Andrea Gabrielli
A. Joseph Layon
Ahamed H. Idris
Jerome H. Modell
Definitions and Descriptions
Several definitions and multiple terminology have appeared during the past half-century regarding the description of victims who suffer a fatal or near-fatal event from being submerged in water and other liquids. Some of the descriptors have had modifications placed on them and, furthermore, their meaning was somewhat lost when translated into some languages other than English. Because drowning is a global problem, at the World Congress on Drowning in Amsterdam, The Netherlands, on June 26 through 28, 2002, a group was convened from multiple countries to develop a definition of “drowning” that would be applicable in multiple languages worldwide (1). Although unanimity may not have been present on every term discussed, there clearly was a consensus to simplify the terminology for international application. What follows is the consensus of that group with comment and, in some cases, slight modification representing the bias of the authors of this chapter.
The Drowning Process
Drowning is the process resulting from primary respiratory impairment from submersion or immersion in a liquid medium. Implicit in this definition is that a liquid-to-air interface must be present at the entrance to the victim's airway, thus precluding the possibility of the victim to breathe air. Although it is possible to suffer a drowning episode in multiple types of liquid, this chapter will be confined to the most common use of the terminology, namely, drowning in water.
The drowning process is a continuum that begins when the victim's airway is initially below the surface of the water. At this time, the victim first will voluntarily hold his or her breath. Some victims will swallow significant quantities of water during this time. This period of voluntary breath holding, which has been found in human volunteers to last an average of 87 seconds at rest and shorter with exercise (2), is followed by an involuntary period of laryngospasm secondary to water in the oropharynx or at the level of the larynx acting as a foreign body (3). During this period of breath-holding and laryngospasm, the patient cannot breathe; therefore, oxygen is depleted and carbon dioxide is not eliminated. This results in the patient becoming hypercarbic, hypoxic, and acidotic (4).
As the levels of carbon dioxide increase in the blood and levels of oxygen decrease, respiratory efforts become very active but no exchange of air occurs because of the obstruction at the larynx. Victims who subsequently recover and recall this period frequently describe it as being quite terrifying and painful as they struggle to create intense negative intrapleural pressure breathing against a closed glottis (5). As the patient's arterial oxygen tension drops further, laryngospasm abates, and the patient then actively breathes water. Further evidence of the magnitude of negative pressure created during laryngospasm is the fact that the lungs of drowning victims frequently demonstrate significant hyperinflation at autopsy (6).
The amount of liquid a drowning victim breathes varies considerably between victims (4). Studies comparing the biochemical changes occurring in humans after a drowning episode with those in experimental animals suggest that, while the volume of liquid actually inhaled varies considerably from one victim to another, only 15% of persons who die in the water aspirate in excess of 22 mL/kg of water (7), and the percentage is considerably less in those who survive (4). Changes occur in the lung, body fluids, and electrolyte concentrations, which are dependent on both the composition and volume of the liquid aspirated (8,9,10).
Resuscitation
A victim can be rescued at any time during the drowning process and given the appropriate resuscitation measures, in which case the process is interrupted. The victim may recover with the initial resuscitation efforts or after subsequent therapy aimed at eliminating hypoxia, hypercarbia, and acidosis and restoring normal organ function. If the patient is not removed from the water, then circulatory arrest will occur, and, in the absence of effective resuscitative efforts, multiple organ dysfunction and death will result, primarily from tissue hypoxia.
Although the tolerance to hypoxia of various tissues is different, it should be noted that the brain is the organ most at risk for permanent detrimental changes from relative brief periods of hypoxia. Frequently, the question is asked, How long can a person be submerged and still be rescued and resuscitated back to a normal life? While, obviously, there are no controlled human studies—nor should there be on this subject—the limiting time factor is likely the duration that cerebral hypoxia can be tolerated before irreversible changes occur. Irreversible damage to brain tissue is reported to begin approximately 3 minutes after the PaO2 falls below 30 mm Hg under normothermic conditions in otherwise normal people (11). Such data suggest that if the victim is rescued and effective resuscitation efforts are applied within 3 minutes of the cessation of respiration (i.e., submersion in water), the vast majority of such victims should be able to be resuscitated and suffer no permanent brain damage. Further, because the period of voluntary breath holding and laryngospasm is thought to last for approximately 1½ to 2 minutes (2,12), persons who are retrieved within that time frame will likely not suffer lung damage secondary to the aspiration of liquid. Once the 3-minute time frame has been exceeded, although some normal survivors are reported, it becomes less likely that normal survival will result from resuscitation efforts. This time frame may be prolonged if hypothermia occurs rapidly because it decreases the cerebral requirement for oxygen.
Trained divers have been shown to be able to voluntarily hold their breath for much longer periods of time, approaching 4 to 5 minutes without complication (13). Persons who become hypothermic due to immersion or submersion in extremely cold water will rapidly develop hypothermia, which protects the brain by decreasing its oxygen requirement, and prolongs survival (14). In the latter case, seemingly miraculous recoveries of patients who have been submerged for over 20 minutes have been reported (15). It should be noted, however, that hypothermia is a two-edged sword; although it can protect the brain from oxygen deprivation, it also can cause death in the water secondary to its effect on the conduction system of the heart, resulting in circulatory arrest either by asystole or ventricular fibrillation (16).
The drowning process can be altered by the initiating event, such as if the victim suffers trauma, develops syncope or unconsciousness, has a circulatory arrest either by asystole or ventricular fibrillation as the precipitating event, hyperventilates prior to breath-holding under water, or has a convulsive disorder that leads him or her to become incapacitated, thereby becoming submerged, or if the victim's judgment and/or motor function is impaired by significant parenteral levels of depressant drugs, including alcohol. For example, in the victim who suffers a concussion from a blow to the head, subsequent recollection of the events is unlikely. If trauma results in a cervical fracture, disastrous damage to the spinal cord may occur acutely and, thus, motor function may be lost below that level. If the victim has a circulatory arrest either by asystole or ventricular fibrillation as a precipitating event, respiration will cease, and it is highly unlikely that significant amounts of water will be breathed into the lung given that active respiration is necessary for this to occur (17). If the victim hyperventilates prior to breath holding under water, it has been shown that the breath-holding breaking point can be extended until the level of hypoxia is so severe that consciousness is lost, and thus the victim actively breathes in water (2,12). The effect of drug usage is variable, depending on the level of depression and the patient's response. There is considerable variation in tolerance to depressant drugs and alcohol and their effects on performance and orientation. To better understand what to expect in each victim, the initiating event should be reported in every case if it is known.
When a person experiences a drowning episode, the result can be death or survival. Furthermore, the victim can survive without residual damage, or with residual damage to varying degrees (e.g., from minor neurologic difficulty to one that has no normal function other than continuation of an effective heart beat with or without spontaneous respiration).
Classifications
The terms drown and near-drown have been used for decades in an attempt to separate these outcomes (18). At the World Congress on Drowning, however, it became apparent that their meaning was not felt to be clear when translated into some languages (1). Furthermore, a victim could have no signs of spontaneous physiologic function and, therefore, be “drowned”; however, once resuscitative efforts were applied, they would respond positively and survive to varying degrees and, thus, the term applied to them would have to be changed to “near-drown” (19). In addition, there is another group who do not die acutely, but die later of complications from their drowning episode. In this case, the question is, Were they “near-drowned” or were they “drowned”?
