Sodium Balance Analysis In The Burn Resuscitation Period

Summary

The different formulae for resuscitation therapy after thermal damage recommend 0.5-0.6 mmol sodium for each % TBSA burned, suggesting fluid requirements from 2-4 ml/kg/% burn because of sodium loss in burned and unburned tissues. There is a gap especially in the recommendations regarding dysnatremia in the burn population. Many studies have focused on calculating amount of resuscitation fluids, avoiding the situation of “fluid creep”, and not on calculating sodium remaining in the body after resuscitation. The goal of this observational study was to provide data for sodium disturbances in the shock period after burns. Our study underscores the challenge of understanding whether there is a relationship between amount of crystalloid fluids given during resuscitation and meeting sodium needs. We set out to examine sodium balance (sodium deficit, received, excreted, and retained) after burns. The area under the ROC curve was performed by analyzing fluid and sodium load. Moreover, we conducted linear regression to analyze if there was a correlation between sodium retained and sodium excreted. Sodium deficit persisted until the second 24h despite resuscitation. Resuscitation was performed using Parkland formula, but urine output (UO) values were higher than expected. The threshold for fluid administration (ml/kg/%) or fluid load in the first 24h and sodium load (mmol/kg/%) for positive state (sodium received >0.5-0.6 mmol/kg/%) was 3.7 ml/kg/%. With linear regression, it was evident that sodium excreted was responsible for sodium retained, indicating a moderate correlation in the first 24h and a strong correlation in the second 24h. Resuscitation with LR did not correct hypoosmolality hyponatremia, which persisted even after the first 24h, especially in patients with burns >60%. If more than 3.7 ml/kg/% of LR is given, a sodium load higher than the normal level will be introduced, leading to increased urinary output, elevated sodium excretion, and non-correction of plasma sodium at the end of resuscitation. What is important for colleagues in clinical practice is that the focus of burn resuscitation should be expanded with data regarding sodium balance and the impact of dysnatremias in morbidity and mortality.

Introduction

The burn patient has many unique features in the burn resuscitation period because of exceptional resuscitation characteristics when compared with other injuries.¹ Despite hourly urine output (UO) and hemodynamic parameters as the main approaches for resuscitation, it is unclear whether these parameters are adequate to monitor resuscitation.^2,3,4 The different formulae for calculation of resuscitation therapy after thermal damage recommend 0.5-0.6 mmol sodium for each % of total body surface area (TBSA) burned, suggesting fluid requirements between two to four ml/kg/% burn because of sodium loss in burned and unburned tissues. The real resuscitation volume depends on the burn patient’s response during fluid resuscitation treatment.⁵

To realize survival from severe burns, patients should be given high amounts of fluids containing sodium salts. Sodium required for resuscitation is 0.5 mmol/kg/% burn.⁵ Fox and Monafo proposed that the amount of sodium managed and held by the body is critical in burn management.^6,7 Generally, in critical patients, including the burned one, the irregular sodium concentrations emerge from crystalloid treatment, certain medicines, or renal function disturbance. ^8,9

The researchers found that extreme alterations of sodium concentrations in critically ill patients have a higher hazard ratio for mortality.¹⁰ Also, incorporating alterations in sodium levels with other clinical indicators of physiologic status may improve the accuracy and precision of determining a severely burned patient’s clinical course.^11,12 Some authors recommended sodium supplementation or even intravenous albumin for optimal Rl resuscitation, but according to our knowledge there is a gap especially in the recommendations regarding dysnatremia in the burn population.¹³ Many studies have focused on calculating the amount of resuscitation fluids, avoiding the situation of “fluid creep”, and not on calculating sodium remaining in the body after resuscitation.

Taking into consideration the fact that the majority of sodium disturbances acquired in hospital are preventable and indicative of substandard care, this study aimed to observe the resuscitation period of 48h after burns.^14,15 Because hyponatremia (Na+< 135 mmol/L) is mainly caused by extracellular sodium consumption taking place after changes in cellular permeability, we set out to examine sodium balance (sodium deficit, received, excreted, and retained) after burns.

The goal of the study was to provide data for sodium disturbances in the shock period after burns. Our study underscores the challenge of understanding whether there is a relationship between the amount of crystalloid fluids given during resuscitation and meeting sodium needs for a successful resuscitation. We think that this study may help to conduct a broader prospective study for plasma sodium correction association with morbidity and mortality.

