PIR_abril_2010

DOI: 10.1542/pir.31-4-135 2010;31;135-144 Pediatr. Rev.

Albert Ross and Neal S. LeLeiko Acute Abdominal Pain

http://pedsinreview.aappublications.org/cgi/content/full/31/4/135located on the World Wide Web at:

The online version of this article, along with updated information and services, is

Pediatrics. All rights reserved. Print ISSN: 0191-9601. Online ISSN: 1526-3347. Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2010 by the American Academy of published, and trademarked by the American Academy of Pediatrics, 141 Northwest Pointpublication, it has been published continuously since 1979. Pediatrics in Review is owned, Pediatrics in Review is the official journal of the American Academy of Pediatrics. A monthly

by Enrique Mendoza-Lopez on April 2, 2010 http://pedsinreview.aappublications.orgDownloaded from

http://pedsinreview.aappublications.org/cgi/content/full/31/4/135

http://http://pedsinreview.aappublications.org

Acute Abdominal PainAlbert Ross, MD,*

Neal S. LeLeiko, MD, PhD*

Author Disclosure

Drs Ross and LeLeiko

have disclosed no

financial relationships

relevant to this

article. This

commentary does not

contain a discussion

of an unapproved/

investigative use of a

commercial product/

device.

Objectives After completing this article, readers should be able to:

1. Understand the principal causes of acute abdominal pain in children.2. Describe the characteristics of visceral versus somatic abdominal pain.3. Be familiar with the differential diagnosis of abdominal pain based on symptoms and

location of pain.4. Discuss the evaluation of acute abdominal pain.5. Distinguish surgical from medical abdominal pain.

The Problem“Hello, Doctor Jones, Billy has an awful tummy ache!” For such a simple statement, so manypossible outcomes exist. Is this an emergency? Does he have appendicitis? Does he need asurgeon? Is this something trivial? Has Billy eaten something harmful? Is he constipated?Acute abdominal pain can be caused by myriad conditions whose outcomes vary from rapidimprovement to surgery, posing a diagnostic Gordian Knot. However, through evaluationof the patient’s history and symptoms and the use of technology, a pediatrician usually canarrive at a reasonable conclusion about the care of the patient, even if the diagnosis still isundetermined.

Acute abdominal pain can be classified according to its location and nature, history, orassociated signs (Table 1).

Location and NatureSome conditions can cause pain in different regions, and it may be difficult to associate thedisease with the location of the pain. Localization of the source of abdominal pain isconfounded by the nature of the pain receptors involved. Further, the type of painassociated with a particular disease may change as the disease process progresses, as inappendicitis. Abdominal pain may be classified as visceral, somatoparietal, and referredpain. Most abdominal pain is associated with visceral pain receptors.

Visceral pain receptors are located in the muscles and mucosa of hollow organs, in themesentery, and on serosal surfaces. These pain receptors typically respond to stretch, suchas when the bowel is distended or mesentery is stretched or torsed. Visceral pain responseis not well localized because the afferent nerves associated with this pain have fewer nerveendings in the gut, are not myelinated, are bilateral, and enter the spinal cord at severallevels. However, there are three broad areas of association. Visceral pain in the stomach,lower esophagus, and duodenum is perceived in the epigastric area. Pain emanating fromthe small intestine is felt in the umbilical area. Colonic visceral pain is felt in the lowerabdomen. The pain can be described as dull, diffuse, cramping, or burning and mayprompt the child to move in an attempt to decrease the pain. Because autonomic nervesmay be involved secondarily in the same pathologic process, patients also may exhibitsweating, nausea, vomiting, pallor, and anxiety.

Somatoparietal pain receptors are located principally in the parietal peritoneum, muscle,and skin. These pain receptors typically respond to stretching, tearing, or inflammation.The nerves that convey somatoparietal pain travel within specific spinal nerves that aremyelinated and numerous and that transmit to specific dorsal root ganglia. The pain ismore localized, associated with one side or the other, more intense, and more often is

*Division of Pediatric Gastroenterology, Nutrition and Liver Diseases, Hasbro Children’s Hospital/Rhode Island Hospital,Providence RI.

Article gastrointestinal

Pediatrics in Review Vol.31 No.4 April 2010 135 by Enrique Mendoza-Lopez on April 2, 2010 http://pedsinreview.aappublications.orgDownloaded from


described as having a sharp quality. Movement usuallyintensifies parietal pain so the child will stay still or splintwhen walking.

Referred pain arises when visceral pain fibers affectsomatic nerve fibers in the spinal cord or central nervoussystem. This pain generally is well localized but distantfrom the affected site. For example, any inflammatoryprocess that affects the diaphragm can be perceived aspain in the shoulder or lower neck because of conver-gence of the nerve pathways of these two regions.

History and SymptomsOften, the most important component needed for diag-nosis is a history. The order of onset of symptoms, theirprogress, characteristics of emesis (Table 2) and stools(Table 3), and application of knowledge of the nature ofpain are important. For example, the pain of appendicitismay have started a day or two before presentation, grad-ually increasing in severity and changing location. Thatpain generally begins as poorly localized and vague (ie,visceral receptors are affected), but as the inflammation

Table 1. Differential Diagnosis Mapped to Location of Abdominal Pain

Epigastric Right Upper Quadrant Left Upper Quadrant

Gastroesophageal reflux Hepatitis SplenomegalyEsophagitis Cholecystitis Splenic infarctionGastritis Cholelithiasis Traumatic spleen injuryGastric ulcer Biliary colic Lower left lobe pneumoniaDuodenal ulcer Cholangitis Kidney diseasePancreatitis Right lower lobe pneumonia Urinary tract diseaseGastric volvulus Kidney diseaseSmall bowel volvulus Urinary tract diseaseErythromycin inducedNon-steroidal inflammatory medication-induced

Hypogastric Left Lower Quadrant Right Lower Quadrant

Constipation Constipation ConstipationColon spasm Colon spasm Mesenteric adenitisColitis Colitis Crohn diseaseBladder disease Ovarian torsion Acute obstructionUterine conditions Ectopic pregnancy Localized perforationPelvic inflammatory disease Testicular torsion Appendicitis

