Christopher Ray, MD
Fellow Physician, PGY-7
Childrens National Health Systems
Washington, District of Columbia
Disclosure information not submitted.
Murray Pollack, MBA, MD, MCCM
M.D.
Childrens National Health Systems, United States
Disclosure information not submitted.
Anita Patel, MD
MD
Childrens National Health Systems, United States
Disclosure information not submitted.
Title:
The Lactate-Albumin Ratio as a Marker of In-Hospital Mortality in Critically Ill Pediatric Patients
Introduction: Serum lactate and albumin levels are independent risk factors for mortality during critical illness. The lactate-albumin ratio (LAR) was recently described in adult literature as a more discriminative predictor of mortality than albumin or lactate alone in ICU patients. The aim of this study was to evaluate the LAR’s ability to discriminate in-hospital mortality in a retrospective pediatric patient population, versus admission lactate or albumin levels alone.
Methods: The Cerner HealthFacts® database was queried for pediatric ICU admissions between January 2009 and June 2018. Inclusion criteria were ages 30 days to 22 years, and measurement of both a serum lactate and albumin level within 6 hours of admission. Patients were excluded if they received Lactated Ringers or albumin infusions during their admission. LAR was derived from admission values for lactate and albumin. Admission lactate, albumin, and LAR were compared between survivors and non-survivors using the student’s t-test. Logistic regression and associated areas under the receiver operating characteristic curves (AUROC) were used to assess the relative discrimination of the three associations: lactate and mortality, albumin and mortality, and LAR and mortality. All results except for AUROC are presented as mean±standard deviation.
Results: A total of 1,318 patient encounters were included with age of 12.9±0.2 years. The patient sample was 54% male, 63% Caucasian, and 16% African American, with hospital length of stay 4.3±0.2 days, and in-hospital mortality rate of 9.0% (n=118). Lactate level (mmol/L) was significantly higher in non-survivors vs. survivors (8.01±0.54 vs. 2.68±0.08, p< 0.05) and albumin level (g/dL) was significantly lower in non-survivors vs. survivors (3.21±0.07 vs. 4.10±0.02, p< 0.05). LAR was significantly higher in non-survivors vs. survivors (2.78±0.22 vs. 0.71±0.03, p< 0.05). Discrimination assessed by AUROC indicated that the highest discrimination of in-hospital mortality was for LAR (0.848), followed by lactate (0.812) and albumin (0.804).
Conclusions: LAR may better discriminate mortality vs. admission lactate or albumin levels alone in pediatric ICU patients. Selection bias is possible for patients who had a serum lactate and albumin within 6 hours of ICU admission and may have influenced these results.