Yu-Wen Tang

Plastic and Reconstructive Surgery and Burn Center, Taichung Veterans General Hospital, Taichung, Taiwan; National Young-Min Medical University, Taipei, Taiwan; Chinese Medicine University, Taichung, Taiwan

SUMMARY. Patients who have both inhalation injury and cutaneous thermal injury require more resuscitation fluid in the acute stage than patients who have cutaneous thermal injury only. In the later interval after burn injury, pulmonary complications (infiltration, oedema, pneumonia, and adult respiratory distress syndrome) are usually noted in septic patients with critical conditions. Fluid restriction with intensive fluid monitoring is suggested in most reports. It is general practice to limit urine output amount to around 0.5 ml/kg/h, which is considered adequate. Since these patients usually present severe sepsis, poor renal and gastrointestinal (GI) functions, and nutritional problems, fluid intake should be increased rather than decreased. We believe that fluid restriction may aggravate renal and cardiac function, or do it no good, and that lung damage is not the cause of the reduced fluid demand. On the contrary, a large fluid supply may help to maintain a more stable and adequate blood supply to each part of the body, especially the GI system, kidneys, and lung. The whole body would benefit from it. In this retrospective study, we did not reduce the patients’ fluid intake volume after pulmonary complications occurring late, and we continued to maintain urinary output at between 1 and 2.5 ml/kg/h or more, as before the onset of pulmonary complications. Intensive fluid monitoring and restriction of the daily fluid balance are important in this period. Compared with our previous cases (fluid restriction to maintain urinary output at around 0.5 mk/kg/h), we found that these patients had better GI function (oral feeding), nutritional support, and renal and liver function, as well as a shorter hospital stay. This method did not aggravate pulmonary complications. In conclusion, we believe that fluid restriction in cases of cutaneous burns with associated later pulmonary complications is not necessary.

Introduction

Patients suffering an inhalation injury associated with a cutaneous burn have increased mortality compared with those presenting a comparable cutaneous burn alone. In the acute phase, they require more resuscitation fluid than patients with cutaneous thermal injury alone in order to achieve adequate resuscitation. Fluid resuscitation with lactated Ringer’s solution (LR) 5.7 ml/kg/%/total body surface area (TBSA) (with inhalation injury), as compared to 3.98 ml/kg/%/TBSA (without inhalation injury), has been suggested in some burns centres.1,2 This can prevent extravascular lung water increase, hypoxia, and reductions in lung function. These additional requirements probably reflect increased severity of the injury, which in turn predisposes the patients to other complications.

Pulmonary complications (infiltration, oedema, pneumonia, and adult respiratory distress syndrome [ARDS]) are usually noted several weeks after burn injury. Fluid restriction with intensive fluid monitoring is suggested in most reports. Urine output around 0.5 ml/kg/h is considered adequate. However, these patients usually present severe sepsis, poor renal and GI function, and nutritional problems, and fluid intake should therefore be increased rather than decreased. We believe that fluid restriction may aggravate renal and cardiac function, or do it no good, and that lung damage is not the cause of reduced fluid demand. A large fluid supply may help to maintain a more stable and adequate blood supply to every part of the body, especially for renal, hepatic, and GI functions. In fact, the whole body would benefit from it.

In order to understand the effect of the “relative” fluid overloading formula for the resuscitation of burn patients with concomitant pulmonary complications in later intervals, we did not restrict our patients’ fluid intake and kept their urine output at 1-2 ml/kg/h or more. In this retrospective study, we evaluated:

1. whether this formula harmed damaged lungs;
2. the functional improvement of vital organs;
3. whether this would lead to a better result for the whole organism;
4. how to prevent fluid overload in the lungs, and whether there were any contraindications to use of the formula.

Methods

A retrospective study was made of 18 patients (12 men and 6 women, aged 18 to 72 yr) with inhalation injuries and cutaneous burns who were admitted to the Taichung Veterans General Hospital Burn Center between 1994 and 2000. The patients had been resuscitated using a modification of the Parkland formula (4 ml/kg/TBSA) with LR solution. Colloid was initiated after the first 16 h and only in patients with > 25% TBSA burns. Urine output of over l ml/kg/h was considered indicative of adequate resuscitation. In most cases, we like to keep urine output over 1 to 2 ml/kg/h in our centre. Central venous pressure (CVP) and pulmonary wedge pressure were routinely assessed during resuscitation. Wounds were dressed with silver sulphadiazine twice daily until excision and grafting had been performed. Enteral feeding was initiated within 24 to 48 h.

