PEDIATRICS Vol. 108 No. 3 September 2001, pp. 769-775

EXPERIENCE AND REASON:
Pulmonary Edema Associated With Child Abuse: Case Reports and Review of the Literature

	ABSTRACT

Top Abstract Introduction Discussion References

Pulmonary edema has been an unreported finding in the evaluation of abused children. We describe 2 cases of pulmonary edema in abused infants, 1 after confessed suffocation and the other after inflicted head injury. A review of the literature regarding postobstructive and neurogenic pulmonary edema suggests useful inferences for the forensic evaluation of maltreated children who present with this finding.

Children who present to the hospital with acute pulmonary edema may have a cardiogenic or noncardiogenic cause of their symptoms. Noncardiogenic pulmonary edema may occur after head injury, prolonged seizure, acute airway obstruction, or ingestion or inhalation of toxic drugs or chemicals. Although much has been written about the cause of noncardiogenic pulmonary edema, there is little literature describing pulmonary edema associated with child maltreatment. We report 2 abused children1 intentionally suffocated and 1 with severe head injurywhose abuse resulted in noncardiogenic pulmonary edema.

	CASE REPORTS

Case 1

A 6-month-old boy presented to Children's Hospital of Philadelphia (CHOP) with acute respiratory distress. The family stated that the infant was in good health on the day of admission. According to the mother, she had fallen asleep with the infant on her lap after he had been fed several ounces of formula. Approximately 15 minutes later, the mother awoke to find the infant on her lap, coughing and in respiratory distress. The mother noted blood-tinged secretions from his mouth and nose. The infant was on an apnea monitor at the time of the incident. His mother sought help from the maternal grandmother, who called emergency medical services. Emergency medical services arrived to find the infant in respiratory distress with copious blood-tinged frothy secretions in his nose and mouth.

The infant was born at term by vaginal delivery with routine discharge from the hospital. He had been hospitalized previously on 3 occasions for acute life-threatening events at 3 weeks, 2 months, and 4 months of age. Each episode was characterized by respiratory distress and nasopharyngeal bleeding without a source. Evaluations, including pneumograms, a nuclear medicine scan for gastroesophageal reflux disease, an electrocardiogram, an echocardiogram, and an electroencephalogram, failed to reveal cause for his symptoms. The infant was discharged from the hospital after the first hospitalization with an apnea monitor. Although the infant was with the mother during each episode, his mother stated that other relatives (grandmother and great-grandmother) were also present.

The infant had no recent history of fever, upper respiratory symptoms, vomiting, or diarrhea. He had no previous history of wheezing, and there was no family history of asthma or sudden death in childhood among the infant's relatives.

On presentation to an outside emergency department, the infant's initial vital signs were a temperature of 36.8°C, heart rate of 170 beats per minute, respiratory rate of 60 breaths per minute, and blood pressure of 102/67 mm Hg. He was in severe respiratory distress and had an oxygen saturation of 92% with 100% supplemental oxygen delivered by nonrebreathing mask. He had frothy, blood-tinged secretions in his nose and mouth. He had severe respiratory accessory muscle use, coarse breath sounds, and occasional expiratory rhonchi. His abdomen was soft and nondistended. His extremities were cool. His mental status was notable for periods of lethargy and irritability; his neurologic examination was otherwise nonfocal. There were no rashes, petechiae, or other dermatological findings.

Initial arterial blood gas on 100% oxygen delivered via a nonrebreathing mask revealed a pH of 7.24, a pCO₂ of 46 mm Hg, a pO₂ of 66 mm Hg, an HCO₃ of 20 mmol/L, and base deficit of 7 mmol/L. Initial white blood cell count was 19 000/ul (53% segmented neutrophils, 17% bands, 17% lymphocytes, 12% monocytes, and 1% atypical lymphocytes). Initial hemoglobin was 11.3 g/dL, and initial platelet count was 446 000 u/L. Prothrombin time was 12.8 seconds, and partial thromboplastin time was 22.1 seconds; fibrinogen was 112 mg/dL, and fibrin split products were negative.

