Immersion Submersion and Drowning
For a critical care college in a country with this much shoreline and such a vibrant beach culture, the CICM is strangely unconcerned with drowning. This is perhaps because much has already been achieved to prevent immersion fatalities. The rate of death from drowning in Australia has fallen from 8.76 to 1.32 per 100,000 according to the MJA (Franklin et al, 2010) although it is impossible to determine the "true denominator" (it would be unfair to include in this statistic all the people who never go anywhere near a body of water). Certainly, the author can conceive of a seaside ICU where perhaps half of the inpatient population is recovering from a near-drowning, but this pathology does not seem to form a very large proportion of our daily workload out here in the Western Suburbs of Sydney, where the people are either good swimmers or too poor to afford pools. Uneducated social remarks aside, the drowning SAQs consist of Question 25 from the second paper of 2012 and Question 1 from the first paper of 2009. Thus far, the issues examined have been acute management of the drowning victim, potential complications of near-drowning, and risk factors for severe neurological injury arising from the hypoxia of submersion.
- Drowning is the process of experiencing respiratory impairment from submersion or immersion in a liquid
- Common complications of drowning include death from hypoxic arrest, laryngospasm, aspiration of water and gastric contents, ARDS and pulmonary oedema due to loss of surfactant, hypothermia and cerebral hypoxia which is the main determinant of long-term morbidity.
- Uncommon complications of drowning include electrolyte derangement, haemolysis, renal failure due to haemoglobinuria, and infection (due to aspiration of unclean water).
- Predictors of poor neurological outcome following drowning include immersion for more than 5 minutes, a delay in CPR longer than 10 minutes, GCS of 3 and fixed dilated pupils on admission, severe acidosis (pH < 7.00) and abnormal neurology during admission (eg. GCS less than 6 and abnormal brainstem function after 48 hours).
Resuscitation of the drowned patient
- Unskilled rescuers should avoid drowning themselves.
- Do not start CPR while still in the water (one should not need to say this)
- CPR should not be of the compression-only variety (you really need the breaths)
- Avoid all active attempts to "force" the water out by placing the person face-down or any sort of abdominal thrusting, as this will only lead to the aspiration of stomach contents.
- Do not stop the resuscitation of the hypothermic drowning victim (the ICU doctors might want to publish another case report of miraculous ECMO-aided survival).
Emergency management issues
- Assessment of the airway and of the need for immediate intubation.
Drowning is associated with a high risk of aspiration (and not just of lake water).
- Ventilation with high FiO2
High PEEP, 12-15
Investigation of possible aspiration with CXR and ABG
- Establishment of IV access and correction of hypovolemia;
drowning victims may become hypovolemic following prolonged immersion due to the hydrostatic effects of water (particularly salt water)
- Investigate causes of drowning related to intracranial events, eg. ICH, or trauma resulting from a fall into submerged obstacles
- Assessment of temperature, and rewarming (the immersed patient is invariably hypothermic, as it is rare to drown in a body of water with an ambient temperature higher than human core body temperature).
ICU management issues
- Assessment of the airway device effectiveness (i.e. is it in the right main bronchus?)
Bronchoscopy and suction as indicated by copious aspirated material.
- Lung protective ventilation; open lung strategy
No benefit in corticosteroids
- Assess the effectiveness of volume resuscitation; give more.
- Sedation as required: no specific recommendations can be made.
If the patient has had a cardiac arrest, therapeutic hypothermia might be worthwhile.
- Electrolytes are unlikely to be deranged by this stage.
- Renal function is unlikely to be impaired
- There is no reason to omit normal nasogastric feeds
- Monitor Hb, and satisfy yourself that there is no haemolysis.
- There is no need for antibiotics.
The best published resource for these questions would have to be the section on drowning from the ARC ALS2 manual (2011, p. 127). If one has not been able to acquire this manual, one may find similar information in the 2012 paper by Szpilman et al (NEJM) or in the much earlier Pearn article (1985, BMJ). The paediatric flavour of Question 25 from the second paper of 2012 may arouse interest in articles such as Austin et al (2013), "Management of drowning in children."
Definitions of drowning, near-drowning, immersion and submersion
Previously, a confusion of definitions had persisted. From 1971 (Modell) the following weird terminology had achieved acceptance:
- Drowning without aspiration: to die from respiratory obstruction and asphyxia while submerged in a fluid medium;
- Drowning with aspiration: to die from the combined effects of asphyxia and changes secondary to aspiration of fluid while submerged;
- Near-drowning without aspiration: to survive, at least temporarily, following asphyxia due to submersion in a fluid medium;
- Near-drowning with aspiration: to survive, at least temporarily, following aspiration of fluid while submerged.