Chest roentgenograms were taken from the first day of hospitalization and reviewed by burn surgeons and a radiologist. The radiographs were classified as normal, localized infiltrate (one lung only or half the lungs on both sides), or diffuse infiltrate (more than half the lungs on both sides, pulmonary oedema). The pulmonary complications of pulmonary infiltration, oedema, ARDS and pneumonia were usually noted 3 to 4 weeks post-burn. Treatment with endotrachal intubation, mechanical ventilation, pulmonary hygiene, suitable antibiotics, and wound care was given.

Before June 1996 (group A), fluid restriction was initiated as soon as there were lung complications (most of these patients had a history of inhalation injury). We reduced the total amount of fluid administered and tried to keep urine output at around 0.5-1 ml/kg/h. Diuresis and colloid replacement were also given. Eight cases from this time period were selected for our study.

In June 1996 (group B) we changed our protocol for fluid supplementation, to the effect that the patients had no strict fluid restriction. Once there were pulmonary complications, we still kept urine output at around 1-2 ml/kg/h or more, i.e. the same as before the onset of pulmonary complications. However, the fluid balance between intake and output was strictly restricted, usually not exceeding 1000 ml/day. Intensive monitoring, colloid replacement, and intermittent diuresis were used to keep the “daily differences” in cumulative fluid intake versus fluid output (IO) between -500 to 1000 ml/day. Daily body weight, CVP, serum electrolyte and albumin levels, and urine specific gravity were also intensively monitored.

The outcome of diuresis/fluid-restriction (group A) versus diuresis/no fluid restriction (group B) was evaluated according to various functions (renal [BUN, creatinine], hepatic [SGPT, SGOT], pulmonary [chest x-ray, PaO2, PaCo2, and O2 saturation], and gastrointestinal [serum albumin and enteral feeding amount]), serum electrolytes (potassium and sodium), morbidity (days on ventilator), and mortality.

In patients with inhalation injury, the damage was confirmed by clinical characteristics and/or the findings of fiberoptic bronchoscopy after injury. Initial management in the acute period included nasotracheal intubation, mechanical ventilation, supplement oxygen, and suitable antibiotics. Some of the patients developed severe lung damage at a later interval (2-3 weeks after the injury). They were included in this study.

Mortality was defined on the basis of in-hospital deaths. The cause of death was secondary to the burn wound, inhalation injury, pneumonia, or multi-organ dysfunction syndrome with or without sepsis. The cause of death as recorded in the medical record was used for data assessment (this could be used to evaluate the response of adequate tissue perfusion).

Statistics

Data were analysed by Student’s t test and reported as the mean ± standard deviation and p value. Any p value < 0.05 was considered significant.

Results

The patient population’s characteristics are summarized in Table I. Most patients were burned by flame due to gas explosion (42%), and in a closed space (68%). Most (66.6%) had a history of inhalation injury that was confirmed by bronchoscopy in 95% of them. These patients most often had a smooth course with their lung condition in the first week. Severe lung damage was normally noted in the following weeks. All cases were managed with intubation and mechanical ventilation. The mean duration of ventilation was 41.3 ± 3.25 days (group A) and 36.28 ± 4.77 days (group B). Three patients were subjected to tracheostomy for prolonged ventilation or tube compression due to marked swelling.

The estimated fluid requirements (Tables II, III) for the pre-pulmonary complication period ranged from 3100 to 6620 ml (mean, 5141 ± 302 ml). After the onset of pulmonary complications (pulmonary infiltration, oedema, pneumonia, and ARDS), the fluid requirements were 2806 ± 283 ml (group A) and 4277 ± 281 ml (group B). In group A (Table III), fluid restriction was enforced after pulmonary complications developed, and urine output was strictly limited to 2.23 ± 0.16 ml/kg/h (before) and 0.83 ± 0.04 ml/kg/h (after). The net balance between fluid in and fluid out was markedly decreased, from 2709.75 ± 315.81 ml/day (before) to -396.87 ± 270.12 ml/day (after). In group B, there was no strict fluid restriction after pulmonary complications, and urine output was unchanged at 2.29 ± 0.13 ml/kg/h (before) and 2.016 ± 0.11 ml/kg/h (after). The net fluid balance was 2700.40 ± 167.2 ml/day (before) and 383.5 ± 125.4 ml/day (after).

Pulmonary complications with bilateral pulmonary infiltration, pulmonary oedema, pneumonia, and ARDS were noted in all cases (Table IV). The overall mortality rate was 50% in group A and 30% in group B. Death was attributed to purely pulmonary complications in most of the patients (5 out of 6), and most of these were associated with sepsis.