Materials and methods

The data used in this study are in the domain of routine medical practice, and the subjects’ data were obtained according to the Declaration of Helsinki. Ethical approval for this observational study was granted by the Ethical Committee of the Ministry of Health and Social Protection in Tirana, Albania.

An observational study was conducted on 155 patients hospitalized in the Burn Intensive Care (ICU) of the Service of Burns at the University Hospital Center “Mother Teresa” in Tirana, Albania during 2016. Of these, 50 patients were included in our study. The study included patients with burns of more than 20% TBSA, patients with less than 20% TBSA but who required resuscitation in the shock period, and children as well as adults and elderly with thermal, flame, chemical, and electrical burns. Exclusion criteria were: the presence of inhalation burns, pregnancy, discharge from the ICU in the first 48h, and death within this period. Moreover, the patients who had unsteady clinical signs and were considered for advanced invasive monitoring were not included because they stayed during this period in the polyvalent ICU. Concretely analyzing patients selected for inclusion, 81 of 150 patients had burns ≤20% TBSA without clinical shock, 12 had inhalation burns (5 were treated primarily in the polyvalent ICU), 5 died within 24 hours, and 7 were admitted more than 8 hours after burn injury. Because of these inclusion and exclusion criteria, we have no possibility to analyze the mortality of all the patients hospitalized during 2016.

Fluid resuscitation protocol

The %TBSA burned was determined using the Lund and Browder chart. Adult patients were resuscitated according to the Parkland formula. In the first 24h, fluid resuscitation was done with Lactate Ringer (LR) solution. The fluid requirements were calculated as 4 ml × weight (kg) × % of TBSA burn = ml/24h. Half of the calculated volume was administered in the first eight hours. The Parkland formula determined the initial rate of fluid administration; afterwards, the attending physicians titrated it to maintain a urine output between 0.5-1.0 ml/kg/h.¹⁶

In the second 24h, basal fluid requirements were calculated 1500 ml × m² while evaporative water loss was calculated (25 + %TBSA) × m² × 24h. From the total amount of the maintenance fluids, colloids (human albumin) were given as 20-60% of the calculated plasma volume and glucose was added in water in amounts required to maintain the urinary output.¹⁷

Children were resuscitated according to the Shriner-Galveston formula, which provides 5000 ml/m² burned TBSA as a resuscitation fluid and 2000 ml/m² TBSA as a maintenance fluid in the first 24h. As with the adult formulas, half is given over the first 8h, and the remainder is given over the next 16h. The fluid utilized for resuscitation was Lactated Ringer’s solution with 12.5 g of 25% albumin per liter + 5% dextrose, as needed. In the second 24h, 3750 ml/m² TBSA burn was used as a resuscitation fluid and 1500ml/m² TBSA as a maintenance fluid.^18,19

The use of human albumin in adults in the second 24 hours and children since the first 24 hours has been based on the respective formulas. We used Albunorm 5% or 20% solution (3.3 mg sodium per ml or 0.14 mmol per ml). After assessing the nutritional requirements, the patients were fed differently. Some patients who were able to take food by mouth, also depending on the last nutrition time before they got burned, received small amounts of oral Supportan or Fresubin (1.5-2.0 kcal/ml) mainly after 4 to 6 hours of hospitalization. In comparison, the rest of the patients were fed through enteral nutrition (via nasogastric tube) within 24 hours of burn injury and simultaneously parenteral nutrition (after 48 hours) using Kabiven 1440 ml.¹⁸

Partial-thickness burns were cleaned with Chlorhexidine or Betadine solution, followed by occlusive and non-adherent dressings with antibiotic ointments (like Bacitracin/Silver Sulphadiazine) or hydrocolloid dressings. Full-thickness burns were treated with early excisions, multiple excisions, escharectomy, and skin grafting. For circumferential deep burns, escharotomy was performed.

Data collection and analysis

For data collection and analysis, we aimed to present general data, fluid resuscitation data and sodium balance data

The baseline data included age, gender (male, female), TBSA (%), depth of burn (partial-thickness burns, full-thickness burns), etiology of burn (scalds, flame, electrical, chemical), mortality, length of stay (LOS) in the ICU and LOS in hospital. Patients were stratified based on age groups and subgroups: children of 0-15.9 years and adults of more than 16 years according to the World Health Organization as follows: infant 0-1.9 years, young child 2-5.9 years, child 6-11.9 years and adolescent 12-18 years, while for adults the age groups were created by decade. Depending on the % TBSA, we classified the patients into groups of 10%, namely 11-20% TBSA, 21-30% TBSA, 31-40% TBSA, 41-50% TBSA, 51-60% TBSA, 61-70% TBSA, 71-80% TBSA, and 81-100% TBSA.

