Prediction of outcome for critically ill patients with unexplained hypotension.

OBJECTIVES

To determine the clinical variables that affect the prognosis of critically ill patients with sustained unexplained hypotension. A further goal was to develop a prognostic scoring system based on clinical data available at the onset of hypotension.

DESIGN

Prospective cohort study.

SETTING

The intensive care units (ICUs) of an academic medical center.

PATIENTS

One hundred one adult ICU patients with sustained (> 60 mins) unexplained hypotension. Using the initial 50 patients (derivation set), a prognostic score was developed that was then tested in the next 51 patients (validation set).

INTERVENTIONS

None

MEASUREMENTS AND MAIN RESULTS

The main outcome variable was death or hospital discharge. The overall hospital mortality in the combined sets was 58%. Using a multivariable model we identified three independent (p < .05) predictors of hospital mortality, including the Acute Physiology and Chronic Health Evaluation (APACHE) II score at the time of hypotension, the time from hospital admission to hypotensive episode, and hospital admission for surgery or treatment of malignancy. These variables were weighted and combined to create a Hypotension Score which separated patients in the combined sets into three prognostic groups: a) Hypotension Score of < 40, mortality 7%, (n = 27); b) Hypotension Score of 40 to 64, mortality 70%, (n = 50); and c) Hypotension Score of > or = 65, mortality 92%, (n = 24). The area under the receiver operating characteristic curve was .85 for the derivation set and .83 for the validation set vs. .76 for the APACHE II score alone.

CONCLUSIONS

The prognosis of hypotension in the critical care setting is highly variable, but can be predicted from patient characteristics.