Thus, one only drowned if one were dead, and one "near-drowned" if one's heart beat even temporarily following one's extraction from the brine. For some reason, this was tolerated until 2002. Fortunately, a panel of experts eventually intervened. In a 2005 WHO bulletin, van Beeck et al describes the consensus procedure which ultimately agreed on the new definition of drowning as “the process of experiencing respiratory impairment from submersion or immersion in a liquid”. This replaces all previous terms including “near drowning”. According to the bulletin, "there was also consensus that the terms wet, dry, active, passive, silent, and secondary drowning should no longer be used". In order to bring some clarity, here are some important definitions:
- Drowning: the process of experiencing respiratory impairment from submersion or immersion in a liquid”
- Submersion: the airway is below the surface of the liquid
- Immersion: the airway is above the surface of the liquid (eg. taking a bath)
Complications of drowning
In order to simplify revision for people whose mind is as unstructured as mine, these complications will be discussed in a familiar "A-B-C" pattern.
Death due to submersion
The Szpilman paper from NEJM goes through the pathophysiology of a drowning death very nicely. The LITFL chapter on this topic offers an excellent point-form summary of Szpilman's article, and represents what the college probably expected from a pass-credit level answer to Question 1 from the first paper of 2009. For a distinction-level answer, one may wish to turn to the 1993 paper by Modell (the same guy who came up with the 1971 definitions). In short, the key feature of death from submersion is hypoxia, which develops fairly rapidly in the struggling swimmer. With this, cardiac activity degenerates into bradycardia, PEA and asystole. This progression may be rapid, or it could take up to an hour in the presence of severe hypothermia. In Australia, extreme cold is uncommon, and survival rates of drowning people suffer as a consequence. Dyson et al (2013) reported on a case series of 336 drowning victims from Victoria, spanning the era which saw the arrival of therapeutic hypothermia (1999-2011). Among these people, the majority (79%) had asystole as their first recorded rhythm. Only 27% survived the ambulance ride to hospital, and only 12% survived to discharge. Those with a shockable rhythm were fifty times more likely to survive.
Laryngospasm is in fact the reflex which prevents foreign material from entering the trachea, and it is therefore appropriate to develop laryngospasm when drowning. Unfortunately, the onset of hypoxia causes the larynx to relax, and water still finds its way into the lung. Furthermore, the incidence of laryngospasm must be fairly low, no greater than 10%. This conclusion is arrived to on the basis of autopsy series, where only around 10% of patients have ended up dead from hypoxia without much fluid in their lungs. Of these patients, it must be assumed, a few died because of obstructive asphyxia, as the consequence of laryngospasm. In years gone by, this was called "dry drowning".
Aspiration of water
Surprisingly, though 90% of drowning victims aspirate water (according to autopsy series), the aspirated volume varies greatly, from lots to virtually none. The vast majority of them aspirate less than 22ml per kg of body weight (Modell, 1969). In short, they don't exactly fill their lungs with it. The aspiration of pure water is actually a rather benign experience, from the standpoint of hypoxia- the volume of fluid is distributed into two tennis-courts of surface area, from where it is rapidly absorbed into the systemic circulation (Tabeling et al, 1983). If one has managed to aspirate a rather more substantial amount, the aspiration may surpass the capacity for such resorption, and contribute to hypoxia purely by clogging the alveoli- but this usually does not happen, because people don't tend to inhale that much water. While conscious you will (hopefully) not be actively inhaling water, and when you become unconscious from hypoxia you will usually also be apnoeic. In fact, according to autopsy analysis by Fuller et al (1962) one is rather more likely to aspirate a lung-full of one's own stomach contents (in 25% of cases) than a substantial amount of water. In any case, the hypoxia which follows drowning is usually due to surfactant failure, and the shunt which is caused by it.
Aspiration of even a small amount of water tends to alter the surface tension properties of alveolar surfactant. The result is collapse of the alveoli. Atelectasis ensues, and this could be very severe. Shunt through these atelectatic regions is responsible for the severe hypoxia seen following drowning. "Pulmonary oedema" which has been described as a complication of drowning is in fact an ARDS due to surfactant loss and the subsequent damage to fragile alveolar membranes. Certainly, the bloodstained froth which issues from such patients closely resembles the pink foam of cardiogenic pulmonary oedema, but in contrast to normal cardiogenic pulmonary oedema, the left atrial pressure is usually normal. Poor lung compliance usually completes the ARDS picture.