Because of impending respiratory failure, the infant was intubated with a 4.0 uncuffed endotracheal tube facilitated with vecuronium and lorazepam before his transport to our hospital. Initial ventilator settings included a peak inspiratory pressure (PIP) of 25 cm H₂O and a positive end-expiratory pressure (PEEP) of 10 cm H₂O, set at a rate of 40 breaths per minute. Copious pink, frothy fluid was suctioned from the endotracheal tube. Despite several IV fluid boluses (only 1 of which occurred before intubation or chest radiograph), his extremities remained cool. Dopamine was started at 5 µg/kg/min. Initial chest radiographs revealed bilateral airspace consolidation consistent with pulmonary infiltrates or edema (Fig 1). Ampicillin (200 mg/kg/d) and cefotaxime (120 mg/kg/d) were given. Nebulized albuterol, intravenous methylprednisolone (1 mg/kg/dose every 6 hours), and a terbutaline infusion of 0.5 µg/kg/min were begun. Cranial imaging was obtained; computed tomography (CT; and later magnetic resonance imaging) revealed no structural abnormality or intracranial bleeding. He was transferred to the intensive care unit at CHOP for additional evaluation and management.

View larger version (91K): [in this window] [in a new window]   Fig. 1.   A 6-month-old infant with postobstructive pulmonary edema after confessed suffocation. A) Radiograph obtained within hours of event reveals bilateral airspace consolidation consistent with pulmonary edema. B) Radiograph obtained 4 days later reveals clearing of lung fields, supporting the diagnosis of resolving pulmonary edema.

After arrival in the intensive care unit, ventilation was accomplished with a Drager Evita 4 ventilator (Drager, Inc, Lubeck, Germany), with initial settings of a fraction of inspired oxygen (FIO₂) of 100%, PIP of 35 cm H₂O, PEEP of 10 cm H₂O, pressure support of 5 cm H₂O, and intermittent mandatory ventilation (IMV) of 30. The infant required a dopamine infusion of 5 µg/kg/min for the first 3 hospital days. During this time period, the terbutaline infusion was discontinued, and a dobutamine infusion of 5 µg/kg/min was provided for inotropic support. The infant also received a fentanyl infusion of 1-3 µg/kg/h with intermittent bolus doses of 0.05 mg/kg of midazolam. An echocardiogram obtained during the period of pressor support was normal. Bacterial cultures of the blood and tracheal secretions were negative. Viral culture of tracheal secretions was negative for respiratory syncytial virus, adenovirus, parainfluenza, and influenza. Stool viral cultures were negative.

The patient's lung disease improved rapidly over the first 3 days of hospitalization. Dopamine and dobutamine were discontinued on the third hospital day. On the fifth hospital day, he was extubated; the chest radiograph was now clear, consistent with the diagnosis of resolving pulmonary edema (Fig 1). Consultation by the child protection team was requested because of concerns that the family was not visiting the patient, and because the grandmother had phoned the hospital to express concern that her daughter, the infant's mother, might have harmed the infant.

During interviews by the child protection team, the family revealed that on all 4 occasions, the mother had been alone with the infant before his acute life-threatening event. On each occasion, other family members noted blood-tinged secretions from the nose and mouth. This young adolescent mother had a significant mental health history (including attempted suicide) and substantial involvement with law enforcement for various crimes. The grandmother stated that on at least 1 occasion, she had suspected that the mother put cleaning powder in the bottle of another infant under the grandmother's care.

The police and child welfare were contacted for concerns of possible intentional suffocation or poisoning. During a police interview, the mother reported that she had placed her hand over the infant's mouth to stop his crying. The infant then choked and stopped breathing. One week after presentation to the hospital, the infant was discharged under protective care of the local social service agency. Except for mild hypotonia, he was well at discharge with a final diagnosis of pulmonary edema and respiratory failure attributable to intentional suffocation.

