Type of aspirated fluid

There are several reports in the literature regarding physiologic differences in electrolytes and blood volume after seawater and freshwater drowning. Aspiration of more than 11 mL of fluid per kg of body weight is required for blood volume to be altered, and aspiration of more than 22 mL/kg is necessary before significant electrolyte changes occur. ^, Drowning victims are often hypovolemic regardless of the type of aspiration because of increased capillary permeability resulting from asphyxia and the loss of protein-rich fluid into the alveoli. In a canine model of drowning, aspiration of cold fresh water led to a 16% increase in body weight, compared with 6% for saltwater aspiration. Reported outcomes in freshwater drowning are worse than in saltwater drowning. In a meta-analysis of seven studies with 2163 drowning victims, saltwater drowning compared to freshwater drowning was associated with a favorable outcome, with a relative risk of 1.16 (95% confidence interval, 1.08–1.24). Also, hypoxia, as assessed by partial pressure of arterial O ₂ /fraction of inspired O ₂ (Pa o ₂ /F io ₂ ) ratios, was worse in a matched cohort of freshwater versus saltwater adult drowning victims upon ICU admission. This is consistent with the known effect of the aspirated fluid on pulmonary surfactant, which is destroyed by fresh water but only diluted by salt water.

Pulmonary effects

Functional residual capacity (FRC) is the only source of gas exchange at the pulmonary capillary level in the submerged state. Increased metabolic demands from struggling, breath holding, a depletion of FRC from breathing efforts, and aspiration of fluid into the lungs all result in seriously compromised O ₂ uptake and carbon dioxide (CO ₂ ) elimination, with consequent hypoxia and hypercarbia. Between 10% and 15% of drowning victims have severe laryngospasm after submersion, resulting in fatal asphyxiation without aspiration of significant water into their lungs (see Fig. 115.1 ). There is often a combined respiratory and metabolic acidosis caused by hypercapnia and anaerobic metabolism. Patients without significant fluid aspiration recover from asphyxia rapidly if they are successfully resuscitated before cardiac arrest or irreversible brain damage occurs. Aspiration of fluid, however, results in persistently abnormal gas exchange. Aspiration of as little as 1 to 3 mL per kg body weight results in profound impairment of gas exchange. ^, Soon after the aspiration of fluid, there is an elevation of partial pressure of arterial carbon dioxide (Pa co ₂ ) and a fall in Pa o ₂ as a result of airway obstruction, hypoventilation, and impaired gas exchange between alveoli and pulmonary capillary blood. With adequate resuscitation, normocapnia or even hypocapnia is usually achieved while hypoxemia persists, indicating a significant ventilation/perfusion mismatch and diffusion defect leading to intrapulmonary shunting and venous admixture.

