Drowning: A preventable cause of death

July 1, 1999

Don't let the joys of summer be destroyed by the tragedy of drowning. Know how to manage this emergency and--better yet--how to teach parents about prevention.

Cover story

Jump to:Choose article section...Drowning: A preventable cause of deathWho's at risk?Where do drownings happen?How submersion killsClinical managementNeurologic outcomePrevention: The only real cureGUIDE FOR PARENTSWater safety

Drowning: A preventable cause of death

By Robert Bolte, MD

Don't let the joys of summer be destroyed bythe tragedy of drowning. Know how to manage this emergency and--betteryet--how to teach parents about prevention.

Drowning claims nearly 4,000 lives a year in the United States,and about half of those who die are children.1 Onlymotor-vehicle accidents exceed drowning as a cause of accidentaldeath for those under 25 years of age. In 18 states, drowningis the leading cause of accidental death in children 1 to 4 yearsof age. For every fatality there are approximately 3.5 hospitalizations.For every hospitalization, four children are evaluated in emergencydepartments and discharged home. About 15% of children who surviveinitially die after admission to the hospital. This rate is remarkablyhigh compared to other injuries. One third of those who are comatoseon admission and survive suffer significant neurologic impairment.The annual cost of care for an impaired survivor in a chroniccare facility is more than $100,000.2,3

The sections that follow provide the information you need tohelp reduce this terrible toll. You will learn which childrenare at highest risk; where the danger spots are--from pools totoilets to hot tubs; just how submersion produces pulmonary andneurologic damage; what's protective about icy water; how to managesubmersion injuries, at the scene and in the emergency department;which children are most likely to survive without neurologic damage;and the paramount importance of starting resuscitation beforethe paramedics arrive. Most important, you will be encouragedto incorporate sound advice on how to prevent drowning into routineanticipatory guidance for parents. That's the simplest and mosteffective intervention you can make.

Who's at risk?

Toddlers are the age group most at risk for drowning. The incidenceof drowning for children younger than 4 years of age is 2.5/100,000,with peak incidence at 2 years of age. Next come teenagers, witha peak incidence of 1.9/100,000 among those 16 to 18 years ofage. Males predominate at all ages, accounting for about 80% ofsubmersion deaths.1 The gender difference among toddlershas been attributed to sociocultural factors, in that boys maybe allowed to explore their surroundings more freely than girls.The gender difference among adolescents seems easier to understandand has been attributed to the three "Ds" that ofteninfluence teenage male behavior: drinking, drugs, and dares.4Nearly one out of every thousand boys will drown before he reacheshis twentieth birthday.2

Seizure disorders are a significant risk factor in all agegroups.5 Estimates range from four- to 10-fold greaterrisk of drowning for children with epilepsy than for the generalpediatric population. The bathtub is a common site, especiallyfor children over 4 years of age, who are beginning to be concernedwith modesty and want to be alone in the bathroom--often withthe door closed. Swimming pools and hot tubs are also frequentsites.

Drowning rates are highest in the southern and western statesand in Alaska. Drowning rates for African-Americans and NativeAmericans are significantly higher than those for Caucasians,except in children 1 to 3 years of age, where this pattern isreversed. Non-pool drowning rates are inversely related to socioeconomicstatus. No such relationship exists for pool drownings.2Risk factors are summarized in Table 1.

Where do drownings happen?