Case 2

An emergency medical team was dispatched to the home of a male babysitter who called emergency medical services to report that a 71/2-month-old boy with no previous history of trauma or seizures had developed seizure activity and stopped breathing. The babysitter reported he had thrown the infant up into the air playfully, the infant vomited once, then became limp and apneic. Two unsuccessful attempts were made to intubate the infant en route to the outside hospital. He therefore received assisted bag-mask ventilation until arriving at the hospital. In the emergency department, he was noted to be apneic and unresponsive with fixed, dilated pupils. His other vital signs were a temperature of 35.9°C, heart rate of 180 beats per minute, and a blood pressure of 75/53 mm Hg. Using a 4.0 uncuffed endotracheal tube, an anesthesiologist intubated the infant without the need for a paralyzing agent or sedative. Initial ventilator settings included an FIO₂ of 100%, IMV of 30 breaths per minute, PIP of 30 cm H₂O, and PEEP of 5 cm H₂O. Arterial blood gas after intubation revealed a pH of 7.61, pCO₂ of 17 mm Hg, and pO₂ of 653 mm Hg; the PIP was reduced to 25 mm Hg, IMV to 25, and FIO₂% to 50%. Other initial laboratory values included a hemoglobin of 8.7 g/dL, a prothrombin time of 14.7 seconds, and a partial thromboplastin time of 49 seconds. A portable chest radiograph confirmed endotracheal tube placement and revealed prominent lung markings but no pulmonary infiltrates (Fig 2). A head CT revealed bilateral subdural hematomae. Initial management also included intravenous administration of 0.5 g/kg of mannitol and 50 mg/kg of ceftriaxone. An isotonic saline infusion was initiated at a rate equal to 1/2 of maintenance fluid. He was then transferred to CHOP for management of acute head trauma.

View larger version (148K): [in this window] [in a new window]   Fig. 2.   A 71/2-month-old boy with severe inflicted head injury and neurogenic pulmonary edema. A) Initial chest radiograph on presentation to the hospital shows relatively clear lung fields. B) Follow-up radiograph several hours later on hospital day 1 reveals evidence of pulmonary edema. C) Radiograph obtained on hospital day 2 shows continued evolution of pulmonary edema. D) Radiograph obtained on hospital day 3 shows clearing lung fields, confirming the diagnosis of pulmonary edema.

On the infant's arrival to CHOP, approximately 3 hours after the 911 call, his vital signs included a temperature of 34.8°C, heart rate of 81 beats per minute, assisted respirations of 25 breaths per minute, and blood pressure of 76/64 mm Hg. General examination revealed an unresponsive infant with no external manifestations of injury. Lung examination revealed coarse breath sounds, and suctioning revealed pink secretions from his nares and endotracheal tube. Pupils were reactive from 4 mm to 2 mm bilaterally. He withdrew to painful stimuli and had spontaneous movement of all 4 extremities. New onset seizure activity was noted, prompting the administration of 20 mg/kg of intravenous dilantin. Ophthalmologic examination showed bilateral diffuse retinal hemorrhages. The patient was admitted to the pediatric intensive care unit for supportive care with a presumed diagnosis of nonaccidental head injury.

On arrival to the intensive care unit, volume ventilation was accomplished using a Servo 300 ventilator (Siemens Elma-AB, Sweden), with the settings adjusted to a tidal volume of 15 mL/kg (PIP of 22-30 cm H₂O), PEEP of 3 cm H₂O, IMV of 15, and FIO₂ of 25% to 30%. The goal of ventilation was to keep the child's pCO₂ between 35 and 40 mm Hg. Chest radiograph at this time revealed bilateral perihilar infiltrates in a butterfly distribution consistent with noncardiogenic pulmonary edema (Fig 2). Heart size was normal. An electrocardiogram was unremarkable. Repeat CT and magnetic resonance imaging of the head, done within 24 hours of admission, revealed bifrontal subdural hemorrhage with hemorrhagic contusions, right vitreal hemorrhage, and mild cerebral edema.

Over the first 24 hours of admission, repeat chest radiographs showed infiltrates that initially worsened, but then resolved by 30 hours after admission. Because of evolving cerebral edema and the development of an asymmetric pupillary examination, the infant was given intermittent intravenous doses of 0.5 g/kg of mannitol followed by 1 mg/kg of lasix for the first 5 days of hospitalization. The goal of this management was to keep the serum osmolality greater than 290 mOsm/kg. Maintenance fluids were continued at a rate equal to that of maintenance, and serum sodium remained stable between 140 and 145 mmol/L. Although intermittent boluses of morphine (0.1 mg/kg) and vecuronium (0.1 mg/kg) were used for agitation on hospital day 0 to 1, they were discontinued by 48 hours to allow for continuous neurologic assessment. The infant also received intravenous dilantin to maintain a serum dilantin level between 15 and 20 µg/mL.

Despite these measures, follow-up CT of the head on day 4 of admission showed diffuse brain edema and uncal herniation. A skeletal survey done 24 hours after admission was negative for fractures. Serial electroencephalogram studies with therapeutic levels of dilantin were negative for seizure activity.