As an indicator of renal function, BUN levels showed a marked increase (Tables V, VII) in group A (63.62 ± 18.12 mg/dl) compared with group B (43.10 ± 19.32 mg/dl), of which Ø 32.5% and p < 0.05. However, there was no significant difference in creatinine levels between the two groups (1.73 ± 0.97 mg/dl and 1.17 ± 0.46 mg/dl, respectively, Ø 32.3% and p > 0.05). Indicators of liver function (SGOT and SGPT) (Tables V, VII) showed a minor elevation in both group A (26.0 ± 12.5 U/l, 38.12 ± 23.40 U/l) and group B (30.0 ± 15.14 U/l, 47.40 ± 37.65 U/l) (of which SGPT ¤ 15% and SGOT ¤ 18.4, but p > 0.01).

The changes in serum albumin levels (Tables VI, VII) in both groups were minimal: 2.16 ± 0.12 g/dl (group A) and 2.25 ± 0.10 g/dl (group B). The difference in gastric feeding amounts was from 832.25 ± 378.00 ml (group A) to 1038.80 ± 399.44 ml (group B), and although there was a 20% variation, p was greater than 0.05.

Both groups required mechanical ventilation in the burn unit. However, the patients in group B (36.28 ± 4.77 days) were ventilated for a significantly shorter period of time than patients in group A (41.3 ± 3.25 days) (p < 0.05) (Table I).

Discussion

Acute inhalation injury results in severe lung damage, i.e., pulmonary infiltration, oedema, pneumonia, and ARDS. However, in burn patients, pulmonary complications are more commonly the result of a systemic process initiated by the products of burn tissue, infection, or inflammation in the weeks following the burn injury. Among our cases, pulmonary complications mostly occurred in patients with a history of inhalation injury (66%), either minor or major. Characteristically, alveolar consolidation with fluid, protein, and inflammatory cells accompanies normal pulmonary artery occlusion pressure (i.e. pulmonary oedema is non-cardiogenic). Increased pulmonary capillary permeability results in increased movement of fluid from the intravascular to the interstitial space.

In the acute period of inhalation injury with cutaneous burns, smoke inhalation injury or chemical burns to the airways lead to increased capillary permeability, mucosal irritation with bronchorrhoea, and cough. Distal airways obstruction and atelectasis may occur, sometimes followed by interstitial oedema and inflammation. With good integrity of host defences and aggressive pulmonary support, secondary pulmonary infection and the subsequent development of ARDS do not usually occur in this period.3 Many papers support the opinion that patients with both types of injury require more fluid during the first 24 h post-burn than patients with burn injuries alone.2,4 The additional volume does not lead to pulmonary oedema.5

In the late interval after burn injury, bacteria and bacterial products derived from burn wound tissue or gut contents initiate a systemic inflammatory response. More cytokines (from macrophages), neutrophil proteases, oxidants, and multiple protein cascades6,7 impair the microcirculatory flow. Increased endothelial permeability and impaired lymphatic function lead to non-cardiogenic pulmonary oedema. Pulmonary venous and capillary congestion, accompanied by endothelial cell swelling and necrosis, is a characteristic feature of early ARDS. Clinical management of non-cardiogenic pulmonary oedema and ARDS includes treatment of the initial disorder, i.e. debridement and skin grafting of burn wounds, volume replacement, nutritional support, and the administration of antimicrobial and/or antifungal agents.

The current method of fluid management with such lung injuries is fluid restriction/diuresis. It is considered to be adequate to maintain urine output at around 0.5 to 0.8 ml/h/kg. Supporters of this method believe that it is harmful to lung tissue if urine output exceeds 1 ml/kg/h. However, our results suggest that fluid restriction with a urine output amount limited to 0.5 ml/kg/h is unnecessary in burn patients with lung complications. Continual maintenance of urine output at over 1.5 ml/kg/h (the same as before the pulmonary complication) does not aggravate lung damage. On the contrary, we believe it leads to better tissue perfusion (i.e. in the GI tract, liver, and kidneys). In our patients, the amount of fluid administered depends on the capacity and tolerance of the vital organs. If the patient has good renal function with normal urine excretion (1 to 2 ml/kg/h), we do not markedly reduce the volume of fluid administered. However, it is important to watch closely and monitor “daily differences” in the cumulative volume of fluid intake versus fluid output. The cumulative input/output should not exceed 500 to 1000 ml/day. Usually, we keep input/output between -500 and 300 ml/day with a urine output of 1 to 2.5 ml/kg/h.