The detailed data on given fluids and urinary output were collected uninterruptedly in real-time during the 48h by nurses taking care of the patient. The fluid resuscitation was recorded on the 24h input-output chart. The resuscitation was evaluated with fluid load (ml/kg/%) and urine output (UO) (ml/kg/h). The sodium balance data was done from laboratory analysis of blood and 24-hour urine collection (electrolytes and osmolality).

From fluid load, urine output, blood and urine analyses, we calculated:

Sodium deficit (Nad) with formula: Nad (mmol) = (140 - Nas) × 0.6 × weight (kg) where serum sodium (Nas) was corrected for glucose in blood, expressed in mmol and in mmol/kg/% burn Sodium received (Nar) with formula: Nar (mmol) = 130.5 × fluid load (ml)/1000, expressed in mmol/kg/% burn

In the sodium received calculations, we also included sodium from albumin and the amount of sodium in Fresubin (2.6 mmol/ml) or Supportan (2.1 mmol/ml) and the sodium received by intravenous therapy.

Sodium excreted (Nae) with formula: Nae (mmol) = Nau × UOP (ml)/1000, expressed in mmol/kg/% burn.

Sodium retained as the difference between sodium administered and sodium excreted, expressed in mmol/kg/% burn.

We calculated water loss (ml/cm² burn area) from the Scott, McDougall, Slade, Pruitt (S.M.S.P.) formula of evaporation, as below:20,21

= [(25 + TBSA % burn) × m2 TBSA × 24] for adults and = [(35 + TBSA % burn) × m2 TBSA × 24] for children.

Moreover, from sodium deficit (mmol), we calculated the sodium loss (mmol/cm² burn area).

Statistical methods

Statistics software SPSS v.23 (IBM Corp.) was used for the statistical analysis. The continuous data were presented by means and SDs (standard deviation scores) and were examined with independent samples t-tests, whereas the discrete variables were presented by the absolute value and percentage, and the Fischer exact test was used for comparison of proportions. Bar and line graphs were employed to compare changes in sodium elements in the age groups. Moreover, a data plot was used to graph the relationship between sodium retained and sodium excreted in a linear regression model. The area under the ROC curve was used to compare the diagnostic performance of the best cut-off of fluid load to give the recommended amount of sodium load. Statistically, P values of 0.05 or less were considered significant.

Results

General data

Cohort baseline characteristics are summarized in Table I. The average patient age was 22.0±23.7 (range 1-80 years, median 5). Twenty-eight of the 50 (56%) patients were children (0-15.9 years), while 22 were adults (44% of the total). We observed that the age group most affected from burn injury was that of young children (2-4.9 years) with a total of 16 patients (or 32% of the total), followed by infants (0-1.9 years) and adults (40-49.9 years) with 9 (18%) and 8 (16%) patients, respectively. The mean age in the children’s group was 2.8±2.6 years, while in adults, the group was 46.4±13.9 years. Approximately 60% of the population was male, with a male to female ratio of 1.5:1.

The average TBSA proportion for all patients was 30.56±16.1%, median 25 (range 12-80%). We initiated resuscitation at burn size 12% for children and 20% for adults. The average TBSA for children was 27.4±17.1%, while for adults it was 34.5±14.1%. Moreover, nine children with burns less than 20% needed resuscitation. In the children group, 25 of 28 patients had burns up to 40% TBSA where 11-20% burns were predominant (15 patients or 53.6% of children), and 10% presented with burns from 40% to 80% TBSA. In the adult group, there were six patients (27.3%) with burns of 20% TBSA, and 18 patients (72.7%) with burns up to 60% TBSA.

Full-thickness burns were present in 13 patients or 26% of the total number, while within the groups, they accounted for 14% of children (4 of 28 patients) and 41% of adults (9 of 22 patients). Scalds caused a significant part of burns in 27 patients (54% of the total), out of which 25 were children. Overall, scalds accounted for 89% of burns in children. Flame was the cause of the burns in 16 patients or 32% of the total, out of which the major part, 13 patients, were adults. Flame caused 59% of the burns in the adult group, whereas electrical burns represented 10% of all burns, and all patients were adults ranging from 40 to 59.9 years. Finally, chemical burns were present only in adults: in 2 patients or 4% of all patients.