By the sixth hospital day, the intensive care unit team discontinued the intermittent mannitol boluses. Enteral feedings were begun. The infant ultimately required intubation and mechanical ventilation for 11 days. His subsequent neurologic examination was notable for the development of spastic quadriplegia. A gastrostomy tube for enteral feedings was placed before discharge, and the infant was transferred to a chronic care facility. Because this was a case of unexplained severe head trauma and child abuse, the case was referred to child welfare and the police.

	DISCUSSION

Top Abstract Introduction Discussion References

Pulmonary Edema After Airway Obstruction

Postobstructive pulmonary edema (POPE) was first reported in 1977 in a series of case reports of hanging, strangulation, and after tracheostomy to relieve an airway obstruction from laryngeal cancer.¹ Since then, POPE has been described after relief of airway obstruction from foreign bodies,^2,3 epiglottitis, croup,^4-6 laryngospasm,⁷ tonsillar and adenoidal hypertrophy,⁸ and near strangulation.⁹ Our first case is the first detailed report of pulmonary edema and respiratory failure after confessed intentional suffocation. It joins a previous letter to the editors of BMJ as the only 2 cases reported in the medical literature associating pulmonary edema with confessed suffocation of a infant.¹⁰

The diagnosis of pulmonary edema in Case 1 is supported by the rapid onset of respiratory distress and copious pink, frothy secretions before initiation of fluid resuscitation. Early chest radiographs revealed perihilar airspace consolidation without cardiomegaly, suggestive of pulmonary edema. The appearance of the infant's secretions, coupled with the rapid clinical and radiographic improvement in this infant's course of illness, was most consistent with the diagnosis of pulmonary edema and helped to exclude the diagnosis of infiltrates because of aspiration, infection, or ingestion of toxic drugs or chemicals.¹¹

Several authors have speculated about the mechanism causing pulmonary edema after relief of the airway obstruction.^14-6,8 Early theory purported that airway obstruction leads to large negative intrathoracic pressures as the victim forcibly tries to inspire against a closed glottis. Hydrostatic forces would then favor transudation of fluid out from the pulmonary vascular bed into the interstitium.^4,6

It is now believed that the large negative intrathoracic pressures generated after inspiration against an obstructed airway favor the accumulation of blood volume within the thoracic cavity and impede flow to the systemic circulation. The large negative intrathoracic pressure favors a large systemic venous return to the right side of the heart.⁵ Several researchers have demonstrated that the same pressures increase the afterload to the left side of the heart through impedance of arterial circulation in the aorta.^12-14 If hypoxia also leads to hypoxic pulmonary vasoconstriction, and metabolic and respiratory acidosis adversely affect contractility of the heart, the result of these interactions would be to favor retention of blood volume in the thoracic cavity leading to pulmonary edema.

It is also possible that airway obstruction may lead to a ball-valve effect, whereby large positive intrathoracic pressures are created during the period of airway obstruction. When the obstruction is relieved, intrathoracic pressure falls, and the balance of pressure between the vascular bed and interstitium favors fluid shifts to pulmonary interstitium. This theory may better explain patients with chronic airway obstruction, such as infants with tonsillar or adenoidal hypertrophy, who develop postoperative pulmonary edema.⁸

The treatment of a patient with postobstructive pulmonary edema depends on the severity of symptoms. If supplemental oxygen administration and fluid restriction are unable to adequately maintain oxygenation and ventilation, continuous positive airway pressure or PEEP may be necessary. Steroids and diuretics have been administered, but it is unclear if this is of any benefit in shortening the course of illness.⁸

Meadow^15,16 described the historical and clinical features of 81 suffocated infants based on perpetrator confessions or verdicts in criminal and family courts. The majority of infants had histories of previous unexplained life-threatening events with one third of infants having >1 previous event. Nearly half the families had previous infants who died suddenly from unexplained circumstances. Although the presence of facial bruising or petechiae is suggestive of strangulation or suffocation, these findings occurred in only 10% to 15% of infants. Nasopharyngeal bleeding was reported in one third of patients. Care was taken to confirm the finding of gross bleeding as compared with the serosanguinous discharge sometimes observed after sudden unexplained infant death. Meadow reported no infants to have an initial clinical presentation with respiratory distress and pulmonary edema. This case report would suggest that in addition to multiple unexplained life-threatening events, a clinician should also suspect suffocation when such infants present to the hospital with unexplained acute respiratory distress and noncardiogenic pulmonary edema.

