Article Text
PDF
Abstract
Background Hypocalcaemia is a common metabolic derangement in critically ill patients. Blood transfusion can also contribute to depleted calcium levels. The aims of this study were to identify the incidence of hypocalcaemia in military trauma patients receiving blood products en route to a deployed hospital facility and to determine if intravenous calcium, given during the prehospital phase, has an effect on admission calcium levels.
Methods This was a retrospective review of patients transported by the UK Medical Emergency Response Team in Afghanistan between January 2010 and December 2014 who were treated with blood products in the prehospital setting. Total units of blood products administered, basic demographics, Injury Severity Score and trauma type were collected. Ionised serum calcium levels on admission to hospital were compared between those who received blood products without prehospital intravenous calcium supplemental therapy (non-treatment) and patients who were treated with 10 mL of intravenous calcium chloride (10%) concurrently with blood products (treatment).
Results The study included 297 patients; 237 did not receive calcium and 60 did. The incidence of hypocalcaemia in the non-treatment group was 70.0% (n=166) compared with 28.3% (n=17) in the treatment group. Serum calcium levels were significantly different between the groups (1.03 mmol/L vs 1.25 mmol/L, difference 0.22 mmol/L, 95% CI 0.15 to 0.27). In the non-treatment group, 26.6% (n=63) had calcium levels within the normal range compared with 41.7% (n=25) in those who received calcium. There was a dose response of calcium level to blood products with a significant decrease in calcium levels as the volume of blood products increased.
Conclusion Trauma patients who received blood products were at high risk of hypocalcaemia. Aggressive management of these patients with intravenous calcium during transfusion may be required.
- military
- major trauma management
- prehospital care
Statistics from Altmetric.com
Key messages
What is already known on this subject
Patients transfused with blood products in hospital are susceptible to hypocalcaemia.
Previous work suggests that 2–15 units of blood are needed to produce a drop in calcium.
There has been little research to confirm these findings in the prehospital environment.
What this study adds
This study confirms that trauma patients who are given blood products in the prehospital phase of care are susceptible to hypocalcaemia.
After only 1 unit of blood, calcium levels drop below the lower limit of normal, suggesting that intravenous calcium should be considered in all patients transfused with blood products in the prehospital environment.
Ten millilitres of calcium chloride administered concurrently with blood products in the prehospital environment appears to reduce the incidence of hypocalcaemia.
Introduction
Haemorrhage is the leading cause of preventable trauma deaths on the battlefield, and blood transfusion can be life-saving in these circumstances. Studies have previously explored the use of blood products and coagulopathy,1–5 but there is little evidence exploring the incidence of electrolyte abnormalities, specifically hypocalcaemia in patients who require blood transfusion due to traumatic injury. Two previous studies6 7 have explored this phenomenon in patients admitted to major trauma centres, but neither were performed in the prehospital environment.
The use of blood products in the prehospital environment has been widely implemented.8 Red blood cells (RBCs), fresh frozen plasma (FFP) and platelets contain approximately 3 g of citrate per unit that binds with calcium, lowering the ionised plasma calcium concentration.9 Intravenous calcium is sometimes, but not consistently, given as an adjunct to blood transfusion. The governance relating to this is generally ad hoc, based on personal experience and anecdotal discussion without any strong evidence base.
The role of calcium is multifaceted; it has an effect on clotting and platelet adhesion as well as contractility of myocardial and smooth muscle. The citrate contained in stored blood products would normally be insignificant, as a healthy patient would rapidly metabolise this in the liver to prevent hypocalcaemia occurring. This would not be the case with a patient who is suffering from hypovolaemic shock as the combination of infused blood products and hepatic dysfunction due to hypoperfusion may impair liver function.10 The use of supplemental calcium is becoming widespread within the EDs of major trauma centres and trauma units and in some establishments it is now part of the major haemorrhage treatment protocol. However, there is limited appetite at present to initiate calcium treatment concurrently with blood product administration in the prehospital setting.
Up to 4 units of RBCs and 4 units of FFP (total 8) can be transfused during the prehospital period (according to military practice) and thus patients can be exposed to up to 24 g of citrate prior to arrival at hospital. Due to the limitations of point-of-care (POC) testing equipment used in the prehospital environment, ionised calcium (iCa) measurement does not occur until arrival at the medical treatment facility. Currently, military prehospital standard operating procedures and some civilian establishments11 recommend calcium chloride as the preferred choice of calcium salt. The military treatment regimen advocates the use of intravenous calcium after 2 units of blood product. However, the use of intravenous calcium when administering blood products is still determined by the clinician.
The aims of this study were therefore to identify the incidence and severity of hypocalcaemia in military trauma patients receiving blood products en route to a deployed hospital facility and to determine if intravenous calcium, given as a supplemental therapy during the prehospital phase, has an effect on admission calcium levels.