We calculated evaporative water loss (ml/cm² burn area) in the first two days. The average water loss was 0.54±0.16ml/cm² burn. This value was greater in children (0.62±0.16 ml/cm² burn area after 24h and 0.65±0.18 ml/cm² burn area after 48h), than in adults (0.4±0.07ml/cm² and 0.44±0.78 ml/cm² burn area after 24 and 48h, respectively). Water loss was greater in partial-thickness burns, i.e., 0.57±0.17 ml/cm² burn area compared with 0.44±0.07 ml/cm² burn area in full-thickness burns (for the first 24h). The water loss data exhibited an increase in values till 48h, more evidently in partial-thickness burns (up to 0.59±0.18 ml/cm² burn area).

We also presented data on sodium (mmol/cm² burn) with an average 0.06±0.03 mmol/cm². In 31 patients (62%), there was a rise in sodium loss per cm² during the first 24h, while in 19 patients (38%), there was not.

First aid was given before hospitalization in 36 patients or 72%, and only 14 patients or 28% presented to our service without having received first aid. Patients were admitted to the burn center 2.4±1.6h after injury (range 1-7h). 80% of the patients presented immediately, up to three hours after burn, and 20% more than three hours after injury. Overall, 42% of patients were treated with fluid therapy in regional hospitals and 58% presented directly to our service.

Outcome data are as follows: 46 or 92% were survivors. Mortality was 2%, while three patients (or 6%) were transferred abroad to complete treatment. LOS was 11.5±9.2 days in the intensive care unit (ICU) and 12.7±7.8 days in the hospital.

Resuscitation data (Parkland formula)

Data for fluid resuscitation for two days are presented in Table II. In the first 24h, the fluid load for adults was 3.1±0.7 ml/kg/% (median 3.2 and range 1.6-4.3 ml/kg/%). The fluid load for children in terms of Parkland formula was 4.6±1.4 ml/kg/% (median 4.9 and range 2.5-7.2 ml/kg/%). The UO in adults was 1.27±0.39 ml/kg/h. Only five patients (22.7%) with ideal diuresis values fell within 0.5- 1.0 ml/kg/h, and the others had higher values. Mean UO in children was 1.71±0.35 ml/kg/h. There were 24 children (85.7%) with values of UO 1.0 to 2.0 ml/kg/h while only four patients (14.3%) had values more than 2 ml/kg/h. In the second 24h, we observed a halving of the fluid load for all patients up to 2.0±0.9 ml/kg/%, which was more evident in adults (1.5±0.5 ml/kg/%). Moreover, an increase in diuresis values was observed both in adults and children. In our study, there were five adult patients with electrical burns, of which two originated from high voltage burns and three from flash. Patients with deep burns were resuscitated with 4.0±1.2 ml/kg/%, while those with flash burn by electrical arc were resuscitated with 3.0±1.6 ml/kg/%. In the first eight hours, when pigmentation was there, the urine output was 1.4-2.8 ml/kg/h, while at the end of the first 24 hours, it was 1.0-1.8 ml/kg/h.

Table I. Characteristic features of the population in the study.

Sodium balance data

The calculated sodium deficit on admission, after the first 24h and after the second 24h after corrections of plasmatic sodium for glycemia, have been presented in Table II. On admission, the mean sodium deficit of all patients was 163.1±130.7mmol, after 24h 193.1±141.3 mmol, and after 48h 195.6±144.4 mmol. Sodium deficit improved in a quarter of patients after the first 24h and in half of the patients after the second 24h.

Table II. Resuscitation data and sodium balance in the first and second 24h of resuscitation.

All input, mainly intravenous, was recorded on the twenty-four-hour input-output chart. From this chart, the total daily sodium intake or sodium load is calculated and is named “sodium received”. The mean sodium received for all patients in the first 24h was 0.51±0.17 mmol/kg/%. Thirty-three patients (18 adults and 15 children), or 66% of the total, received 0.5-0.6 mmol/kg/%, while only 17 patients (4 adults and 13 children), or 34%, received more than 0.6 mmol/kg/%. The mean sodium received was greater in children than in adults (0.6±0.1 mmol/kg/% vs 0.4±0.09 mmol/kg/%). In the second 24h, sodium received was 0.3±0.3 mmol/kg/%.