Most pediatric drownings (over 95%) occur in fresh water. Evenin coastal areas, ocean drownings are uncommon (about 10% of thetotal), presumably because most ocean swimming is supervised bylifeguards. Overall, private swimming pools are the most commonsite of drownings--especially in children 1 to 4 years of age.Not surprisingly, considering the ubiquity of private pools, pooldrownings are particularly common in the sunny states of California,Arizona, and Florida. In California in 1993, pool immersion incidentswere the leading cause of injury death among preschoolers, withfatality rates of 3.2 per 100,000. Only about 5% of drowningsin preschoolers occur in above-ground pools, even though theyoutnumber in-ground units. About half of all pool drownings occurin the child's own home pool with the other half in pools of neighborsor relatives.2

A 1987 study by the Consumer Product Safety Commission on residentialpool drownings in children less than 5 years of age revealed thatabout 70% of the victims were at home with one or both parentsat the time of the accident. Only 30% of the children were wearingswimsuits when they drowned. Most of them were last seen in thehome with only a momentary lapse of supervision. Most enteredthe pool through the unprotected (unfenced) side of the pool facingthe house. In the vast majority of cases the drowning was "silent,"with no screams or splashing heard.6

Geographic and environmental factors affect the settings inwhich drowning accidents occur most often. In Minnesota, oftencalled "the land of 10,000 lakes," lake drownings arerelatively frequent. In Utah, which might well be called the "landof 10,000 irrigation canals," canals are a common drowningsite. As hot tubs have become increasingly popular, they alsohave become common drowning sites, accounting for up to 10% ofdrownings in some series.

The bathtub is the most common site of drowning for infantsless than 12 months of age, and a frequent site for toddlers 12to 24 months of age. Drowning is exceedingly rare in infants lessthan 9 months of age, who are seldom left alone in the tub. Butonce the child has learned to sit up and has fairly good grossmotor control, some parents think it's safe to leave the childin the tub--however briefly--without direct adult supervision.Sometimes the baby is in the tub with an older sibling; a Utahstudy reported that all their fatal bathtub immersions in childrenless than 2 years of age (7 of 7) occurred when the victim wasbathing with a young sibling (ages ranging from 10 months to 7years) while the parent or the babysitter was out of the room.7Not all bathtub drownings are attributable to lack of supervision,however. Child abuse must also be considered.8 In a10-year retrospective study of near-drownings (the victim survivedfor at least 24 hours) in Philadelphia bathtubs, the physicalfindings or the history were suspicious for abuse in 67% of thecases.9

Young children can also drown in buckets, especially large,stable ones that a top-heavy toddler can easily flip into headfirst.10 A study of Chicago-area drowning deaths inchildren younger than 3 reported that 24% drowned in 5-gallonplastic buckets.11 Toilets, a source of fascinationfor many toddlers, are also an occasional site of drownings.

Teenage boys drown in a greater variety of places. Rivers,lakes, canals, and beaches are common locations, swimming poolsless so. Even in coastal states, almost all adolescent drowningsoccur in fresh water. Boating accidents start at this age, withinjuries related to personal watercraft (Jetskis and Waverunners)emerging as a new problem.12 Diving mishaps with associatedcervical spine trauma (about 800 per year in the US), and occasionalsubmersions while surfing are also seen in this age group.13Cervical spine injuries occur most often the first time a boydives in a new area, unaware of the shallow depth. The "threeD's," especially drinking, play a major role. Up to halfthe drownings in adolescents are associated with alcohol. Sitesin which most pediatric drownings occur are listed in Table 2.

How submersion kills

The first response a person makes to being under water is,of course, to hold his breath. But in a relatively short time,as the victim feels the effects of progressive hypoxia and hypercapnia,air hunger and panic ensue. The victim struggles violently toreach the surface. Finally, reflex efforts to breathe lead eitherto aspiration of water (in 90% of cases, generally referred toas "wet" drowning), or intractable laryngospasm ("dry"drowning, in 10%).

A phenomenon called "shallow water blackout" occursin some swimmers who hyperventilate before they go under water.This lowers the pC02 enough that susceptible individualsbecome confused and lose consciousness from hypoxia before sufficienthypercapnia develops to overcome the "breath holding breakpoint."Hyperventilation can also increase the propensity for seizures.

Survivors of submersion incidents in either fresh or salt waterseldom aspirate more than 4 mL/kg. For that reason, changes inelectrolytes, hemoglobin, and volume that result from aspirationare rarely of clinical significance. Significant hyperkalemiamay be seen in submersion victims, but this is the result of acidosisand hypoxia, not aspiration.

