This study compares the effect of the modified early warning score (MEWS) versus a new early warning system (Niguarda MEWS) for detecting instability and criticality in hospital medical departments. A retrospective observational study was conducted in the Internal Medicine ward of Niguarda Ca' Granda Hospital in Milan between November 2013 and October 2014. MEWS and Niguarda-MEWS were gathered using: systolic blood pressure, respiratory frequency, heart rate, temperature, level of consciousness, oxygen saturation, creatinine level, hematocrit level and age. In order to determine if the patient was critical or not the MEWS criticality cut-off value chosen was 3, while in the Niguarda MEWS it was 6. The primary outcome was the correlation between the critical level of the two scores and in-hospital mortality. The secondary endpoint was the correlation between a specific disease and the two scores. In the study, 471 patients were included, using both the MEWS and the Niguarda MEWS score at admittance: 33.4% of patients turned out to be critically ill using the former, 40.98% when using the latter. Therefore, the specificity of scores was 70% for MEWS and 73% for Niguarda MEWS, the sensitivity 58% for MEWS and 63% for Niguarda MEWS, Niguarda MEWS area under the curve (AUC): 0.736, MEWS AUC: 0.670. For the secondary outcome, the new score is higher for genitourinary and respiratory diseases. Niguarda-MEWS could be an optimal tool to detect criticality and instability in order to address the patient to the right level of care.