Hernia IntussusceptionSigmoid volvulus Ovarian torsion

Ectopic pregnancyTesticular torsionHernia

Periumbilical “All Over” “All Over”

Functional disease Gastroenteritis PorphyriaConstipation Perforation Sickle cell crisisGastroenteritis Constipation VolvulusEarly appendicitis Functional disease Abdominal migrainePancreatitis Colic Cyclic vomiting syndromeSmall bowel volvulus Streptococcal pharyngitis Lead poisoningHenoch-Schonlein purpura Intussusception Iron ingestionIncarcerated umbilical hernia Inflammatory bowel disease Familial Mediterranean fever

Henoch-Schonlein purpura Angioneurotic edemaDiabetic ketoacidosis Venomous bite

Location Varies

TraumaInfarctionGluten-sensitive enteropathy

gastrointestinal acute abdominal pain

136 Pediatrics in Review Vol.31 No.4 April 2010 by Enrique Mendoza-Lopez on April 2, 2010 http://pedsinreview.aappublications.orgDownloaded from


continues and the appendix swells, pain fibers in the pari-etal peritoneum are stretched, and the pain becomes morelocalized toward the right lower quadrant. The classic pat-tern in intussusception is intermittent crampy pain.

Besides location, associated signs and symptoms canhelp determine the cause of the pain. The color of emesiscan be a useful clue, as can the appearance of the stool.

Laboratory DiagnosisIn perplexing cases, laboratory studies frequently arerequested and, with notable exceptions, are remarkablyunhelpful. Studies generally include a complete bloodcount, erythrocyte sedimentation rate, and urinalysis.Adolescent females should have a pregnancy test (re-gardless of whether they have experienced menarche).

The white blood cell count can be misleading andonly can confirm the examiner’s suspicions; it cannot berelied on to exclude serious illness. The sedimentationrate, when elevated, can indicate the presence of aninflammatory process. However, like the white blood cellcount, it cannot be diagnostic.

The urinalysis is relatively easy to obtain and can revealthe presence of a urinary tract infection, diabetes, nephri-tis, and sometimes, chronic kidney disease.

Other specific laboratory studies may be appropriatebased on the conditions being considered, but generallythey are not helpful in the immediate acute setting.

General Symptoms and Assessmentof SeverityAcute abdominal pain can be caused by an easily reme-diable problem or by a serious condition that requiressurgery. Confounding the presentation is the variety ofpatient responses to pain, from the stoic to the hysterical.Some of the loudest patients have functional pain. Un-fortunately, no signs of illness are absolute, but certainwarning signs can suggest more serious illness.

First, does the patient look ill? If the child is bouncingaround the room, it is easier to provide reassurance andavoid excessive testing. When the patient appears ill, it isimportant to distinguish whether the child is improvingor worsening. Observing the patient in the office oremergency department for several hours allows serialexaminations. A rapid revisit or even admission may beprudent until serious disease is excluded. History ofabdominal trauma, pain worsening with movement, in-voluntary guarding, rebound tenderness, and tendernesswith percussion are indications for prompt surgical eval-uation. Are there signs of bleeding or significant volumeloss or dehydration? These clues may not lead to surgerybut need to be addressed to restore well-being. When indoubt, the pediatrician should detain the patient, per-form serial examinations, and ask the surgeon for help.

GastroenteritisProbably the most common cause of acute abdominalpain is infectious gastroenteritis. Obtaining a history ofrecent travel, ill contacts, and diet (food-borne patho-gens) is important. Viruses are the more common cul-

Table 2. Differential Diagnosis Basedon the Color of the Vomitus

Emesis Suggested Diagnosis

Bile-colored ObstructionMidgut volvulus

Coffee ground- Esophagitiscolored Gastritis

Gastric ulcerTrauma from forceful vomiting

Bright red blood, Esophagitissmall volume Gastritis

Bright red blood, Esophageal tear (Mallory Weiss tear)large volume Gastric ulcer

Duodenal ulcerEsophageal varices

Food or gastric Infectious gastroenteritiscontents Obstruction

Fecal appearance Obstruction

Table 3. Appearance of StoolStool Suggested Diagnosis

Watery diarrhea InfectionBacterialViralParasitic

Appendicitis with perirectalabscess

Hard or large stool ConstipationDecrease in stool Constipation

frequency ObstructionMucus-containing Colitis (can be normal)Bright red blood, Constipation

small volume FissureHemorrhoid, suggesting

constipationColitisHenoch-Schonlein purpuraPolyp

Bright red blood, Colitislarge volume Polyp

“Currant jelly stool” IntussusceptionMelena Gastric ulcer

Duodenal ulcerPale, acholic stools Biliary or hepatic disease




prits, but bacteria and parasites also can produce acuteillness. Clinical findings vary, based on the infectiousorganism, but many of these agents cause fever, vomit-ing, and diarrhea along with pain. The pain usually isnonspecific in location, and the child also can have diffusetenderness, although “guarding” is unlikely. If bloodydiarrhea is present, stool cultures and stool examinationfor parasites should be requested. Antibiotics can worsenserious illness such as hemolytic-uremic syndrome andshould not be used without clear indication. An acutepresentation with blood in the stool is more likely a signof infectious colitis but may be the initial presentation ofinflammatory bowel disease. Positive bacterial culturesmust be reported to appropriate authorities.

Most causes of acute abdominal pain that requiresurgery do not present in this manner. Generally, fever,vomiting, and diarrhea indicate acute-onset infectionrather than surgical disease. In some cases, particularlywhen the child looks ill, making the distinction can bedifficult.

Rehydration is beneficial. Oral rehydration is pre-ferred, but intravenous fluids may be used until oraltherapy can be started. Rehydration during acute gas-troenteritis usually makes the child feel much better.Improving appearance with rehydration is reassuring.

Acute AppendicitisInflammation of the appendix results in distention lead-ing to ischemia. Necrosis, perforation, and peritonitis orabscess may ensue. It is not known why the appendixbecomes inflamed, but a fecalith or lymphoid tissue

obstructing the lumen may be the precipitating cause.Acute appendicitis is the most common reason for emer-gent abdominal surgery in children.