The definition of “drowned” we believe to be fairly clear—namely, death secondary to undergoing the drowning episode. “Near-drowned” presents a significantly greater problem of understanding. We believe that the term “drowned” should be retained for both those who die acutely in the water and those who die later of consequences directly resulting from the submersion episode. However, we agree with the consensus of the World Congress members that “near-drowned” may lead to unnecessary confusion and, therefore, should be replaced by terminology such as “the victim survived the drowning episode” and then describe the ultimate condition of the victim.
Other terms that have appeared in the literature over the past few decades that we believe are confusing and should be abandoned are as follows:
· Dry versus wet drowning: Because all drowning occurs in liquid, by definition, they are all wet. This terminology has been used by some to categorize drowning victims into those who aspirate liquid into the lungs and those who do not. Frequently, it is not possible to determine at the scene of the accident whether the victim actually did aspirate water. This is particularly true when the quantity of water aspirated is small. Further, if evidence of fluid aspiration is not detected in the victim who dies or is discovered dead in the water, the diagnosis may be suspect (20). In these cases, one should look for other explanations such as acute mechanical standstill of the heart, from asystole or ventricular fibrillation, or, for that matter, whether the victim actually was alive when he or she first became submerged.
· Active versus passive versus silent drowning: This terminology has been used by some to separate those victims who are observed to be struggling at the surface of the water from those who are first discovered when they are actually submerged and motionless. It has been shown with underwater cameras that even victims who were not seen to be in difficulty on the surface of the water by observers may have had unrecognized active motion while submerged. We believe, therefore, that these terms should be abandoned in favor of the terms “witnessed,” when the episode is witnessed from the onset of submersion/immersion to the time of rescue, or “unwitnessed,” when a body is found in the water without anybody seeing how it got there.
· Secondary drowning: This terminology has been used by some to describe a situation when a precipitating event from another origin (e.g., syncope) causes a victim to be below the surface of the water, and then he or she drowns. On the other hand, some use this terminology to describe a victim who appears to be recovering from a drowning episode in the hospital and then develops adult respiratory distress syndrome. Not only is this terminology confusing but also, in the latter instance, a patient does not experience a second submersion or drowning episode, and therefore, this terminology should be abandoned.
Pathophysiology
There have been extensive studies both in animals (7,8,9,14,18,21,22,23,24,25,26,27,28,29,30,31,32,33,34) and in humans (4,6,20,35,36,37,38,39,40) over the past century in an attempt to quantitate the changes that occur as a result of a drowning episode. What has consistently been shown over and over again is that, acutely, drowning produces asphyxia (i.e., hypoxia, hypercarbia, and acidosis). The hypercarbia is due to absent or ineffective ventilation, and is readily correctable when aggressive mechanical ventilation is instituted. The hypoxia that occurs initially is not as readily correctable and may be persistent for long periods of time (8,9,10). This hypoxia is first due to apnea, and then primarily to intrapulmonary shunting from alveoli that are perfused but not being ventilated, or not being ventilated adequately (32). The acidosis is mixed, and the respiratory component rapidly disappears with effective ventilation. The patient is, however, frequently left with significant metabolic acidosis due to anaerobic metabolism during the period of time that profound tissue hypoxia secondary to absent or ineffective respiration and cardiac output was present. The hallmark of this high anion gap metabolic acidosis is an increased level of serum lactic acid.
Pulmonary
While intrapulmonary shunting occurs after both fresh water and sea water aspiration, the etiology is different (33). In the case of fresh water, the aspirated water alters the surface tension properties of pulmonary surfactant. Thus, the alveoli become unstable and do not maintain their normal shape or patency, resulting in an increase in both absolute and relative intrapulmonary shunt (32,33). Sea water does not change the surface tension properties of pulmonary surfactant but, because it is hypertonic, it pulls fluid from the circulation into the alveoli, thus producing obstruction to gas exchange at the alveolar level. Bronchoconstriction also has been reported after aspiration of even small quantities of water (28).
Fresh water, being hypotonic, is absorbed very rapidly into the circulation and, because of the transient hypervolemia that occurs and the change in the surface tension properties of pulmonary surfactant, pulmonary edema results. The pulmonary edema is most commonly described as frothy or foamy and blood-tinged. This coloring is secondary to the presence of free plasma hemoglobin from the rupture of some red blood cells due to the absorption of hypotonic fluid into the circulation in the face of hypoxia (41). Pulmonary edema also occurs when sea water is aspirated, secondary to a semipermeable membrane effect because the sea water is hypertonic compared to plasma. Even though the etiology of the hypoxia is different between fresh water and sea water aspiration, the result of both is to increase intrapulmonary shunt, which requires aggressive therapy (32,42,43).
Extensive studies of serum electrolyte concentrations after drowning have shown that only 15% of victims who die in the water aspirate more than 22 mL/kg of water. In patients who survive, the percentage is much less and, thus, significant changes in serum electrolyte concentrations that require treatment are rarely observed (7), with the only exception being, perhaps, victims of drowning in the Dead Sea (44).
The treatment of the respiratory lesion requires providing mechanical ventilatory support in a fashion that will restore an adequate functional residual capacity and keep the alveoli open during all phases of the respiratory cycle, thus decreasing the intrapulmonary shunt. Obviously, if foreign material such as sand, silt, or plant life is aspirated into the lung, it may produce obstruction, and it should be removed via bronchoscopy.
Cardiovascular
The cardiovascular changes that occur during a drowning episode can best be ascribed to inadequate oxygenation. Although fatal arrhythmia such as ventricular fibrillation is rarely documented in human drowning victims, ventricular fibrillation can occur with profound hypoxia, especially if very significant changes in serum potassium and serum sodium result from the movement of fluid and rupture of red blood cells. Although a wide variety of cardiac arrhythmias have been reported (45), particularly in animal models, rarely do they require specific therapy other than improving oxygenation and correcting severe metabolic acidosis. More common problems are profound hypoxia and the leak of fluid into the lung as pulmonary edema, resulting in a relative hypovolemia in the patient. It has been shown by multiple investigators that to treat this hypovolemia, it may be necessary to infuse significant amounts of intravenous fluid to maintain an adequate effective circulating blood volume, even in the face of pulmonary edema (26,43). Without such therapy, even though the arterial oxygen tension might have improved with mechanical ventilatory support, the delivery of oxygen to the tissues remains compromised and incompatible with supplying adequate tissue oxygenation (43). Use of vasopressors and other pharmacologic agents may be indicated as a temporary crutch, so to speak, in these patients; however, they are not a substitute for providing adequate oxygenation and adequate intravascular fluid volume. If the latter are established, it is highly unlikely that pharmacologic support of the heart will be necessary.