The surfactant system of the lung is affected differently in freshwater versus seawater aspiration. Freshwater aspiration results in marked disruption of the surfactant system of the lung, leading to alveolar instability and atelectasis. Sea water, because of its hypertonicity, draws water into the alveoli. Although surfactant may be diluted by the presence of sea water in the alveoli, its surface tension properties are not significantly altered. Zhu and colleagues examined serum levels of pulmonary surfactant-associated protein (SP-A) and lung weights in 53 victims of fatal drowning. They showed significantly heavier lungs in those who drowned in sea water versus fresh water, suggesting an osmolar effect. Although they found no difference in serum SP-A, intraalveolar aggregates of pulmonary surfactant-associated protein were noted more frequently in those who drowned in fresh water. This is likely related to the disruption of surfactant noted in freshwater drowning. Karch demonstrated marked changes in the pulmonary vasculature in rabbits within 30 minutes of aspiration of both fresh water and salt water. Mitochondrial swelling and disruption of pulmonary vascular endothelial cells were consistently observed in these experiments. Clinically, pulmonary abnormalities are encountered in both freshwater and seawater aspiration. These are consistent with pronounced injury to alveoli and pulmonary capillaries resulting in increased membrane permeability, exudation of proteinaceous material in alveoli, pulmonary edema, decreased lung compliance, and increased airway resistance. The extent of these abnormalities may not be manifested fully for several hours after the submersion episode and may be progressive in nature. Acute respiratory distress syndrome (ARDS) is the hallmark of delayed pulmonary insufficiency resulting from aspiration in drowning . This is characterized by progression of ventilation/perfusion mismatch, alveolar-capillary block, increased capillary permeability, and pulmonary edema. Reduced FRC and diffusion barrier resulting from accumulation of fluid and inflammatory cells in the alveoli and interstitium further accentuate hypoxemia. Aspiration of stomach contents and other debris such as sand, mud, and algae may also impair gas exchange. Bacterial pneumonia resulting from aspiration of contaminated water may further contribute to pulmonary insufficiency. Factors that contribute to drowning-associated pneumonia include aspiration of vomitus or aspiration of contaminated material from water that may contain sewage. Water temperature plays a role, with warmer temperatures predisposing to a higher number of organisms. The chemical composition of the water—such as pH, salt content, and presence of organic and inorganic substances—influences bacterial growth as well. Implicated organisms include aerobic gram-negative bacteria, such as Enterobacter spp., Klebsiella spp., and Pseudomonas spp., as well as gram-positive bacteria, including Streptococcus pneumoniae and Staphylococcus aureus . Oropharyngeal flora, as well as pathogens present in the body of water that the child drowned in, have been implicated in drowning-associated pneumonia. ^, There have been reports of multidrug-resistant bacteria and fungal pneumonia as well. The diagnosis of pneumonia is based on clinical parameters such as fever, leukocytosis, and new infiltrates on chest radiographs. When pneumonia is suspected, respiratory cultures are a valuable tool to guide antimicrobial therapy.

Understanding the alterations in pulmonary mechanics is important in order to provide the necessary support in the least injurious fashion. Predominant manifestations are those of ARDS complicating the drowning event. Although pulmonary involvement is often bilateral and diffuse, there is considerable inhomogeneity, with some areas more affected than others. Overall, lung compliance is reduced, necessitating higher inflation pressures to maintain adequate tidal volume (V t ). Low V t at low FRC leads to a vicious cycle of atelectasis, decreased compliance, and further reduction in V t . Critical opening pressure necessary to begin alveolar inflation is increased. Appropriate positive end-expiratory pressure (PEEP) needs to be administered to maintain the necessary FRC for adequate oxygenation and ventilation above the critical opening pressure. Airway resistance is relatively less affected or only minimally elevated unless there is airway obstruction from aspirated debris. Time constant, a product of compliance and resistance, reflects the time needed for pressure equilibration between the proximal airway and alveoli to occur. In ARDS, time constant is decreased, allowing for quicker approximation of pressures at these sites during the inspiratory and expiratory phases of the mechanical ventilation. Relatively large V t (10–12 mL/kg) is associated with greater ventilator-associated lung injury in ARDS, whereas smaller V t (6 mL/kg) is associated with less volutrauma. Because of the short time constant, prolongation of inspiratory time to improve oxygenation and increasing the respiratory rate for CO ₂ elimination often are effective options during mechanical ventilation. A more detailed description of management of ARDS in drowning appears later in this chapter.

Cardiovascular effects

Profound cardiovascular instability is often encountered after a severe drowning event, which poses an immediate threat to survival after the initial rescue. The hypoxemia that occurs due to ventilation-perfusion mismatch can cause life-threatening dysrhythmias, such as ventricular tachycardia, ventricular fibrillation, and asystole. The two determinants of O ₂ delivery—namely, cardiac output and arterial O ₂ content—can be adversely affected by the drowning event. A decrease in Pa o ₂ , if sufficiently severe, decreases O ₂ saturation and therefore arterial O ₂ content. This decrease in arterial O ₂ content can cause a decrease in myocardial O ₂ delivery, which contributes to worsening cardiac output and decreased myocardial perfusion pressure. Smooth muscle contraction banding in the media of major coronary arteries and local ventricular myocytes with focal myocardial necrosis have been described after a drowning episode. ^, Cytosolic calcium overload and O ₂ -derived free radicals have also been implicated in the mechanism of myocardial injury after resuscitation following cardiac arrest. Cardiogenic shock may result from hypoxic damage to the myocardium. Metabolic acidosis may further impair myocardial performance. Additionally, therapeutic application of PEEP decreases venous return and right and left ventricular preload. Right ventricular afterload is also increased by structural pulmonary microvascular damage and humoral inflammatory mediators involved in ARDS. The right ventricle is anatomically designed to tolerate increased preload, but it is not as tolerant of high pressures and afterload as the left ventricle. If pulmonary hypertension is severe enough, left ventricular preload further decreases owing to right ventricular failure. These factors, as well as the excessive permeability of pulmonary and systemic capillaries, result in hypovolemia and decreased left ventricular filling pressures. The end result is that of inadequate supply of O ₂ to tissues to meet their metabolic demands.