Unfortunately, it still is difficult to be certain of adiagnosis of acute appendicitis. Timely diagnosis is criti-cal, but it can be extremely challenging, especially inyoung children. As the inflammation starts, the visceralnerves send a message of general unease, which maymanifest as pain referred to the umbilical region, thenanorexia, typically followed by nausea. A young child hasa hard time explaining this feeling and may show onlyanorexia and decreased activity.

Vomiting, fever, guarding, and abdominal pain withany movement (especially walking) are important signswhen present. Requesting the patient to hop off of theexamination table or hop up and down usually is refusedor elicits a dramatic increase in abdominal pain.

As inflammation increases and the parietal perito-neum becomes irritated, the somatic nerves begin tosignal that something is wrong. This pain usually isappreciated in the area two thirds of the distance from theumbilicus to the anterior superior iliac spine (McBurneypoint). Pain and tenderness in this location are sensitivesigns for appendicitis but, unfortunately, are not specificfor appendicitis (Table 4).

If the appendix ruptures, a child can show clinicalimprovement as the pressure in the organ is released, thusdecreasing pain. Over the next day, the child may worsendue to peritonitis; sometimes, a localized abscess formsinstead. With abscess formation, the right lower quad-

Table 4. Signs of AppendicitisTenderness at McBurney point Percussion or palpation in the right lower quadrant results in pain in an area

approximately two thirds of the distance from the umbilicus to the anteriorsuperior iliac spine

Involuntary guarding Abdominal wall muscle spasm to protect inflamed abdominal organs from motionPain on movement Significant increase in pain with walking, hopping off of the table, or jumping up

and downRovsing sign Pressure in the left lower quadrant results in pain in the right lower quadrantRebound tenderness Sudden release of deep palpation of the abdomen results in a large increase in

pain. (Save this test for the end of the examination to stay in the child’s goodgraces.)

Psoas sign With the patient on his or her left side, extend the right thigh while applyingstabilizing resistance to the right hip. This maneuver should cause an increase inpain due to the location of the appendix over the iliopsoas muscle.

Obturator sign Increased pain with passive flexion and internal rotation of the right thighAnorexiaNauseaVomitingFeverBent knees The child is most comfortable while lying with knees bent




rant pain can continue, and a tender mass becomespalpable.

Many maneuvers can elicit pain associated with ap-pendicitis, and the clinician should be familiar with atleast some of them (Table 4).

Diagnostic laboratory tests for appendicitis includethe white blood cell count, which typically is mildlyelevated and may have a shift toward neutrophils but isnot a reliable diagnostic test. Radiologic studies havebecome more helpful in determining the presence ofappendicitis and can help define an abscess and demon-strate other causes of pain such as a renal stone, Crohndisease, or gynecologic problems. Right lower quadrantultrasonography often can show enlargement of theappendix as well as changes in the wall, presence ofincreased fluid around the appendix, or an abscess if theappendix has ruptured. Because ultrasonography doesnot expose a child to radiation or contrast, it is pre-ferred to computed tomography (CT) scan, althoughCT scan may be necessary when the physical findings areuncertain and an experienced ultrasonographer is notavailable.

Because there is no perfect test for appendicitis otherthan the pathology report, the best diagnostic instru-ment is the examiner. Appendectomy is the appropriatetreatment.

Small Bowel VolvulusVolvulus is a surgical emergency; delay in surgical inter-vention can cause short gut or death. Incomplete rota-tion of the embryonic bowel results in the vascularsupply of the small intestine flowing through a narrowpedicle of mesentery, which can twist about its base,cutting off blood flow. Dull, aching abdominal pain maybe the first symptom, but more dramatic pain also may bethe presentation. Obstructive symptoms are followed byan acutely inflamed abdomen.

Volvulus typically presents early, before 1 year ofage, but it can occur at any age. The obstruction resultsin bile-stained emesis and pain, although the pain canbe hard to detect in infants. Bile-stained emesis signals asurgical emergency. Rectal bleeding is a late sign indicat-ing vascular compromise to the mucosa.

A plain radiograph may show a dilated stomach andproximal duodenum, but the primary test for a volvulusis a contrast upper gastrointestinal study. Recently,Doppler ultrasonography has been used to detect volvu-lus and malrotation.

The bowel must be untwisted before vascular necrosisoccurs. An appendectomy typically is performed be-cause the appendix would be left in an abnormal loca-

tion, which would make diagnosing appendicitis moredifficult.

IntussusceptionProbably the most frequent cause of intestinal obstruc-tion in children is an intussusception. This conditionoccurs when part of the intestine is pulled antegrade intothe adjacent part of intestine, trapping the more proximalbowel in the distal segment. The most common site is thejunction of the ileum and colon, where the ileum ispulled into the colon. In some cases, a lead point such asa polyp, tumor, or Meckel diverticulum is pulled down-stream. The cause in infants typically is unknown. Somehave suggested hypertrophy of mesenteric lymph nodescaused by a viral infection.

Like a volvulus, intussusception occurs more com-monly in infants than in older children. The signs andsymptoms include abdominal pain, lethargy, vomiting,pallor, and if the obstruction is prolonged, abdominaldistention and rectal bleeding. The bloody bowel move-ment in this illness often is described as looking like redcurrant jelly. Such a stool, however, is not seen com-monly, but when seen, it suggests vascular compromise.

The child may show signs of crampy pain when peri-stalsis occurs and causes additional stretching andsqueezing of the trapped intestine. The child may liequietly between the peristaltic waves.

Older children often localize the pain to the perium-bilical region, but it can be in the right lower quadrant.Appendicitis may be suspected, but the pain often isintermittent in intussusception rather than continuous.

In the most common form of intussusception, ileo-colic, a sausage-shaped mass may be palpable on the rightside or in the right upper quadrant of the abdomen.Abdominal radiographs may show obstruction, and amass also may be visible. Ultrasonography demonstratesbowel within bowel or a “target.” Ultrasonography isvery accurate in detecting intussusceptions and is consid-ered the test of choice.

Treatment (and confirmation of the intussusception)is with an air contrast enema. Air is safer and cleaner thanliquid and is more effective. If the enema fails, surgerymust be performed to reduce the intussusception.

Henoch-Schonlein PurpuraBecause the rash of Henoch-Schonlein purpura (HSP)may present after the onset of abdominal pain, severeacute pain can be the initial sign of the condition. HSP isa vasculitis that can be triggered by infection, medica-tions, or even insect bites. The rash begins on the but-tocks or extensor surfaces of the legs and may spread




peripherally. It can start as urticaria but progresses to“palpable purpura.”

