Abstract
Drowning is defined as a process resulting in primary respiratory impairment from submersion/immersion in a liquid medium. Worldwide, drowning is the 11th most frequent cause of death in the 0–4 years age group, the third most frequent cause of death in children aged from 5 to 14 years, and the second leading cause of injury related death in childhood. The vast majority of drowning deaths occur in non-Western countries, and as with the majority of accidental deaths, there is a strong male preponderance. The drowning sequence has both pulmonary injury and nonpulmonary sequelae such as hypothermia, electrolyte imbalance, trauma, and hypoxic-ischemic damage. Pulmonary management ranges from the need for some supplemental oxygen to intubation and management of severe acute respiratory distress syndrome. Initial management is often in the field, and care should be taken to evaluate endotracheal tube sizes and ventilation practices as soon as feasible. The strongest predictor of good outcome is duration of immersion. Poor outcomes are observed in 60%–100% of subjects immersed for more than 10 minutes. Predictors of poor outcome include the presence of cardiac (as opposed to respiratory) arrest and the need for prolonged resuscitation defined as more than 20–25 minutes. Medically induced hypothermia has not been shown to change neurological outcome. Drowning disproportionately affects children, and most instances are preventable. Prevention measures through legislative and public health interventions have had modest success.
Keywords
immersion, submersion, epidemiology of drowning, drowning sequence, temperature control, induced hypothermia, hypoxic-ischemic injury, electrolyte imbalance, acute respiratory distress syndrome, pulmonary injury, pulmonary sequelae, sepsis, prediction of outcome, prevention
Definitions
A uniform definition of drowning was agreed upon during the World Congress on Drowning in Amsterdam, The Netherlands, in 2002. Drowning is now defined as “a process resulting in primary respiratory impairment from submersion/immersion in a liquid medium.” The term “drowning” now encompasses both fatal and nonfatal outcomes of immersion, and “near-drowning” is no longer used.
Epidemiology
Drowning occurs in all age groups and is responsible for approximately 4000 deaths per annum in the United States, with a mortality frequency of 12–18 deaths per million person years. The highest mortality rates of approximately 30 deaths per million person years have been observed in the 0–4 and 15–19 year age groups, although there is some evidence that both mortality and hospitalization rates have declined over the last 2 decades. In the first year of life, drowning mortality in the United States is 63 per 100,000 live births. Worldwide, drowning is the 11th most frequent cause of death in the 0–4 years age group, the third most frequent cause of death in children aged from 5 to 14 years, and the second leading cause of injury-related death in childhood. The vast majority of drowning deaths occur in non-Western countries; in Bangladesh, more children in the 1–4 years age group die from drowning than from diarrhea or respiratory infection. Even in a well-developed country such as Australia, where drowning accounts for less than 1% of all reported deaths, there is a disproportionate rate of death in children. The rate of drowning is 4.6 per 100,000 per year for children under 5 years of age, three times the rate for adults. In addition, worldwide, drowning episodes are believed to occur around 3–10 times more frequently than drowning.
As with the majority of accidental deaths, there is a strong male preponderance with male-to-female incidence ratios ranging from 2 : 1 to 10 : 1. Approximately one in three drowning fatalities occur in accomplished swimmers. Children can and do drown in any receptacle containing water, from buckets to bathtubs to the ocean. The majority of drowning events occur in swimming pools (usually the child’s home pool) for children aged under 4 years and in open water for older children.
Drowning is most frequently a primary event. However, the presence of underlying disease such as epilepsy or cardiac arrhythmia should always be considered along with the possibility of drug or alcohol intoxication in older children. Approximately 6%–10% of drowning victims have a previous history of a seizure disorder, and it has been estimated that children with epilepsy have a relative risk for drowning of 96 in the bathtub and 23.4 in a swimming pool compared to nonepileptic subjects. A primary arrhythmia such as prolonged QT syndrome should always be considered, particularly in subjects who are capable swimmers. Approximately 30%–50% of adolescents who drown are intoxicated with drugs or alcohol.
Drowning Sequence
Based upon studies in animals, the sequence of events in drowning has been reported as follows:
- 1.
Immediate struggle
- 2.
Suspension of movement with frequent swallowing
- 3.
Violent struggle
- 4.
Convulsions and spasmodic inspiratory efforts
- 5.
