Severe Burn Injuries – The Day the Sodium Starts Rising

Abstract

Background/Aim

The current study was designed to evaluate the etiologies of hypernatremic episodes in patients with severe burn injuries in comparison to critically ill non-burn patients.

Patients and Methods

The retrospective data acquisition was limited to the first 14 days and to patients with at least 20% total body surface area (TBSA) 2^nd degree burn injuries or more than 10% TBSA when including areas of 3^rd degree burn injuries. The results were compared to the results of a previously published study that analyzed the risk factors for hypernatremia in 390 non-burn intensive care unit patients.

Results

In total, 120 patients with a total of 50 hypernatremic episodes were included. Compared to non-burn injury patients, no significant differences were detectable except for a lower rate of hypokalemia and a higher rate of mechanical ventilation. The main trigger for hypernatremic episodes was the loss of free water, while 24% of the hypernatremic episodes seemed to be at least partly triggered by a surplus sodium influx. Patients with hypernatremic episodes had a significantly higher mortality rate. However, in none of the cases was hypernatremia the decisive cause of death.

Conclusion

Besides the unique phenomenon of high volume internal and external volume shifts, the overall risk factors and etiologies of hypernatremia in patients with severe burn injury do not seem to significantly differ from other ICU patient collectives. Remarkably, a surplus of sodium influx and therefore a modifiable factor besides the specific burn injury volume resuscitation had an impact on the hypernatremic episodes in 24% of cases.

Hypernatremia represents a major clinical burden among critically ill patients and seems to be associated with an increased mortality in this patient collective (1). A patient subgroup which seems to be especially susceptible for this complication are patients with third degree burn injuries and high percentages of burned skin (2-4). Considering the pathophysiological processes associated with this disease, like high volume internal and external fluid shifts, high rates of infection and renal dysfunction, this observation does not seem surprising. The relevance of this complication was demonstrated in previous studies, which identified serum sodium (s-Na) levels above 150 mmol/l as an independent risk factor for higher mortality rates in critically ill patients (5,6). Accordingly, it seems crucial to identify modifiable risk factors for this complication in order to optimize the treatment of patients with severe burn injuries. For non-burn injury patients these risk factors were already identified in several previous studies, among those the study of Hoorn et al., which analyzed the risk factors for hypernatremia in a retrospective study of 390 non-burn injury patients (7). However, taking the unique pathophysiology of severe burn injuries into account, it is questionable whether those results are also applicable to this specific patient subgroup. Therefore, the current study was designed to verify the transferability of these results to patients with severe burn injuries by analyzing the clinical and biochemical course around the day the s-Na started rising in a cohort of 120 patients with severe burn injuries.

Patients and Methods

After approval from our institutional review board, we retrospectively reviewed all admissions to our burn center from 01/2017 till 06/2023. All patients aged >18 years who suffered from at least 20% total body surface area (TBSA) 2nd degree burn injuries or more than 10% TBSA when including areas of 3rd degree burn injuries were screened. The review of the electronic medical records included the complete stay at the intensive care unit; the evaluation of the data was limited to the first 14 days.

After identifying all episodes with a S-Na above 145 mmol/l, the following data were recorded from day -2 till day +1 with day 0 being defined as the first day the sodium level reached a level above 145 mmol/l: age, sex, TBSA, inhalation trauma, percentage of 3rd degree burn injury, Abbreviated Burn Severity Index (ABSI), medication, electrolytes, blood sugar, serum creatinine/urea, serum albumin, serum lactate, sepsis-related organ failure assessment (SOFA) score, ventilation parameters, and fluid balances. All data were carefully recorded in an automated data management system by experienced nursing staff. The blood-gas parameters were documented as mean values of all available blood-gas-analyses, which varied in number from 3 to 12 per day and were all carried out on the same blood gas analyzer (ABL800 BASIC, Radiometer).