The intestine also shows purpuric lesions, and theedema and inflammation result in colicky pain. Childrenalso may vomit with HSP, and the abdomen can betender to palpation. The lesions can lead to gastrointes-tinal bleeding or complications of intussusception orperforation. HSP usually affects children younger than10 years of age, but it is rare in infants. HSP recurs inabout one third of cases.

Arthralgias or arthritis are seen in most cases, with thelower extremity large joints affected most often. HSPalso can lead to nephropathy in up to 50% of the children.The renal involvement usually is mild and may presentweeks after the abdominal pain.

If the typical rash is seen, no testing is indicated.Ultrasonography or contrast radiographs of the intes-tines show the edematous lesions in the gut. Endoscopydemonstrates purpuric lesions. The white blood cellcount can be increased, and markers of inflammationsuch as the sedimentation rate usually are elevated. Oc-casionally, there are no other signs of HSP apart from theabdominal pain, and the diagnosis may be made afterobserving purpuric lesions of the gastrointestinal tract onendoscopy.

Treatment is supportive. In the case of severe joint orabdominal pain, prednisone can be used to decreasesymptoms.

PancreatitisUpper abdominal pain and tenderness, especially whenassociated with vomiting, are typical features of pancre-atitis as well as many other diseases. To determine ifpancreatitis is present, serum amylase or lipase must bemeasured. If concentrations of these enzymes are greaterthan three times the upper limit of normal, pancreatitismost likely is the cause of the symptoms. Normal valuesdo not exclude the diagnosis.

Pancreatitis arises from many different infections,medications, or trauma. Other causes include gallstones,abnormal ductular anatomy, systemic illness, and meta-bolic problems. The cause in any specific patient can behard to determine, and finding the cause can be expen-sive. Therefore, in an isolated case, an exhaustive search isnot necessary. Most children who experience acute pan-creatitis do not suffer additional episodes.

CT scan or ultrasonography can help diagnose pan-creatitis as well as look for anatomic causes or gallstones.In recurrent episodes of acute pancreatitis, pancreatic-sufficient cystic fibrosis should be excluded, along withgenetic forms of pancreatitis. Magnetic resonance chol-

angiopancreatography or endoscopic retrograde chol-angiopancreatography should be considered.

Treatment is supportive. The patient may eat if fooddoes not cause pain. Narcotics should be used for severepain. Intravenous fluids and intravenous acid suppressionare used. If vomiting continues with gut “rest,” a naso-gastric tube can be used to decompress the stomach. Insevere cases, patients require intensive care due to thefluid shifts and hypotension accompanying necroticpancreatitis.

Ulcer DiseaseEpigastric or right upper quadrant pain can signify apeptic ulcer. These lesions usually occur in the distalstomach or proximal duodenum. In severe cases, bleed-ing or perforation can occur. Ulcer symptoms are com-mon because many children have nonulcer dyspepsia, inwhich there is pain similar to that caused by an ulcer, butno ulcer is present.

Nonsteroidal anti-inflammatory drugs (NSAIDs)such as ibuprofen are an important cause of ulcers anddyspepsia in children. Some ulcers are caused by infectionwith Helicobacter pylori. H pylori ulcers are less commonin children than in adults, and more ulcers fall into theidiopathic category. Eosinophilic gastroenteritis, Crohndisease, and any severe illness can be associated with ulcerdisease as well.

Ulcers are diagnosed with upper gastrointestinal en-doscopy. Biopsies should be taken to look for H pylori aswell as other causes of ulcers.

Treatment is with acid suppression, typically withproton pump inhibitors (PPIs). Histamine-2 receptorantagonists (H2RAs) also are used but are not as effec-tive as PPIs at suppressing acid production. Antacidpreparations can help with symptoms and provide addi-tional buffering. If H pylori is found, antibiotics also arenecessary.

Bleeding ulcers can be treated endoscopically withcautery, injection, and mechanical methods. Surgery isused when endoscopic therapy and medications fail orwhen there is a perforation.

GastritisGastritis can feel the same as an ulcer, and diagnosis ismade by endoscopy. Gastritis has many different causes,with acute infectious gastritis and NSAID therapy beingtwo of the more frequent. Treatment is to remove anyprecipitating agent, provide acid suppression, and givesupportive care.




NSAID-induced DyspepsiaPatients who must be treated with NSAIDs are at risk ofdeveloping NSAID-related gastrointestinal dyspepsia,which may be a manifestation of gastritis or gastric ulcer.The risk of NSAID-related complications is increasedby a history of ulcer disease or bleeding or by use ofcorticosteroids. In patients receiving NSAIDs, com-plications can be lessened by providing adequate acidsuppression. Standard doses of H2RAs do not preventmost NSAID-related gastric ulcers. Doubling the dosemay be effective, but single daily doses of a PPI aresuperior to H2RAs and other treatments (including mi-soprostol) in reducing ulcers and NSAID-associateddyspepsia.

EsophagitisGastroesophageal reflux disease (GERD) and esophagitiscan present acutely as epigastric abdominal pain. GERDis present when gastric contents move into the esophagusand produce symptoms or damage. Esophagitis can re-sult from GERD or from other inflammation such aseosinophilic esophagitis or from infections such as herpesor Candida.

Treatment is with antacids and acid suppression forthe pain and ongoing therapy for reflux. If the childdoes not improve, endoscopy to look for inflammation,infection, or possibly even a foreign body should beperformed.

HepatitisInflammation of the liver can cause right upper quad-rant pain. Anorexia, nausea, and vomiting also are com-mon in hepatitis. The liver can be inflamed due toinfection, reactions to medications or chemicals, orautoimmune hepatitis. Clues to hepatitis being thesource of the acute pain include jaundice, hepatomegaly,and liver tenderness.

Liver enzyme concentrations are elevated in acutehepatitis. The child also should have direct hyperbili-rubinemia. Urinalysis can be a screening test for liverdisease by detecting bilirubin and urobilinogen.

Acute infectious hepatitis is treated with supportivecare and is prevented best by immunization for hepatitisA and B and avoiding behaviors that can lead to hepatitisC or E infection.