Death
Loss of consciousness is thought to be related to hypoxia rather than hypercarbia. Some observers have reported that human victims stop moving suddenly after swimming underwater and then float motionless on the surface of the water and subsequently disappear quietly. The scenario of drowning without a struggle is probably due to a primary loss of consciousness secondary to other factors such as hypothermia or cardiac arrhythmia.
Sequelae of Submersion/Immersion Events
The drowning sequence has both pulmonary and nonpulmonary sequelae.
Pulmonary Injury
The majority of drowning victims aspirate water (salt or fresh) at the time of drowning. However, in about 10% of cases, laryngospasm prevents the entry of water into the lungs. The quantity of fluid aspirated is usually less than 22 mL/kg, a volume that approximates the functional residual capacity (FRC). In cases where aspiration occurs, local insult arises secondary to infection, surfactant depletion, aspiration of debris, and fluid shifts that depend on the relative tonicity of body fluids and aspirated fluid. While radiological pulmonary edema is the most common finding, the incidence and degree appear to be the same irrespective of saltwater or freshwater immersion, although the mechanisms may be different. Seawater aspiration results in an osmotic gradient with fluid shifts into the alveolar spaces, whereas aspirated freshwater is rapidly absorbed into the systemic circulation. Pulmonary edema may arise in both seawater and freshwater aspiration secondary to neurogenic causes, forced inspiration against a closed glottis, and altered surfactant or pulmonary capillary permeability.
Animal data suggests that 0.225% and 0.45% saline solutions are least injurious to the lungs in terms of gas exchange, possibly because they are rapidly absorbed from the alveolar spaces into the circulation along an osmotic gradient. Freshwater will also be rapidly absorbed but causes rapid inactivation of surfactant and is probably the most injurious fluid to aspirate, followed closely by seawater, which is approximately 3% saline. The presence of chlorine at 1–2 ppm in freshwater (typical of chlorinated domestic pools) does not affect the pulmonary injury.
Although it has been postulated that drowning in hypertonic fluid may lead to hypovolemia secondary to fluid shifts into the alveolar spaces, animal data indicates that hemodynamic changes following drowning are entirely attributable to hypoxia and are independent of the tonicity of the aspirated fluid.
Sepsis may occur, including with unusual and or atypical organisms, and significant allergic reactions to aspirated materials have been described. The incidence of pneumonia in adult patients requiring mechanical ventilation, as deduced from a retrospective study in the Netherlands, is around 50%.
Pathological findings are inconsistent and nonspecific. The most common finding in cases where the drowning medium has entered the lungs is the presence of reactive edema, with hyperinflation of the lungs and increase in lung weight (emphysema acquosum); however, these findings may also be seen in deaths from other causes including asphyxia and drug overdose. Even where no water has been aspirated into the lungs, neurogenic pulmonary edema can occur.
In severe cases of immersion/submersion accidents (with or without aspiration of fluid), some patients will develop pneumonia and pediatric acute respiratory distress syndrome (PARDS). The management of this disorder is discussed elsewhere in this textbook (see Chapter 38 ). However, of the patients admitted to the Intensive Care Unit at Children’s Hospital Los Angeles over the past 30 years after drowning events, less than 10% developed this serious complication. Figs. 41.1 and 41.2 show the typical radiological course of pulmonary injury in a drowning patient.
Nonpulmonary Sequelae
Hypothermia
Hypothermia is a common manifestation of drowning in water of almost any temperature, and there is anecdotal evidence that rapid hypothermia in a submersion incident is neuroprotective, particularly in children. Conductive losses through the skin are compounded by rapid heat exchange across the pulmonary capillaries if a significant volume of water is inhaled. In a canine model, dogs breathing water at 4°C demonstrated a decrease in carotid artery blood temperature of 8°C within 5 minutes. Cooling occurs most rapidly in small infants who have a relatively large surface area. In cases of extreme hypothermia, rewarming will be essential to allow return of cardiac function. Hypothermia can also play a major role in facilitating aspiration in immersion victims. As the core temperature drops below 35°C, muscular incoordination and weakness occur, which can interfere with swimming. As the core temperature decreases further, obtundation develops. At core temperatures below 30°C, unconsciousness can occur and the myocardium becomes irritable. Atrial fibrillation can occur, and at temperatures below 28°C, ventricular fibrillation is likely.