Submersion injury is a multisystem syndrome with hypoxia asthe most important physiologic consequence. Hypoxia and the directeffect of aspirated fluid combine to produce pulmonary dysfunction.The final common pathway for pulmonary injury in both fresh- andsalt-water submersion is the destruction of surfactant and alveolarcapillary membranes, resulting in noncardiogenic pulmonary edema(adult respiratory distress syndrome, or ARDS). The syndrome ischaracterized by microatelectasis, poor lung compliance, ventilation-perfusionmismatch, and right-to-left intrapulmonary shunting resultingin hypoxemia and hypercarbia. Aspiration of gastric contents canexacerbate the pulmonary injury. Fluid-induced bronchospasm mayalso occur. Aspiration of contaminated foreign material may leadto pneumonia, abscess, or sepsis. Pneumothoraces, oxygen toxicity,and barotrauma resulting from mechanical ventilation may alsocontribute to pulmonary dysfunction. The acidosis seen in criticalsubmersion victims has both respiratory and metabolic componentsand can be profound.14

Hypothermia (core temperature of less than 35° C or 95°F, measured by tympanic thermometry or rectal, bladder, or esophagealprobe) is frequent in pediatric submersion victims, for severalreasons. Conductive loss of body heat is about 30 times greaterin water than in air. Heat loss also accelerates as the drowningvictim struggles, swims a few strokes, and aspirates water. Infantsand young children develop hypothermia especially rapidly, becausethey have less cutaneous fat than older children and teens, andbecause of their high body surface to mass ratio. A child whois suddenly immersed in icy water may go into immediate ventricularfibrillation (prior to core cooling) and die, although this israre.

Dramatic incidents of victims surviving submersion in icy waterbecause of the so-called diving reflex (bradycardia with shuntingof blood to the central circulation in response to cold watersubmersion) have been widely publicized in the press. The protectiveeffect of this reflex has undoubtedly been overemphasized. Recentresearch calls into question the idea that reflexive central circulatoryshunting occurs in humans. The special character of these submersionsis discussed in "Icy water submersion: An exception to therules". Ultimately, hypoxia and ischemia irreparably damagethe central nervous system. Cerebral edema, perivascular hemorrhage,infarction, degenerative changes, and seizures may result fromthe initial insult. Direct injury to the cervical spine can alsooccur, usually related to a diving accident.

Clinical management

The initial management of drowning must be divided into twoparts: what's done at the scene of the accident and what happensonce the child is brought to the emergency department.

At the scene. This is the most crucial time for clinical interventionfor any child showing signs of impending or actual cardiopulmonaryfailure. No therapeutic maneuver is more important than promptresuscitation, especially the institution of effective ventilation.Should you be in the area when a child is pulled from the water,make sure that 911 has been called but don't wait for emergencyservices personnel to arrive before starting bystander CPR. Ifthe events surrounding the accident are unknown, if the victimwas diving, or if there are signs of trauma, maintain the neckin a neutral position to protect the cervical spine. Don't wastetime with the Heimlich maneuver or postural drainage; these proceduresincrease the likelihood of aspiration and make cervical spinecontrol difficult. Use the Heimlich maneuver (or back blows/ chestthrusts in an infant) only if you suspect airway obstruction froma foreign body.16

Wet clothing should be removed and the child covered with ablanket to limit further heat loss. A core temperature shouldbe obtained, ideally with a tympanic or rectal thermometer. WhenEMS personnel arrive, they should begin aggressive initial resuscitation--evenif the victim is in cardiac arrest after a warm water submersion.Standard pediatric advanced life support (PALS) protocols shouldbe implemented.

