A score to predict medical emergencies in hospitalized patients

Claudia Cofré, Gabriel Cavada, César Maquilón, Paula Daza, Ángel Vargas, Antonio Vukusich

Research output: Contribution to journalArticlepeer-review


All right reserved. Background: The medical alert system (MAS) was created for the timely handling of clinical decompensations, experienced by patients hospitalized at the Medical Surgical Service (MSS) in a private clinic. It is activated by the nurse when hemodynamic, respiratory, neurological, infectious or metabolic alterations appear, when a patient falls or complains of pain. A physician assesses the patient and decides further therapy. Aim: To analyze the clinical and demographic characteristics of patients who activated or not the MAS and develop a score to identify patients who will potentially activate MAS. Material and Methods: Data from 13,933 patients discharged from the clinic in a period of one year was analyzed. Results: MAS was activated by 472 patients (3.4%). Twenty two of these patients died during hospital stay compared to 68 patients who did not activate the alert (0.5%, p < 0.01). The predictive score developed considered age, diagnosis (based on the tenth international classification of diseases) and whether the patient was medical or surgical. The score ranges from 0 to 9 and a cutoff ≥ 6 provides a sensitivity and specificity of 37 and 81% respectively and a positive likelihood ratio (LR+) of 1.9 to predict the activation of MAS. The same cutoff value predicts death with a sensitivity and specificity of 80% and a negative predictive value of 99.8%. Conclusions: This score may be useful to identify hospitalized patients who may have complications during their hospital stay.
Original languageAmerican English
Pages (from-to)156-163
Number of pages8
JournalRevista Medica de Chile
Issue number2
StatePublished - 1 Jan 2017


  • Emergencies medical services
  • Hospital rapid response team
  • Patients’ rooms
  • Risk assessment

Fingerprint Dive into the research topics of 'A score to predict medical emergencies in hospitalized patients'. Together they form a unique fingerprint.

Cite this