Purpose: Injured patients who require aggressive resuscitation with intravenous (IV) fluids and blood products will frequently acquire low levels of serum calcium (CA) and albumin (ALB) in the intensive care unit (ICU) as result of this therapy. The purpose of this longitudinal study was to determine the time course of CA and ALB during ICU admission in survivors (S) compared to nonsurvivors (N) after major trauma. The study design is to verify if CA, ALB, or albumin-corrected CA can be used as indicators of patient survivability after critical injury.
Materials and methods: CA and ALB values were retrospectively recorded in 64 random subjects (S= 32 and N= 32) admitted to the Trauma ICU for 3 or more days. CA and ALB data points were partitioned into 6 time frames of ICU care. Mean values and standard error of the mean for each frame were obtained to depict parametric differences in the time profiles for S versus N. Subgroup analysis was used to determine the impact of blood transfusions on CA and ALB levels. Albumin-corrected CA was computed for every patient at each measurement point and then partitioned into the 6 time frames of ICU care. Parametric t-test and nonparametric rank sum analysis were used to evaluate the ability of CA, ALB, and ALB-corrected CA at discriminating S from N. Each predictive covariate was ranked, divided into quartiles (grades = normal, mild, moderate, severe), and correlated with patient survival likelihood (viz., ratio of S to N in each quartile).
Results: Parametric and non-parametric analysis of collected data indicates that the response patterns of CA were significantly different ( P<.00005 ) in S versus N. Time profiles of CA and ALB exhibited similar reductions in both S and N during the resuscitation phase (ie, "hypocalcemia of trauma"). But from these nadir points, CA response patterns in S tended to steadily elevate toward normal levels (ie, "responders"), while N exhibited no such increase in CA values (ie, "nonresponders"). Data revealed that survival likelihood in trauma patients after 3 ICU days is proportional to the upward response of CA from depressed values present after the initial resuscitation. Decreased CA levels after 3 ICU days were associated with decreased survival (Table 1). Rank sum testing showed that values of CA corrected for ALB creates less obvious difference in S and N than uncorrected CA. Subgroup analysis showed a linear decrease in CA and ALB levels with increasing units of blood transfused during treatment for trauma.
Conclusions: CA changes during ICU care demonstrate distinct response patterns (P <.00005) for survivors versus nonsurvivors. The magnitude of upward response in CA after the fluid resuscitation phase is a marker that correlates with a patient's ability to withstand the physiologic stresses encountered during ICU treatment after major trauma. Our findings indicate that uncorrected CA values are a better guide for calcium replacement therapy in trauma patients than albumin-adjusted CA. This study suggests that response patterns of CA can be a useful reference to aid in monitoring the progress of critically injured patients.