The collection of 24h urine is generally considered a gold standard for the assessment of sodium intake. Assuming all sodium is from dietary intake, about 90% is excreted in 24h urine, and the remaining 10% is excreted through sweat and feces, which can vary in particular conditions. Using urinary sodium, we calculated “sodium excreted” and the difference as “sodium retained”. The mean sodium excreted for all patients in the first 24h was lower (0.3±0.2 mmol/kg/%) compared with the second 24h (0.30±0.3 mmol/kg/%). In the case of sodium retained, the values were higher in the first 24h (0.2±0.2 mmol /kg/%) vs the second 24h (0.04±0.21 mmol /kg/%).

We observed that a particular amount of sodium was retained after rehydration even though its deficiency persisted in the body. After the first 24h, sodium retained was 0.2±0.2 mmol/kg/%, while sodium deficit in the burn wound was still present in values of 0.2±0.1 mmol/kg/%. After the second 24h, the sodium retained was very low (0.06±0.2 mmol/kg/%), and the deficit of sodium was in the same values (0.2±0.1 mmol/kg/%).

The sodium deficit (derived from blood analysis of plasma sodium after the first 24h and the second 48h) with respect to corresponding time has been presented in Fig. 1. We observed that sodium retained and sodium deficit had a close relationship with each other. In situations where the retaining sodium increased, the plasma sodium deficit was reduced and vice versa. Furthermore, to see if there was a correlation of sodium retained with sodium excreted, we performed a linear regression. With linear regression, it was seen that sodium excreted was responsible for sodium retained in 43% of the variance in the first 24h and for 83% of the variance in the second 24h. Concretely, for the first 24h, F (1,49) = 36.3, p<0.001, R2=0.431 and for the second 24h, F (1,49) = 237.8, p<0.001, R2= 0.832. The correlation was stronger, especially in the second 24h (Fig. 2).

Fig. 1. Presentation of sodium retained (mmol/kg/%) in the body and sodium deficit (mmol/kg/%) during the 1st and the 2nd 24 hours of resuscitation.

Fig. 2. Linear regression of sodium retained as dependent value from predictor sodium excreted (R2=0.43 in 1st 24h and R2=0.83 in 2nd 24h).

In Table III, we presented the changes in plasmatic values of electrolytes sodium and potassium (mmol/L), osmolality (mOsm/kg), hematocrit (HCT %), and albumin (g/dL) on admission, after the first 24h and the second 24h. Patients were classified according to their burn sizes. On admission, patients with more than 70% TBSA had extreme hyponatremia values, hyperkalemia, hypoosmolality, and a decrease in hematocrit values. Other patients indicated typical values of osmolality, potassium and the presence of hyponatremia and an increase in hematocrit values. After 24h of resuscitation, we observed a decrease of osmolality, sodium and potassium, especially in patients with burns of more than 60% TBSA, a trend which continued even after 48h. Albumin values were lower than expected on admission; they decreased after 24h while after 48h, the values partially improved.

Table III. Plasmatic values of osmolality, sodium, potassium, hematocrit, albumin in different % TBSA during 48 hours.

Sodium deficit indicated the sodium missing in situations of hyponatremia. The prevalence of hyponatremia on admission was 80%, with only 20% of patients having normal sodium values. The prevalence of mild hyponatremia (plasmatic sodium 131-135 mmol/L), moderate hyponatremia (plasmatic sodium 126-130 mmol/L), and severe hyponatremia (plasmatic sodium: ≤125 mmol/L) was 58%, 16%, and 6%, respectively. To determine the hyponatremia with hypoosmolality, which is characteristic of dehydration conditions, we first analyzed sodium, then plasma osmolality, and in the second stage, we analyzed urinary osmolality. It was evident that the possibility of developing hyponatremia with hypoosmolality increased during the first 24h, regardless of resuscitation (Fig. 3).

Fig. 3. Different types of hyponatremia in the 1st hour, 1st 24 hours and 2nd 24 hours of resuscitation.

On admission, the mean plasma sodium was 132.5±3.1 mmol/l, and osmolality was 286.5±8.7 mOsm/kg. Of the total, there were 40 patients with hyponatremia (80%), while only ten patients (20%) had normal values. In the hyponatremic patients, eight patients had hyperosmolality, and 20 patients had normal osmolality; so, both of these cases were classified as cases with pseudo hyponatremia. The other 12 patients (30% of cases with hyponatremia or 24% of the total number) were classified in hyponatremia with a hypoosmolality group.