There are many theories and studies supporting the belief that diuresis and/or fluid restriction are helpful in cases of pulmonary oedema and ARDS,8-10 as these measures reduce hydrostatic pressures. The justification for restricting fluid administration or, more directly, for actively trying to lower pulmonary capillary pressure during pulmonary oedema is embodied in the familiar “Starling equation”.11 This model predicts that pulmonary oedema will develop if lymph flow or changes in other so-called “safety factors” cannot compensate for increases in pulmonary capillary pressures. Numerous experimental studies support the logical extension of this paradigm, namely that reduced capillary pressures and/or reduced perfusion to acutely injured lung units will result in reduced extravascular lung water accumulation.12 A strategy of fluid restriction/diuresis should be pursued during the first few days of ARDS, while carefully monitoring and supporting the perfusion of vital organs. However, the strategy of relative or absolute fluid restriction for these patients can certainly be misunderstood and abused.

In burn patients, these lung complications have commonly been noted with “shock” (septic shock) or near shock conditions in the later interval. This can be minimized by adequate fluid resuscitation to maintain tissue perfusion, early excision of burn wounds, and rapid wound coverage. These measures, combined with antibiotic coverage and nutritional support in the form of early enteral tube feeding, will decrease the hypermetabolic response and the incidence of sepsis, which can lead to haemodynamic instability and organ failure.13 Treatment for patients with severe inhalation injury involves supportive care with conventional volume-cycled positive-pressure ventilators, supplemental oxygen, and an intensive tracheobronchial toilet. The first major improvement in the treatment of burn injury came with the recognition of the importance of fluid resuscitation in the prevention of shock and renal failure.

Fluid restriction would aggravate this associated shock condition, whereas intravascular volume expansion could improve cardiac and renal function. Some specialists believe that fluid restriction and/or wedge pressure reduction in manifestly hypovolaemic patients (such as in sepsis) may invite clinical disaster, with vital organ hypoperfusion as the expected adverse result.8,14

Our hypothesis is that the maintenance of adequate tissue perfusion and oxygenation can prevent tissue hypoxia and acidosis in pulmonary, peripheral, and splanchnic microcirculations, and that “adequate tissue perfusion” will not aggravate lung complications. It is important in cases with associated cutaneous burn injuries to maintain good tissue perfusion. Experimental evidence suggests that hypoxic, acidotic endothelium stimulates the release of cytokines, kinins, and other mediators that can induce ARDS and other forms of organ failure.6 One report suggests that fluid restriction for ARDS is not necessary,9 and that it would aggravate or be of no benefit to cardiac function, in addition to aggravating renal function.15 These studies support the concept that a positive fluid balance per se is at least partially responsible for a poor outcome in patients with pulmonary oedema, and defend the strategy of attempting to achieve a negative fluid balance if tolerated haemodynamically.14

In our retrospective study, we found that maintaining good tissue perfusion with a urine output of 1.5 ml/h/kg or more does not aggravate lung complications. Rather, it leads to better results by shortening the course of such complications. We believe that this is due to improved tissue perfusion and function of the kidneys, liver, and GI system. In cases of lung complications associated with shock, it is important to maintain a good haemodynamic condition, including cardiac output, renal perfusion, and intestinal and hepatic function. Although progress has been made in the treatment of lung complications associated with cutaneous burn injury, the pathophysiology of these complications is still incompletely defined.16

In conclusion, we can say that even with the limited number of cases in our retrospective study we noted that in patients with pulmonary complications (pulmonary infiltration, oedema, pneumonia, and ARDS) in the later interval after burn injury, the maintenance of urine output at around 1 to 2 ml/kg/h or more did not endanger pulmonary function. On the contrary, it improved tissue perfusion, leading to better renal function (BUN) and a lower mortality rate. However, strict control of the fluid balance was necessary. The small case number in our study necessitates further investigation with a greater number of cases.