Methods
A retrospective database review using the UK Joint Theatre Trauma Registry (JTTR) was performed in order to identify patients who had been treated with blood products en route to a military medical treatment facility. Inclusion criteria were all patients transported by the UK Medical Emergency Response Team (MERT) between January 2010 and December 2014 in Afghanistan. The study population included UK and coalition military, Afghan security forces and local nationals. A retrospective chart review was then undertaken involving anonymised manual data collection from scanned medical notes at the Central Health Records Library, a Ministry of Defence establishment where military health records are stored. Data collected included basic demographics, Injury Severity Score (ISS), trauma type, baseline vital signs on arrival at the ED and iCa results. These results were further examined in order to identify two cohorts; patients who had blood products without prehospital intravenous calcium supplemental therapy (non-treatment) and patients who were treated with 10 mL intravenous calcium chloride minijet (10%) concurrently with blood products (treatment). Subgroup analysis was performed on a small sample of patients who did not have any fluid resuscitation (blood products or crystalloid) on MERT (figure 1) to compare against patients receiving blood products. Patients were excluded if they did not have an iCa recorded within 30 min of arrival at hospital, had intravenous calcium for treatment other than supplementary to blood products (eg, during cardiac arrest) or if iCa results were unobtainable. The sample size was based on the total number of patients transported throughout the study period.
Consort diagram. JTTR, UK Joint Theatre Trauma Registry; iCa, ionised calcium; IV, intravenous; MERT, UK Medical Emergency Response Team.
Ionised hypocalcaemia has been defined as serum iCa <1.12 mmol/L;6it has been further defined as severe at levels <0.9 mmol/L, and this has been associated with increased mortality in critically ill adults.12 Levels <0.8 mmol/L have been associated with adverse cardiac effects, and Kraft et al 13 proposed 0.9 mmol/L as a trigger for supplementary calcium therapy. In line with Giancarelli et al’s recommendation, a working definition of hypocalcaemia in this study was a blood assay of <1.12 mmol/L.6 This lower limit of normal is also the recommended limit for the i-stat blood monitoring system14 used at the time of the study, which recommended a normal value range of 1.12–1.32 mmol/L.
All statistics were analysed using SPSS V.23 (IBM). Due to the non-normal distribution of the data, non-parametric Spearman’s correlation coefficient was used to assess relationships between variables. Mann-Whitney and Kruskal-Wallis tests are used to explore differences between groups when a distinction can be meaningfully made between the independent and dependent variable. Data are presented as medians with IQRs for continuous variables or as number (n) and percentage for categorical variables. Two-tailed tests were used to determine statistical significance and a P value of <0.05 was considered significant. A Hodges-Lehmann test was performed to determine the CI between the main outcomes. The study was registered as a service evaluation at the Royal Centre for Defence Medicine (RCDM/Res/Audit/1036/17/0472); due to the nature of the study, ethical review was not required.
Results
A total of 450 patients were identified as having blood products on MERT during the study period, and 297 were included in the final analysis. One hundred and fifty-three patients were excluded (figure 1). The treatment group consisted of patients having had supplemental intravenous calcium (n=60) and the non-treatment group was categorised as patients whom were managed without intravenous calcium (n=237). Patients were administered FFP and RBCs within the transport phase of treatment with a median total blood product administration of 4 units in both groups (table 1). The overall incidence of hypocalcaemia in the non-treatment group was 70.0% (n=166), compared with 28.3% (n=17) in the patients treated with intravenous calcium with a significant difference (P<0.001) between the two groups (1.03 mmol/L vs 1.25 mmol/L, difference 0.22 mmol/L, 95% CI 0.15 to 0.27). In addition, 26.6% (n=63) had normal iCa levels in the non-treatment group compared with 41.7% (n=25).
Study population characteristics
Blood product administration in the non-treatment group was inversely related to calcium levels (figure 2). Further to this, an independent samples Kruskal-Wallis test indicates a significant difference in the distribution of first calcium levels by number of blood products (P<0.001). The small sample of patients who did not have blood products on MERT had a higher initial calcium level (figure 2) with a score of 1.2 (0.8–1.4) for this subgroup.
Blood product administration (in units) in the non-treatment group in relation to first recorded serum calcium measurement. Dotted line signifies lower limit of reference range (1.12 mmol/L).
Discussion
This study is the first to explore the relationship between hypocalcaemia and blood product transfusion in patients who have received blood in the prehospital phase of treatment. It demonstrates that patients who are treated with blood products in this phase of care are susceptible to hypocalcaemia and that after just 1 unit (blood product), the calcium levels drop below the normal range. This study also suggests that supplemental therapy with 10 mL of intravenous calcium chloride given concurrently with blood products reduces this risk. This supports the findings of previous in-hospital studies that have demonstrated that blood product administration reduces calcium levels.6 7 These studies disagree on the cut-off point with Giancarelli et al advocating that over 5 units of blood is needed to produce a clinically relevant reduction,6 while others argue that even 1 unit of blood reduces calcium levels below the accepted normal value, consistent with our findings.7
A significant difference was seen between patients who had no blood products and those who were given blood. Lier et al have suggested that an ionised calcium of <1 mmol/L significantly increases patient mortality regardless of ISS, and <0.88 increases mortality threefold.12 Our study suggests that 1 unit of blood would reduce calcium levels below the lower limit of the normal range of 1.12 mmol/L, while 2 units of blood would reduce calcium levels to <1 mmol/L and 5 units of blood to <0.88 mmol/L.