Renal
Detrimental changes in renal function are rarely seen in persons recovering from near-drowning. However, when present, they likely are the result of inadequate perfusion and oxygenation rather than anything specifically related to the drowning episode per se. Some have emphasized the need for the kidneys to clear free plasma hemoglobin after fresh water drowning; however, significant levels of free plasma hemoglobin have rarely been reported in such patients. This is likely due to the fact that for red blood cells to rupture and release enough hemoglobin into the plasma during a drowning episode to require specific therapy for its clearance, it requires transfer of substantial volumes of free water into the circulation in the face of hypoxia (41). As stated above, this rarely occurs.
Initial Rescue and Resuscitation
To ensure survival after a drowning episode, it is imperative that one never lose sight of the fact that time is of the essence. The longer a person is without the ability to breathe air, the more profound are the hypoxia and permanent damage to vital tissues. Thus, those who are entrusted with guarding swimming facilities must never lose sight of the fact that continual vigilance is required to recognize a victim in distress, and that the victim must be removed from the water and resuscitative measures begun in a timely fashion. Frequently, bodies are discovered motionless in a pool, without anyone in attendance being able to pinpoint the length of time that the victim was submerged. In many cases, lifeguards report that they thought the victim was “fooling around” and, therefore, they did not effect a timely rescue. Also, lifeguards may be assigned other duties such as pool maintenance and tending to concession stands, thus precluding them from timely recognition of a victim in trouble and prompt rescue.
Frequently, bathers are permitted to deliberately hyperventilate on the side of the pool before becoming submerged to see “how long they can hold their breath” or “how far they can swim under water.” These practices are to be condemned. If an individual is not noted to be making purposeful movements for more than 10 seconds, rescue attempts should be initiated (46). The individual responsible for safety at the pool should always be in proper attire and in position to effect such a rescue and complete it within 20 seconds of the recognition of the problem.
While removing the victim from the water, care should be taken to avoid complicating neck injuries when they are suspected. Routine stabilization of the neck is unnecessary unless the circumstances leading to the drowning episode suggest that trauma was likely (47). These circumstances include a history of diving, use of a water slide, signs of injury, or evidence of alcohol intoxication. If neck injury is suspected, gentle immobilization of the head should be accomplished, securing it in a neutral position. However, if the neck appears to be obviously deformed and the patient has pain with neck movement, the neck should be immobilized in the existing position.
If the victim is apneic, the airway should rapidly be cleared of foreign material, a patent airway secured, and mouth-to-mouth resuscitation started immediately. It is preferable to begin artificial ventilation in the water if it can be accomplished without jeopardizing the safety of the rescuer. It should be remembered that not all victims are in a state of cardiac arrest when the rescue attempt begins. They may be in a state of vasoconstriction or have a significant bradycardia, in which case, if effective ventilation is started, the myocardium will be reoxygenated, and increased cardiac activity will result in improved tissue perfusion.
Upon removing the victim from the water, he or she should rapidly be assessed for the presence of both spontaneous respiration and cardiac activity. In the absence of these, the airway should be inspected rapidly to ensure that there is no mechanical obstruction, and artificial respiration and cardiac compression should be instituted without delay. Chest compression alone—without artificial respiration—is an alternate resuscitation method that has been proposed for victims of dysrhythmic cardiac arrest. It must be emphasized that these recommendations do not apply to the drowning victim because the pathophysiologic lesion in the lungs requires active attempts at reinflation and stabilization of the alveoli. Therefore, cardiac arrest following drowning is more likely due to asphyxia, and thus, immediate provision of ventilation is recommended (47).
If equipment is available at the site for administering supplemental oxygen, it should be delivered in the highest concentration possible under the existing conditions. Electrical activity of the heart should quickly be evaluated and an automatic defibrillator applied if indicated. A pulse oximeter will frequently be of assistance in determining the effectiveness of oxygenation. However, many pulse oximeters do not work well if the victim is cold and vasoconstricted or if there is excessive movement.
Although the airway should rapidly be inspected for the presence of obstructing material, the abdominal thrust maneuver, which had been advocated by some in the past, has been thoroughly debated and found not to be of value in treating a drowning victim unless solid material is actually blocking the conducting airway (47,48). If the victim is apneic and a pulse is absent, chest compression should be initiated promptly (47). Chest compression alone can be effective in relieving airway obstruction.
Patient Transport and Emergency Medical Services
Neither equipment nor properly trained personnel are usually available at the site to provide advanced cardiac life support, including endotracheal intubation, intravenous access, drug therapy, and electrical defibrillation. However, these measures should be instituted when indicated and when the proper equipment and properly trained personnel are available. It is crucial that someone other than the individual rescuing and resuscitating the patient contact emergency medical services (EMS) as rapidly as possible so that they can respond in a timely fashion and perform advanced cardiac life support treatment on the victim.
Whenever a drowning victim has to be transported to a location or facility such as a hospital emergency room, it is important that a call be made promptly to inform the emergency room personnel of the exact circumstances, type of treatment instituted, and condition of the patient en route so that they will be prepared to accept the patient and render appropriate therapy immediately upon arrival.
When moving a critically ill drowning victim, it is imperative to remember the fragility of such patients because they can decompensate in a matter of a few seconds or minutes if appropriate therapy is withdrawn. Examples of such situations are movement (a) from the scene to the EMS vehicle, (b) from the EMS vehicle to the hospital emergency department, or (c) from the hospital emergency department to other hospital locations for testing, such as radiology, or for treatment, such as the intensive care unit (ICU). Thus, every attempt should be made to continue essential therapy at all times.
Treatment in the Emergency Department
In the emergency department, a thorough evaluation of the patient should be performed, keeping in mind that the most serious problems that require immediate therapy are pulmonary insufficiency and cardiovascular instability, which result in inadequate delivery of oxygen to vital tissues. If the victim is responding fully, does not require respiratory or cardiovascular support, and has a normal oxyhemoglobin saturation while breathing room air, it is unlikely that the victim has aspirated a significant amount of water, and observation may be all that is necessary. At the other extreme is the patient who is still unconscious and requires extensive pulmonary and cardiovascular support in an attempt to normalize vital signs and produce adequate cardiac output and tissue oxygenation. Thus, a cookbook-type treatment that would apply to every victim cannot be prescribed. However, the treating physician should keep in mind that increased intrapulmonary shunt and poorly matched ventilation-to-perfusion ratios are the rule rather than the exception for the victim who has aspirated a significant quantity of water.
Therapy must be aimed at improving ventilation-to-perfusion ratios and restoring adequate residual lung volume to optimally oxygenate the blood. A relative hypovolemia frequently is present due to fluid shifts between the lung and the circulation. These can be accentuated by the increase in mean intrathoracic pressure that occurs with mechanical ventilatory support. Thus, evaluation of effective circulating blood volume and replenishment of intravascular fluid volume to physiologic levels is important as a primary concern.
While currently some controversy exists regarding when to treat metabolic acidosis in persons who have suffered a cardiac arrest, we believe that the adverse effect of acidosis on the pulmonary vasculature and cardiac function is sufficient so that metabolic acidosis producing a pH of less than 7.20 should be treated with intravenous sodium bicarbonate. If the patient is a victim of sea water drowning and has aspirated sufficient water to produce hypernatremia, we might be better advised to use an agent such as tris (hydroxymethyl) aminomethane (Tris) buffer; 2-amino-2-hydroxyl-1.3-propandiol (THAM) to avoid compounding the hypernatremia. However, once again, it should be noted that the quantity of water aspirated is seldom sufficient to produce such significant changes in serum electrolyte concentrations, except perhaps when the drowning occurs in water of extreme hypersalinity such as the Dead Sea (44).