If signs of respiratory distress are present, the child shouldbe taken to an appropriate medical facility. Transport for furtherevaluation is imperative if the child has stopped breathing atany point or if CPR has been administered--even if the child appearsto be stable. Transport is essential in such cases because assessingmild respiratory distress in a young child is difficult, and postsubmersionpulmonary dysfunction can be progressive.

At the ED. The top priority in the ED is reassessment of ventilationand circulation. If the patient was intubated at the scene, theED physician should assume the tube has been dislodged until itis proven to be in place. Correct endotracheal tube placementcan generally be confirmed by clinical examination. End-tidalC02 monitoring devices can also help in this assessment.If there is reason to suspect a cervical spine injury, protectthe spine and order radiographs. Use pulse oximetry to guide managementdecisions in all cases, and obtain arterial or capillary bloodgases when significant respiratory distress is present. Get achest radiograph if there is clinical evidence of aspiration orrespiratory distress, but be aware that initial radiographic findingsmay underestimate the degree of pulmonary injury. In the morecritical submersion victim, obtain reliable venous access andplace nasogastric and urinary catheters.

Core temperature should be measured with a low-reading thermometer.Excellent discussions of hypothermia management can be found inthe American Heart Association's 1992 Guidelines for Special ResuscitationSituations17 as well as in a 1992 review by HowardCornelis.18 All significantly hypothermic patientsbenefit from humidified air (delivered by mask or endotrachealtube) and a 20 mL/kg intravenous bolus of normal saline, bothwarmed to approximately 43° C. If the child's temperatureis at least 30° C (86° F), external rewarming with heatingpacks or pads, heat lights, or air warmers, and less invasivecore rewarming techniques such as warm gastric and bladder lavagemay also be useful. If the temperature is less than 30° C,use only core rewarming techniques such as heated humidity, warmedintravenous fluid, gastric/ peritoneal/thoracic lavage, or cardiopulmonarybypass. At these lower temperatures, avoid active external rewarmingtechniques, which can can exacerbate shock, acidosis, and cutaneousthermal injury and further decrease core temperature--a phenomenonknown as "afterdrop." When core temperatures are verylow, the heart is highly susceptible to ventricular fibrillation,which can be precipitated by rough handling--particularly at thescene of the drowning. Intubation, however, rarely causes fibrillation.In general, cardiopulmonary by-pass should be considered onlyfor potentially salvageable submersion victims who are profoundlyhypothermic (temperature lower than than 30° C) and also lackan organized cardiac rhythm (QRS complex). The hypothermic heartis not very responsive to pharmacotherapy or countershock. Ifthe core temperature is lower than 30° C, limit countershockattempts to three and use only one dose of epinephrine; furtherattempts should wait until the patient has been rewarmed to atemperature above 30° C. Dopamine, however, may improve cardiacoutput in the severely hypothermic victim. Lidocaine and bretyliummay be useful in both prophylaxis and treatment of hypothermicventricular fibrillation. Do not correct arterial blood gas resultsfor temperature; base ventilation adjustments on the nonadjustedvalues that the laboratory provides.

In addition to pulse oximetry and blood gases, other laboratorytests to consider include glucose, electrolytes, blood urea nitrogen,creatinine, urinalysis, urine myoglobin, clotting studies, andblood alcohol level. These additional tests are generally notnecessary in patients who are not critical. Computerized tomographyof the head should be ordered if a decreased level of consciousnesspersists after metabolic correction or if there is clinical evidenceof significant closed head injury.

Drowning victims should not receive routine prophylactic steroids;steroids can impair normal healing after pulmonary injury anddo not prevent postsubmersion cerebral edema. Routine use of prophylacticantibiotics is generally not recommended, either. At least 40%of pediatric submersion victims develop a fever, with a mediantime of onset at five hours after the drowning incident. The feverusually lasts less than eight hours. Bronchospasm should be treatedwith b-agonist aerosols. Bronchoscopy is indicated for a suspectedforeign body aspiration.

