest Syndrome for induced Therapeutic hypothermia (CAST) score to predict the neurologic prognosis prior to initiation of induced hypothermia.ConclusionsThe CAST score was developed to predict the neurological outcomes of PCAS patients treated by induced hypothermia. The likelihood of good recovery at 30 days was extremely low in PCAS patients with a CAST score of ≤15%. Prospective validation of the score is needed in the future.Electronic supplementary materialThe online version of this article (doi:10.1186/s13049-017-0392-y) contains supplementary material, which is available to authorized users.

Prediction of poor neurological outcome in comatose survivors of cardiac arrest: a systematic review

“…The characteristics of the 94 included studies (30,200 patients) are listed in Table 2. TTM was used in 100% of patients in 68/94 (72%) studies, while 22 studies reported a percentage of TTM treatment ranging from 17.5% and 94%.…”

“…Target temperature varied across studies (ESM Table E3). Fifteen studies [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36] were included for clinical examination, 12 for biomarkers [37][38][39][40][41][42][43][44][45][46][47][48], 26 for electrophysiology , and 15 for imaging [75][76][77][78][79][80][81][82][83][84][85][86][87][88][89], while 26 studies were included for more than one prognostic modality.…”

Variable Selection for Nonparametric Learning with Power Series Kernels

“…In section 5.2, we analyze three real datasets, PCAS dataset [11], diabetes dataset [22] and Wisconsin breast cancer dataset. The latter two datasets are published at the UCI machine learning repository.…”

“…Here we employ the logistic regression to learn a classification model. In this study, we analyzed four medical datasets: post-cardiac arrest syndrome (pcas) [11], chronic kidney disease (ckd), cervical cancer (ccancer) and cortex nuclear (cnuclear) [5]. The characteristics of each dataset are summarized in Table 2.…”

Accuracy of the first interpretation of early brain CT images for predicting the prognosis of post-cardiac arrest syndrome patients at the emergency department

“…A definitive estimation of the prognosis of PCAS patients undergoing targeted temperature management (TTM) should be performed 72 h after the return to normal body temperature according to the guideline [ 3 ]. But a few previous studies have reported the usefulness of early estimation of the prognosis in PCAS patients at the time of the arrival at the emergency department (ED) [ 4 – 6 ].…”