Changes in serum electrolyte concentrations and hemoglobin and hematocrit of sufficient magnitude to justify specific therapy are rare, as are alterations in renal function other than those that might be expected in the hypovolemic, hypoxic, or markedly acidotic patient.
The patient's level of consciousness on admission to the emergency room has been shown to markedly influence outcome (49,50). The most important consideration here is to provide adequate oxygenation and perfusion and to avoid producing increased intracranial pressure if possible. Treatments aimed specifically at preservation of cerebral function have not been shown to be particularly beneficial to date (50,51).
If the patient requires diagnostic testing in a distant location such as the radiology department, it is imperative that adequate personnel and equipment accompany the patient to ensure that optimum therapy is not interrupted at any time during transport or when performing the procedure. Likewise, transportation to the intensive care unit should be done with a “full team approach.” Should optimum therapy be interrupted during any of these time periods, adverse consequences should be anticipated.
Drowning episodes in cold water may produce significant hypothermia. There are several methods of rewarming that have been recommended including, but not necessarily limited to, heating blankets, warmed intravenous fluids, warmed humidification of breathing circuits, gastric lavage, and cardiopulmonary bypass. The method used should be tailored to the resources available and the condition of the patient. It must be remembered, however, that rewarming peripheral tissues before the patient's circulation is capable of supplying adequate amounts of oxygenated blood can compound the situation and increase the degree of metabolic acidosis.
In-Hospital Therapy: Postresuscitation Care
Expert intensive care is vital to survival once optimal prehospital and emergency department management have been performed. Hemodynamic instability after cardiac arrest, respiratory insufficiency, and severe neurologic impairment are all criteria for admission to the intensive care unit. The administrative structure of the hospital's critical care service dictates the setting to which the patient is admitted. A recent attempt to classify survivors of drowning based on the severity of symptoms on a scale of 1 to 6 recommends ICU admission for all pediatric patients requiring high concentrations of oxygen, with or without the need for invasive ventilation (52).
Respiratory Support
Although the degree of intrapulmonary shunting after drowning is variable from one patient to the next, if the patient is breathing adequately to clear carbon dioxide, the single most important method of treatment in reversing hypoxemia is the application of continuous positive airway pressure (CPAP). The amount of CPAP applied must be individualized because the degree of atelectasis, the amount of pulmonary edema, and the magnitude of the intrapulmonary shunt varies between patients. In great measure, this will depend on the type and quantity of the water aspirated. Although the mechanism for producing the intrapulmonary shunt is different between fresh water and sea water (33), Lee found no statistically significant difference between the PaO2/FiO2 ratio in patients after the two types of aspiration (53).
The pathophysiologic mechanism involved in fresh water drowning is lowering of the sodium concentration in the alveolus, thus changing the surface tension characteristics of pulmonary surfactant (33,54). The alteration in the surface tension properties of pulmonary surfactant increases alveolar surface tension upon compression of the surfactant layer and results in alveolar volume loss. Also, pulmonary capillaries become more permeable, resulting in an increase in interstitial lung water that eventually compresses alveoli and promotes volume loss and causes pulmonary edema. Based on the severity of the acute respiratory derangement, this “abnormal surfactant state” has been termed acute lung injury (ALI) or acute respiratory distress syndrome (ARDS) (55).
ALI and ARDS represent a final common pathway that accompanies a number of physiologic insults that may occur after drowning, including respiratory obstruction, aspiration of water or gastric contents, and global hypoxemia from cardiovascular insufficiency or cardiac arrest. Unfortunately, ALI and ARDS often can be clinically and radiologically confused with acute pulmonary edema from left ventricular dysfunction or fluid overload of different etiologies.
Both CPAP and positive end-expiratory pressure (PEEP) have the capability to restore lung volume and improve oxygenation in many patients with decreased lung volume, especially functional residual capacity. However, there are some differences in their function. By definition, CPAP means that airway pressure remains positive during all phases of the respiratory cycle. With PEEP, during the inspiratory phase of a spontaneous breath, circuit pressures drop to zero or become negative as a result of a vigorous inspiratory effort by the patient. Because PEEP with spontaneous ventilation increases the work of breathing, it may increase pressure gradients between the pulmonary vasculature and the alveoli, thereby leading to more pulmonary edema. Also, it does not forcibly inflate alveoli with abnormal surfactant after fresh water drowning (32). Thus, CPAP is more beneficial than PEEP for spontaneously breathing drowning victims (42,56).
Both CPAP and PEEP increase expiratory pressure; thus, air is trapped within the lungs during the expiratory phase of respiration. This results in an increase in residual lung volume in many patients with ARDS. As alveolar units re-expand, intrapulmonary shunt decreases, and improvement is seen in oxygenation and compliance. The increase in compliance decreases the work of breathing (57). The degree of lung volume restoration roughly correlates with the improvement in oxygenation. As lung volume increases toward normal, gas exchange continues to improve. It has been shown, however, that while the above beneficial effect is found with CPAP in many victims of both fresh and sea water drowning (56,58), unless mechanical breaths are added, PEEP does not improve the ventilation-to-perfusion ratio after fresh water drowning (32,42,56). Also, in some fresh water drowning victims, CPAP alone does not produce an adequate response, and hence, mechanical breaths should be added (42).
When ARDS develops and oxygen desaturation occurs, an FiO2 of 1.0 is recommended to attempt to restore adequate oxygenation. Increased work of breathing, severe hypoxemia, and hypercarbia are all indications for instituting mechanical ventilation. Ordinarily, CPAP is titrated to achieve an oxygen saturation greater than 95%, with the lowest possible inspired oxygen (FiO2) levels down to an FiO2 of 0.5 or less. We routinely increase CPAP at the bedside in increments of 3 to 5 cm H2O in an attempt to achieve an oxygen saturation of 95%, and subsequently, the FiO2 is gradually decreased to reach a PaO2/FiO2 of greater than 300 mm Hg. Increased dead-space ventilation and decreased preload are the two most important adverse effects that can limit the use of CPAP. Once adequate PaO2/FiO2 has been achieved, CPAP can slowly be weaned based on improvement of patient lung compliance and general clinical conditions.
Mechanical Ventilation
CPAP therapy alone is not sufficient in the case of the patient who is apneic, hypoventilating, or hypercarbic or shows little to no improvement in ventilation-to-perfusion matching while breathing spontaneously. In these patients, mechanical ventilatory breaths must also be provided. In general, mechanical ventilation in patients with ALI or ARDS can be applied either noninvasively or invasively (i.e., face mask vs. endotracheal tube, respectively). Noninvasive ventilation is reserved for milder cases of ARDS or pulmonary edema when the patient is awake, cooperative, triggering spontaneous ventilation, and has his or her swallowing and protective laryngeal reflexes intact. Although successful experience with noninvasive positive pressure ventilation (NPPV) for patients with respiratory failure other than from chronic obstructive pulmonary disease (COPD) is growing (59), potential complications include gastric distention, nasal congestion, regurgitation and aspiration of stomach contents, nasal bridge ulceration, and eye irritation (60). Several modes of mechanical ventilation and adjunct therapies are available; while not specifically used in drowning, their use has proven valuable in the ventilatory support of any patient with ALI or ARDS. A list of the most commonly used forms in drowning victims follows.
