CORRECTION OF HYPERNATREMIA IN SEVERELY BURNED PATIENTS Samatov I.Yu, Veynberg A.L., Mezhin A.V., Streltsova E.I.,Vereshchagin E.I.
Novosibirsk State Medical University,
State Novosibirsk Regional Clinical Hospital, Novosibirsk, Russia
Hypernatremia (HN) presents the serious problem in therapy for patients with severe burn injury. It is associated with high rate of HN in these patients and with high mortality in patients with burn disease, which is complicated by HN [1, 2, 3]. The hypernatremic condition was registered in 37.5 % of patients with severe burns in a study [3]. The authors proved that HN worsened the prognosis and increases the risk of lethal outcome in burn patients. So, HN appeared on the day 5 (+/- 1.4). The mortality was 20 %. The infusion/diuresis ratio was several times lower than in the patients without HN.
The authors indicate that HN in burn patients may have the iatrogenic origin because of inadequate and excessive crystalloid infusion therapy and high fluid losses [4, 5]. Therefore, patients with severe burn injury have serious water-electrolyte disbalance. The statistical analysis showed the significantly higher volumes of discharged fluid in burn patients on the days 3-7 after burn injury as compared to patients with normal level of serum sodium.
Generally, HN develops in two conditions: increasing sodium and water deficiency. The first group of the causes includes the excessive administration of sodium solutions and primary or secondary hyperaldosteronism [6]. The decrease in general level of water is one of the causes of HN, but not single one. The other promoting factors are: 1) iatrogenic: excessive use of osmo- and saluretics; 2) water transition into the cell; 3) plasma loss in burn shock and water evaporation from the wound; 4) intestinal and pulmonary losses; 5) renal (tubular insufficiency); 6) fluid sequestration into “the third space” etc. [1, 6]. Therefore, according to these authors’ opinion, the optimal strategy for HN correction is internal administration of hyposmolar solutions [3]. HN was corrected with 5 % glucose. The authors mention the possible use of 0.45 % sodium chloride.
The basic principles of water-salt metabolism are considered during estimation of ways for prevention and correction of HN in critically ill patients [8]:
1. Normal kidneys reabsorb or discharge the water for maintenance of normal osmolarity of plasma (275-290 mOsm/l). Vasopressin is a plasma osmolarity regulator. Osmoreceptors regulate its production [9, 10]. Hypotension, hypervolemia, pain and hunger are the triggers for vasopressin release [11].
2. For maintenance of osmotic balance, water freely moves between intracellular and extracellular sectors under influence of osmolarity.
3. Fast intercellular movement of water created the cellular injury. The compensatory mechanisms of maintenance of normal volume of intercellular fluid activate only after 48-72 hours. They include the accumulation of intracellular electrolytes (fast adaptation) and organic substances with osmotic activity (slow adaptation) [5].
Therefore, the decrease in sodium is recommended with not higher than 0.5 mmol/l/h for prevention of cerebral edema, since in state of HN, the cerebral cells are dehydrated, and fast fluid losses can lead to cellular hyperhydratation [12].
The main recommendations for HN correction in burn patients are related to calculation of water and electrolyte deficiency, the rat and duration of infusion, and monitoring. There are acute and chronic disorders of water-salt metabolism: chronic condition is confirmed by more than 24 hours of water-electrolytic disbalance. Certainly, correction of water disbalance is the priority task. The water deficiency is calculated as: percentage of theoretical level of water × body mass × (plasma [Na] / normal [Na] – 1). The very important task is Na monitoring each 4-8 hours. The common recommendations are hypotonic solutions (0.45 % NaCl), bolus administration of saluretics; in some publications, spironolactone is mentioned, but without clear recommendations [1, 3, 5].
However the safety of intravenous administration of hyposmolar solutions is unclear. The case is that intravenous administration of hyposmolar solutions is most dangerous despite of apparent simplicity. In compensatory sense, CNS cells accumulate the electrolytes (fast adaptation) and osmotic active substances (slow adaptation). Therefore, fast correction of HN and hyposmolar solutions can worsen the neurological status with possible cerebral edema [5].