Biliary Tract DiseaseCholelithiasis and Cholecystitis

Right upper quadrant abdominal pain can result fromgallstones and cholecystitis. Pain, fever, vomiting, andjaundice often are present. The pain occasionally radiates

to the back. A positive Murphy sign is strongly suggestiveof gall bladder disease. The examiner palpates the rightupper quadrant at the costal margin while the patientbreathes in. The sign is considered to be positive if thepatient feels pain. Gallstones are seen more frequentlywith hemolytic disorders, such as sickle cell disease, andin infants and children who have received peripheralalimentation. Acalculous cholecystitis typically occursduring a significant systemic illness such as sepsis or anillness requiring a stay in the intensive care unit.

Ultrasonography can show the presence of stones anda thickened gall bladder wall with possible gall bladderdilatation. The ultrasonographer can produce a positiveMurphy sign with the transducer, which helps to diag-nose cholecystitis.

Laboratory tests should show an elevation in liver en-zymes, especially gamma-glutamyltranspeptidase (GGTP)and alkaline phosphatase. The white blood cell count iselevated, and direct bilirubin is increased. The amylasevalue can be elevated, making it harder to know if theproblem is cholecystitis or pancreatitis.

Treatment consists of bowel rest, intravenous paincontrol, and intravenous fluids. If fever is present or thechild looks ill or unstable, antibiotics are needed forenteric bacteria. The timing of curative cholecystectomyis determined best with the surgeon. Complications ofcholecystitis include perforation of the gall bladder, withperitonitis or abscess formation.

Acute hydrops of the gall bladder can look like acal-culous cholecystitis, but the gall bladder wall is notinflamed. The symptoms usually are the same, but ultra-sonography shows an enlarged gall bladder without wallthickening. Treatment is supportive rather than surgical,but perforation can result, which requires surgery.

Stones in the Bile DuctCholedocholithiasis, or bile duct stones, can presentmuch like cholecystitis, but jaundice is more likely to bepresent. Right upper quadrant pain, fever, and tender-ness are consistent with impacted stones.

GGTP, alkaline phosphatase, and conjugated bili-rubin concentrations are elevated, as might beaminotransferase values. Because duct obstruction cancause pancreatitis, amylase and lipase values should beassessed.

Ultrasonography usually demonstrates the stone, butsometimes stones can be hard to see. A dilated duct alsomay be present.

If fever is present, antibiotics should be started. Thechild is given nothing by mouth but should receiveintravenous fluids and narcotic analgesics.




The gastroenterologist and surgeon should be con-sulted if the stone does not pass spontaneously becauseeither surgery or an endoscopic retrograde cholangio-pancreatography with stone removal may be necessary.

ConstipationOne of the more common treatable causes of acuteabdominal pain is constipation, which can cause severepain in some children and raise concerns about moreserious illness. Constipation can follow a social changesuch as toilet training, starting school, changing the diet,or taking a trip. The child frequently does not know thathis or her pattern of stooling has become abnormal, andthe parent is not aware of a change. Nausea can accom-pany constipation, but other symptoms are rare.

Examination may show distention, a mass in the leftlower quadrant or low mid-abdomen, and mild tender-ness when the mass is palpated. The rectal examinationusually demonstrates a full rectal vault in contrast toHirschsprung disease, in which the rectum contains littlestool. Guarding is not typical. An abdominal radiographshould show a full rectal vault and fecal loading, but signsof obstruction are absent.

Treatment varies, depending on the age of the childand the degree of constipation.

Incarcerated Inguinal HerniaSigns of intestinal obstruction with abdominal pain ac-company an incarcerated inguinal hernia. Examinationshould reveal a groin mass that may be tender andsometimes can be red due to the underlying inflamma-tion. An abdominal radiograph may show obstruction oran air bubble in the groin.

The best therapy is early repair. Therefore, healthsupervision examinations should include an evaluationfor hernias. Emergent surgery is required to treat anincarcerated hernia.

Urinary Tract DiseaseUrinary tract infections and renal stones can present asabdominal pain. Vomiting may be present and mask thediagnosis, especially in small children. A urinalysis isnecessary, and if results are suggestive of infection, urinecan be sent for a culture. Acute pyelonephritis often isaccompanied by costovertebral angle tenderness; supra-pubic tenderness may be elicited in a child who haslocalized cystitis.

Reproductive Tract DiseasesDisorders of the reproductive system can cause abdomi-nal pain. Ovarian or testicular torsion and ectopic preg-

nancy are not rare events. Ovarian cysts and sexuallytransmitted infections can cause abdominal pain.

Testicular torsion manifests as a tender scrotum withan enlarged testis. Pain radiates into the abdomen. Nau-sea may accompany the pain and sometimes progressesto vomiting. Adolescent boys may be embarrassed todescribe testicular pain and instead report hip or ab-dominal pain. This condition reinforces the importanceof the physical examination when evaluating a child foracute abdominal pain. If in doubt, ultrasonography canevaluate blood flow to the testicle. Emergent surgery isnecessary to save the affected testicle.

Ovarian torsion is harder to differentiate from othercauses of acute abdominal pain due to the location ofthe ovaries. The pain is in the lower abdomen. Besidespain, nausea and vomiting can be present. As with manyother conditions, infants who have this problem simplymay be fussy, feed poorly, and vomit. The torsed organswells, resulting in a palpable mass. Ultrasonography isneeded for this diagnosis. As with testicular torsion,emergent surgery is necessary to preserve the organ.

Ovarian cysts are common in postmenarchal adoles-cents and usually cause acute pain only if there is hem-orrhage into the cyst or the cyst ruptures and releasesblood into the abdomen. Analgesics and time may beappropriate treatment. In cases of complicated cysts,surgery sometimes is required. A pregnancy test shouldbe performed.

An ectopic pregnancy must be considered in anypostmenarchal female presenting with lower abdominalpain. Because adolescents sometimes are confused oruntruthful about their sexual histories, every female ad-olescent presenting with acute abdominal pain shouldundergo a pregnancy test.