After the first 24h, the mean sodium was 130±2.8 mmol/L, and osmolality was 276.5±11.1 mOsm/kg. Careful examination showed that there were 46 cases with the presence of hyponatremia, and only four patients had normal values of sodium (even in the first hour). On observing osmolality in 46 patients with hyponatremia, it was noticed that there were two patients with hyperosmolality and seven patients with normal osmolality, which were classified as cases with pseudo hyponatremia. The other 37 patients (80.4% of the cases with hyponatremia or 74% of the total) were classified in the hyponatremia with hypoosmolality group. Mean urine osmolality was 594.94±181.95 mOsm/kg.

After the second 24h, the mean sodium was 131.6±4.53mmol/L, and the mean osmolality was 277.91±9.95 mOsm/kg. Forty-five cases had hypona-tremia, and only five patients had normal values of sodium even though they belonged to the hyponatremia group in the previous measurements. Out of these 45 patients with hyponatremia, there were 26 patients with normal osmolality or with pseudo hyponatremia. The other 19 patients (42.2% of hyponatremia cases or 38% of the total number) were classified in the hyponatremia with the hypoosmolality group. Urine osmolality was 524.33±188.97 mOsm/kg.

The evolution in serum sodium, after observing the changes in serum sodium values during the 48h of resuscitation referencing the values on admission, was as follows. Patients starting with values of 120-125 mmol/L (severe hyponatremia) on admission showed an increase of serum sodium in 67% of cases by 10 mmol/L, and in 33% of cases, values did not change. Patients with values of 125-130 mmol/L (moderate hyponatremia) on admission maintained steady values in 62% of cases, and in 38%, the values increased by 5 mmol/L. Among those having values of 130-135 mmol/L (mild hyponatremia) on admission, 59% of the patients maintained the serum sodium values, in 17% the values increased by 5 mmol/L, in 21% the values decreased by 5 mmol/L, while in 3% of cases sodium reduced by 10 mmol/L. Patients having normal values on admission (135-145 mmol/L) showed a decrease of 5 mmol/L in 90% of cases and in 10% of cases a decrease of 10 mmol/L.

We wanted to set the threshold for fluid administration (ml/kg/%) or fluid load in the first 24h and sodium load (mmol/kg/%) classification positive state (sodium received >0.5-0.6 mmol/kg/%) and negative state (sodium received <0.5-0.6 mmol/kg/%). In Fig. 4, the ROC curve suggested that the cut-off was 3.7 ml/kg/% and the values greater than this cut-off illustrated sodium load more than 0.5-0.6 mmol/kg/% (AUC = 0.822, 95% CI 0.678; 0.966, p<0.000).

Fig. 4. Area under the ROC curve of testing the threshold of fluid administration (cut-off = 3.7 ml/kg/%) for giving adequate sodium load (0.5-0.6 mEq/kg7%.

Discussion

Our data for the prevalence of severe hyponatremia agreed with other reports.^12,14,22 Liquid shifts are fast during the early period of burn shock (24-72h).^6,23,24 Therefore, monitoring of hematocrit, serum electrolytes, osmolality, calcium, glucose and albumin is necessary to decide the suitable rehydration strategy.²⁵ In major burns, the intravascular volume is misplaced in burned and unburned tissues, and hyponatremia is frequent, primarily due to extracellular sodium exhaustion because of changes in cellular penetrability. Moreover, hyperkalemia is also a characteristic of this period. The restoration of sodium in this period is fundamental.^16,26,27,28

For interpreting hyponatremia, we measured plasma osmolality to distinguish three possibilities: pseudohyponatremia (hyponatremia with normal osmolality), hyponatremia with hypoosmolality with normal renal water excretion, and hyponatremia with hypoosmolality with impaired renal water excretion. Since a considerable number of hospitalization cases have pseudohyponatremia, more careful monitoring of them is needed in terms of the number of resuscitation fluids. Concretely, we observed 28 cases of pseudohyponatremia on admission, 9 cases after 24h, and 26 cases after the second 24h.