Controlled Mechanical Ventilation
Controlled mechanical ventilation (CMV) provides total ventilation, and it does not permit spontaneous breathing. It usually is indicated only in patients who are apneic, deeply comatose, deeply sedated, or paralyzed. All breaths delivered with CMV are positive pressure breaths; therefore, mean intrathoracic pressure is increased with potential deleterious hemodynamic effects. Most notable of these is the impedance of venous return, thus effectively causing a relative hypovolemia and decreased cardiac output (43).
Intermittent Mandatory Ventilation
Intermittent mandatory ventilation (IMV) combines mechanical ventilatory breaths with spontaneous breathing, and is better tolerated than CMV by most patients (61). Allowing some spontaneous breathing reduces mean intrathoracic pressure, which increases venous return and maintains better cardiac output. It also may reduce the incidence of barotrauma. The numbers of mechanical breaths used are those necessary to supplement the patient's own spontaneous ability to maintain adequate minute ventilation. As the patient is recovering, the ventilator rate is gradually reduced by one to two breaths per minute down to a minimum of two breaths per minute. IMV remains the mainstay of our ventilator support. It may be coupled with other modes such as pressure support ventilation (PSV).
Pressure Support Ventilation
The primary benefit of this ventilatory mode is to reduce the inspiratory work of breathing. The patient maintains control of the inspiratory-to-expiratory ratio, inspiratory time, and frequency during the spontaneous efforts. The mechanical breath delivered during PSV usually discontinues once flow decreases to 25% of peak inspiratory flow. Adjustable pressure support parameters include the time necessary to reach maximal flow or rate of rise of PSV. A shorter pressure rise time is generally used to reduce work of breathing in patients with the highest inspiratory flow demand.
PSV has the capability to reduce or eliminate both imposed (apparatus and airway resistance) and physiologic (lung and chest wall static compliance) work of breathing. Therefore, by choosing the appropriate level of PSV, the clinician may reduce or eliminate the extra imposed work of breathing and keep the physiologic work of breathing within tolerable limits.
In spite of a careful, stepwise approach to mechanical ventilation in ALI and ARDS, iatrogenic complications are frequent. Several potentially protective measures have been evaluated to reduce the incidence of barotrauma from increased peak airway pressure in patients with severely reduced total lung compliance; however, strong evidence in favor of their use is still lacking.
Nitric Oxide
Inhaled nitric oxide (NO) appears to act selectively on the pulmonary vascular bed and only in those areas associated with adequate ventilation, locally reversing hypoxic pulmonary vasoconstriction and increasing oxygenation. However, outcome in terms of mortality or number of days alive and off mechanical ventilation between patients treated with NO and those not treated has not changed when the effect of NO is studied in a prospective randomized fashion (62). Nevertheless, reducing the level of mechanical ventilatory support or FiO2 needed to achieve adequate oxygenation is a potential benefit that could reduce barotrauma and the side effects of treatment.
Prone Positioning
Rotation of patients from supine to prone may cause rapid improvement in oxygenation that may last for up to 12 hours (63). With this maneuver, there is a relatively high risk of inadvertent extubation and removal of invasive monitors; nonetheless, oxygenation improves mainly because the nondependent dorsal portion of the lung has a higher air-to-tissue ratio (64). Obviously, the risks and benefits need to be considered before using this technique in any specific patient.
Bronchodilator Therapy
Small airway closure has been shown to occur even with aspiration of relatively small amounts of water (25). Thus, bronchodilator therapy should be considered in patients when bronchospasm is thought to be present.
Corticosteroids
The rationale for use of corticosteroids in ARDS seems to be limited to the fibroproliferative phase to reduce the incidence of pulmonary fibrosis (65). However, its efficacy for use in drowning victims has not been shown either in large retrospective clinical studies (4) or in prospective animal studies (66). Corticosteroids can interfere with normal pulmonary healing and increase the rate of sepsis. Corticosteroids have been associated with higher mortality in one study, probably due to the immunosuppressant effect in patients with sepsis (67). In another, their use has shown, after aspiration of gastric contents, to increase pulmonary granuloma formation (68,69).
Surfactant
ARDS from drowning involves both quantitative (sea water) and qualitative (fresh water) alterations in lung surfactant (33,70). Although the use of exogenous surfactant has been shown to lower mortality in neonates with respiratory distress syndrome (71), this effect in adults has been disappointing, and its prohibitive cost makes its use infrequent (72).
Prophylactic Antibiotics
The use of broad-spectrum antibiotics may enhance the emergence of resistant organisms. An exception represents survival from drowning in heavily contaminated water such as stagnant ponds or public spas, where Pseudomonas species are endemic. Our initial choice in this situation is usually a fourth-generation cephalosporin with broad Gram-negative coverage. In other patients, antibiotics are not recommended unless the patient develops evidence of infection, in which case cultures and sensitivities will guide the choice of antibiotics to be given.
Cardiovascular Support
By the time a drowning victim reaches the intensive care unit, cardiac arrhythmias are rarely a problem. If witnessed in the emergency department or the ICU, the most common cause of arrhythmias is severe hypoxia, and providing adequate ventilation and oxygenation will usually restore a normal rhythm. If not, drug therapy or, in the case of severe ventricular arrhythmias, electrical intervention is appropriate.
Hypotension may require initial pharmacologic support, but it should be remembered that the hypotension seen in drowning victims is predominantly due to fluid shifts resulting in hypovolemia (26,43). This hypovolemia may be accentuated when mechanical ventilatory techniques that increase mean intrathoracic pressure are used (43).
Experimental studies have shown that, whereas mechanical ventilation and CPAP will decrease intrapulmonary shunt and increase PaO2, because of the detrimental effect on cardiac output, tissue perfusion is compromised. In one study, attempting to increase oxygen delivery by use of vasopressors and inotropes was not productive, but fluid administration to increase blood volume resulted in an increased cardiac output and oxygen delivery (43).
Precise fluid replacement is dependent on an accurate assessment of effective circulating blood volume. To this end, monitoring the patient with a pulmonary artery catheter or transesophageal echocardiography is extremely helpful.
Central Nervous System Support
The two most important factors influencing morbidity and mortality in victims surviving drowning are severe respiratory insufficiency and permanent neurologic impairment secondary to cerebral hypoxia. Despite improvement in emergency and intensive pulmonary and cardiovascular care, neurologic outcome in drowning patients is directly related to the initial duration of hypoxia from the onset of submersion until effective cardiopulmonary resuscitation (CPR) is provided. The Glasgow coma scale (GCS) score mirrors this relationship during the first few hours after submersion.