Neurogenic Pulmonary Edema (NPE)

NPE, described in our second case, in which a central nervous system (CNS) insult leads to pulmonary edema in the absence of cardiac or other disease, is a phenomenon that has been well documented for almost a century. In 1908, Shanahan¹⁷ first described postictal pulmonary edema in 11 patients (ages 9-36) with epilepsy. In 1918, Moutier¹⁸ described a case of pulmonary edema after a gunshot wound to the head, and in 1939, Weisman¹⁹ described pulmonary edema coexisting with traumatic and spontaneous intracranial bleeds in a large series of patients. Since then, NPE has been described in children and adults after seizures, closed head injury, intracranial hemorrhage, penetrating head trauma, brain tumors, and induction of general anesthesia.^20-27 Despite NPE being well described in the medical literature, there has only been 1 previous report of NPE in a young child with inflicted head injury.²⁸

Signs and symptoms of NPE include tachypnea, dyspnea, chest pain, decreased breath sounds, "bubbly rales," and frothy pink pulmonary secretions. When head injury is severe, the pulmonary edema can be massive, pouring out of the mouth and nares of the patient.²⁹ Chest radiographs confirm the diagnosis, showing diffuse, often butterfly shaped pulmonary infiltrates.

Uncertainty exists regarding the pathophysiology of NPE. Some researchers believe that a rapid increase in intracranial pressure leads to a catecholamine surge important for the development of pulmonary edema.^2130-33 This "blast injury" theory of NPE proposes that an acute rise in intracranial pressure secondary to the CNS insult leads to a massive catecholamine release and a brief but dramatic rise in peripheral vascular resistance. This rise leads to a redistribution of blood from the systemic circulation to the lower resistance vascular bed of the lungs. Coupled with vasoconstriction of the pulmonary bed, the ultimate effect is a transient increase in pulmonary capillary wedge pressure favoring the development of pulmonary edema.³² These physiologic interactions have been demonstrated in a patient whose pulmonary capillary wedge pressures were monitored during the development of an acute case of NPE secondary to intracranial hemorrhage.³⁴

Although it is generally accepted that there is redistribution of blood to the pulmonary circuit and an increase in pulmonary vascular resistance, these forces alone cannot explain the persistence of pulmonary edema in patients with NPE. Proponents of the blast theory have theorized that sympathetic discharge may directly impair endothelial integrity through increased hydrostatic pressures.^32-34 Another emerging theory, however, is that pulmonary edema in many patients with NPE is less explained by a transient sympathetic discharge, but more by centrally mediated changes in capillary permeability in the lungs. Investigators who favor this theory have argued that the prolonged duration of pulmonary edema in many patients with NPE whose pulmonary pressures are measured as normal and the high protein content of the fluid would lend support to this capillary permeability model.^35-37 This model has also been used to explain the minority of adult patients who present with a delayed onset of NPE without evidence of increased pulmonary hydrostatic pressures.^37-38

Ultimately, it is likely that a combination of a large centrally mediated sympathetic discharge and a sustained increased capillary permeability explains the findings in most patients with NPE. Experimental research with a canine model failed to isolate a single mechanism for NPE, suggesting that both models may explain the development of pulmonary edema.^39-41 Despite several investigators showing evidence of a transient sympathetic discharge after an acute rise in intracranial pressure, a recent study revealed that even among head-injured victims with spinal cord transection, there was evidence of pulmonary edema at autopsy.⁴² This led the investigators to hypothesize that a humorally mediated change in capillary permeability must in some way contribute to the findings in patients with NPE.

Clinically, the onset of pulmonary edema is sudden and massive in most cases of NPE. Experimental data in animals confirm that inflicted brain injury can lead to instantaneous pulmonary edema.³² Clinical studies and case reports in humans support the conclusion that NPE develops acutely. In cases of fatal head injury, large retrospective studies of autopsy data suggest the pulmonary edema occurs within hours of the injury. Simmons²⁷ (1969) examined 56 cases of fatal head wounds in soldiers in Vietnam and found that 17/20 patients whose deaths could be classified as instantaneous had massive pulmonary edema at autopsy. Weisman¹⁹ (1939), analyzing autopsy data of patients dying of traumatic intracranial bleeds, found that 17/23 patients who died within a half hour of the trauma had pulmonary edema. Ducker²² (1968) presented a case series of young (ages 12-44), otherwise healthy patients with CNS insults (trauma, bleeds, tumor, seizure) who developed acute pulmonary edema within 2 hours of their event.