Patients with hyponatremia and hypoosmolality on admission were mainly children with partialthickness burns. Depending on the severity of the burn shock, they exhibited gastrointestinal symptoms (nausea, vomiting) to mild neurological symptoms (alteration in consciousness) together with hemodynamic alterations. This finding is in accordance with some authors who concluded that hyponatremia is more likely to be a marker of illness rather than having a casual effect on mortality (Chawla) and did not find a significant association of hyponatremia with mortality.^12,29

Based on our data, hyponatremia with hypoosmolality with impaired renal water excretion was present in 12 patients on admission, in 37 patients after the first 24h despite resuscitation, and in 19 patients after the second 24h. Resuscitation with LR did not correct hyponatremia with hypoosmolality, which persisted even after the first 24h, especially in patients with burns with TBSA >60%. Following this finding, we also measured urine osmolality and made a clinical assessment for evaluating the low effective circulating volume or antidiuretic hormone (ADH) effect. Urine osmolality was more than 100 mOsm/kg, and more precisely 594.94±181.95 after 24h and 524.33±188.97 after the second 24h.

During the whole period of the study (48 hours), we did not identify patients with hypernatremia. On the other hand, we have records of patients with hyponatremia without significant clinical signs because hyponatremia occurred gradually throughout 48 hours. Only two patients presented neurological symptoms such as lethargy and confusion, which lasted until they passed the burn shock phase.

Sodium deficit in patients with hyponatremia estimates the total amount of sodium that needs to be replaced. Upon correction, the rate should be 6-12 mmol/L in the first 24h. We calculated it on admission, after the first 24h, and after the second 24h. In 67% of patients with severe hyponatremia on admission, values increased by 10 mmol/L only after the second 24h of resuscitation. In 38% of cases of patients with moderate hyponatremia on admission, values increased by 5 mmol/L. In contrast, in patients with normal values or with mild hyponatremia on admission, values may have decreased after the second 24h.

The resuscitation of a patient with an extensive burn requires crystalloid fluid administration during the resuscitation period. The number of resuscitation fluids can be reduced by using colloids or hypertonic solutions according to the respective protocols of the burn centers.³⁰ On the other side, 40.6% of burn centers in the UK and Ireland frequently consider changing the intravenous fluid during the resuscitation period from Human Albumin to Ringer Lactate and from LR to 0.9% NaCl.³¹

There is an inverse relationship between the salt concentration and the volume of fluid used for resuscitation. ³² Generally, the optimum sodium load was 0.5-0.6 mmol/kg/% burn, and the total volume of fluid required ranged from 2-4 ml/kg/% burn.⁵ In order to weaken edema, it is crucial to establish the ideal amount of sodium inside the slightest commendable fluid volume. Habib et al. has studied levels of electrolytes, albumin and proteins following RL, arriving at a conclusion that sodium supplementation may be required to correct hyponatremia. Colloids, preferably intravenous albumin, should be added, as advised by the original Parkland Formula.³³

We calculated the sodium received by the volume of administered LR. It was noticed that specifically in the first 24h, all patients received sodium 0.5±0.17 mmol/kg/% burn while only 17 patients (34%) received more than 0.6 mmol/kg/% burn (mainly children). We also calculated sodium excretion from urinary sodium, which increased in the second 24h mainly due to the rise in diuresis. The sodium retained due to our treatment was 0.22±0.19 during the first day and slightly less on the second day. Based on our data, rehydration with LR did not correct natremia even after 48h. It remained in moderate values, and sodium deficiency in the body was improved only in a quarter of patients during the first 24h and in half of the patients after 48h. From the study, it may be derived that sodium retained in the body influences the sodium deficit values negatively. If there are higher values of sodium retained, the sodium deficit calculated from plasma sodium is lower. Based on the linear regression, sodium retained is strongly related to excreted sodium and urinary output.

The water loss and sodium loss values per cm² burn tissue for our patients whose wounds were covered with the occlusion method were parallel to earlier reports.^20,34 From our modest calculations, the water loss was greater in partial-thickness burns and in children rather than adults. This could be attributed to the fact that children had more scald burns while adults had full-thickness burns caused by flame.

From our data, the cut-off for giving sodium load >0.5-0.6 mmol/kg/% with greater sensitivity and lower (1-specificity) was 3.7 ml/kg/%. If more than these amounts of fluids are administered during resuscitation, a higher than needed sodium load may increase urinary output and cause higher sodium excretion and non-correction of plasma sodium at the end of resuscitation.