Central nervous system effects

Owing to a lack of metabolic substrate reserves, the brain depends on continuous O ₂ delivery. Metabolic failure can begin within seconds after an immediate disruption in circulation. Hypoxia, if sufficiently prolonged, causes profound disturbances of central nervous system (CNS) function. The severity of brain injury depends on the magnitude and duration of hypoxemia and cerebral hypoperfusion, as well as on mechanisms of secondary brain injury. The most important components of brain-specific management include prompt rescue, early and successful resuscitation, and mitigation of secondary injuries.

Following restoration of adequate cerebral O ₂ delivery after the initial hypoxemia and/or hypoperfusion after drowning, there are several mechanisms of secondary brain injury at the tissue, synaptic, and cellular level. At the tissue level, increased intracranial pressure (ICP) and alterations in cerebral blood flow (CBF) can adversely impact local tissue O ₂ delivery. At the synaptic level, the excitotoxic neurotransmitter glutamate can cause an imbalance in the neuronal supply and demand. Delayed neuronal death occurs owing to cellular responses to hypoxia and subsequent reperfusion, leading to complex pathways of pro-survival and pro-death signals. Intracellularly, accumulation of cytosolic calcium, neuronal energy failure associated with DNA damage and repair, and generation of O ₂ -free radicals can all contribute to secondary neuronal death.

There are several developmental factors that render the neurologic effects of pediatric cardiac arrest different from that of adults. Brain injury from sudden cardiac arrest, which more commonly occurs in adults, results when CBF suddenly stops. In contrast, brain injury from drowning, along with most causes of pediatric cardiac arrest, is caused by asphyxia. Cerebral O ₂ delivery continues for some time and progressively decreases because of hypoxemia and decreased CBF. Cardiac arrest may occur but is usually preceded by this period of decreased O ₂ delivery. Neurologic injury is similar to that of other causes of pediatric asphyxia, including foreign body airway obstruction, apnea, asthma, and suffocation. Approximately 46% of all children with cardiopulmonary arrest survive to hospital discharge and only 38% if the arrest occurred outside the hospital. Studies in humans and animals indicate that there are several developmental differences in excitotoxic pathways. In the neonatal period, there is increased vulnerability to N -methyl- d -aspartate receptor activation and glutamate toxicity, as well as heightened capability for apoptosis, which, albeit an important process of normal brain development, may render the immature brain vulnerable to neuronal loss after an insult. In addition, a developmental difference exists in CBF after cardiac arrest. In adult models of cardiac arrest, global hyperemia is present for 15 to 30 minutes after return of spontaneous circulation (ROSC) followed by delayed hypoperfusion that persists for several hours. In an immature animal model of a brief asphyxial cardiac arrest, hyperemia was observed for 10 minutes after ROSC followed by restoration of baseline CBF, whereas prolonged cardiac arrest was followed by hypoperfusion and blood pressure–dependent CBF.

Effects on other organ systems

Drowning victims are at significant risk for the development of hypoxemia and/or ischemia, particularly in the face of cardiac or respiratory arrest. These factors increase the risk for multiple-organ dysfunction syndrome (MODS). In addition to neurologic injury and acute lung injury, hepatic and renal injuries have been observed. Typpo and colleagues reported on the incidence of MODS on day 1 in children admitted to the PICU. Of the 18% of children who developed MODS on the first day, 30% to 35% had an unfavorable neurologic outcome. In 2015, Mtaweh et al. reported on the patterns of MODS in children after drowning. Pediatric Logistic Organ Dysfunction Score-1 (PELOD-1) and Pediatric Multiple Organ Dysfunction Score (P-MODS) were calculated for the first 24 hours of admission. Of the 60 children, 39 (65%) were in respiratory arrest and 21 (35%) were in cardiac arrest at the scene. Seventeen of 60 patients had severe neurologic injury. Of the children who developed MODS, approximately half had been in respiratory or cardiac arrest. Children who had been in cardiac arrest had much more significant organ dysfunction, which included neurologic, respiratory, hepatic, and renal systems. Of those children who developed acute kidney injury (3 in 60), one child required continuous renal replacement therapy and hemodialysis.