The most common cerebral lesion results from cytotoxic injury due to global central nervous system (CNS) hypoxemia. Cerebral edema, which usually is not clinically evident or is mild on presentation, reaches its peak by day 2 to 3 after the submersion event. It is understood that successful intensive care management of these patients reflects the ability to control the intracranial pressure (ICP) and limit secondary brain injury from inadequate cerebral perfusion and hypoxia through standard protocols. Therefore, monitoring of the intracranial pressure is often recommended in patients with a GCS score compatible with severe central nervous system injury (8 and below), in conjunction with what is described in detail in the neurosurgical guidelines for traumatic brain injury (73). Unfortunately, monitoring of ICP has not been shown to increase normal survival after drowning.
Seizure prophylaxis is immediately initiated in patients with CNS compromise, and ventilatory rate is titrated to achieve a PaCO2 of 35 to 40 mm Hg. Although a chronic lower level of PaCO2 has been often used in the past to decrease intracranial pressure, it is no longer recommended except for only a short period of time and only in patients with an acute increase of ICP refractory to pharmacotherapy or ventriculostomy drainage while a definitive imaging diagnosis is in process. In fact, chronic hyperventilation, while decreasing the ICP, can be accompanied by a reduction in cerebral blood flow (74), which can result in worsening cerebral ischemia (75). There are no data to support the use of barbiturates or steroids to lower refractory ICP (51). Despite adequate control of the intracranial pressure and maintenance of the cerebral perfusion pressure with aggressive brain resuscitation modalities, the majority of patients who were severely comatose upon arrival to the ICU died or left the ICU in a persistent vegetative state, because the damage from the initial event was so severe that it was irreversible (49,50).
Control of Blood Glucose Levels
Aggressive blood glucose control (less than 110 mg/dL) with insulin infusion has been associated with a reduced mortality, from 8% to 4.3%, when compared with intermittent doses of subcutaneous regular insulin in a heterogeneous large group of critically ill patients that were prospectively randomized (76). While this study included a variety of patients admitted to the intensive care unit with hypoxic or hypercapnic respiratory failure, the reduction in mortality from multiple organ failure suggests a possible benefit in patients surviving episodes of drowning who require prolonged ICU hospitalization. Interestingly, critical illness polyneuropathy was reduced 44% in the insulin infusion group. It is our practice to control the blood sugar level with an insulin infusion in any critically ill patient with a level above normal. A glucose-based crystalloid infusion is used when blood sugar is below 200 mg/dL to limit the risk of hypoglycemia. Blood glucose level is usually checked every 1 or 2 hours. Recent data (77), however, suggest that there may not be a positive difference if the glucose level is controlled at 110 mg/dL as compared to 150 mg/dL. Rather, the more intensive glucose control may lead to more episodes of hypoglycemia. While not advocating out-of-control glucose values, we are actively re-evaluating the appropriate level of control to values between 130 and 150 mg/dL.
Renal Support
Albuminuria, hemoglobinuria, oliguria, and anuria, while rare, have all been described in drowning victims secondary to acute tubular necrosis from hypoxemia, rhabdomyolysis, or both. Hypothermia leads to reduced blood flow to the skin and muscle, preserving core temperature and central organ perfusion. The acute pathophysiology of acute rhabdomyolysis is probably secondary to tissue hypoxia from acute vessel constriction due to the competitive need for heat conservation. Skeletal myolysis and increased circulating myoglobin will result. Acute renal failure may be aggravated by acute tubular necrosis secondary to hemodynamic instability.
Acute tubular necrosis and rhabdomyolysis require early and vigorous treatment directed at correcting hypovolemia, improving oxygenation, and enhancing heme protein elimination. Volume replacement therapy aims to restore normal blood flow and enhance renal oxygen supply. The medullary ascending limb of Henle loop is most vulnerable to hypoxic injury. Invasive monitoring may be necessary to provide adequate intravascular volume. A central venous pressure or pulmonary wedge pressure around 15 mm Hg is a reasonable hemodynamic goal if ventricular function is normal. Higher pressures may be necessary in patients with a significant increase in mean intrathoracic pressure. Right ventricular ejection fraction, a pulmonary artery catheter, or transthoracic or transesophageal echocardiography can be used if the interpretation of preload by invasive monitoring is difficult, as often is the case in patients requiring major ventilator support. The window of opportunity for restoration of intravascular volume and volume expansion is likely within 6 hours or less of the acute event.
If rhabdomyolysis is present, enhancing the elimination of heme protein helps to limit tubular damage. Systemic alkalinization of the urine with sodium bicarbonate increases the solubility and, therefore, the elimination of heme protein (78). A urine pH between 7 and 8 produces a myoglobin solubility of around 80% and is a reasonable goal. However, in a patient with low urine output, massive doses of sodium bicarbonate may be associated with volume overload secondary to an acute increase in intravascular osmolarity (79). In these cases, when the hemodynamic goal is mild hypervolemia, the weak diuretic acetazolamide may be a valid alternative. Acetazolamide increases the excretion of bicarbonate in urine as a result of the inhibition of the carbonic anhydrase enzyme; for regulatory reasons, the drug will no longer be available in the United States within the next year. However, diuretics, particularly in patients on significant ventilatory support, may adversely affect venous filling and cardiac output. Three other therapeutic agents have been used successfully to preserve renal function in patients with acute rhabdomyolysis: dopamine, loop diuretics, and mannitol. All three drugs enhance recovery of renal function by optimizing the relationship between renal oxygen supply and demand after a hypoxic insult (80).
Manipulating the renal output by means of significantly altering the effective circulating blood volume in drowning victims frequently has a detrimental effect on pulmonary and cardiovascular function. Therefore, a fine-tuned balancing act is frequently required to not adversely affect one organ system while treating another.
Other Concerns
Severe metabolic acidosis from low systemic oxygen delivery and resulting anaerobic metabolism should be corrected. We recommend correction of the base deficit with bicarbonate or acetate solutions to maintain a pH no lower than 7.20. Mechanical ventilation is adjusted frequently with the help of arterial blood gas determinations to maintain a PaCO2 between 35 and 40 mm Hg. Lactic acid levels are checked frequently for a few hours after resuscitation. In fact, while base deficit and single absolute levels of lactic acidosis do not necessarily correlate with the development of multiple organ failure and survival, the rate of lactic acid clearance does (81). Because significant electrolyte abnormalities requiring specific therapy rarely are observed in the drowning victim, normal saline is given as replacement fluid in drowning victims. Isotonic solution also provides less chance of aggravating cerebral edema.
Summary
An awareness of the hidden dangers of recreational activities in and around water, and close supervision of infants, children, and adolescents are the secrets to preventing a significant number of drowning incidents. Swimming pools should be enclosed by security fences to prevent small children from entering the water inadvertently or unsupervised. By identifying age-related drowning risks, communities can reduce drowning rates. Effective CPR and water safety skills should be encouraged in the community, particularly for parents with small children who own home pools. Furthermore, children who can swim should never do so alone or without adult supervision. Everyone participating in water sports should wear an approved personal flotation device. Adolescents need to be taught to swim and informed about the dangers of alcohol and other drug consumption during water sport activities. Between 13 and 19 years of age, risk-taking behavior increases significantly in boys; therefore, extra counseling is warranted. Alcohol should never be consumed, regardless of age, while swimming or engaging in water sports. Swimming with a partner is particularly important for individuals with medical conditions that may abruptly alter their level of consciousness, such as seizure disorders, cardiac disease, and several metabolic diseases. Emergency gear for rescuing and resuscitating drowning victims should be readily available at the poolside. The specific gear required may vary with the size, access, and ownership of the facility.