Giancarelli et al 6 postulated that a threshold level of 0.9 mmol/L of iCa should be used to start calcium therapy. This is a useful strategy in hospital with accurate testing capability, but in the prehospital environment, without POC testing facilities, this cut-off is of no practical use. Based on the evidence to date, it would seem clinically sensible to suggest that intravenous calcium is commenced while giving blood products in the prehospital phase of treatment, particularly if there is a prolonged prehospital phase.
This study also explored the effect of intravenous calcium on trauma patients who had blood products. This subgroup was given 10 mL (1 g or 13.2 mEq) of 10% intravenous calcium (via minijet) en route to hospital (according to military protocols), and when compared with the non-treatment group, iCa was significantly higher, suggesting that the treatment used in this cohort had a therapeutic effect.
There are limited studies regarding the dose response of calcium supplementation needed after administration of blood products.6 Elmer et al suggest the use of 2 g of calcium gluconate for every 2 to 4 units of blood,9 and another13 suggested a strategy of 290 mg of calcium gluconate or 96 mg of calcium chloride for every 100 mL of blood received, although the effects of these two strategies have not been reported. In another study, patients were given 1 g of calcium gluconate (4.65 mEq elemental calcium),15 and this did not produce a therapeutic effect of increasing calcium levels above the normal threshold of 1.12 mmol/L. Studies by Dickerson et al examined the replacement of calcium gluconate in critically ill trauma patients; they found that 4.65 to 9.3 mEq of elemental calcium (1.4 g of calcium chloride) was adequate to produce a therapeutic effect.16 17 Giancarelli et al’s study cohort received 2 g of calcium chloride after 4 units of blood transfusion, and their findings supported the use of calcium chloride as the preferred salt with a dose of 2 g or more;6 this is three times the recommendation of Elmer et al and twice that prescribed in our study. There does not appear to be any rationale regarding the choice of salts, with the majority of guidance suggesting calcium chloride as the preferred choice. This may be due to the increased amount of calcium contained in the chloride solution compared with gluconate (13.6 mEq vs 4.65 mEq).
A recently published study18 with similar patient characteristics to ours demonstrated a mean iCa level of 1.1 mmol/L in patients who received up to 5 units of RBCs in the prehospital environment. This compares favourably with the study by Webster et al,7 who provided evidence of a similar mean calcium level following transfusion in the ED (mean 0.95 mmol/L).
Civilian and military prehospital emergency medical services that use blood products are starting to include intravenous calcium as an adjunctive therapy; the massive haemorrhage protocols for MERT have recently changed to include 10 mL intravenous calcium chloride minijet (10%), but there is still some discussion regarding when it should be commenced. Calcium therapy is normally started when a patient’s levels drop below 1 mmol/L. Based on studies to date,6 7 18 2 units of blood would be required to meet this criteria, but it could be argued that in the prehospital environment with no POC testing available, a simpler approach would be to administer intravenous calcium concurrently after 1 unit of blood.
Limitations
This study has certain inherent limitations due to the nature of a retrospective database review. The analysis of military injury data can be difficult when the data are collected in the often-austere conditions of the battlefield. Limitations of this study are inherent in the data contained within the JTTR. The number of patients included in this study was small, and it is therefore difficult to ascertain whether interventions such as intravenous calcium supplementation had a survival benefit. Twenty-eight per cent of cases (n=126) were excluded due to lack of complete data on the JTTR and in clinical notes.
Conclusions
This study has shown that the prehospital administration of blood and blood products is associated with hypocalcaemia on admission to the hospital, and that this can be ameliorated by the administration of 10 mL of intravenous calcium in the prehospital phase. The study also found that calcium levels dropped below the normal range after 1 unit of blood. These findings add weight to a growing body of evidence suggesting that adjunctive calcium therapy during blood transfusion may have the required effect of increasing serum calcium levels. A prospective randomised controlled trial is now required to definitively answer this research question.
Acknowledgments
The Royal Centre for Defence Medicine Clinical Information Exploitation Team, Defence Statistics Health and RCDM Blood Supply Team are thanked for collecting, collating and identifying the appropriate data for this paper. We would also like to acknowledge all those individuals who have been involved in data entry to the Joint Theatre Trauma Registry on deployed operations. The authors would finally like to thank Lt Col Paul Hunt for his comments on the manuscript.
References
Footnotes
Contributors The following individuals have contributed to the article: TK—contributed on all aspects of the article including literature search and review, data collection, data analysis and write-up, and is the guarantor for the overall content of the manuscript. IG—contributed to writing the initial and subsequent versions of the manuscript. AB—contributed to data collection and writing the manuscript. VW—statistician involved in all elements of statistical analysis. MB—biomedical scientist involved in writing initial and subsequent versions of the manuscript. JS—research adviser involved in project supervision, writing and editing the manuscript.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Additional unpublished data from the study are only available to the research team and are stored on a password-protected computer.