RESUME. Les patients atteints soit de lésions d’inhalation soit de lésions thermales cutanées nécessitent une quantité majeure de liquides pour la réanimation dans la phase aiguë par rapport aux patients atteints seulement de lésions thermales. Dans la période tardive après la brûlure, les complications pulmonaires (infiltration, œdème, pneumonie, syndrome de difficulté respiratoire dans les adultes) se manifestent normalement dans les patients septiques en condition critique. La limitation des liquides accompagnée par un monitorage intensif des liquides est recommandée par la plupart des spécialistes. Généralement il est conseillé de limiter la diurèse à environ 0,5 ml/kg/h, quantité considérée suffisante. Puisque ces patients sont normalement atteints aussi de sepsis sévère, de fonctionnement rénal et gastrointestinal (GI) insuffisant et de problèmes de nutrition, l’apport des liquides devrait augmenter, plutôt que diminuer. Selon l’opinion de l’Auteur de cet article, la limitation des liquides détériorerait la fonction rénale ou cardiaque ou ne donnerait aucun bénéfice; en plus, les dégâts pulmonaires ne sont pas la cause de la demande réduite des liquides. Au contraire, un apport abondant de liquides contribuerait à maintenir une bonne distribution sanguine dans tout le corps, et en manière particulière dans le système GI, les reins et les poumons. Tout le corps en bénéficierait. Dans cette étude rétrospective, l’Auteur n’a pas réduit le volume des liquides administrés aux patients après la période tardive et a continué à maintenir la diurèse à des valeurs entre 1 et 2,5 ml/kg/h ou encore de plus, comme avant le commencement des complications pulmonaires. Cependant, il est important de continuer un monitorage intensif des liquides et de limiter le bilan quotidien des liquides. Par rapport à ses cas précédents (limitation des liquides pour maintenir une diurèse d’environ 0,5 mg/kg/h), l’Auteur a trouvé que ces patients présentaient une fonction GI meilleure (alimentation moins prolongée). En outre, cette méthode n’aggrave pas les complications pulmonaires. L’Auteur croit que la limitation des liquides dans les cas de brûlures cutanées avec des complications associées pulmonaires successives ne soit pas nécessaire.

Bibliography

1. Hughes K.R., Armstrong R.F., Brough M.D., Parkhouse N.: Fluid requirements of patients with burns and inhalation injuries in an intensive care unit. Intensive Care Med., 15: 464-6, 1989.
2. Dai N.T., Chen T.M., Cheng T.Y., Chen S.L., Chen S.G., Chou G.H., Chou T.D., Wang H.J.: The comparison of early fluid therapy in extensive flame burns between inhalation and non-inhalation injuries. Burns, 24: 671-5, 1998.
3. Herndon D.N.: “Total burn care”. W.B. Saunders Company Ltd, Philadelphia, 321-9, 1996.
4. Cioffi W.G., Jr, Rue L.W., III, Graves T.A., McManus W.F., Mason A.D., Jr, Pruitt B.A., Jr: Prophylactic use of high-frequency percussive ventilation in patients with inhalation injury. Ann. Surg., 213: 575-82, 1991.
5. Ramzy P.I., Barret J.P., Herndon D.N.: Thermal injury. Crit. Care Clin., 15: 333-52, 1999.
6. Thangathurai D., Charbonnet C., Wo-C.C., Shoemaker W.C., Mikhail M.S., Roffey P., Roessler P., Kuchta K., Zelman V., De Meester T.R., Katz R.: Intraoperative maintenance of tissue perfusion prevents Adult Respiratory Distress Syndrome. New Horiz., 4: 466-74, 1996.
7. Steffes C.P., Bender J.S., Levison M.A.: Blood transfusion and oxygen consumption in surgical sepsis. Crit. Care Med., 19: 512-7, 1991.
8. Schuster D.P.: Fluid management in ARDS: “Keep them dry” or does it matter? [editorial]. Intensive Care Med., 21: 101-3, 1995.
9. Barie P.S.: Organ-specific support in multiple organ failure: Pulmonary support. World J. Surg., 19: 581-91, 1995.
10. McGrath A.M., Conhaim R.L., Myers G.A., Harms B.A.: Pulmonary vascular filtration of starch-based macromolecules: Effects on lung fluid balance. J. Surg. Res., 65: 128-34, 1996.
11. Civetta J.M.: A new look at the Starling equation. Crit. Care Med., 7: 84-91, 1979.
12. Schuster D.P.: The case for and against fluid restriction and occlusion pressure reduction in adult respiratory distress syndrome. New Horiz., 1: 478-88, 1993.
13. Nguyen T.T., Gilpin D.A., Meyer N.A., Herndon D.N.: Current treatment of severely burned patients. Ann. Surg., 223: 14-25, 1996.
14. Schuller D., Mitchell J.P., Calandrino F.S., Schuster D.P.: Fluid balance during pulmonary edema. Is fluid gain a marker or a cause of poor outcome? Chest, 100: 1068-75, 1991 [see comments].
15. Ronco J.J., Fenwick J.C., Tweeddale M.G.: Does increasing oxygen delivery improve outcome in the critically ill? No. Crit. Care Clin., 12: 645-59, 1996.
16. Sheridan R., Remensnyder J., Prelack K., Petras L., Lydon M.: Treatment of the seriously burned infant. J. Burn Care Rehabil., 19: 115-8, 1998.