In critical cases, arrange early psychosocial consultationfor the family, without waiting for a definitive prognosis. Parentsand caretakers often feel tremendous guilt and anger in thesesituations, because of real or perceived negligence. They willneed support.

A child who has major symptoms of respiratory distress or neurologicimpairment after stabilization in the ED should be hospitalized.So should victims who may have been abused and children whosehome situation is not reliable for follow-up.

An interesting study of noncritical pediatric submersion victimsby Noonan and colleagues gives a basis for ED disposition in mildlysymptomatic (mild tachypnea, abnormal breath sounds, minimal supplementaloxygen requirement, mildly sleepy) or asymptomatic patients.19The study shows that children who are going to develop symptomsdo so within a few hours of their arrival in the ED; 98% of Noonan'spatients who developed symptoms did so within five hours of thesubmersion and all within eight hours.

Some 70% of patients in the Noonan study (seven of 20) whohad no symptoms when they came into the ED, as well as 57% ofthose who had symptoms initially (35 out of 62), had no symptomseight hours after the submersion and remained normal thereafter.By 18 hours after the event, all patients who were intially asymptomaticand 72% of those who had had symptoms earlier were normal. Thirty-fivepercent of the children with abnormal initial physical exams andan abnormal chest X-ray had a normal exam by eight hours afterthe event, and 80% of those patients were normal within 18 hours.Noonan and colleagues concluded that patients who had no symptomsafter six to eight hours of observation in the ED could be safelydischarged home, with follow-up in 24 hours. They also concludedthat a chest X-ray is not indicated in patients with no initialsymptoms, that chest X-rays are not reliable predictors of theclinical course, and that an abnormal chest X-ray alone shouldnot mandate admission to the hospital. Implicit in these findingsis a valuable role for 24-hour observation units for noncriticalpediatric submersion victims.

Neurologic outcome

The final measure of success in treating the pediatric submersionvictim is ultimate neurologic function. Fortunately, the neurologicprognosis is excellent for the vast majority of children who survivesubmersion. In a 1995 study, Graf and colleagues from Seattleanalyzed data for 194 children hospitalized after warm (non-icy)water submersions.20 All 91 patients who were alerton admission were neurologically normal when they were dischargedfrom the hospital. Nearly all patients (29 of 31) who were lethargicor confused on admission were also normal at hospital discharge.One child was discharged in a vegetative state and one other hadmild neurologic impairment. Several other studies, including a1991 series of 188 children from Britain,21 confirmthat poor outcomes are very unusual among childhood submersionvictims who are conscious when admitted to the hospital.

Unfavorable outcomes in the Graf study were almost exclusivelylimited to the 72 patients who were comatose (not responsive tonoxious stimuli) at the time of their hospital admission. Amongthese 72 children there were 38 deaths (53%) and 14 (19%) weredischarged in a vegetative state. Nine (13%) children had mildto moderate neurologic impairments and 11 (15 %) were neurologicallynormal at hospital discharge. Using logistic regression, Graffound that, among the 72 comatose children, the single variablemost strongly correlated with unfavorable outcome (vegetativestate or death) was absence of the pupillary light reflex in theED. The specificity of this single variable was 72% with a sensitivityof 92%. Somewhat surprisingly, an elevation of the initial bloodglucose over 200 mg/dL was the variable next most strongly correlatedwith an unfavorable outcome. A model combining these two variablessuccessfully predicted which comatose children did well, but miscategorized58% of those with ultimately unfavorable outcomes. Unfortunately,no single clinical parameter or system of parameters has yet beenfound that predicts outcome with absolute reliability.

Aggressive resuscitation before arrival at the hospital, especiallyCPR by family and bystanders, is crucial to the outcome of pediatricsubmersion. A 1994 study from southern California found that childrenwho received CPR from bystanders right away, without waiting forEMS personnel to arive, had significantly better neurologic outcomesthan children who did not--even if the technique used was lessthan perfect.22 And a large 1999 study of out-of-hospitalpediatric cardiopulmonary arrests had similar findings for a subsetof 65 children with serious submersions who received CPR frombystanders before paramedics arrived.23 Forty-one ofthese children (63%) were breathing and had pulses when the paramedicsarrived, and all 41 survived to hospital discharge with full neurologicrecovery. None of the children in the study who did not receivebystander CPR survived.