Mammalian diving reflex

A certain degree of protection against hypoxia in drowning may occur in the form of a response similar to the diving reflex observed in seals and other air-breathing diving mammals. While the heart, brain, and lungs remain adequately perfused, blood flow to tissues resistant to hypoxia (i.e., gastrointestinal tract, skin, and muscle) is markedly reduced. Significant bradycardia occurs with a reduction in cardiac output. The mammalian diving reflex acts as an O ₂ -conserving adaptation in response to submersion. It has been proposed that this reflex is most active in infancy and is potentiated by fear and low water temperature. A combination of marked bradycardia and impalpable pulses resulting from vasoconstriction may make the victim appear dead at a time when mouth-to-mouth resuscitation could be life-saving. Clinical studies involving children and adults have failed to demonstrate an efficient diving reflex in response to cold-water submersion. ^,

Preexisting associated conditions

An underlying etiologic mechanism should be explored depending on a given clinical scenario involving unexplained drowning. Children with seizure disorders are at greater risk for submersion accidents. Similarly, an occurrence of vasovagal syncope or a hypoglycemic episode during swimming may be the underlying factor responsible for drowning. Occult cardiomyopathy or a cardiac arrhythmia should also be considered in an unexplained drowning event. An episode of drowning might be the first manifestation of LQTS. Ackerman et al. studied blood samples or archived autopsy tissue samples in 35 cases of autosomal-dominant LQTS. Six of these patients had a personal history or extended family history of near-drowning. All of these patients were found to have LQTS causing mutations in KVLQT1 , whereas such an abnormality was found in only 3 of the remaining 29 patients who did not have a personal history of submersion episode. Thus, swimming appears to be a gene-specific ( KVLQT1 ) arrhythmogenic trigger for LQTS. Diagnosis of inherited LQTS allows for identification of other family members with similar affliction. Yoshinaga et al. showed that face immersion in cold water results in abnormal lengthening of the QT interval in children identified with nonfamilial LQTS and that such children could potentially be at risk of a life-threatening arrhythmia during swimming. More recently, Albertella et al. described the presentation of outcome of water-related events in children with LQTS. Of the 10 children identified with LQTS and who had a documented history of water-related syncope, nine had LQTS type 1. Seven were swimming at the time of the syncopal event. Factors leading to arrhythmias during swimming in patients with LQTS are thought to include facial immersion in cold water and the dive reflex, which causes a fall in heart rate, competing with epinephrine release from exercise. Patients with LQTS should avoid swimming and diving and any medications that can prolong the QT interval. It has been shown that β-blockers reduce the risk of sudden death in LQTS by 75%.

A significant risk factor for drowning is epilepsy. Accidental drowning causes 8% of all deaths in children and young adults with epilepsy. These patients have a four-fold higher prevalence of drowning than those without the condition. Unwitnessed bathtub drownings occur in roughly 30% of these cases and are more common in children age 5 years and older. ^, In older children and adolescents with epilepsy who are able to bathe independently, showers are advised over baths to reduce this risk.

Children with intellectual delays may be at reduced risk of drowning because they are more likely to be supervised in pools and tubs. However, there remains the risk of ambulatory children wandering off. One study of children with autism spectrum disorder who drowned found that 74% of these deaths occurred in natural bodies of water (pond, river, lakes) in close proximity to home.