Electrolyte Imbalances
Electrolyte imbalances may arise if a significant amount of nonisotonic water is aspirated, although this is unusual in regular seawater. Although freshwater immersion victims have decreased serum sodium, and saltwater immersion victims have elevated serum sodium and chloride levels, these are rarely substantial or clinically significant. Even in the Dead Sea, which has electrolyte concentrations approximately 10 times higher than seawater, immersion victims rarely have severe abnormalities of sodium or chloride, although hypercalcemia and hypermagnesemia are common. Hemolysis due to aspiration of hypotonic or hypertonic fluids appears to be an extremely infrequent complication.
Trauma
Traumatic injuries resulting from a fall into water must be considered but are generally of lesser importance than the immersion itself. Cervical spine injuries are the most critical to consider but are uncommon, occurring in only 0.5% of all nonfatal drowning cases, and only then in cases with a clear history of diving, motorized vehicle crash, or fall from a height.
Hypoxic-Ischemic Damage
All organs are susceptible to hypoxic-ischemic injury following prolonged low cardiac output, and inadequate oxygenation and multiorgan failure is an almost inevitable consequence of severe submersion/immersion injury. Clinically, the brain is particularly susceptible, with the liver and the gastrointestinal tract being the most resistant.
Management
Management of the pulmonary injury will usually require supplemental oxygen, diuretic administration for pulmonary edema, and the most severe cases will require support with intubation and mechanical ventilation. Many of these serious immersion accidents occur in relatively isolated locales, and often children are intubated in the field or in outside facilities where there may be little experience managing children. Endotracheal tube sizes can be too large and sufficient to cause significant damage to the larynx if not recognized and promptly downsized after arriving at the receiving pediatric institution. In addition, with severe cases, the lung injury will comprise all the features of PARDS. The management of this disorder is addressed in Chapter 38 .
Instillation of surfactant has been reported and is an appealing therapeutic intervention given that the majority of victims aspirate a quantity of fluid that will denature and wash out existing surfactant. However, the temptation to administer surfactant should be considered in the context of recent randomized controlled trials demonstrating a lack of efficacy of surfactant in PARDS. Administration of steroids appears to be effective in animal models of seawater aspiration, and there is evidence to support the use of steroids in PARDS, although this is by no means universally accepted.
Broad-spectrum antibiotics should be administered to treat likely bacterial contamination of the lungs, such as after drowning in stagnant water. The incidence of neurological infection is stated to be high, with a number of case reports in children.
Most drowning victims will be hypothermic at the time of presentation. In cases of extreme hypothermia, rewarming will be essential to allow return of cardiac function, and if the core temperature is below 26°C–28°C, or the patient is in cardiac arrest, rewarming is probably best achieved using cardiopulmonary bypass. Given the potential benefits of hypothermia on hypoxic CNS injury, it is our view that modest hypothermia (core temperature 32°C–34°C) should be maintained immediately following the injury until the patient reaches the receiving hospital if there is any suspicion of the patient having sustained a hypoxic brain injury. While duration of submersion, but not water temperature, is reported to be more associated with drowning outcome, excellent neurological outcomes have been reported after prolonged immersion in very cold water, with several case reports indicating full neurological recovery after periods of up to 66 minutes in near-freezing water. Children lose body heat more rapidly than adults, and if significant brain cooling occurs prior to cessation of circulation, then some degree of neuroprotection may occur. It has been estimated that brain temperature needs to fall by at least 3°C within the first 5 minutes of immersion for cerebral protection to be effective.
