BOSTON—APACHE scores may be helpful for assessing illness severity and predicting mortality in patients with known or suspected sepsis, but they were developed for use in the ICU. “So many things go into calculating an APACHE score that it is too much of a hassle in the emergency department,” says Nathan I. Shapiro, MD, an Instructor at Harvard Medical School in Boston.

Dr. Shapiro and colleagues are developing a quicker, simpler alternative—the Mortality in Emergency Department Sepsis (MEDS) score. This prediction rule, which incorporates nine easily assessed clinical factors, can be used at the bedside to rapidly identify the potentially septic patients at highest risk of death.

The MEDS score was recently validated in a large population at Dr. Shapiro’s institution.[1] The next step will be to validate it at other centers to ensure wide applicability. However, even at its current stage of development, the model may save lives, said the authors of an accompanying editorial.2 By allowing physicians to accurately gauge a patient’s mortality risk in the ED, the MEDS score may increase awareness of the acuity of the patient’s condition, which, in turn, could prompt more appropriate treatment.

VALIDATION PROCESS

To validate the MEDS score, the Harvard group prospectively studied consecutive adult ED patients at risk for systemic infection. Of the 3,301 patients who were eligible for the study, 3,179 were enrolled.

About two thirds of those enrolled were randomized to the “derivation set.” In this group, investigators identified and analyzed nine clinical factors that correlated with 28-day mortality. They then assigned points to each correlate to stratify the derivation set into five risk categories.

The nine correlates (and assigned points) were terminal illness (6 points); age older than 65 (3), tachypnea or hypoxia (3), septic shock (3), low platelet count (3), bandemia (3), nursing home residence (2), lower respiratory infection (2), and altered mental status (2). A score of 0 to 4 points indicated very low risk; 5 to 7 points, low risk; 8 to 12 points, moderate risk; 12 to 15 points, high risk; and 16 or more points, very high risk.

Dr. Shapiro and colleagues then applied the MEDS scoring system to the remaining one third of patients (the validation set). In this group, the actual 28-day mortality for those risk categories was 1.1%, 4.4%, 9.3%, 16%, and 39%, respectively.

In a statistical analysis, the area under the received operating characteristic curve for the MEDS score was 0.82 in the derivation set and 0.76 in the validation set, which suggests a high level of clinical utility. However, an even better indicator of the MEDS score’s clinical utility may be its ease of use, said Dr. Shapiro.

“If you had the calculation for the MEDS score in a handheld computer, you could easily whip up a score in a second,” he explained. “Patients with a higher score require more intensive therapy, such as fluid resuscitation or early antibiotics.”

—Timothy Begany

References
1. Shapiro NI, Wolfe RE, Moore RB, et al. Mortality in Emergency Department Sepsis (MEDS) score: a prospectively derived and validated clinical prediction rule. Crit Care Med. 2003;31:670-675.
2. Rivers EP, Nguyen HB, Amponsah D. Sepsis: a landscape from the emergency department to the intensive care unit. Crit Care Med. 2003; 31:968-969.

Return to table of contents