Cold water drowning

Proposed protective mechanisms of hypothermia for the injured brain include a decrease in cerebral metabolic rate, ICP, excitotoxic neurotransmission, cytotoxic edema, generation of deleterious O ₂ -derived free radicals, and cerebral hyperemia, which can cause ischemia/reperfusion injury and further increase ICP. The effects of hypothermia may apply to drowning injury, as good outcomes have been reported in children submerged in ice water (<5°C) for prolonged periods. ^, It appears that if hypothermia is protective in drowning, rapid cooling in ice water is necessary, as a lesser degree of hypothermia in warmer water does not offer cerebral protection. It is also possible that a publication bias exists in reporting of cold water drowning, as cases with good outcome are more likely to be published than cases with poor outcome. Indeed, more recent studies call into question the protective effects of hypothermia in drowning. Quan et al. studied more than 1000 child and adult open-water drowning victims and did not find a protective effect of cold water but rather estimated that submersion duration was the most powerful predictor of outcome. In this study, young age was also a predictor of good outcome, which may reflect the capacity of children to become hypothermic faster and possibly benefit from its neuroprotective effects. A meta-analysis of 24 cohort studies by the same author again reported no change in outcome based on water temperature. Children who were submerged for less than 5 minutes generally had good outcome, and those submerged for more than 25 minutes invariably died. Shorter emergency medical services arrival time was also associated with favorable outcome. This meta-analysis, however, found no correlation between age and outcome.

Management at the scene

Ensuring the adequacy of the airway, breathing, and circulation is the goal of basic life support after the initial rescue. In cases of inadequate airway and cardiopulmonary status, CPR must be instituted immediately. The fundamentals of basic life support are the same after drowning as for any other situation requiring CPR; however, some practical aspects are worth considering. Unlike the out-of-hospital cardiac arrests in adults, cardiac arrests from drowning in children are initiated by hypoxia due to disconnection of atmosphere from the airways and lungs. Therefore, it is important that CPR is performed in the traditional airway-breathing-circulation (ABC) rather than circulation-airway-breathing (CAB) sequence. Although in recent years there has been a push for chest compression–only CPR for out-of-hospital cardiac arrests, rescue breathing as well as chest compressions is essential for successful resuscitation. The aim of resuscitation at the scene is to prevent irreversible tissue injury from prolonged hypoxia and ischemia. After ensuring a safe environment for a rescue attempt and calling for help, the victim should be removed from the water as soon as possible. Wet clothing should be removed to prevent heat loss. Less than 0.5% of drowning victims sustain cervical spine injuries; thus, application of cervical spine precautions is not routinely indicated. Cervical spine immobilization in the water is indicated when head and neck injury is strongly suspected, such as in accidents involving diving, water-skiing, and surfing. In such situations, the preferred airway opening maneuver is anterior displacement of the jaw, rather than extension of the neck. If breathing is either absent or agonal, rescue breathing should be performed even while the victim is in the water if possible, either via bag-valve-mask if available or by mouth-to-mouth. Chest compressions should not be attempted in water because they are ineffective and waste valuable time. After the victim is removed from the water and placed on a hard board, chest compressions should be given at a rate of 100 per minute with a ratio of 15:2. If available, an automatic defibrillator should be applied to detect a shockable rhythm and defibrillate if necessary. Prolonged attempts to remove water from the lungs are futile and may hinder ongoing ventilatory support. Heimlich and Patrick have recommended the use of subdiaphragmatic pressure in drowning victims to remove water from the airway, but this practice is ill advised. In addition to the fact that most drowning victims aspirate relatively small amounts of water, there is no evidence to suggest that the Heimlich maneuver can remove aspirated fresh water or pulmonary edema fluid. On the other hand, such patients frequently swallow large amounts of water. Consequently, increased abdominal pressure may result in regurgitation of gastric contents into the oropharynx and aspiration into the tracheobronchial tree. Any debris observed in the oropharynx should be removed before initiation of mouth-to-mouth breathing. Presence of airway obstruction caused by a foreign body should be suspected if effective chest expansion cannot be accomplished with appropriate ventilatory technique. A subdiaphragmatic thrust in such a situation would be indicated. As soon as the equipment becomes available, ventilation with 100% oxygen via bag-valve mask device should be initiated for patients who are not breathing adequately. Pressure used during resuscitation to inflate the lungs of drowning victims may have to be higher than anticipated because of reduced compliance of the edematous lungs. A PEEP valve should be used if available. Overinflation should be avoided, however, because this can lead to pulmonary barotrauma and overdistention of the stomach with regurgitation and aspiration of gastric contents. In patients who are too obtunded to maintain airway protection, exhibit hypoxia, or otherwise unable to maintain oxygenation despite bag-valve mask ventilation, endotracheal intubation should be performed.

