Prediction of Critical Illness During Out-of-Hospital Emergency Care

In a retrospective study of emergency patients without traumatic injury or cardiac arrest, researchers were able to develop a critical illness score to reasonably predict which patients would require services of a regional critical care center.

Hospitals vary in the quality of critical care services available. Policy-makers have recommended creation of a network of regional critical care centers, similar to that established for trauma, so that patients in most need of “high intensity acute care” receive that care at hospitals best able to deliver it. So, how do you identify those patients most likely to need such care? Right now, triage of noninjured, critically-ill patients relies on a nonstandardized variety of objective and subjective information. This study examined 144,913 records of adult patients transported by Kings County, Washington EMS from 2002 to 2006 with the purpose of developing predictors of critical illness. Trauma and cardiac arrest patients, subject to different triage standards, were excluded.

Key Findings:

  • Five percent of patients developed critical illness during hospitalization.
  • By examining 60 percent of the patient records, researchers developed a predictive Critical Illness Score using a small number of objective out-of-hospital variables, such as systolic blood pressure, heart rate, respiratory rate, Glasgow Coma Scale score and pulse oximetry.
  • They then applied this score to the remaining 40 percent of patient records to test its veracity.
  • The Score demonstrated “satisfactory discrimination,” but tended to over-identify the critically ill among high-risk patients and under-identify critical illness among those considered at low risk.

Although more work is needed to refine and validate the predictive rule, this study demonstrates that simple objective physiologic assessments during EMS transport can help predict whether a patient will develop critical illness, and may help determine where a patient should be taken for best care.