Abstract
BACKGROUND: The Zwolle Risk Score (ZRS) identifies primary percutaneous coronary intervention (PPCI) patients at low mortality risk, eligible for early discharge. Recently, this score was improved by adding baseline NT-proBNP. However, the optimal timepoint for NT-proBNP measurement is unknown. METHODS: PPCI patients in the On-Time 2 study were candidates. The ZRS and NT-proBNP levels on admission, at 18-24 h, at 72-96 h, and the change in NT-proBNP from baseline to 18-24 h (delta NT-proBNP) were determined. We investigated whether addition of the different NT-proBNP measurements to the ZRS improves the prediction of 30-day mortality. Based on cut-off values reflecting zero mortality at 30 d, patients who potentially could be discharged early were identified and occurrence of major adverse cardiac events (MACE) and major bleeding until 10 d was registered. RESULTS: 845 patients were included. On multivariate analyses, NT-proBNP at baseline (HR 2.09, 95% CI 1.59-2.74, p < 0.001), at 18-24 h (HR 6.83, 95% CI 2.94-15.84), and at 72-96 h (HR 3.32, 95% CI 1.22-9.06) independently predicted death at 30 d. Addition of NT-proBNP to the ZRS improved prediction of mortality, particularly at 18-24 h (net reclassification index 29%, p < 0.0001, integrated discrimination improvement 17%, p < 0.0001). Based on ZRS (<2) or NT-proBNP at 18-24 h (<2500 pg/ml) 75% of patients could be targeted for early discharge at 48 h, with expected re-admission rates of 1.2% due to MACE and/or major bleeding. CONCLUSIONS: NT-proBNP at different timepoints improves prognostication of the ZRS. Particularly at 18-24 h post PPCI, the largest group of patients that potentially could be discharged early was identified.