Emergency department evaluation and stabilization

As with any form of accidental injury, other forms of associated trauma must be considered. Children who slip and fall into the pool may sustain external head injury, such as abrasions, lacerations, and contusions. Occasionally, profuse bleeding from scalp lacerations may be sufficient to aggravate hypovolemic shock. In bathtub drowning, or in instances in which child abuse is suspected, fractures and other evidence of previous injury should be looked for. In adolescent victims, drowning is frequently associated with illicit drug or alcohol use. When appropriate, urine and blood toxicology tests should be performed. Spinal injuries should be suspected in diving, water-skiing, and surfing accidents involving young adults.

The need for hospitalization should be determined by the severity of the drowning episode and clinical evaluation. All patients with a history of drowning should be observed in the emergency department for at least 4 to 6 hours in case they were to develop respiratory distress. Those with insignificant history and normal physical examination may be safely released. Patients with respiratory symptoms, decreased O ₂ saturation indicated by pulse oximetry or blood gas determination, and altered sensorium should be hospitalized. Although the Glasgow Coma Scale was originally designed for use in head trauma, it has been routinely used to assess hypoxic encephalopathy in drowning.

Maintaining adequate airway, respirations, and peripheral perfusion with continued attention to oxygenation, ventilation, and cardiac performance should take priority. Electrocardiographic monitoring and arterial blood gas determination should be performed as soon as possible. Ventricular dysrhythmias, asystole, and hypotension may be encountered during the early resuscitation phase. The standard CPR techniques also apply to the drowned child. Patients with respiratory acidosis and hypoxemia, and those who are unconscious with significant respiratory distress or poor respiratory efforts, require intubation and mechanical ventilation. Early use of PEEP is effective in reversing hypoxemia. Because pulmonary edema is not caused by hypervolemia in drowning, diuretics are not helpful. In addition, they may exacerbate the prevalent hypovolemia. Therefore, pulmonary edema after near-drowning is best treated mechanically with positive-pressure breathing and PEEP rather than diuretics. Hypovolemia is commonly encountered in the early resuscitation phase. Isotonic crystalloids (20 mL/kg) or colloids (10 mL/kg) infused over 15 to 20 minutes should be used for intravascular volume expansion. Additional volume expansion can be carried out based on clinical and hemodynamic status. Administration of large amounts of hypotonic fluid is contraindicated because such solutions are ineffective for intravascular volume expansion. Furthermore, the resultant decrease in serum osmolality may exacerbate cerebral edema. In the face of continued hypotension and/or impaired peripheral perfusion after appropriate intravascular volume expansion, inotropic/vasopressor support may be necessary. Central venous pressure monitoring is extremely helpful for ongoing assessment and management of intravascular volume. Metabolic acidosis resolves with improvement of oxygenation and tissue perfusion. Radiologic studies should include a chest radiograph to determine the presence or absence of pneumothorax or pneumomediastinum. Unless head injury is suspected, a computed tomography scan of the head is usually not necessary because early findings are often normal even in the face of severe hypoxic damage.

Prevention and treatment of hypoglycemia is extremely important. It is reasonable to maintain blood sugar between 80 and 160 mg/dL. The issue of controlling body core temperature also deserves mention. Therapeutic hypothermia after pediatric cardiac arrest has been proven to be of no benefit. However, there may be special considerations in drowning victims since most are relatively hypothermic at the time of rescue. After return of spontaneous circulation, a reasonable approach is to not rewarm the victim aggressively as long as the core temperature is between 32°C and 34°C. Rewarming should occur at a rate no greater than 0.5°C per hour. Hyperthermia (>37°C), on the other hand, must be prevented or treated aggressively if present. ^, Severe bradycardia and intense vasoconstriction associated with marked hypothermia (<32°C) may make drowning victims appear dead. However, resuscitative efforts should be continued while normalizing body temperature.