Lower and upper limits were defined as follows: Hyperlactemia (>2 mmol/l), hypercalcemia (>1.29 mmol/l), hyperglycemia (>160 mg/dl), serum urea (>40 mg/dl), leukocytes (4-10 *1,000/μl), respiratory insufficiency (Horowitz indices <200), hypercapnia (>45 mmHg), hypokalemia (<3.5 mmol/l), hypoalbuminemia (<2.5 g/dl), elevated base excess (>2), elevated CPR (>5 mg/dl), rise of the SOFA Score ≥2 points and of the serum creatinine ≥0.3 mg/dl within 48 h. Due to the massive internal and external fluid shifts during the early stages of the burn disease the urine production is not a reliable indicator when evaluating the renal function and therefore, was not taken into account. When comparing our data to the results of Hoorn et al. the limits were accordingly adjusted (7). The population was subdivided for further analyzes based on the time-point and the level of sodium increase (≤7 days vs. >7 days; 145-150 mmol/l vs. >150 mmol/l).

Statistical analysis was performed by using GraphPad Prism 5 (GraphPad Software Inc., La Jolla, CA, USA) and SPSS (SPSS Statistics 29, IBM, Armonk, NY, USA). The data collected were analyzed by using the Mann-Whitney U-Test and the Fisher’s Exact Test.

Results

Patient characteristics. A total of 120 patients met the inclusion criteria. Out of these 120 patients, 52 patients (43.33%) developed a hypernatremia of 145-150 mmol/l (Na145) and 33 patients (27.5%) a hypernatremia >150 mmol/l (Na150) during their complete stay at the intensive care unit. Limiting the analyses to the first 14 days in order to focus on the impact of the severe burn injuries on those hypernatremic episodes, we identified 16 patients (13.33%) with a total of 19 episodes of a hypernatremia between 145-150 mmol/l and 27 patients (22.5%) with 32 episodes of a hypernatremia >150 mmol/l. Thirty six hypernatremic episodes occurred during the first seven days after the burning, 14 episodes from day eight till 14. Accordingly, seven patients had hypernatremic episodes during both time periods, the first seven days and from day eight till 14. Only one patient presented hypernatremia already on admission (1/43 patients; 2.33%); this hypernatremic episode was excluded from further analyses. The median duration of the hypernatremic episodes was in the Na145 group five days and in the Na150 group 10 days. The age and sex distribution showed no differences between the Na- and the Non-Na groups. The TBSA, Abbreviated Burn Severity Index (ABSI), mortality rate, number of inhalation trauma and tracheotomy and frequency and amount of 3˚ burning were significantly higher in the Na-group compared to those in the Non-Na counterparts (Table I). The overall mortality rate was 10.83% (13/120) whereas the mortality rate in the Na-group reached a rate of 20.93% (9/43). Of those nine patients only six patients were hypernatremic when they died (6/120 patients; 5%). Causes of death in these patients were: 2× septic shock, 1× thromboembolic cerebral ischemia, 1× abdominal compartment syndrome, 1× acute respiratory distress syndrome, 1× palliative treatment approach due to a patient decree. None of the casualties were associated with a drop in the s-Na >6 mmol/l/24 h, generalized seizures or a radiological confirmed brain swelling.

Table I. Patient characteristics subdivided into three groups: all patients, patients without any hypernatremia during the first 14 days, and patients with hypernatremia (>145 mmol/l) during the first 14 days.

The Fisher Exact Test was used for all nominal variables, for all other variables the Mann-Whitney U-test was used. A p-value <0.05 was considered as significant. Significant p-values are shown in bold.

