Drowning is the third most common cause of accidental death in children (1). The various predictors for survival or neurological outcomes include the age of victim, submersion time, salt versus fresh water, temperature of the water, cardiopulmonary resuscitation at the scene, and time required for hospital arrival (2). Recent reviews also describe a cardiac arrhythmogenic response termed “autonomic conflict” that contributes to sudden cardiac arrest following a cold-water drowning (3).

A 10-year-old boy presented to our emergency department for an accidental drowning, which occurred while playing near a recreational pond in July. He was rescued from the pond 10-15 minutes after he went missing. The exact submersion time was unknown. He had received basic cardiopulmonary resuscitation at the site by a doctor who happened to be there, with no automated external defibrillator used. Following the resuscitation, he was transported to our hospital.

On arrival, the boy was obtunded with a Glasgow Coma Scale of 9 (eye-opening, 3; verbal response, 3; and motor response, 3), heart rate (HR) of 70 beats/minute, blood pressure (BP) of 117/71 mmHg, oxygen saturation of 89% on room air with 94% on oxygen support with a non-rebreathing mask, and temperature of 36.7 °C with the cold and clammy extremities and normal capillary refill time. Respiratory examination showed tachypnea with coarse crepitation in bilateral basal lung fields. Neurological examination was suggestive of altered cortical function and brisk deep tendon reflexes, as well as positive plantar reflexes. No signs of trauma were immediately noted and the rest of the systemic examination was non-contributory.

The initial chest radiograph showed the presence of pulmonary edema (Fig. 1). A blood gas analysis showed a severe respiratory and metabolic acidosis with pH 7.10, bicarbonate of 24.3 mmol/L, carbon dioxide of 89 mmHg, base deficit of 0.3 mmol/L, and lactate of 1.2 mmol/L. Primary laboratory tests showed normal electrolyte concentrations with no evidence of sepsis or multiorgan dysfunction. Within the first hour of hospitalization, his mental and respiratory statuses deteriorated to a Glasgow Coma Scale of 6 and type 2 respiratory failure, leading to endotracheal intubation, mechanical ventilation, and subsequent transfer to the pediatric intensive care unit.

Three hours following the hospitalization (day 1), he started showing a symptomatic sinus bradycardia with hypotension which was managed with a bolus dose of epinephrine and atropine, as well as epinephrine infusion starting at 0.2 μg/kg/minute. The pupils were both reactive to light. The deep tendon and plantar reflexes remained normal with no evidence of increased intracranial pressure. We considered the worsened vital signs a possible cardiogenic shock although the initial cardiac markers were within normal ranges: creatine kinase, 146 U/L (reference value, 50-200 U/L); creatine kinase-myocardial band fraction, 11 U/L (0-25 U/L); and lactate dehydrogenase, 286 U/L (220-440 U/L). Two-dimensional echocardiography was not performed due to the unavailability of the device at that time. Inotropic support had to be escalated due to persistent hypotension, the maximum requirement of epinephrine and norepinephrine was 0.4 μg/kg/minute alike.

On day 2, HR variability was noticed, ranging 70-130 beats/minute which increased to 60-165 beats/minute on day 3 (Fig. 2). The HR varied between the second and 99th percentiles for his age and sex, which occurred irrespective of change in the dose of inotropes or sedatives. There was neither fever nor implementation of therapeutic hypothermia. On day 4, follow-up laboratory findings remained normal as to the concentrations of electrolytes, aminotransferases, and creatinine, with a sepsis screening test negative. The pulmonary edema improved, and ventilator settings were tapered by day 3. However, variations in BP were noticed on day 4, ranging 50th-95th percentiles for his age and sex on the inotropic support which persisted till day 7. The inotropes were tapered and stopped on day 8 when the BP and HR remained stable for more than 24 hours, and the boy was extubated. No further events of autonomic instability were noted. At discharge on day 10, he had a normal neurological status and echocardiogram.

In this case of accidental drowning, the case patient experienced profound autonomic dysfunction with wide variations in HR and BP in the absence of other causes for the dysfunction, such as dyselectrolytemia, hypoxemia, hypothermia, myocardial dysfunction, or cerebral dysfunction.

Drowning activates 2 powerful autonomic responses, i.e., cold shock response and diving response. The cold shock response is sympathetically mediated and characterized by tachycardia, gasping respiration, uninhibited hyperventilation, peripheral vasoconstriction, and hypertension (4). The diving response is mediated by the parasympathetic division and constitutes substantial sinus bradycardia, expiratory apnea, and vasoconstriction. This simultaneous activation of the autonomic nervous system (ANS) contributes to the autonomic conflict, particularly in the case of drowning in cold water. Although the reciprocal divisions were previously thought to have different triggers of activation, it has now been postulated to be activated simultaneously due to some specific triggers like drowning (3). The autonomic conflict can predispose to dysrhythmias, such as supraventricular, junctional, or ventricular tachycardia (5,6). The cold shock response is known to be most pronounced within the first 20-90 seconds of cold-water immersion, and subsides within 3 minutes (4). Thus, we cannot completely attribute the autonomic dysfunction in our case to the cold shock response. It remains unanswered whether the dysfunction was due to the prolonged effect of this response. However, it is important that autonomic dysfunction has a favorable outcome if managed appropriately, irrespective of the underlying mechanisms.

An increase in HR variability in a critically ill patient has also been proposed as an index of efficient ANS whereas a low variability indicates a low adaptability or diseased state (7). Studies on the HR variability in pediatric drowning and traumatic brain injury concluded that a lower ratio of measured low-to-high frequencies correlates with poor neurological outcomes (8-10). In this current case, the favorable neurological outcome was probably due to the bystander resuscitation. The autonomic dysfunction also reflected an efficient ANS and might correlate with the outcome.

Autonomic dysfunction may contribute not only to morbidity and mortality in patients undergoing drowning but also to their favorable outcomes. Although the literature describes HR variability and the autonomic conflict causing dysrhythmias, severe and prolonged autonomic dysfunction as in our case has not been reported to the authors’ knowledge.