Most of the morbidity from drowning is the consequence of cerebral hypoxia. In general, if you were submerged, lost consciousness and developed cerebral hypoxia, then you're likely to have also had a cardiac arrest, and most of the data regarding neurological outcomes after drowning comes to us from research on drowned patients who have arrested. Apart from the temperature of the water, the most important factor is the duration of submersion, which determines the duration of hypoxia and the extent of injury. Outside of near-Arctic environments, virtually none of the patients who remained submerged for longer than 25 minutes go one to have a satisfactory neurological outcome, and there is little difference between this sort of cerebral hypoxia and the cerebral hypoxia of any other sort of cardiac arrest.
Electrolyte derangement due to aspirated water
Jerome Modell certainly seems to be the guru of drowning. His 1969 paper on electrolyte changes in drowning victims is still being quoted over forty years since its publication. In this paper, Modell and Davis discussed the serum samples they collected from autopsies of 118 drowning victims (and 23 people who died of something else), as well as samples collected from puppies whom Modell and Davis had personally drowned for scientific reasons. In short, fresh water drowning does not pose a substantial risk to one's electrolyte mileu; or if it does, then by the time the patient ends up in emergency (or in the morgue) the redistribution of water has already compensated for the electrolyte shifts. In any case that water is hypotonic, so it should not contribute very much to the electrolyte load.
Even sea water does not tend to produce wild excesses of electrolyte imbalance. Sea salt makes salt water hyperosmolar (usually about 1000 mOsm/L); in case you wanted to know it owes most of its salinity to common table salt. There is a lot of it there (Na+ 459 mmol/L, Cl- 538mmol/L). Other ions include K+ (10 mmol/L), Ca++ (10mmol/L) and Mg++ (53 mmol/L). However, these millimoles do not tend to make their way into the systemic circulation, largely because the sea water is so hyperosmolar that it attracts water out of the circulation instead. The consequence of this is actually hypovolemia and concentration of preexisting serum electrolyte levels. (Modell et al, 1967)
Hypothermia due to immersion and submersion
The rapid exchange of body temperature with extremely cold water is a major protective factor in neurological survival from drowning. It typically results in very rapid decrease of body temperature, particularly in children and skinny adults (who have a low surface area to mass ratio). Under ideal circumstances, the drowning person ends up being cooled so rapidly that the cerebral metabolic rate is decreased to its minimum early in the hypoxic process. The consequence of this are case reports such as Romlin et al (2015), where an asystolic child is fished out of the icy Swedish sea with a core temperature of 13.8°, is then subjected to ECMO via conventional CPB, and goes on to make a complete neurological recovery. This has implications for lazy rescuers. The seven year old girl in the abovementioned case report had approximately 1 hour downtime and then CPR for a further 64 minutes before the commencement of cardiopulmonary bypass.
On the basis of this, one might expect the outcomes to be better in countries with colder climates, but in actual fact this is not the case. Masahiko et al (2013) published a paper detailing the outcomes of Japanese drowning victims with an even higher mortality rate than steaming hot Victoria (survival to hospital admission was only 12.5% in Japan vs. 27% in Australia, and only 0.7% had a favourable neurological outcome.
Renal effects of drowning
One might expect to have been asystolic and hypoxic, with all the associated unpleasantness, including multiorgan system failure. Apart from that, drowning rarely has much of an effect on renal function. One can make an exception in cases where truly vast amounts of hypotonic water have somehow been aspirated. In this case, the sudden dilution of the intravascular volume with pure water results in the osmotic lysis of red cells (i.e. they swell and burst). The consequences of this are haemoglobinuria and tubular damage from toxic haem moieties.
Haemolysis due to haemodilution
As a consequence of fresh water drowning, this is far from common. Case reports are known. For example, in 1964 US Airforce Captain William D. Munroe reported on a case of a small boy who survived his drowning and went on to develop massive intravascular haemolysis with haemoglobinuria. Such a thing calls for the reversal of a normal trend in drowning, where cerebral anoxia is responsible for the major post-resuscitation morbidity. It appears that one needs to inhale a larger amount of fresh water than is routine for a fatal drowning, and yet to remain alive without severe hypoxic brain injury. One might accomplish this by intermittently coming up for air, then submerging to inhale more water. These days haemolysis is used more as a forensic corroboration of freshwater drowning as the cause of death (particularly, haemolytic staining of the aorta).