Burn injury vs. non-burn injury hypernatremia. Hoorn et al. identified the following possible causes for hypernatremia >150 mmol/l: hypokalemia, hypercalcemia, renal dysfunction, hypoalbuminemia, and hyperglycemia. Further risk factors included the application of sodium bicarbonate and mannitol as well as sepsis as reasons of admission and mechanical ventilation (7). Except for a higher rate of mechanical ventilation and a lower rate of hypokalemia no relevant differences were detectable between the burn injury collective and the case collective of Hoorn et al. (7) (Table II). The higher frequency of hypoalbuminemia in the burn injury group can been neglected since the recommended lower limit of albumin for patients with severe burn injuries is 2.5 g/dl in the German guidelines, while the lower limit in the Hoorn et al. collective was defined as <3.5 g/dl. Adjusting the cut off for hypoalbuminemia to 2.5 g/dl in the burn injury group changes the difference between the Hoorn et al. collective and the burn injury collective into “not significant”. Except for the frequency of hyperglycemia, all risk factors were significantly higher in the burn injury cases group compared to the Hoorn et al. control group (7) (Table II).

Table II. Potential factors contributing to hypernatremia. Comparing burn injury patients to the non-burn injury data of Hoorn et al. (7). Patients with hypernatremic episodes above 150 mmol/l were defined as “cases”, patients without hypernatremic episodes as “controls”.

The Fisher Exact Test was used for all nominal variables. For all other variables the Mann-Whitney U-test was used. A p-value <0.05 was considered as significant. Significant p-values are shown in bold.

Subgroup analyses. The ABSI score was significantly higher in the >150 mmol/l group whereas no significant differences were detectable when comparing the subgroups ≤7 days vs. day 8-14 regarding the ABSI score. Otherwise, no differences were detectable concerning the frequency of hyperlactemia, hypercalcemia, hyperglycemia, serum urea, leukocytes, Horowitz indices <200, hypercapnia, hypokalemia, hypoalbuminemia, elevated base excess values, elevated CPR values, age and TBSA as well as the rise of the serum creatinine ≥0.3 mg/dl and of the SOFA Score ≥2 points within 48 h on day -1 and day 0 between the ≤7 days vs. >7 days and the 145-150 mmol/l vs. >150 mmol/l groups (Table III and Table IV).

Table III. Subgroup analyses after dividing the data into hypernatremic episodes from day 1 till 7 vs. day 8 till 14.

The Fisher Exact Test was used for all nominal variables. For all other variables the Mann-Whitney U-test was used. A p-value <0.05 was considered as significant.

Table IV. Subgroup analyses after dividing the data in hypernatremic episodes from 145-150 mmol/l vs. >150 mmol/l.

The Fisher Exact Test was used for all nominal variables. For all other variables the Mann-Whitney U-test was used. A p-value <0.05 was considered as significant. Significant p-values are shown in bold.

Loss of free water. The mean rise of sodium from day -1 till day 1 was 4.13%, the mean rise of the chloride 3.95%. Therefore, a difference between the percentual rise of sodium and chloride of more than 2% was defined as an “isolated” sodium rise, indicating a surplus sodium influx as part of the hypernatremic etiology. Overall, 12 episodes (24%) met this criterion with an equal distribution within the four subgroups (≤7 days: 8/36, 22.2%, >7 days: 4/14, 28.6%; 145-150 mmol/l: 4/18, 22.2%, >150 mmol/l: 8/32, 25%; Fisher exact test ≤7 days vs. >7 days and 145-150 mmol/l vs. >150 mmol/l: ns). The “combined” increase in sodium and chloride in 76% of the hypernatremic episodes implicates the loss or shift of free water as a major reason for the rise in sodium. This interpretation was confirmed by an increase in the hematocrit in 40/50 episodes (80%) from day -1 to day 0.

Base excess. Another noticeable finding was a high proportion of positive base excess values during the hypernatremic episodes (25/50, 50%). Twenty-four out of this elevated values were based on a metabolic alkalosis or a partially compensated metabolic alkalosis. Possible causes for a metabolic alkalosis are emesis, hypochloremia, hypokalemia, and hypovolemia. Since none of the 25 patients had emesis, hypochloremia or hypokalemia, the most likely etiology of these metabolic alkaloses is hypovolemia, which further supports the previous findings.