The treatment of burns in our service has been a challenge that has been gradually overcome, accompanied by the improvement of our patients’ prognosis, especially in the last decade. The medical staff has paid attention to conducting observational clinical studies to optimize resuscitation regarding a more qualified attitude towards the burned patient. Initially, the studies compared rehydration with and without urinary output monitoring, and then a study was performed on rehydration with Hypertonic Lactate Saline as a mild hypertonic solution (1.5%).³⁵ Rehydration with hypertonic solutions realizes the delivery of a particular sodium load through a smaller hydric load than isotonic rehydration.

Sodium disorders in the ICU are associated with morbidity and mortality. Nevertheless, little is known about how fluids can cause alterations in hormonal responses to injury and inflammation, so it is crucial to monitor and adjust sodium values. Checking the sodium levels can help in understanding the severity (mild, moderate or severe), the time-interval of development (acute or chronic), the tonicity (isotonic, hypertonic and true hypotonic), hyponatremia, and volume status of patients (hypovolemia, hypervolemia or euvolemia) and can help in minimizing fluid overload.

The concentration of sodium in Ringer’s lactate is less than the physiologic range in plasma, so rehydration with a high amount of RL can drive sodium levels to hyponatremic ranges up to 130 mmol/L. However, sodium levels drop gradually, and symptoms are moderate. The serum sodium estimation can help in noticing earlier the situations of in-hospital hypernatremia as a sign of systemic dehydration with an impact on the outcome. What is important for other colleagues in clinical practice is that the focus of burn resuscitation should be expanded with data regarding sodium balance and the impact of dysnatremias in morbidity and mortality.

Further studies should be done to provide evidence of the impact of sodium correction in fluid resuscitation data (ml/kg/%), in diuresis and, what is more important, in morbidity and mortality.

Conclusion

Our patients were resuscitated during the first 24h with 3.97±1.36 ml/kg/% LR, receiving the sodium load of 0.5±0.17 mmol/kg/% burn. UO of the patients was 1.5±0.4 ml/kg/h for this period. In the second 24h, fluids given were 2.04±0.9 ml/kg/% and UO values were 2.3±0.6 ml/kg/h. The prevalence of hyponatremia on admission was 80%, with only 20% of patients having normal sodium values. The prevalence of severe hyponatremia (plasmatic sodium: ≤125 mmol/L) was 6%. Resuscitation with LR did not correct hypoosmolality hyponatremia, which persisted even after the first 24h, especially in patients with burns >60%. In 67% of patients with severe hyponatremia on admission, values increased by 10 mmol/L only after the second 24h of resuscitation.

In critical burns, the restoration of sodium losses in the burn tissue is essential. By keeping diuresis under strict control (considering the risk of acute kidney injury), it is possible to deliver a proper supply of fluids, not eliminate sodium from the urine, ensure better sodium retention in the body, and less sodium deficit. We agree with the burn experts regarding the fluid load of 2 ml/kg/% TBSA burned for 24h. This paper presented our patients’ real fluid load, which was 3.1 ml/kg/% TBSA for adults.

The cut-off for giving sodium load >0.5-0.6 mmol/kg/% with greater sensitivity and lower (1-specificity) was 3.7 ml/kg/%. If we give more than these amounts of fluids during resuscitation, we will introduce a higher sodium load above the normal values, which would lead to increased urinary output, elevated sodium excretion, and non-correction of plasma sodium at the end of resuscitation.

From the statistical analysis, we concluded that during resuscitation, the values of the fluid load should not exceed the value of 3.7 ml/kg/% TBSA in order to only gain the positive aspects of the treatment (giving the right amount of sodium load without high fluid load). In patients with electrical burns, sodium levels were slightly reduced in terms of severity, and the osmolarity was normal. Thus, all the patients had isotonic hyponatremia. We think that patients with electric burns and those with intense burns should be resuscitated very carefully and monitored hourly to avoid over and under hydration as well as maintain renal function. In the future, a prospective study may be performed regarding cases with electrical burns and serious injuries to understand the causes of hyponatremia if it will be evident. Albumin can be used even in the first 24h if blood values are critical, considering it may help maintain osmolality and reduce possible diuresis during this period.

The strengths and limitations of the study

This is a study that is focused not only on fluid load (ml/kg/%) and urine output (ml/kg/h) but also on sodium load (ml/kg/%) while analyzing them during monitoring of resuscitation of critically burned patients.

Limitations of the study consist in the small number of patients and exclusion of inhalation burns. Moreover, it would be better if we had a comparator group of patients with the correction of plasmatic sodium during resuscitation.