Too Little, Too Late? Analysis of Calcium Administration and Outcomes in Trauma Patients with Hemorrhagic Shock

Loading...
Thumbnail Image
Penn collection
Interdisciplinary Centers, Units and Projects::Center for Undergraduate Research and Fellowships (CURF)::Fall Research Expo
Degree type
Discipline
Adult and Continuing Education
Subject
Transfusion
Trauma
Calcium
Funder
Grant number
Copyright date
2025-10-16
Distributor
Related resources
Author
Yendru, Avi
Mani, Ashika
Anandalwar, Seema
Contributor
Abstract

Introduction: Trauma induced coagulopathy is a significant contributor to mortality in patients with hemorrhagic shock. Although the benefits of calcium in all aspects of coagulation are well known, there remains a lack of standardized guidelines regarding the timing and indications for calcium administration. This study aimed to characterize current calcium administration practices in trauma patients with hemorrhagic shock and assess its associations with hemorrhage control and mortality.

Methods: Retrospective, single-center cohort study at a Level 1 trauma center of trauma patients > 18 years of age requiring at least one transfusion within the first two hours of arrival. Baseline demographics, injury severity markers, and calcium administration timing and dosage were collected. Outcomes included 24-hour mortality, 24-hour transfusion requirement, and time to hemorrhage control. Patients were categorized based on calcium administration and timing. Propensity score weighting was used to adjust for differences in illness severity, and outcomes were compared using multivariable regression.

Results: 502 patients were included in the analysis, of which 287 (57.2%) received calcium and 215 (42.8%) did not. Of those who received calcium, 143 (49.8%) received calcium within 60 minutes of arrival. Average time to calcium administration was 114.7 (SD 186.9) minutes. Calcium was more frequently given to patients with higher ISS (23.9 vs 14.7, p<0.001), lower pH (7.2 vs 7.3, p<0.001), and higher base deficit (8.7 vs 5.7, p<0.001), suggesting selective use in more critically ill patients. This effect was more pronounced in patients receiving calcium within the first 60 minutes. In unadjusted analysis, calcium administration was associated with significantly higher 24-hour transfusion requirements (12.4 units vs 2.0 units, p<0.001) and 24-hour mortality (14.6% vs 3.3%, p<0.001). After propensity weighting, mortality differences were no longer significant and even trended towards a higher mortality in the patients who did not receive calcium (11.5% vs 14.5%, p=0.70). However, transfusion requirements remained higher in the calcium group (11.0 units vs 2.3 units, p<0.001), possibly reflecting residual confounding due to clinician-assessed severity not captured in available variables. In a subgroup analysis of patients with an injury severity score (ISS) </= 15, mortality (6.6% vs 3.7%, p=0.5) and time to hemorrhage control (321 minutes vs 320 minutes, p=0.96) were similar between groups, despite increased transfusion requirement in the calcium group (7.1 units vs 1.6 units, p<0.001).

Conclusion: Calcium is currently administered to the most severely injured trauma patients, likely in response to physiologic derangement and clinical gestalt. Even after adjusting for objective markers of severity, patients receiving calcium required more transfusions, suggesting residual clinical-driven triage not captured in models. Importantly, mortality differences were mitigated after adjustment, particularly among less severely injured patients, raising the possibility that broader calcium use may offer benefit. These findings support further investigation into the effects of broadening calcium administration beyond just the sickest patients.

Advisor
Date of presentation
2025-09-15
Conference name
Conference dates
Conference location
Date Range for Data Collection (Start Date)
Date Range for Data Collection (End Date)
Digital Object Identifier
Series name and number
Volume number
Issue number
Publisher
Publisher DOI
Journal Issue
Comments
This project was supported with funding from the Penn Undergraduate Research Mentoring (PURM) program.
Recommended citation
Collection