Discussion

The current study was designed as a supplement to the retrospective study of Hoorn et al., who analyzed the risk factors and mechanisms of hypernatremia in a mixed patient population of 390 medical, surgical, and neurological ICU patients, by adding the missing subgroup of patients suffering from severe burn injuries (7). Due to the large skin defects that are associated with an extraordinary high internal and external fluid shift and a high rate of severe infections and renal dysfunctions, this patient collective might take a special position within the overall ICU population in concerns of the etiology of the hypernatremic episodes. This theory is supported by the noticeably higher incidences of hypernatremic episodes compared to other critically ill patient cohorts (Hoorn et al. 7.6%) with incidences reaching up to 24-35% (2-4,8). Hoorn et al. proposed a three-step hypothesis for the pathogenesis of the ICU-acquired hypernatremia. First, most of the identified risk factors for ICU-acquired hypernatremia share the ability to promote renal water loss: hypokalemia (9), hypercalcemia (10) and renal dysfunction (11-13) can cause a urinary concentrating defect, whereas hyperglycemia (14) can cause osmotic diuresis. Vasopressin deficiency can develop in late stages of sepsis and therefore also contribute to renal water loss (15). Regardless of the mechanism, the rise in s-Na is known to produce a strong thirst stimulus (16). The second step to a hypernatremia therefore consists of an inability to express thirst and access water, as it is the case in mechanically ventilated or sedated patients, and finally an inadequate IV-fluid administration, which does not prevent or even aggravate hypernatremia. In line with this hypothesis, Hoorn et al. demonstrated a significantly higher number of mechanical ventilations, hypokalemia, hypercalcemia, renal dysfunction, hyperglycemia, and hypoalbuminemia in the hypernatremic patient collective (7).

The treatment of hypernatremic episodes follows a standardized protocol in our institution. Balanced crystalloids (Jonosteril^®) are primarily used for burn resuscitation in patients with physiological sodium levels. Jonosteril^® contains electrolytes in physiologic concentrations as well as acetate (Na⁺ 137 mmol/l, K⁺ 4 mmol/l, Ca²⁺ 1.65 mmol/l, Mg²⁺ 1.25 mmol/l, Cl⁻ 110 mmol/l, acetate⁻ 36.8 mmol/l). If the sodium level reaches values between 145-150 mmol/l, Jonosteril^® is diluted by adding D5W in equal parts. If the sodium level lies above 150 mmol/l, Jonosteril^® gets completely replaced by D5W. Due to the fluid loss through the burned skin areas, an accurate liquid balance sheet recording is often hard to achieve. The assessment of the volume status was therefore based on a combination of different diagnostic tools including sonographic evaluation of the heart, lungs and vena cava, the pulse index continuous cardiac output (PiCCO) system, volume challenge maneuvers, the hematocrit, as well as the urine output.

The data acquisition of the current study was limited to the first 14 days after the burn injury, which cover both the shock stage as well as the inflammatory stage after a severe burn injury. Based on previous data, which demonstrated an increased risk for hypernatremia in patients with third degree injury and high TBSA percentages (2-4), the inclusion criteria were limited to patients with at least 20% TBSA 2nd degree burn injuries or more than 10% TBSA when including areas of 3rd degree burn injuries. The TBSA percentage does not only seem to play an important role in concerns of hypernatremic episodes but also in a variety of other medical complications. Especially patients with a TBSA of more than 30% demonstrated in previous studies a considerable hypovolemia coupled with formation and release of inflammatory mediators, which inter alia results in a distinctive cardiovascular dysfunction in those patients (17-20). The consequences of this pathophysiological reaction include a diminished plasma volume, cardiac and urine output as well as an increased systemic vascular resistance resulting in a consecutive reduced peripheral blood flow (17,18,20-23).

The results of the current study demonstrated four major findings: 1) concerning the above mentioned risk factors no significant differences were detectable between the non-burn and the severe burn injury collectives except for a lower rate of hypokalemia and a higher rate of mechanical ventilation, 2) main trigger for the hypernatremic episodes was the loss of free water, 3) 24% of the hypernatremic episodes were at least partly triggered by a surplus sodium influx, 4) despite an increased mortality rate in the hypernatremic group, the hypernatremia seems to represent only a surrogate parameter for an unfavorable clinical course, but not the decisive cause of death.

