Use of Physician-Estimated and Patient Self-Reported Weights to Guide Initial Fluid Resuscitation in Emergency Department Patients With Suspected Sepsis

Background: Knowledge of patient weight is required to guide initial intravenous fluid therapy for patients with sepsis-associated hypotension or elevated lactate. Previous studies have shown patients are better estimators of their weight than medical providers are; critically ill patients, however, may be unable to provide this information. Objectives: This study compares the accuracy of physician-estimated and patient self-reported weights to subsequent inpatient bed/stretcher scale weights for guiding initial protocol-based intravenous fluid therapy in the treatment of emergency department patients with suspected sepsis. Methods: Adult patients presenting with a suspected diagnosis of severe sepsis to a large, urban, academic emergency department had either physician-estimated or patient self-reported weights recorded on presentation. All patients had subsequent inpatient bed/stretcher scale weights recorded on the first day of hospitalization. Results: Physician-estimated and patient self-reported weights linearly correlated (P <.001) with inpatient bed/stretcher scale weights. Median accuracy error for physicians (5.4% [2.0-10.1]) and patients (3.9% [1.6-6.4]) was not significantly different (P =.28). Physician-estimated and patient self-reported weights accuracy was determined at multiple levels: within 5% (46%, 57%, respectively), 10% (75%, 90%), 15% (90%, 95%), and 20% (100%, 95%) error tolerances, as well accurate estimates within 5 kg (69.2%, 70.0%). Conclusions: Both physician-estimated and patient self-reported weights are reliable when calculating initial protocol-based intravenous fluid resuscitation for emergency department patients with sepsis.