Athough the classic presentation for NPE in adults with nonfatal head injury is acute, there have been infrequent descriptions of a delayed-onset of pulmonary edema hours to days after the inciting event.^37,38,42 More recent research, however, has suggested that subclinical pulmonary edema may in fact develop acutely in more patients than previously recognized. Despite a normal appearing early chest radiograph, many head-injured adults with elevated intracranial pressure have evidence of pulmonary dysfunction long before the development of overt pulmonary edema.^37,42 Given that many of these patients do not have significant elevations of their pulmonary pressures, it can be hypothesized that a humorally or hormonally mediated change in capillary permeability accounts for this early pulmonary dysfunction. This subclinical pulmonary edema may subsequently predispose head-injured patients to a higher incidence of adult respiratory distress syndrome or pneumonia during their course.⁴²

Despite a significant body of evidence supporting an acute presentation of NPE in most adults with nonfatal head injury, there are limited reports describing the acute timing of NPE after nonfatal head injury in children.²⁴ Despite the few reports, some meaningful conclusions can be drawn from children who survived other neurologic conditions that can lead to NPE. For instance, children with epilepsy who develop NPE after seizures develop pulmonary edema suddenly with a relatively quick resolution of symptoms. On presentation, these children have diffuse inspiratory rales, pink frothy pulmonary secretions, and chest radiographs with infiltrates. The chest radiographs and symptoms clear within 48 hours.²⁵ NPE is often misdiagnosed in these children as aspiration pneumonitis, but these entities can be distinguished by their radiographic features: aspiration pneumonitis takes longer to present and resolves more slowly.²⁵ Four reports of other nonfatal cases of NPE occurring in children after hemorrhage from an arteriovenous malformation, anesthesia induction, blunt head trauma, and aneurysmal hemorrhage are also similar in the acuteness of presentation and rapid resolution of respiratory symptoms.^20,24,26,29

Although delayed-onset NPE has been described in a minority of adult patients with NPE,^37,38 there are no substantiated reports of a delayed-onset NPE in children. In 1 case, Lear²³ (1990) reported an 8-year-old patient with an arteriovenous malformation and subarachnoid hemorrhage who developed NPE 48 hours after presenting to the emergency department. However, CT scan confirmed a rebleeding hemorrhage at 48 hours. The delay in development of pulmonary edema may have been attributable to the second neurologic insult.

To date, NPE has been unrecognized as a clinical tool in evaluations of inflicted head injury. Clinically, the development of NPE in a child with inflicted head injury may be helpful in timing an alleged assault and thereby identifying a perpetrator. This may be particularly important when an infant presents with severe head injury and an initial clear chest radiograph, which on repeat evaluation, reveals the rapid development of pulmonary edema. In this case, it is likely that the pulmonary edema is a marker for very acute head injury. Furthermore, diagnosing NPE, if only retrospectively, can help to refute the argument that the child's symptoms resulted from aspiration pneumonia. Finally, pulmonary edema may also result from concomitant suffocation of the child, in which case the pulmonary edema would be immediate, and the initial chest radiograph supportive of that diagnosis.

Useful Inferences for Child Abuse Work

We have presented 2 cases of noncardiogenic pulmonary edema associated with child maltreatment. Although both POPE and NPE have been previously reported in the literature, this case series is the first to review the association of pulmonary edema with cases of child maltreatment.

The lessons from such a review of the literature have implications in the field of child maltreatment. For instance, it is well known that victims of intentional suffocation have characteristic family and medical histories that should alert the physician to consider the diagnosis during evaluation of unexplained acute life-threatening events. However, our first case suggests that, on occasion, infant victims of suffocation may also have a presenting sign of postobstructive pulmonary edema. Although the differential diagnosis of acute noncardiogenic pulmonary edema in a infant is extensive, a recent history of wellness should alert the physician to consider exogenous causes (ie, ingestions, suffocation) as the cause of the infant's illness. Furthermore, given the lower likelihood for an ingestion in a nonambulatory infant, the possibility of suffocation would be even greater in this age group.