Some issues include the bolus administration of high dosages of saluretics. The most common technique of control of hydrobalance is furosemide. It is known that furosemide is loop diuretic with natriuretic and chloride uretic effects and blocking reabsorption of Na+ and Cl−. In the period of activity, Na+ output significantly increases, but after that, the output rate decreases below the basic level (“rebound” syndrome). The phenomenon is determined by fast activation of renin-angiotensin and other anti-natriuretic neurohumoral links of regulation in response to massive diuresis; it stimulates arginine-vasopressor and sympathetic system. Therefore, due to “rebound” phenomenon in bolus introduction, furosemide can cause HN. However there are some findings that show provision of natriuretic effect with small dosages of furosemide [13]. As result, the rational thing is estimation of furosemide influence on development and the course of HN with use of small dosages and titrated administration.
The underestimated technique is use of high dosages of spironolactone. In extreme conditions, HN can be caused both by dehydration and sodium retention owing to activation of renin-angiotensin system. Some studies of hypernatremia with CNS injury confirmed the role of pain, blood loss, hypoxia and injury in subsequent activation of renin-angiotensin system and sodium retention even with limitation of infusion therapy and refusal from diuresis augmentation [5]. Renin-angiotensin- aldosterone system is activated by the vascular effects of sympathetic nervous system as response to pain, hypovolemia, injury and/or renal blood flow changes in critical states. Besides, in response to stressor devastation of glucocorticoid fraction, the mineral corticoids partially replace the function. So, secondary aldosteronism syndrome can develop in critically ill patients [7].
Therefore, some issues remain: (1) is HN an iatrogenic effect or a consequence of hyperaldosteronism after burn shock; (2) each technique of HN correction is efficient and safe: sodium output or water introduction?
Another important issue is appropriateness and timeliness of continuous renal replacement techniques (CRRT) in acute burn injury according to extrarenal indices and HN particularly.
Objective – to develop the efficient algorithm for hypernatremia (HN) in burn patients.
Tasks:
1. Estimate the efficiency of various conservative techniques of HN correction.
2. Estimate the efficiency of continuous renal replacement therapy in HN correction and to clarify the indications for CRRT in burn disease on the basis of analysis of efficiency.
MATERIALS AND METHODS
The retrospective study included 165 patients (men and women) with severe burn injuries admitted to the burn ICU of State Novosibirsk Regional Clinical Hospital. The study was approved by the local ethical committee and corresponded to the Rules for Clinical Practice in RF (the Order by Russian Health Ministry, June 19, 2003, No.266).
The inclusion criteria were the age of 15-70, total square of burns (degrees 2-3) > 40 %, or the degrees 2-4 > 20 %, or the degrees 2-3 > 20 % + upper airways burn, more 3 days of stay in the burn ICU. HN was evident at Na > 150 mmol/l.
All patients received the respiratory support, 92 % received inotropes and vasopressors. PiCCO-monitoring was used for 18 patients for additional control of central hemodynamics, volemic status and estimation of fluid accumulation in pulmonary interstitial tissue.
The conservative techniques of HN correction in the group 1 (2009-2011) included the hyposmolar solutions (5 % glucose) with calculation of necessary volume: % of theoretical level of water in the body × present body mass × (plasma [Na] / normal [Na] – 1).
Considering the high risk of increasing neurological dysfunction and gas exchange disorders in the patients with HN, intravenous introduction of hyposmolar solutions was limited from 2012. However administration of small portions of water into the stomach has shown its safety. Therefore per os administration of drinking water is the first main technique of conservative therapy (20 ml/kg/24 h).
At the same time, considering the leading role of hyperaldosteronism in development of HN, we used spirolactone (250-300 mg/day). Then furosemide was prescribed with the titrated dose of 0.5-1.5 mg/kg/day. Similar introduction allowed preventing “rebound” effect, which increases sodium accumulation in the body. Moreover, small titrated dosages of furosemide prevent dehydration and control hydrobalance with accuracy of +/- 1 ml/kg/day. In some cases with inefficiency or impossibility of these two methods within 24 hours, the use of small titrated doses of furosemide was the single conservative technique for HN correction.