The role of induced hypothermia for neuroprotection following drowning remains less certain. Some recent studies have highlighted the beneficial effects of hypothermia on a variety of hypoxic CNS injuries in humans. However, these studies had control arms of “usual care,” and some patients became hyperthermic with potential harmful effects and made the hypothermia group outcomes appear better. In a Canadian trial using hypothermia in drowned children to reduce intracranial pressure and limit brain injury, the death rate in the hypothermic group was higher than in the normothermic group, with most deaths attributed to neutropenic sepsis. However, this trial was relatively small, and the hypothermia group was also managed with hyperventilation and high-dose phenobarbitone, which may have influenced the outcomes. A randomized, controlled trial of therapeutic hypothermia versus normothermia after cardiac arrest outside of the hospital has been undertaken by 38 pediatric centers in the United States and Canada under the auspices of the National Institutes of Health. This study showed no benefit of hypothermia over rigorously controlled normothermia in either mortality or neurodevelopmentally in the 1-year follow-up of 295 survivors. However, this study excluded patients with cardiac arrest secondary to drowning in ice water who had a core temperature of ≤32°C on presentation. In a subsequent report of the subset of the 74 pediatric cardiac arrest patients due to drowning from this study, it was concluded that hypothermia did not result in a statistically significant benefit in survival with good functional outcome or mortality at 1 year, as compared to normothermia. Current Pediatric Advanced Life Support guidelines no longer recommend consideration of cooling to 32°C–34°C for 12–24 hours in comatose children following cardiac arrest, but support rigorous temperature control to avoid heating.
Outcome of Pulmonary Injury
Routine tests of pulmonary function have been reported as normal in adults following a drowning accident. However, in a series of 10 functionally normal children studied 6 months to 8.5 years (mean 3.3 years) after the submersion incident, only one had completely normal pulmonary function. Seven had abnormal methacholine challenges demonstrating a high incidence of bronchial hyperreactivity, and five had clear evidence of peripheral airways disease. Whether these abnormalities were related solely to aspiration inherent in the drowning episode or further related to PARDS was not addressed. It is possible these children are at risk for developing chronic lung disease, especially if exposed to further airway or parenchymal irritants.
Outcome Prediction of Neurological Injury
The best prognostic indicators are observed in the field. The strongest predictor of good outcome is duration of immersion. Poor outcomes are observed in 60%–100% of subjects immersed for greater than 10 minutes, compared to 0%–30% of those immersed for ≤5 minutes. Good outcomes are also associated with the presence of sinus rhythm, reactive pupils, and neurologic responsiveness at the scene. The presence of a detectable heartbeat and hypothermia (<33°C) on arrival to the Emergency Department discriminates intact survivors from those with persistent vegetative state or death. Predictors of poor outcome include the presence of cardiac (as opposed to respiratory) arrest and the need for prolonged resuscitation (more than 20–25 minutes) or the requirement for more than two doses of epinephrine.
The prediction of the neurologic outcome of those children who survive the initial resuscitation event and arrive in the intensive care unit in a comatose state is highly relevant for parents and caregivers. It is difficult to provide early and accurate prognostic information on comatose children, especially if brainstem functions are intact. Severity of illness scores such as the Pediatric Index of Mortality (PIM) and Pediatric Risk of Mortality (PRISM) scores were developed to predict the risk of death in groups of patients and not individuals. However, PRISM has recently been applied to individuals, and it enables the prediction of either absence or presence of serious neurological impairment or death in pediatric drowning patients if they present at extreme values on this scale. In patients with intermediate PRISM scores, though, it is not possible to establish a reliable prognosis. Electrophysiological investigations, such as brainstem auditory evoked potentials and short-latency somatosensory evoked potentials (SSEP), are helpful to assess the likelihood of a permanent vegetative state or a higher level of cognition. The bilateral absence of SSEP is an established predictor for a worse clinical outcome after cerebral hypoxia. Diffusion-weighted magnetic resonance imaging (DWI) provides a quick and reliable tool to detect early tissue injury in acute cerebral ischemia, because it is sensitive to water shifts between the extracellular and intracellular compartments, which conventional MRI often cannot detect. Preliminary results suggest that the extent of diffusion-weighted MRI pathology may serve as a reliable predictor for neurological outcome after cerebral hypoxia ( Fig. 41.3 ). Magnetic resonance spectroscopy also shows promise as a diagnostic and prognostic tool in hypoxic/ischemic cerebral injury, with reduced N-acetyl aspartate (NA) and elevated lactate both associated with poor neurological outcomes ( Fig. 41.4 ). Although computed tomography (CT) is not a sensitive test for ischemic injury, the presence of any CT abnormalities suggestive of ischemia (typically loss of gray-white differentiation and/or basal ganglia edema or infarction) within the first 3 days of drowning is strongly correlated with poor neurological outcome, and the presence of CT abnormalities in the first 24 hours is associated with a very high risk of mortality.