After warm water submersion, even a child with documented cardiacarrest should receive initial aggressive prehospital resuscitation.Quan and her Seattle colleagues analyzed the outcome for 38 pediatricwarm water submersion victims documented by EMS personnel to bein cardiopulmonary arrest (55% asystole, 29% ventricular tachycardiaor fibrillation, 16% profound bradycardia) before they arrivedat the hospital.24 EMS personnel had implemented PALSfor these children in the field. One third of the children inthis "arrested" group survived and, impressively, 67%of the survivors had a good neurologic outcome.

Prehospital resuscitation is not the end of the story, however.The outcome is also improved by aggressive initial resuscitationon arrival at the ED. A 1993 study from Philadelphia reports intactneurologic survival in three of 24 (13%) pediatric warm watersubmersion victims requiring CPR in the ED.25 Kyriacouin 1994 reported a similar rate of good neurologic outcome infour of 33 (12%) pediatric warm-water submersion victims fromsouthern California who required cardiopulmonary resuscitationin the ED.22

Effective resuscitation efforts in the ED tend to be brief.In the Philadelphia study, the three intact survivors had receivedadvanced cardiac life support (ACLS) in the ED for a maximum of12 minutes. Many studies suggest that the outcome of prolongedED resuscitation is dismal. Quan and her colleagues reported on37 pediatric warm water submersion victims who required more than25 minutes of ACLS, including efforts at the scene as well asin the ED.24,26 They found that all 37 either diedor suffered severe neurologic impairment. In a study of 27 Torontochildren, victims of both warm and cold water submersions, whorequired continued ACLS in the ED, 18 died and five remained ina permanent vegetative state.27 All four intact survivorshad experienced cold water submersions and had initial core temperaturesless than 33° C. A 1999 review looked at 44 different studiesof pediatric CPR after arrests from various causes (includingbut not restricted to submersion injury).28 In eightof these studies, with a total of 347 patients, continuing CPRfor more than 20 to 30 minutes in patients who were not hypothermicdid not produce additional survivors. In four other studies, mostsurvivors required CPR for 15 minutes or less. The authors concludedthat continuing resuscitation for longer than 30 minutes and administeringmore than two doses of epinephrine are, in themselves, indicatorsof poor outcome.

Prevention: The only real cure

"After the fact" treatment has been discussed indetail here, but in an ideal world there would be no need forsuch treatment. Most deaths and injuries from submersion accidentsneed not have happened, and pediatricians should direct most oftheir energies to improving preventive measures. At present, fewpediatricians routinely provide information on drowning preventionto patients and families.4 This situation needs tobe rectified by making water safety advice part of routine preventivecounseling. Encourage parents and teenagers to learn basic pediatricCPR and stress the importance of resuscitation efforts at thescene of the accident. This is crucial, since currently familymembers initiate CPR in only 20% of pediatric out-of-hospitalarrests.

It would be impossible to overemphasize the need for constantadult supervision when young children are playing in or near water.This principle applies with particular force to the bathtub, forchildren younger than 36 months of age. Older siblings or playmatescannot play this supervisory role. Parents also need to be forcefullyreminded that toddlers can and do drown in 5-gallon buckets andin toilets. Advise parents not to allow young children to playalone in the bathroom and to use toilet lid locks.

