Variation in crystalloid administration: an analysis of 6248 patients undergoing major elective surgery.

BACKGROUND

Large variations exist regarding the type and volume of fluid to be administered to patients. This study aimed to quantitate variations in the administration of crystalloid fluids at the level of the patient, provider, and procedure at a large, tertiary care center.

METHOD

Patients who underwent major cardiac, thoracic, or abdominal procedures between 2011 and 2014 were identified. Variations in crystalloid administration were compared by procedure and provider using a coefficient of variation (CV). Multivariable hierarchical linear modeling was performed to identify factors predictive of fluid administration and quantitate variation at the level of the patient and provider.

RESULTS

Among 6248 patients who met inclusion criteria, the average crystalloid volume was 25.8 mL kg(-1) m(2) h(-1), corresponding to a CV of 55%. Patients who underwent pancreatectomy received the highest corrected crystalloid volume (32.7 mL kg(-1) m(2) h(-1)), whereas those who underwent coronary artery bypass grafting received the lowest corrected crystalloid volume (14.7 mL kg(-1) m(2) h(-1)). Variations in fluid practices were noted between providers (corrected CV; 14.7%-97.1%) and within the practices of the same provider (corrected CV range; 24.1%-87.9%). On multivariable analysis, age and changes in hemoglobin concentration were associated with a higher crystalloid volume (both P < 0.05). Although over 90% of the variation was attributed to patient-level factors, approximately 10% was due to factors at level of the provider (surgeon: 5.8% versus anesthesiologist: 3.4%).

CONCLUSIONS

Wide variations were noted in crystalloid administration between procedures, providers, and within providers. Evidence-based practices and goal-directed therapies should be incorporated to avoid unwanted variations.