Management in the intensive care unit

The drowning victim can sustain insults to many organ systems, including the brain, lungs, cardiovascular system, liver, and kidneys. Since the major contributor to mortality and long-term morbidity is cerebral hypoxic-ischemic injury, management in the ICU should focus on supporting these organ systems as they relate to the brain. Continued attention to oxygenation and ventilation status and cardiac performance is essential. ARDS can occur as a result of aspiration injury, pneumonia, and surfactant deficiency or dysfunction. Both hypoxemia and hyperoxemia have been associated with poor outcome after cardiac arrest. Maintenance of O ₂ saturation between 94% and 98% and Pa o ₂ between 70 and 100 are reasonable goals. The cerebral vasculature of the injured brain is reactive to changes in Pa co ₂ . Hypercapnia increases cerebral blood volume and can raise ICP. Conversely, arterial hypocapnia decreases CBF and therefore cerebral O ₂ delivery. Thus, ventilation should be titrated to achieve Pa co ₂ around 40 torr. To monitor these effects, arterial and central venous pressure monitoring are necessary in most patients who require intensive care. A useful parameter to monitor is mixed venous O ₂ saturation (Sv o ₂ ). Provided that arterial O ₂ content and O ₂ consumption remain constant, Sv o ₂ is a useful indicator of changes in cardiac output.

The need for endotracheal intubation and different ventilatory strategies should be determined on an individual basis and by clinical judgment. Respiratory acidosis, PaO ₂ less than 60 torr in F io ₂ greater than 0.5, clinical signs of impending respiratory fatigue, and depressed level of consciousness are the most common indications for mechanical ventilation. Lung injury may not peak until at least 24 hours; thus, weaning from mechanical ventilation should not occur before then. Early use of PEEP and supplemental O ₂ are extremely effective in reversing hypoxemia.

The goal of mechanical ventilation is to provide adequate gas exchange to ensure tissue viability while minimizing the inevitable ventilator-associated injury from oxytrauma, barotrauma, volutrauma, and ineffective tracheobronchial toilet. Ventilatory strategy should take into account the major alterations in pulmonary mechanics. As noted earlier, most children who drown have decreased FRC, compliance and time constant and increased critical opening pressure. Salutary effects of PEEP are from maintaining alveolar stability, alveolar recruitment, and increasing FRC. It stabilizes the relatively softer chest wall of a child, thus minimizing chest wall recoil and further decrease in FRC. PEEP also displaces intraalveolar water into interstitial and perilymphatic spaces, resulting in decreased venous admixture and improved compliance. Excessive PEEP can result in decreased venous return and cardiac output, pulmonary overdistension and decreased compliance, and barotrauma. Maintenance of normovolemia is an important consideration in patients receiving PEEP. Furthermore, excessive PEEP can decrease cerebral venous drainage and increase cerebral blood volume and ICP. Overdistention can also impair ventilation, and the subsequent increase in partial pressure of carbon dioxide (P co ₂ ) can increase ICP. Thus, PEEP should be titrated to maintain lung recruitment, ventilation, and oxygenation while maintaining hemodynamics.

The pulmonary management of pediatric drowning victims should follow the same standard approach as in other children with acute lung injury and ARDS: low V t (6–7 mL/kg), lung recruitment with PEEP, and titration of F io ₂ to achieve adequate O ₂ delivery. We recommend pressure-controlled ventilation with a relatively low peak airway pressure and prolonged inspiration while still allowing adequate time for complete exhalation. Alternatively, pressure-regulated volume control mode can also be used to deliver a preset V t with minimum possible inflation pressure. The level of PEEP can be optimized by gradual increments depending on its effects on increasing Pa o ₂ /F io ₂ ratio. Optimal PEEP as evidenced by improvement in dynamic compliance can be determined by measuring exhaled V t at varying levels of PEEP. When PEEP exceeds critical opening pressure or the lower inflection point on the pressure/volume curve, dynamic compliance improves. Ventilatory rate, inspiratory/expiratory times, and peak airway pressures can also be adjusted according to their effects on dynamic compliance and by ascertaining the return of expiratory flow to baseline. In patients without CNS injury and intracranial hypertension, the technique of permissive hypercapnia can be used to minimize barotrauma in a patient with ARDS. This involves using lower inflation pressures or V t and accepting higher levels of P co ₂ as long as pH remains near normal.