Families who own swimming pools (or have neighbors or relativeswho do) should be specifically targeted. Effective pool fencingcan significantly decrease the risk of swimming pool drownings,but unfortunately, only about 15% of in-ground pools in the USare adequately fenced. Residential swimming pools should be fencedon all four sides, and the fence should be at least four feethigh. Pool covers and pool alarms alone do not provide sufficientprotection. Pool fences should have self-closing, self-latchinggates. Hot tubs, streams, ponds, and irrigation or drainage canalsin residential areas also need effective barriers to prevent unsupervisedaccess. Children on boats or playing near bodies of water shouldalways wear approved personal flotation devices and be closelysupervised by an adult.

Make sure that parents of children with seizure disorders areaware of the high risks that swimming pools and hot tubs posefor their children. Recommend showers, rather than tub baths,for older children with seizure disorders.

Tell parents that swimming lessons--however valuable they maybe--can never make a child "drownproof." Discuss safediving and boating with teenage patients, and stress the paramountimportance of not mixing alcohol with swimming or water sports.Finally, encourage teens to learn basic CPR techniques. The topten ways to prevent drowning are listed in Table 3 and a set ofguidelines you can copy and distribute to parents is providedin the Guide for Parents.

The American Academy of Pediatrics published detailed, age-specificrecommendations to prevent submersion injuries in its 1993 policystatement, Drowning in Infants, Children, and Adolescents.3These excellent recommendations are specifically designedto be incorporated into office well-child counseling. In addition,the Academy encourages physicians to work for legislation mandatingpool fencing for new and existing residential pools, and has developedmodel legislation (the "Swimming Pool Safety Act").Both the policy statement and the model legislation are availableat the AAP's website, www.aap.org.

THE AUTHOR is Director, Emergency Services, Primary Children'sMedical Center, and Professor of Pediatrics, University of UtahSchool of Medicine, Salt Lake City, UT.

REFERENCES

1. National Safety Council (1997): Accident Facts, 1997. NationalSafety Council, Itasca, IL, pp. 8, 18­19

2. Winternute GJ: Childhood drowning and near-drowning in theUnited States. Am J Dis Child 1990;144:663

3. Committee on Injury and Poison Prevention, American Academyof Pediatrics: Drowning in infants, children, and adolescents.Pediatrics 1993;92:292

4. O'Flaherty JE: Prevention of pediatric drowning and neardrowning: A survey of members of the American Academy of Pediatrics.Pediatrics 1997;99:169

5. Diekema DS, Quan L, Holt VL: Epilepsy as a risk factor forsubmersion injury in children. Pediatrics 1993;91:612

6. Present P: Child drowning study: A report on the epidemiologyof drownings in residential pools to children under age five.Washington, DC: Directorate for Epidemiology, US Consumer ProductSafety Commission, 1987

7. Jensen LR, Williams SD, et al: Submersion injuries in childrenyounger than 5 years in urban Utah. West J Med 1992;157:641

8. Griest KJ, Zumwalt RE: Child abuse by drowning. Pediatrics1989;83:41

9. Lavelle JM, Shaw KN, Seidl T, et al: Ten-year review ofpediatric bathtub near-drownings: Evaluation for child abuse andneglect. Ann Emerg Med 1995;25:344

10. Mann NC, Weller SC, Rauchschwalbe R: Bucket-related drowningsin the United States, 1984­1990. Pediatrics 1992;89:1068

11. Jumbelic M, Chambliss M: Accidental toddler drowning in5-gallon buckets. JAMA 1990;263:1952

12. Shatz DV, Kirton OC, et al: Personal watercraft injuries:An emerging problem. J Trauma 1998;44:198

13. Kluger Y, Jarosz D, Paul DB, et al: Diving injuries: Apreventable catastrophe. J Trauma 1994;36:349

14. Modell JH: Drowning. N Engl J Med 1993;328:253

15. Bolte RG, Black PG, Bowers, et al: The use of extracorporealrewarming in a child submerged for 66 minutes. JAMA 1988;260:377

16. Quan L: Drowning issues in resuscitation. Ann of EmergMed 1993;22:363

17. Emergency Cardiac Care Committee and Subcommittees, AmericanHeart Association: Guidelines for cardiopulmonary resuscitationand emergency cardiac care, IV: Special resuscitation situations.JAMA 1992;268:2244