Infection due to the aspiration of contaminated material
Fresh water (i.e. pond water) is actually more harmful than the supposedly more "noxious" salt water. The hyperosomolarity of salt water has been thought to produce some additional "oedemagenic" effects, attracting extra fluid into the alveoli. This opinion was formed largely as a consequence of cruel rat experiments. For example, Halmagyi (1961) squirted 1ml/kg of water into rat lungs and reported that there were no local histopathological changes with fresh water aspiration, whereas salt water produced boggier lungs and intra-alveolar haemorrhages.
Overall, the long-term damage from fresh water turns out to be greater, and this is largely due to the presence of microorganisms. A lungfull of Horrendomonas sp. is going to do more harm than a lungfull of sterile sea water. Approximately 60% of autopsy lungs revealed mud, sand or fragments of aquatic vegetation (Fuller et al, 1962). Conventional resources (such as the Sanford Guide) do not have any specific antibiotic recommendations, except to say that antibiotic therapy is not necessary except when the aspirated water is known to have been grossly contaminated (eg. sewer). In their discussion of pneumonia following near drowning, Ender et al (1997) report that fresh water aspiration produces particularly nasty lung infection by organisms such as Aeromonas species, Burkholderia pseudomallei and the fungus Pseudallescheria boydii. The case fatality rate in this pre-ARDSNET pneumonia series was around 60%. The authors did not feel too stroingly about any specific empiric antibiotic shoices, but ventured that a broad spectum β-lactam such as Tazocin might be reasonable. P.boydii is not susceptible to amphotericin, and is near-impossible to identify microbiologically unless you're specifically looking for it, which makes routine antifungal therapy impossible to recommend.
Though fresh water is clearly much filthier and more dangerous, sea water is not exactly sterile either. One may recall that the majority of the Earth's biomass is actually suspended in sea water, and so It stands to reason that one might inhale some of that biomass in the process of drowning in it. Kakizaki et al (2008) were actually able to isolate colonies of bioluminescent seawater organisms (Photobacterium, Vibrio, Shewanella, Psychrobacter) from the lungs of drowned seamen. Fortunately, this is probably only of forensic interest. The pathogenicity of aspirated plankton must be low. The conditions which are optimal for their growth (4°C, 4% NaCl, abundant sunlight) are typically absent in the lungs of the ventilated ICU patient, rendering that environment a hostile medium.
The only possible exception to this generalisation is Vibrio vulinificus which seems to be a common source of marine bacterial infection, although information about vibrionic pneumonia are limited to a single letter to Annals of Internal Medicine. (Sabapathi, 1986). The correspondent described a case of V.vulnificus pneumonia where the culture medium was a 60 year old alcoholic extracted from the Inner Harbour of Baltimore. It is unremarkable that this organism was a cause of infection in this case, because it seems to love human tissues, and takes every opportunity to invade them. Koenig et al (1991) report numerous possible pathological scenarios, including meningitis, uveitis, epiglottitis, endocarditis, fasciitis and endometritis due to sexual intercourse while immersed in sea water. To cut a long story short, if one were concerned that sea water aspiration was causing a clinically significant pneumonia, one may choose doxycycline as the agent of choice (100mg daily appears effective).
Predictors of outcome following drowning
The Szpilman paper from NEJM contains within it Table 2, which describes the most important predictors of outcome. However, as far as exam preparation goes, this box is superceded by the contents of Box 80.1 (page 820) from the "Submersion" chapter by Cyrus Edibam and Tim Bowles (Oh's Manual).
Paraphrased, the box contains the following risk factors for death or severe neurological injury:
Factors at the site of submersion:
- Immersion for more than 5 minutes
- Delay in CPR of more than 10 minutes
- Warm water drowning
- Submersion time > 10 minutes (a good independent predictor of neurological outcome)
Factors on presentation to the ED
- Fixed dilated pupils
- GCS of 3
Factors after admission to the ICU:
- GCS less than 6
- Arterial pH less than 7.00 upon arrival to ICU
- No spontaneous purposeful movement and the abnormal brainstem function after 48 hours
- Abnormal CT within 36 hours