Pelvic inflammatory disease can produce acute ab-dominal pain with rebound tenderness that can be diffi-cult to distinguish from a surgical abdomen. The painusually is in the lower abdomen, but sexually transmittedinfections also can cause a perihepatitis that leads topain in the right upper quadrant. Fever may be present.If pelvic inflammatory disease is suspected, gynecologicevaluation by appropriate colleagues may be critical,along with assuring appropriate follow-up and protec-tion of the child where necessary.

PneumoniaBecause of visceral innervation, a lower lobe pneumoniacan present as abdominal pain. In the febrile child whohas abdominal pain, the lung fields must be auscultatedand a chest radiograph considered if the findings aresuspicious.




Streptococcal PharyngitisMany children who have streptococcal pharyngitis expe-rience abdominal pain due to the pharyngitis. The paincan mimic appendicitis. Some surgeons require a rapidstreptococcal screen as part of the evaluation for appen-dicitis. Why streptococcal pharyngitis causes abdominalpain is not known.

Diabetic KetoacidosisAcute abdominal pain can be the initial presentation ofdiabetes mellitus as a feature of diabetic ketoacidosis(DKA). The review of systems should be positive forpolyuria or the parent may relate increased urinary fre-quency, which should prompt a urinalysis that can lead tothe correct diagnosis. Weight loss and thirst (polydipsia)also are common complaints in those who have diabetes.The serum amylase value can be elevated, but true pan-creatitis is rare.

The pain resolves with appropriate treatment forDKA. Therefore, if the pain remains despite improve-ment in the ketoacidosis, the child should be evaluatedfor other causes of abdominal pain. On occasion, DKA isprecipitated by the stress of another condition that mayaccount for the abdominal pain (such as a urinary tractinfection.)

Sickle Cell CrisisThe vascular occlusion of sickle cell crisis can result in asurgical abdomen due to infarction as well as to gallstoneformation. In the child who has sickle cell disease, thediagnosis may be difficult. The venous occlusive diseasein affected patients is more likely to be accompanied bychest pain or limb pain due to the same sludging in bloodvessels that causes the abdominal pain. The pain in sicklecell crisis improves with oxygen and hydration.

Functional DiseaseAlthough functional abdominal pain more often ischronic, the initial presentation can be the complaint ofacute pain. Objective signs of pain are less likely to bepresent.

Functional pain usually is felt at the umbilicus, but itcan be epigastric, as in nonulcer dyspepsia. Either diar-rhea or constipation can be present. The child can havederangements in the autonomic nervous system, withflushing or pallor. The gait more often is normal com-pared with the stooped, guarded posture of a patientwho has a surgical abdomen.

Functional disease results from complex biopsycho-social features that are beyond the scope of this article.

MalingeringMalingering must be differentiated from organic andfunctional pain. The child does not necessarily con-sciously seek secondary gain from the abdominal pain.Malingering is a complex and potentially serious condi-tion that requires evaluation by a team of specialists,including an expert in child behavior.

Suggested ReadingBlakelock RT, Beasley SW. Infection and the gut. Semin Pediatr

Surg. 2003;12:265–274Bundy DG, Byerley JS, Liles EA, Perrin EM, Katznelson J, Rice HE.

Does this child have appendicitis? JAMA. 2007;298:438–451Cervero F, Laird JMA. Visceral pain. Lancet. 1999;353:2145–2148Erkan T, Cam H, Ozkan HC, et al. Clinical spectrum of acute

abdominal pain in Turkish pediatric patients: a prospectivestudy. Pediatr Int. 2004;46:325–329

Green R, Bulloch B, Kabani A, Hancock BJ, Tenenbein M. Earlyanalgesia for children with acute abdominal pain. Pediatrics.2005;116:978–983

Hayes R. Abdominal pain: general imaging strategies. Eur Radiol.2004;14:L123–L137

Justice FA, Auldist AW, Bines JE. Intussusception: trends in clinicalpresentation and management. J Gastroenterol Hepatol. 2006;21:842–846

Kharbanda AB, Taylor GA, Fishman SJ, Bachur RG. A clinicaldecision rule to identify children at low risk for appendicitis.Pediatrics. 2005;116:709–716

Kwok MY, Kim MK, Gorelick MH. Evidence-based approach tothe diagnosis of appendicitis in children. Pediatr Emerg Care.2004;20:690–698

Scholer SJ, Pituch K, Orr DP, Dittus RS. Clinical outcomes of childrenwith acute abdominal pain. Pediatrics. 1996;98:680–685

Williams H. Green for danger! Intestinal malrotation and volvulus.Arch Dis Child Ed Pract. 2007;92:ep87–ep91

Williams NMA, Johnstone JM, Everson NW. The diagnostic valueof symptoms and signs in childhood abdominal pain. J R CollSurg Edinb. 1998;43:390–392

Summary• Although acute abdominal pain usually is self-

limited, there are serious consequences tooverlooking conditions that require surgery.

• The pediatrician should use the examination todecide if the child is likely to have appendicitis orother surgically treated disease, and when suspicious,consult a surgeon early in the process.

• Vomiting bile is a sign that requires consultationwith a surgeon.

• If in doubt about the seriousness of the illness,detain the child and perform serial examinations.Ask another physician to provide an opinion becauseexperience is one of the most sensitive toolsavailable for evaluating acute abdominal pain.




PIR QuizQuiz also available online at http://pedsinreview.aappublications.org.

1. What is the correct approach to an 8-week-old infant who has bile-colored (bilious) vomiting?

A. Admit for observation.B. Obtain surgical consultation immediately.C. Perform an urgent upper gastrointestinal endoscopy.D. Provide intravenous fluids to maintain hydration.E. Provide reassurance as long as there is no blood.

2. Inflammation of the abdomen that involves the diaphragm may be referred to the:

A. Inguinal region.B. Lower back.C. Shoulder.D. Sternum.E. Testicle.

3. Which of the following conditions should be excluded in a child who has recurrent episodes of acutepancreatitis?

A. Cow milk protein allergy.B. Crohn disease.C. Cystic fibrosis.D. Helicobacter pylori infection.E. Malrotation of the intestine.

4. The diagnostic test of choice for peptic ulcer disease is:

A. Abdominal ultrasonography.B. Computed tomography scan of the abdomen.C. Upper gastrointestinal endoscopy.D. Upper gastrointestinal series (barium radiograph).E. Urea breath test for Helicobacter pylori.