18. Corneli HM: Accidental hypothermia. J Pediatr 1992; 120:671

19. Noonan L, Howrey R, Ginsburg CM: Freshwater submersioninjuries in children: a retrospective review of seventy-five hospitalizedpatients. Pediatrics 1996;98:368

20. Graf WD, Cummings P, Quan L, et al: Predicting outcomein pediatric submersion victims. Ann Emerg Med 1995;26:312

21. Kemp AM, Sibert JR: Outcome in children who nearly drown:A British Isles study. BMJ 1991;302:931

22. Kyriacou DN, Arcinue EL, Peek C, et al: Effect of immediateresuscitation on children with submersion injury. Pediatrics 1994;94:137

23. Sirbaugh PE, Pepe PE, Shook JE, et al: A prospective population-basedstudy of the demographics, epidemiology, management, and outcomeof out-of-hospital pediatric cardiopulmonary arrest. Ann EmergMed 1999;33:174

24. Quan L, Wentz KR, Gore EJ, et al: Outcome and predictorsof outcome in pediatric submersion victims receiving prehospitalcare in King County, Washington. Pediatrics 1990;86:586

25. Lavelle JM, Shaw KN: Near drowning: Is emergency departmentcardiopulmonary resuscitation or intensive care unit cerebralresuscitation indicated?Crit Care Med 1993;21:368

26. Quan L, Kinder D: Pediatric submersions: Prehospital predictorsof outcome. Pediatrics 1992;90:909

27. Biggart MJ, Bohn DJ: Effect of hypothermia and cardiacarrest on outcome of neardrowning accidents in children. J Pediatrics1990;117:179

28. Young KD, Seidel JS: Pediatric cardiopulmonary resuscitation:A collective review. Ann Emerg Med 1999;33:195

GUIDE FOR PARENTS

Water safety

About 2,000 children in the US die every year from drowning.Parents can prevent these needless tragedies by knowing wherethe danger spots are and taking appropriate precautions. Here'swhat you need to do.

  • Don't leave babies or toddlers alone in the bathroom; more children under the age of 1 year drown in the tub than anywhere else. Toilets are dangerous too; curious toddlers can fall in. To keep children safe, use toilet lid locks and never, never leave a baby or toddler alone in the tub, not even for a second. An older brother or sister can't be trusted to keep a baby safe in the tub; only an adult will do. Large buckets of water are also drowning hazards; don't leave them where toddlers can investigate and fall in.
  • If you have a swimming pool, fence it on all four sides and persuade your neighbors to do the same. Make sure the gate locks automatically and the sides are at least four feet high. Most pool drownings happen in the family's own pool, when parents are home. They happen fast, and silently. Proper fencing is the only way to make sure your child is safe. If you have a hot tub, fence that
    as well.
  • Insist that children wear life jackets when they are playing around water or in boats, make sure an adult is supervising when they go swimming, and don't let them dive in unfamiliar surroundings. Swimming lessons are a good idea, but don't assume that your child is "drownproof" just because she or he has passed a swimming test. There is no such thing as a drown-proof child.
  • Children with seizure disorders need special protection. They should take showers rather than tub baths. Don't allow them to swim without adult supervision.
  • Teach teenagers that water sports can be dangerous and precautions aren't just for sissies. Adolescents--especially boys--are the second most vulnerable age group for drowning, right after toddlers. Nearly one out of every 1,000 boys will drown before he reachers his 20th birthday. Warn your adolescent children to beware of the "three D's" that can lead them into danger when they're around water: drinking, drugs, and dares.
  • Learn CPR and encourage your adolescent children to learn, too. Drowning victims who get immediate CPR from family members or bystanders are much more likely to survive the accident without permanent damage. EMT personnel will take over when they arrive, but knowing what to do before the ambulance comes can save lives.