Therefore, HN correction in the group 2 included:
1. Introduction of water into the stomach (20-30 ml/kg/day, 4-6 times).
2. Spironolactone, 250-300 mg/day.
3. Furosemide, i.v., titrated dose of 0.5-1.5 mg/kg/day.
If the conservative techniques for HN correction were inefficient (Na > 160-163 mmol/l), then continuous renal replacement therapy was initiated, and HN was the main extrarenal indication for its initiation and conduction.
For prolonged types of renal replacement therapy, MultiFiltrate and the following techniques were used: HV-CVVH, CVVH, CVVHDF, paed-CVVH. Sodium profiling in the substitute solution with the allowable difference in plasma Na / substitute Na not more than 10 mmol/l was used. For hybrid technology (SLEDD, sustained low efficiency daily dialysis), Fresenius 5008 (artificial kidney) and HDF were used. Ultrafiltration for negative water balance was used in dependence on hyperhydration degree with the mean rate of 1-3 ml/kg/h. The convection dose of hemofiltration was 35-60 ml/kg/h. UltraFlux AV hemofilters and bicarbonate solutions as the substitute (HF-23 and HF-42) were used. The duration of a single procedure of SLEDD was 6-10 hours. Also bicarbonate buffer and hi-flax membranes were used. Prolonged anticoagulation was achieved with infusion of heparin with titration of the rate with targeted 1.8-2-fold increase in APTT. The main criteria for initiation of CRRT were provisionally extrarenal, i.e. the increase in hypernatremia > 160-163.
The statistical analysis was conducted with STATISTICA v.10 (StatSoft, USA). Fisher’s exact test and χ2-test with Yates correction were used for the comparative analysis. The critical level of significance was 0.05 (p < 0.05).
RESULTS
Totally, 92 patients were included into the study in 2009-2011 (the group 1). Hypernatremia was corrected with the hyposmolar solutions (5 % glucose) with the formula: percentage of probable level of water in the body × present body mass × (plasma [Na] / normal [Na] – 1). 41 patients died. The general mortality was 45 % (the table).
Table 1
Mortality after severe burn injury in dependence on techniques of hypernatremia correction
|
Groups |
General mortality (%) (n=92/73) |
HN-associated mortality (%) (n=30/20) |
HN-associated mortality within 14 days (%) (n=30/20) |
|
Group1 |
45 % |
73 % |
60 % |
|
Group 2 |
41 % |
60 % |
25 % |
|
p Fisher’s exact test |
p= 0,6382 |
p= 0,3662 |
p= 0,0213 |
|
p χ2 (chi-square) test with Yates correction |
p = 0,7165 |
p= 0,4960 |
p= 0,0321 |
Note: differences between the groups are reliable for p<0.05
HN was in 33 % of the patients. It appeared on the days 4-6 after severe burn injury. The mortality in the patients with HN was 73 %. It corresponds to the data from other authors. The proportion of the patients with HN who died within 14 days after trauma (i.e. when a lethal outcome is associated with HN) was 60 %. In the later period, the lethal outcome was determined by predominantly septic complications of burn disease and development of multiple organ disorders.
Totally, 73 patients with severe burn injuries were included in 2015-2017 (the group 2). The general percentage of the patients with HN was 27 %. This complication also appeared on the days 4-6 after severe burn injury. So, the differences between the groups were almost absent.
There were not any reliable differences in the general mortality in the analyzed groups. However the patients with HN (33 and 27 % correspondingly) showed the decrease in mortality (73 % in the group 1, 60 % in the group 2). Moreover, we noted a reliable (more than 2 times) decrease in the mortality (60 and 25 % correspondingly) in the patients with HN who died within 14 days after the injury, i.e. when a lethal outcome could be related to HN. With such conservative treatment of HN, the second group did not show any cases of polyuria, hypopotassemia or disorders of creatinine or urea clearance.