5. A 15-year-old girl presents with a 36-hour history of worsening right lower quadrant pain. Her lastmenstrual period was 2 weeks ago. She has tenderness to palpation. Which of the following conditions isthe most likely cause of her pain?

A. Cholelithiasis.B. Ovarian torsion.C. Pyelonephritis.D. Right lower lobe pneumonia.E. Small bowel volvulus.





Albert Ross and Neal S. LeLeiko Acute Abdominal Pain

& ServicesUpdated Information

http://pedsinreview.aappublications.org/cgi/content/full/31/4/135including high-resolution figures, can be found at:

Permissions & Licensing

http://pedsinreview.aappublications.org/misc/Permissions.shtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://pedsinreview.aappublications.org/misc/reprints.shtml

Information about ordering reprints can be found online:


http://pedsinreview.aappublications.org/misc/Permissions.shtml

http://pedsinreview.aappublications.org/misc/reprints.shtml



Megan H. Bair-Merritt Intimate Partner Violence







Intimate Partner ViolenceMegan H. Bair-Merritt,

MD, MSCE*

Author Disclosure

Dr Bair-Merritt has

disclosed no financial

relationships relevant

to this article. This

commentary does not


of an unapproved/


commercial product/

device.


1. Describe the prevalence of intimate partner violence and childhood exposure tointimate partner violence.

2. Identify risk factors associated with intimate partner violence.3. Explain the relationship of child maltreatment in the setting of intimate partner

violence.4. Recognize the impact of intimate partner violence exposure on children’s social-

emotional and physical health and health-care use.5. Discuss strategies for screening and responding to intimate partner violence in the pe-

diatric setting.

Case StudyYou are seeing a healthy 4-month-old infant for a health supervision visit. As part of a routinesocial history, you inquire about intimate partner violence (IPV). The infant’s mother disclosesthat her partner frequently yells at her, pushes her, and makes her feel afraid. On additionalquestioning, she describes the infant as “fussy.” The baby’s physical examination findings areunremarkable, but you note that he missed his 2-month visit and is behind on his immuniza-tions. How do you proceed?

DefinitionFamily violence and domestic violence often are used synonymously and refer to violenceoccurring between any family member dyad, including parent-child, intimate partner-intimate partner, or sibling-sibling. IPV refers specifically to violence perpetrated betweenromantic partners and has been defined by the Family Violence Prevention Fund as “apattern of purposeful coercive behaviors that may include inflicted physical injury, psycho-logical abuse, sexual assault, progressive social isolation, stalking, deprivation, intimidationand threats. These behaviors are perpetrated by someone who is, was, or wishes to beinvolved in an intimate or dating relationship with an adult or adolescent victim and areaimed at establishing control of one partner over the other.” (1)

EpidemiologyOver the course of a lifetime, between one fourth and one third of women in the UnitedStates are abused by an intimate partner. Every year, 2 to 4 million women and 0.5 to1 million men report IPV victimization. IPV occurs in families of all races, ethnicities, andsocioeconomic classes. Certain sociodemographic factors, however, have been associatedwith increased risk of IPV, including young age (with the highest rates in women 18 to24 years old), lower socioeconomic status, mental health problems, and substance abuse.Separated or divorced women report higher rates of IPV than married women.

Male-perpetrated IPV initially was recognized as a significant public health problem inthe 1960s. Since that time, experts have achieved a greater understanding of the epidemi-ology and health consequences of male-perpetrated IPV. However, recent peer-reviewedpapers and population-based surveys have reported that women perpetrate IPV as often asor more frequently than do men (2). These studies often are limited by failure to identifyacts of self-defense and to recognize that men are much more likely than women to incite

*Assistant Professor of Pediatrics, Associate Director, Primary Care Fellowship Program, Johns Hopkins School of Medicine,Baltimore, Md.

Article psychosocial issues



fear. (3) Despite the prevalence of female-perpetratedIPV, women still remain significantly more likely thanmen to be injured or killed by their partners.

Fifteen million children in the United States are ex-posed to IPV each year. (4) Almost 50% of these childrenare exposed to severe IPV, such as one parent beating upanother parent or one parent using a knife or gun againstanother parent. (4) Rates of IPV are disproportionatelyhigh in families who have children younger than 5 yearsof age. Childhood exposure may include being present inthe room while the violence is occurring, overhearingthe violence from another room, or observing the after-math of a violent argument, including parental injuries ordestruction of property. It is unclear whether exposureto male- or female-perpetrated IPV affects child healthdifferentially, and most of the literature on the IPV-childhealth relationship focuses on the impact of male-perpetrated IPV. Therefore, the remainder of this reviewfocuses predominantly on male-initiated violence.

Clinical AspectsChild Abuse

In 1998, the American Academy of Pediatrics (AAP)recommended routine IPV screening during health su-pervision visits, stating that “identifying and interveningon behalf of battered women may be one of the mosteffective means of preventing child abuse.” (5) Thisstatement was based on the well-documented overlapbetween child maltreatment and IPV. Although the as-sociations between physical IPV and child maltreatmentare most prominent in the literature, risk for child mal-treatment also is elevated in families in which intimatepartners engage in psychological abuse.

In populations of families either reported to ChildProtective Services (CPS) for child maltreatment or indomestic violence shelters, the co-occurrence of childmaltreatment and IPV ranges from 30% to 60%. (6)(7)Rates of overlap between IPV and child maltreatment incommunity-based samples are notable, although lowerthan rates in high-risk samples. Poverty, parental de-pression, and substance abuse increase the risk of co-occurrence.

Children living in homes where IPV occurs also maybe injured inadvertently by being “caught in the cross-fire” of parental altercations. (8) Infants and toddlersmay be injured if they are being held in a parent’s armsduring a physical altercation. Older children, on theother hand, may be injured while trying to intervene tostop a fight.

Social-Emotional HealthSince the 1980s, increasing attention has been focusedon infants and children as the “silent victims” of IPV,and research related to the social-emotional health con-sequences of childhood IPV exposure has grown expo-nentially. Childhood IPV exposure likely leads to adversechild health through a number of pathways, includingtrauma, altered stress physiology, and disruption ofthe caregiver-child attachment relationship. Childrenexposed to IPV also frequently perceive the world ashostile and unsafe and learn via social modeling thataggression is an acceptable means through which toresolve conflict.

