817

Abstract

Introduction: The pathophysiology of critical care unit acquired anemia is not well understood. The ubiquitous drop in hemoglobin concentration (Hb) seen in critical care units is likely multifactorial and the pattern is not well described. The goal of our study is to better characterize the phenomena of early critical care anemia and to identify an expected Hb change. This may be useful to prevent unnecessary evaluation for hemorrhge. Hypothesis: An expected Hb drop can be identified in critical care patients. Methods: This is a retrospective, observational study of critical care unit patients admitted to a 14 bed ICU over a 3-month period. All patients in the cohort were admitted to the ICU within 48 hours of hosptial admission. Patients with age < 18; hemodynamic instability; active blood loss; use of vasopressors; or history of documented chronic anemia were excluded. The mean and standard deviation of key descriptors of anemia (Hb, Hct, MCHC) at 6, 12, 24, 36, 48 hrs and day 10 in the ICU were recorded and compared. Results: Thirty five out of 52 consecutive patients met inclusion criteria. The mean admission (baseline) Hb was 13.1 g/dL (SD 2.1). The mean fluid volume infused over the first six hours of ICU admission was 2.0 L, 1.8 L from 7-12 hrs, and 2.5 L from 13-24 hrs. The mean cumulative fluid infused by the end of 24 hrs was 6.4 L. The greatest rate of decline in Hb consistently occurred within 24 hrs of ICU admission. The mean nadir Hb and hct levels occurred at 48 hrs after ICU admission and were 10.2 (SD 1.7) and 30.2 (SD 5.4) respectively despite continuous IV fluid input. After 48 hrs, Hb levels began to rise and eventually stabilized throughout the 10 day ICU period but did not return to baseline levels. The mean drop in Hb and hct during the first 24 hrs were 2.2 g/dL and 7.4% respectively. Mean Hb, hct, and MCHC levels closely followed each other throughout the 10 day period. Conclusions: There was a mean drop of 2.2 g/dL in Hb that occurred within the first 24 hours of ICU admission and a maximum drop of 2.9 g/dL. This drop is consistent and predictable and need not be interpreted as an indicator of hemorrhage in the absence of other signs.