Another lesson from this review is that pulmonary edema, whether occurring in the suffocated or head-injured infant, usually has a rapid onset. This overlooked finding, by itself, may have important implications in timing an injury. Although there are a few reports in adults to suggest a delayed onset of NPE, most of the medical literature regarding POPE and NPE supports a rapid onset of pulmonary edema after airway obstruction or CNS insult. Although there has been considerable debate regarding the accurate timing of inflicted head injury by clinical or radiographic data, the rapid development of pulmonary edema in a head-injured infant can add additional evidence of a very recent injury and may help identify the perpetrator of the trauma by restricting the time of the injury.

David M. Rubin, MD^*,
Clare O. McMillan, MD
Mark A. Helfaer, MD^{*, §}
Cindy W. Christian, MD^*,
* University of Pennsylvania School of Medicine
 Division of General Pediatrics
^§ Division of Anesthesiology and Critical Care Medicine
Children's Hospital of Philadelphia
Philadelphia, PA 19104

	FOOTNOTES

Received for publication Jan 18, 2002; accepted Mar 22, 2001.

Reprint requests to (D.M.R.) Division of General Pediatrics, Children's Hospital of Philadelphia-Room 2423, 34th Street and Civic Center Blvd, Philadelphia, PA 19104. E-mail: rubin{at}email.chop.edu

	ABBREVIATIONS

CHOP, Children's Hospital of Philadelphia; PIP, peak inspiratory pressure; PEEP, positive end-expiratory pressure; CT, computed tomography; FIO₂, fraction of inspired oxygen; IMV, intermittent mandatory ventilation; POPE, postobstructive pulmonary edema; NPE, neurogenic pulmonary edema; CNS, central nervous system.