The community expects the government to enforce safety rules, promote health education through medical and nonmedical personnel, and punish individuals who transgress basic safety rules and regulations. Despite recent advances in cardiopulmonary resuscitation and more sophisticated intensive care medicine, drowning victims with poor Glasgow coma scale scores have a high likelihood of living in a vegetative state as a result of the initial injury. When this occurs, making life or death decisions regarding withdrawal of life support by relatives and health professionals represents a significant stressful event. At the time of this writing, prevention is still the most fundamental way to limit neurologic disasters from drowning.
References
1. Idris AH, Berg R, Bierens J, et al. Recommended guidelines for uniform reporting of data from drowing; the “Utstein style”. Circulation. 2003 Nov 18; 108(20):2565–2574.
2. Craig AB Jr. Causes of loss of consciousness during underwater swimming. J Appl Physiol. 1961;16:583–586.
3. Swann HG. Resuscitation in semi-drowning. In: Whittenberg JF, ed. Artificial Respiration: Theory and Application. New York: Harper and Roe; 1962:202–224.
4. Modell JH, Graves SA, Ketover A. Clinical course of 91 consecutive near-drowning victims. Chest. 1976;70:231–238.
5. Lowson JA. Sensations in drowning. Edinburgh Med J. 1903;13:41–45.
6. Fuller RH. The 1962 Wellcome prize essay. Drowning and the post-immersion syndrome. A clinicopathologic study. Milit Med. 1963;129:22–36.
7. Modell JH, Davis JH. Electrolyte changes in human drowning victims. Anesthesiology. 1969;30:414–420.
8. Modell JH, Moya F, Newby EJ, et al. The effects of fluid volume and seawater drowning. Ann Intern Med. 1967;67:68–80.
9. Modell JH, Moya F. Effects of volume of aspirated fluid during chlorinated fresh water drowning. Anesthesiology. 1966;27:662–672.
10. Modell JH, Gaub M, Moya F, et al. Physiologic effects of near-drowning with chlorinated fresh water, distilled water and isotonic saline. Anesthesiology. 1966;27:33–41.
11. Leach RM, Treacher DS. ABC of oxygen: oxygen transport—2. Tissue hypoxia. BMJ. 1998;317:1370–1373.
12. Craig AB Jr. Underwater swimming and loss of consciousness. JAMA. 1961;176:255–258.
13. Ferrett G, Costa M, Ferrigno M, et al. Alveolar gas composition exchange during deep breath-hold diving and dry breath holds in elite divers. J Appl Physiol. 1991;70:794–802.
14. Gray SW. Respiratory movement of rat during drowning and influence of water temperature upon survival after submersion. Am J Physiol. 1951;167:95–102.
15. Kvittingen TD, Naess A. Recovery from drowning in fresh water. BMJ. 1963;1:1315–1317.
16. Movritzen CV, Andersen MN. Myocardial temperature gradients and ventricular fibrillation during hypothermia. J Thorac Cardiovasc Surg. 1965;49:937–944.
17. Cot C. Les asphyxies accidentecelles (submersion, electrocution, intoxication, oxycarbonique) etude clinique, therapeutique et preventive. Paris: Editions Medicales N. Maloine; 1931.
18. Modell JH. The Pathophysiology and Treatment of Drowning and Near-drowning. Springfield, IL: Charles C. Thomas; 1971:9.
19. Modell JH. Drown vs. near-drown: a discussion of definitions, editorial. Crit Care Med. 1981;9:341–352.
20. Modell JH, Bellefleur M, Davis JH. Drowning without aspiration: is this an appropriate diagnosis? J Forensic Sci. 1999;44(6):119–123.
21. Loughead DW, Janes JM, Hall GE. Physiological studies in experimental asphyxia and drowning. Can Med Assoc J. 1939;40:423–428.
22. Swann HG, Spafford NR. Body salt and water changes during fresh and sea water drowning. Texas Rep Biol Med. 1951;9:356–382.
23. Swann HG, Brucer M, Moore C, et al. Fresh water and sea water drowning. A study of the terminal cardiac and biochemical events. Texas Rep Biol Med. 1947;5:423–437.
24. Halmagyi DFJ, Colebatch HJH. Ventilation and circulation after fluid aspiration. J Appl Physiol. 1961;16:35–40.
25. Colebatch HJH, Halmagyi DFJ. Lung mechanics and resuscitation after fluid resuscitation. J Appl Physiol. 1961;16:684–696.
26. Redding JS, Voight GC, Safer P. Treatment of sea water aspiration. J Appl Physiol. 1960;15:1113–1116.
27. Colebatch HJH, Halmagyi DFJ. Reflex pulmonary hypertension of fresh water aspiration. J Appl Physiol. 1963;18:179–185.
28. Colebatch HJH, Halmagyi DFJ. Reflex airway reaction to fluid aspiration. J Appl Physiol. 1962;17:787–794.
29. Fainer DC, Martin CG, Ivy AC. Resuscitation of dogs from fresh water drowning. J Appl Physiol. 1951;3:417–426.
30. Redding JS, Cozine RA. Restoration of circulation after fresh water drowning. J Appl Physiol. 1961;16:1071–1074.
31. Redding JS, Voight GC, Safer P. Drowning treated with intermittent positive pressure breathing. J Appl Physiol. 1960;15:849–854.
32. Modell JH, Moya F, Williams HD, et al. Changes in blood gases and AaDO2 during near-drowning. Anesthesiology. 1968;29:456–465.
33. Giammona ST, Modell JH. Drowning by total immersion. Effects on pulmonary surfactant of distilled water, isotonic saline and sea water. Am J Dis Child. 1967;114:612–616.
34. Spitz WV, Blanke RV. Mechanism of death in fresh-water drowning. I. An experimental approach to the problem. Arch Path (Chicago). 1961;71:661–668.
35. Modell JH, Davis JH, Giammona ST, et al. Blood gas and electrolyte changes in human near-drowning victims. JAMA. 1968;203:337–343.
36. Fainer DC. Near-drowning in sea water and fresh water. Ann Intern Med. 1963;59:537–541.
37. Hasan S, Avery WE, Fabian C, et al. Near-drowning in humans. A report of 36 cases. Chest. 1971;59:191–197.
38. Fuller RH. The clinical pathology of human near-drowning. Proc Roy Soc Med. 1963;56:33–38.
39. Moritz AR. Chemical methods for the determination of death by drowning. Physiol Rev. 1944;24:70–88.
40. Butt MP, Jalowayski A, Modell JH, et al. Pulmonary function after resuscitation from near-drowning. Anesthesiology. 1970;32:275–277.
41. Modell JH, Kuck EJ, Ruiz BC, et al. Effect of intravenous vs. aspirated distilled water on serum electrolytes and blood gas tensions. J Appl Physiol. 1972;32:579–584.