A clinical case
The patient D., was admitted to the burn ICU of Novosibirsk Regional Clinical Hospital on October 1, 2010. The diagnosis was: “A burn injury of degrees 2-3-4, square of burns – 60 %”. The sodium level was 162 mmol/l on the 6th day, despite of the conservative methods for prevention and treatment of HN (water 20 ml/kg per os, spirolactone 300 mg/day). The titrated administration of furosemide (100 mg/day; 1.3 mg/g/day) was initiated. After 24 hours, the sodium level in the blood serum was near optimal one (154 mmol/l), and the rate of decreasing sodium – 8 mmol/24 hours. Urea excretion was 611 mmol/day (the normal value – 250-570), the daily diuresis – 1,500 ml (2 ml/kg/day), i.e. within the normal values. The glomerular filtration rate was 96 ml/min, tubular reabsorption rate – 98.9 %. HN was corrected within subsequent 2 days.
Therefore, the use of small titrated dosages of furosemide (at least 1.5 mg/kg/day) was accompanied by increasing sodium excretion without diuresis forcing and with preservation of optimal concentration function. The rebound effect was not noted.
If conservative techniques were inefficient, CRRT was performed with the protocols described in Materials and Methods section. In all cases of CRRT, the blood sodium level decreased to the normal values. It was accompanied by regression of signs of MODS. There was a possibility for full volume of nutritive support and the control of electrolyte and hydrobalance in both CRRT and SLEDD.
In 2009-2011, CRRT was initiated for 10 patients (among 30) with HN (33 %). 7 patients survived. All three lethal outcomes were associated with late initiation of CRRT both for terms of burn disease and plasma sodium (Na > 170 mmol/l). In 2015-2017, CRRT was initiated early, including one-third of the patients (6 cases of 20) with impossible conservative therapy of HN (30 %). 3 patients died within 14 days after injury as result of septic complications of burn disease.
The analysis of each clinical case has shown the best results in early activation of RTT within 7 days after injury. The best results were achieved in the patients with large square of dermal burns and critically increasing HN (plasma Na > 163 mmol/l). The control of central hemodynamics and volemic status with PiCCO is also optimal. SLEDD-technology is the economical alternative of CRRT.
However any positive results of CRRT were not found in later initiation (Na > 170 mmol/l, the days 12-14 and more after burn disease) with sepsis development and impossibility for surgical removal of incrustation.
The general mortality in the patients with HN in the group 2 decreased by 13 % as compared to the group 1. The most important argument in favor of the offered strategy is more than two-fold decrease in the mortality (60 and 25 % correspondingly) in the patients with HN within 14 days after burn injury, i.e. the period with the most intense negative influence of HN. In most cases, the cause of death in these patients was late purulent septic complications in the period when water-salt disbalance is already removed.
CONCLUSION
1. The conservative techniques for hypernatremia correction (water load 20-30 ml/kg/day per os + spirolactone 200-300 mg/day + furosemide 0.5-1.5 mg/kg/day i.v., titration) were safe and efficient in timely correction (145 mmol/l < Na < 163 mmol/l). Small titrated dosages of furosemide (60-100 mg/day or 0.5-1.5 mg/kg/day) promoted the decrease in sodium level in blood serum (8-10 mmol/l/day) in HN. The “rebound” effect does not appear, the serum potassium level, reabsorption degree and diuresis do not change.
2. The use of CRRT in burn patients with HN is the most optimal extrarenal indication for this technique. The highest efficiency of RRT is noted in early initiation of the procedure (Na < 163 mmol/l not later than 7 days after injury). At the same time, late initiation (Na > 170 mmol/l and development of purulent septic complications of burn disease) did not cause any positive results of CRRT.
Information on financing and conflict of interests
The study was conducted without sponsorship.
The authors declare the absence of any clear or potential conflicts of interests relating to publication of this article.