High-frequency ventilation is another strategy that can be used in the management of hypoxic respiratory failure. This mode of ventilation uses a relatively high mean airway pressure while minimizing excessive fluctuations in pressures during the respiratory cycle. High-frequency ventilation is a safe and effective modality in the treatment of severe acute respiratory failure that is unresponsive to conventional mechanical ventilation. ^,

While extracorporeal life support (ECLS) has been used for rewarming in patients with severe hypothermia following drowning in cold water, the routine use of ECLS for the treatment of ARDS associated with drowning is less clear. The presumed benefit of ECLS is the provision of lung rest to mitigate barotrauma and O ₂ toxicity in patients who do not improve despite maximum ventilatory support. Criteria for ECLS in drowning with refractory cardiovascular or respiratory dysfunction should be similar to other disease states.

There is no evidence to suggest that “prophylactic” antibiotics help prevent drowning-associated pneumonia; in addition, they may lead to the selection of resistant organisms. Furthermore, drowning in swimming pools rarely results in pneumonia. However, fulminant S. pneumoniae bacterial sepsis and pneumonia have been described shortly after a severe drowning injury. Therefore, it is reasonable to institute broad-spectrum antibiotic therapy in patients with positive sputum cultures, fever, sepsis, or severe cardiopulmonary deterioration, especially when this occurs after a period of stability.

CNS injury is by far the most important cause of death and long-term functional impairment among the immediate survivors of drowning accidents. Studies have failed to demonstrate beneficial effects in improving outcome of various cerebral protective strategies, such as therapeutic hypothermia, ICP control, corticosteroids, and barbiturate coma. Our experience suggests that significant intracranial hypertension is not commonly encountered in the early postdrowning period, whereas late, uncontrollable intracranial hypertension carries an unfavorable prognosis. Additionally, satisfactory control of intracranial hypertension is not necessarily associated with improved outcome. It appears that the occurrence of cerebral edema and intracranial hypertension 2 to 3 days after drowning is a reflection of the early hypoxic injury rather than a manifestation of a reversible process. Late, persistent intracranial hypertension associated with a comatose state is of ominous significance and is almost always associated with an unfavorable outcome.

The most practical tool of neuromonitoring is the standard neurologic examination. The vast majority of patients who are awake and interactive in the emergency department survive neurologically intact. Admission neurologic examination by itself is not predictive of outcome. However, persistent absence of cognitive function 48 to 72 hours after the drowning episode is of grave prognostic significance. Large and unreactive pupils portend a severe hypoxic ischemic injury and a poor neurologic prognosis. Components of the examination that should be performed serially include pupillary reactivity, level of consciousness, brainstem reflexes, and motor function. The use of ICP monitoring in children with hypoxic-ischemic encephalopathy after drowning is not recommended. The emphasis of management of a comatose child in the immediate post-drowning period should be on maintaining adequate oxygenation/ventilation, O ₂ delivery, and avoidance of hypotonic fluid overload. Pathophysiologic changes from asphyxia—as well as various nonindicated therapies aimed at cerebral salvage, such as barbiturates and osmotic diuresis—adversely affect myocardial performance. Cardiovascular support with maintenance of intravascular volume and the use of inotropic agents is often necessary to maintain optimum organ perfusion in patients who have sustained a significant hypoxic-ischemic insult.

The neuroprotective properties of hypothermia that have been extensively demonstrated in laboratory studies suggest potential merits as a therapy in some children, including victims of drowning. A randomized controlled trial of therapeutic hypothermia has shown benefits in neonatal hypoxic-ischemic encephalopathy. Although therapeutic hypothermia has been shown to be effective in neonatal hypoxic-ischemic injury, studies in pediatric drowning showed that not only did hypothermia not improve outcome, it increased infectious complications, and improved survival to merely a vegetative state. ^{, ,} In cases of cold-water drowning, although the patient should be actively warmed to prevent arrhythmias and secondary infections, once a core body temperature of 30°C is achieved, warming should not exceed 0.5°C per hour to prevent rises in CBF and ICP, ischemia/reperfusion injury, and fever.