The value of shock index in prediction of cardiogenic shock developed during primary percutaneous coronary intervention

Shock index(SI) is a conventional predictive marker for haemodynamic state. Its breakpoint varies by different conditions according to previous studies. The current study was performed to evaluate the capability of SI in prediction of cardiogenic shock(CS) developed during primary percutaneous coronary intervention (pPCI). Total 870 patients of ST segment elevation myocardial infarction(STEMI) who were haemodynamic stable before pPCI were involved in the study. In this cohort, 625 consecutive patients composed analysis series and 245 consecutive patients composed validation series. Multivariate regression analysis was used to evaluate whether SI was a significant predictor of developed CS and Hosmer-Lemeshow test was used to assess the goodness of model fitness. Receiver-operating characteristics (ROC) analysis was used to compare the predictive capability of SI with other predictors. The sensitivity, specificity, accuracy, positive and negative predictive values of SI at different cutoff values was compared to identify a best breakpoint. In the analysis series, SI and Killips classification were identified as independent predictors. ROC analysis demonstrated the diagnostic capability of SI was superior to pre-procedural systolic blood pressure(SBP) or heart rate(HR) alone (0.8113 vs 0.7582, P = 0.04 and 0.8113 vs 0.7111, P < 0.001). The diagnostic capability of SI was equivalent to that of combination of SBP, HR and Killips claasification(0.8133 vs 0.8137, P = 0.97). SI had a high specificity and low sensitivity. When the cutoff value was set at 0.93, the positive predictive value, negative predictive value and diagnostic accuracy was 42.6%, 95.1% and 87.4% respectively. In validation series, the area under ROC curve was 0.8245, which was similar to that in the analysis series. The positive predictive value, negative predictive value and diagnostic accuracy at the cutoff value of 0.93 was 53.8%, 93.2% and 88.9% respectively. SI has a high predictive accuracy for developing CS during pPCI in STEMI patients. It is an excellent exclusion diagnosis index rather than confirmative diagnosis index.