Despite representing a very specific patient collective, burn injury patients do not seem to significantly differ from the non-burn injury patients concerning the risk factors and mechanisms of hypernatremia. Except for a lower number of hypokalemia and a higher rate of mechanical ventilation no relevant differences were detectable when compared to the case group of Hoorn et al. The significantly higher rate of hypoalbuminemia is explainable by the recommended lower limit of albumin of 2.5 g/dl for patients with severe burn injuries in the German burn injury guidelines, while the lower limit in the Hoorn et al. collective was defined as <3.5 g/dl. After adjusting the cut off for a hypoalbuminemia to 2.5 g/dl in the burn injury group, the difference between the Hoorn et al. case collective and the burn-injury collective changed into “not significant”. Regarding the small difference between the overall very high rates of mechanical ventilation in the non-burn injury collective (92%) (7) and our burn injury collective (100%), the mathematically significant higher rate of mechanical ventilation in the burn injury collective does not seem to represent a characteristic feature when searching for relevant differences between these two groups. Except for the frequency of hyperglycemia, all risk factors had significantly higher incidences in the burn injury collective compared to the non-hypernatremic control group of the Horn et al. collective (7), which supports the above mentioned theoretical expectations.

The differentiation between free water loss and surplus sodium intake was based on the increase in the sodium-chloride-ratio. The mean rise of the sodium from day -1 till day 1 was 4.13% and the mean rise of the chloride was 3.95%. Therefore, a difference between the percentual rise of sodium and chloride of more than 2% was defined as an “isolated” sodium rise, which implicates a surplus of sodium influx as part of the hypernatremic etiology. Based on this assumption, 24% of the documented hypernatremic episodes were at least partly triggered by a sodium surplus. Possible sodium sources are drugs like certain antibiotics, fresh frozen plasma, and sodium bicarbonate. Due to the use of Jonosteril^® and D5W as primary fluid substitutes, the fluid resuscitation itself is not considered as a risk factor for hypernatremic episodes. Nevertheless, the loss of free water remains the major cause for hypernatremia in burn injury patients. This seems reasonable due to the massive internal and external fluid shift in context of the pathophysiological processes associated with severe burn injury, which makes it difficult to attain a balanced fluid status during the early stages of this disease (17,18,20-23).

The increased mortality rate in the hypernatremic patient collective is in line with previous findings (5,6). Nevertheless, in none of the analyzed cases hypernatremia was the direct cause of death. None of the cases was associated with a drop in the s-Na >6 mmol/l/24 h, generalized seizures or a radiological confirmed brain swelling. Accordingly, the hypernatremic episodes seem to represent only a surrogate parameter for an overall unfavorable clinical course, but not the decisive cause of death.

Conclusion

The major trigger for the hypernatremic episodes in patients with severe burn injury is the internal and external loss of free water. However, besides this unique pathophysiological phenomenon, the overall risk factors and etiologies of hypernatremia in patients with severe burn injury do not seem to significantly differ from other ICU patient collectives. Remarkably, a surplus of sodium influx and therefore, a modifiable factor besides the volume resuscitation had an impact on the hypernatremic episodes in 24% of all cases. Although hypernatremia was associated with an increased mortality rate, hypernatremia itself only seems to represent a surrogate parameter for an unfavorable clinical course, but not the decisive cause of death.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflicts of Interest

The Authors have no conflicts of interest to declare in relation to this study.

Authors’ Contributions

Maximilian Menger: data curation, formal analysis, writing - original draft. Holger Wenz: data curation, formal analysis, writing - original draft. Maximilian Bamberg: data curation, formal analysis, validation. Sabrina Krauß: data curation, validation, writing - review & editing. Henrik Lauer: data curation, validation, writing - review & editing. Tim Viergutz: writing - review & editing, validation. Johann Fontana: conceptualization, formal analysis, writing - review & editing, supervision.