	REFERENCES

Top Abstract Introduction Discussion References

1. Oswalt CE, Gates GA, Holmstrom FMG Pulmonary edema as a complication of acute airway obstruction. JAMA 1977; 238:1833-1835 [Abstract/Free Full Text]
2. Boykett M Pulmonary oedema after acute asphyxia in a child. BMJ 1989; 298:928
3. Izsak E Pulmonary edema due to acute upper airway obstruction from aspirated foreign body. Pediatr Emerg Care 1986; 2:235-237 [CrossRef][Medline]
4. Travis KW, Todres ID, Shannon DC Pulmonary edema associated with croup and epiglottitis. Pediatrics 1977; 59:695-698 [Abstract/Free Full Text]
5. Galvis AG, Stool SE, Bluestone CD Pulmonary edema following relief of acute upper airway obstruction. Ann Otol 1980; 89:124-128
6. Kanter RK, Watchko JF Pulmonary edema associated with upper airway obstruction. Am J Dis Child 1984; 138:356-358 [Abstract/Free Full Text]
7. Lee KWT, Downes JJ Pulmonary edema secondary to laryngospasm in children. Anesthesiology 1983; 59:347-349 [Medline]
8. Brandom BW Pulmonary edema after airway obstruction. Int Anesthesiol Clin 1997; 35:75-84 [Medline]
9. Shumaker D, Kottamasu S, Preston G, Treloar D Acute pulmonary edema after near strangulation. Pediatr Radiol 1988; 19:59-60 [CrossRef][Medline]
10. Marcovitch H Inexplicable pulmonary oedema [letter]. BMJ 1989; 298:1383
11. Davis SL, Furman DP, Costarino AT Adult respiratory distress syndrome in children: associated disease, clinical course, and predictors of death. J Pediatr 1993; 123:35-45 [CrossRef][Medline]
12. Peters J, Fraser C, Stuart RS, Baumgartner W, Robotham JL Negative intrathoracic pressure decreases independently left ventricular filling and emptying. Am J Physiol 1989; 257:H120-H131 [Abstract/Free Full Text]
13. Peters J, Kindred MK, Robotham JL Transient analysis of cardiopulmonary interactions. II. Systolic events. J Appl Physiol 1988; 64:1518-1526 [Abstract/Free Full Text]
14. Robotham JL, Scharf SM The effects of positive and negative pressure ventilation on cardiac performance. Clin Chest Med 1983; 4:161 [Medline]
15. Meadow R Suffocation, recurrent apnea, and sudden infant death. J Pediatr 1990; 117:351-357 [CrossRef][Medline]
16. Meadow R Unnatural sudden infant death. Arch Dis Child 1996; 75:57-61 [Abstract/Free Full Text]
17. Shanahan WT Acute pulmonary edema as a complication of epileptic seizures. N Y Med J 1908; 37:54-56
18. Moutier F Hypertension et mort par oedeme pulmonaire aigu, chez les blesse cranio-encephaliques. Presse Med 1918; 26:108-109
19. Weisman SJ Edema and congestion of the lungs resulting from intracranial hemorrhage. Surgery 1939; 6:722-729
20. Braude N, Ludgrove T Neurogenic pulmonary edema precipitated by induction of anaesthesia. Br J Anaesth 1989; 62:101-103 [Abstract/Free Full Text]
21. Casey WF Neurogenic pulmonary oedema. Anaesthesia 1983; 38:985-988 [Medline]
22. Ducker TB Increased intracranial pressure and pulmonary edema; Part 1: clinical study of 11 patients. J Neurosurg 1968; 28:112-117 [Medline]
23. Lear GH Neurogenic pulmonary oedema. Acta Paediatr Scand 1990; 79:1131-1133 [Medline]
24. Milley JR, Nugent SK, Rogers MC Neurogenic pulmonary edema in childhood. J Pediatr 1979; 94:706-709 [CrossRef][Medline]
25. Mulroy JJ, Mickell JJ, Tong TK, Pellock JM Postictal pulmonary edema in children. Neurology 1985; 35:403-405 [Abstract/Free Full Text]
26. Pender ES, Pollack CV Neurogenic pulmonary edema: case reports and review. J Emerg Med 1992; 10:45-51 [CrossRef][Medline]
27. Simmons RL, Martin AM, Heisterkamp CA, Ducker TB Respiratory insufficiency in combat casualties (pulmonary edema following head injury). Ann Surg 1969; 170:39-44 [Medline]
28. Ducker TB, Simmons RL, Martin AM Pulmonary edema as a complication of intracranial disease. Am J Dis Child 1969; 118:638-641 [Abstract/Free Full Text]
29. Kosnik EJ, Paul SE, Rossel CW, Sayers MP Central neurogenic pulmonary edema: with a review of its pathogenesis and treatment. Child Brain 1977; 3:37-47 [CrossRef]
30. Carlson RW, Schaeffer RC, Michaels SG, Weil MH Pulmonary edema following intracranial hemorrhage. Chest 1979; 75:731-734 [Abstract/Free Full Text]
31. Rippe B, Townsley M, Thigpen J, Parker JC, Korthuis RJ, Taylor AE Effects of vascular pressure on the pulmonary microvasculature in isolated dog lungs . J Appl Physiol 1984; 57:233-239 [Abstract/Free Full Text]
32. Theodore J, Robin ED Speculations on neurogenic pulmonary edema . Am Rev Respir Dis 1976; 113:405-411 [Medline]
33. Wray NP, Brooke N Pathogenesis of neurogenic pulmonary edema. Am Rev Respir Dis 1978; 118:783-786 [Medline]
34. Fein IA, Rackow EC Neurogenic pulmonary edema. Chest 1982; 81:318-320 [Abstract/Free Full Text]
35. Colice GL, Matthay MA, Bass E, Matthay RA Neurogenic pulmonary edema. Am Rev Respir Dis 1984; 130:941-948 [Medline]
36. Colice GL Neurogenic pulmonary edema. Clin Chest Med 1985; 6:473-489 [Medline]
37. Mackersie RC, Christensen JM, Pitts LH, Lewis FR Pulmonary extravascular fluid accumulation following intracranial injury. J Trauma 1983; 23:968-975 [Medline]
38. Fisher A, Aboul-Nasr HT Delayed nonfatal pulmonary edema following subarachnoid hemorrhage. J Neurosurg 1979; 51:856-859 [Medline]
39. Maron MB A canine model of neurogenic pulmonary edema. J Appl Physiol 1985; 59:1019-1025 [Abstract/Free Full Text]
40. Maron MB Analysis of airway fluid protein concentration in neurogenic pulmonary edema. J Appl Physiol 1987; 62:470 [Abstract/Free Full Text]
41. Maron MB Pulmonary vasoconstriction in a canine model of neurogenic pulmonary edema. J Appl Physiol 1990; 68:912 [Abstract/Free Full Text]
42. Rogers FB, Shackford SR, Trevisani GT, Davis JW, Mackersie RC, Hoyt DB Neurogenic pulmonary edema in fatal and nonfatal head injuries. J Trauma 1995; 39:860-868 [Medline]