42. Bergquist RE, Vogelhut MM, Modell JH, et al. Comparison of ventilatory patterns in the treatment of fresh water near-drowning in dogs. Anesthesiology. 1980;52:142–148.
43. Tabeling BB, Modell JH. Fluid administration increases oxygen delivery during continuous positive pressure ventilation after freshwater near-drowning. Crit Care Med. 1983;11:693–696.
44. Yag IR, Stalnikowicz R, Michaeli J. Near-drowning in the Dead Sea—electrolyte imbalances and therapeutic implications. Arch Intern Med. 1985;145:50–53.
45. Modell JH. The Pathophysiology and Treatment of Drowning and Near-drowning. Springfield, IL: Charles C. Thomas; 1971:61–66.
46. Ellis J. National Pool and Water Park Lifeguard Training Manual. Boston: Jones & Bartlet; 2000.
47. 2005 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2005;112(24):IV-133–IV-135.
48. Rosen P, Stoto M, Harley J, et al., eds. The Use of the Heimlich Maneuver in Near-Drowning. Washington DC: Institute of Medicine Committee on the Treatment of Near-Drowning Victims.; 1994.
49. Conn A, Montes J, Barker G. Cerebral salvage in near-drowning following neurologic classification by triage. Can J Anaesth. 1980;27:201–210.
50. Modell JH, Graves SA, Kuck EJ. Near-drowning: correlation of level of consciousness and survival. Can Anaesth Soc J. 1980;27:211–215.
51. Bohn DJ, Biggart WD, Smith CR, et al. Influence of hypothermia, barbiturate therapy and intracranial pressure monitoring on morbidity and mortality after near-drowning. Crit Care Med. 1986;14:529–534.
52. Orlowski JP, Szpilman D. Pediatric critical care: drowning. Rescue, resuscitation and reanimation. A new millennium. Pediatr Clin North Am. 2001;48:627–646.
53. Lee KH. A retrospective study of near drowned victims admitted to the intensive care unit. Ann Acad Med Singapore. 1998;27:344–346.
54. Goodwin SR. Aspiration syndromes. In: Civetta JM, Taylor RW, Kirby RR, eds. Critical Care. Philadelphia: Lippincott-Raven; 1997:1861–1875.
55. Bachofen M, Weibel ER. Alterations of the gas exchange apparatus in adult respiratory insufficiency associated with septicemia. Am Rev Respir Dis. 1977;116:589–615.
56. Ruiz BC, Calderwood HW, Modell JH. Effect of ventilatory patterns on arterial oxygenation after near-drowning with fresh water. A comparative study in dogs. Anesth Analg. 1973;52:570–576.
57. Suter PS, Fairley HB, Isenberg MD. Optimum end-expiratory airway pressure in patients with acute pulmonary failure. N Engl J Med. 1975;292:284–289.
58. Modell JH, Calderwood HW, Ruiz BC, et al. Effects of ventilatory patterns on arterial oxygenation after near-drowning in sea water. Anesthesiology. 1974;40:376–384.
59. Antonelli M, Conti G, Rocco M, et al. A comparison of noninvasive positive-pressure ventilation and conventional mechanical ventilation in patient with acute respiratory failure. N Eng J Med. 1998;339:429–435.
60. Rabatin JT, Gay PC. Noninvasive ventilation. Mayo Clin Proc. 1999;74:817–820.
61. Downs JB, Klein EF Jr, Desautels D, et al. Intermittent mandatory ventilation: a new approach to weaning patients from mechanical ventilators. Chest. 1973;64:331–335.
62. Dellinger RP, Zimmerman JL, Taylor RW et al. Effects of inhaled nitric oxide in patients with acute respiratory distress syndrome: results of a randomized phase II trial. Crit Care Med. 1998;26:15–23.
63. Jolliet P, Bulpa P, Chevrolet J. Effects of the prone position on gas exchange and hemodynamics in severe acute respiratory distress syndrome. Crit Care Med. 1998;26:1977–1985.
64. Pelosi P, Tubiolo D, Mascheroni D, et al. Effects of the prone position on respiratory mechanics and gas exchange during acute lung injury. Am J Respir Crit Care Med. 1998;157:387–393.
65. Meduri GU, Headley AS, Golden E, et al. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: A randomized controlled trial. JAMA. 1998;280:159–165.
66. Calderwood HW, Modell JH, Ruiz BC. The ineffectiveness of steroid therapy for treatment of fresh-water near-drowning. Anesthesiology. 1975;43:642–650.
67. Bone RC, Fisher CJ Jr, Clemmer TP et al. Early methylprednisolone treatment for septic syndrome and the adult respiratory distress syndrome. Chest. 1987;92:1032–1036.
68. Wynne JW, Modell JH. Respiratory aspiration of stomach contents. Ann Intern Med. 1977;87:66–474.
69. Wynne JW, Reynolds JC, Hood CI, et al. Steroid therapy for pneumonitis induced in rabbits by aspiration of food stuff. Anesthesiology. 1979;51:11–19.
70. Petty TL, Reiss OK, Paul GW, et al. Characteristics of pulmonary surfactant in adult respiratory distress syndrome associated with trauma and shock. Am Rev Respir Dis. 1977;115:531–536.
71. Corbet A, Bucciarelli R, Goldman S, et al. Decreased mortality among small premature infants treated at birth with a single dose of synthetic surfactant: a multicenter controlled trial. J Pediatr. 1991;118:227–234.
72. Anzueto A, Baughman RP, Guntupalli KK, et al. Aerosolized surfactant in adults with sepsis-induced acute respiratory distress syndrome. N Engl J Med. 1996;334:1417–1421.
73. Bullock RM, Chestnut RM, Clifton GL, et al. Guidelines for the management of severe traumatic brain injury. J Neurotrauma. 2000;17:451–627.
74. Fortune JB, Feustel PJ, deLuna C, et al. Cerebral blood flow and blood volume in response to O2 and CO2 changes in normal humans. J Trauma. 1995;39:463–471.
75. Weckesser M, Posse S, Oltholf U, et al. Functional imaging of the visual cortex with bold contrast MRI: hyperventilation decreases signal response. Magn Reson Med. 1999;41:213–216.
76. Van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in critically ill patients. N Engl J Med. 2001;345:1359–1367.
77. Brunkhorst FM, Engel C, Bloos F, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med. 2008;358:125–139.
78. Better OS, Stein JH. Early management of shock and prophylaxis of acute renal failure in traumatic rhabdomyolysis. N Engl J Med. 1990;322:825–829.
79. Eneas JF, Schoenfeld PY, Humphreys MH. The Effect of infusion of mannitol-sodium bicarbonate on the clinical course of myoglobinuria. Arch Intern Med. 1979;139:801–805.
80. Gelman S. Preserving renal function during surgery. In: International Anesthesia Research Society Review Course Lectures. Baltimore: Williams and Wilkins; 1992:88–92.
81. Bakker J, Gris P, Coffernils M, et al. Serial blood lactate levels can predict the development of multiple organ failure following septic shock. Am J Surg. 1996;171:221–226.