Strong evidence links childhood IPV exposure witha wide variety of adverse social-emotional health out-comes in childhood. (9) Specific problems vary accord-ing to age and developmental stage but include develop-mental delay, depression, anxiety, peer aggression, andposttraumatic stress disorder/hypervigilance. Infantsmay display symptoms of trauma, including excessivecrying and resisting comfort. IPV also may disturb infantroutines such as sleeping and feeding and may affectmaternal-infant attachment. Common symptoms in tod-dlers include extreme separation anxiety, excessive tan-trums, and aggression with peers.

School-age children living in homes in which IPVoccurs are more likely than their peers to exhibit aggres-sive and antisocial behaviors and are more likely to beanxious, fearful, and hypervigilant. IPV exposure inschool-age children also has been linked with poor peerrelations, perhaps due to poor self-esteem and sensitiza-tion to hostility. Adolescents in homes where IPV ispresent have higher rates of school failure, substanceabuse, and risky sexual behaviors. These adolescents aremore likely than their peers to enter into a violent datingrelationship.

Physical Health and Patterns ofHealth-care Use

Research about the impact of IPV on children’s physicalhealth and health-care use is in its infancy, although somepatterns are emerging. Compared with peers, childrenexposed to IPV are less likely to attend health supervisionvisits and to be fully immunized and are more likely tohave emergency department visits. Children exposed toIPV also may incur greater overall health-care costs thantheir peers. Less is known about the impact of IPV onchildren’s physical health, although recent studies haveassociated IPV exposure with a higher incidence of child-hood asthma and with infant failure to thrive and child-hood malnutrition. (10)

psychosocial issues intimate partner violence



Childhood IPV exposure also affects adult healthadversely. Felitti and colleagues (11) published a seriesof articles examining the association between “adversechildhood experiences,” including exposure to a bat-tered mother, and a host of adult health outcomes. Theauthors enrolled more than 20,000 men and women andfound a consistent, graded relationship between thenumber of adverse childhood experiences and poor adulthealth outcomes. When analyzed separately, exposure toa battered mother was consistently linked with adverseadult health sequelae ranging from depression to sub-stance use.

DiagnosisScreening for IPV

Screening caregivers for IPV may improve the pediatricclinician’s understanding of the child’s current illness,home environment, and ability to follow physician rec-ommendations. Screening also communicates that IPVis a common public health problem, reduces the isolationcommonly associated with IPV, and conveys that IPVexposure affects children. In addition, abused womenmay be more likely to seek health care for their childrenthan for themselves, making the pediatric setting animportant site for screening. Mothers generally supportscreening for IPV in the pediatric setting and feel thatthe pediatric office is a safe location to discuss IPV andits consequences. Most major medical organizations(AAP, American College of Obstetrician-Gynecologists,American Medical Association) recommend routine IPVscreening.

It is unclear, however, whether screening leads toshort- or long-term benefits for women. MacMillan andcolleagues (12) recently found that female patients ran-domized to IPV screening, compared with nonscreenedcontrol women, did not report decreased IPV or in-creased health or quality of life over time. In 2004, theUnited States Preventive Services Task Force publishedguidelines stating that there was not sufficient empiricevidence to recommend either for or against routine IPVscreening. (13) Criticisms of these guidelines include:1) Asking about IPV should be viewed as a routinebehavioral health inquiry rather than as a “screeningtest”; 2) Benefits of screening may extend beyond de-creasing violent episodes, and these benefits should beconsidered in evaluating the merits of screening; and3) IPV is a complex issue, and qualitative research orstudies outside of the traditional medical literatureshould be considered. Benefits and risks for childrenrelated to screening mothers for IPV remain uninvesti-gated.

Options to routine screening include no screening orperforming targeted screening. In the pediatric setting,targeted screening includes children in whom concernexists for child abuse and neglect. Pediatric clinicians alsoshould be concerned about IPV based on findings forboth the mother and child. They should be concernedif the patient’s mother appears depressed or overly anx-ious, has obvious physical injuries, repeatedly cancels ormisses appointments or has difficulty following throughwith medical advice, or is abusing alcohol or other drugs.In addition, concern is raised if the child has emotionalproblems, including depression, anxiety, or attention-deficit/hyperactivity disorder, and has frequent nonor-ganic symptoms, such as headaches or stomachaches. Inaddition, certain situational factors may precipitate IPV,including pregnancy, parental separation or divorce, orefforts to leave the current relationship. The risk of atargeted screening approach, however, is that manywomen affected by IPV will not be identified.

When screening patients’ mothers, the topic of IPVcan be introduced with a statement such as: “The safetyand well-being of mothers affects the safety and well-being of children, so I ask all mothers a number ofquestions about themselves and about their safety.”

The psychometric properties of numerous IPVscreening questions have been tested in health-care set-tings. The general consensus is that precise, behaviorallyanchored questions are more sensitive and specificthan are general questions such as “Do you feel safe athome?” or “Are you a victim of IPV?” A comprehensivecompendium of IPV screening questions, publishedby the Centers for Disease Control and Prevention, canbe found at http://www.cdc.gov/NCIPC/pub-res/images/IPVandSVscreening.pdf. Notably, most of thevalidated IPV screening instruments have been testedexclusively on women; two sets of screening questionshave been validated in a pediatric setting.

When a child 3 years of age or older is in the exami-nation room, abused women have expressed concernabout clinicians using behaviorally anchored questions.(14) When these older, verbal children are present,women worry that the child may be traumatized or laterinadvertently relay the conversation about IPV to theperpetrator. If the child and mother can be separated(such as during hearing and vision screening), clinicianscan screen the mother when she is alone. If not, theclinician should use more general questions such as“How do you and your partner work out arguments?”and “In general, how would you describe your relation-ship with your partner: a lot of tension, some tension, orno tension?” Attention to the response and to nonverbal




cues is important. If the response or the reaction to thesequestions is a concern, the clinician should attempt tointerview the woman alone to pursue more specific ques-tions. A woman may answer “no” many times before shediscloses.

ManagementDocumentation

In the pediatric setting, the child is the patient, and bothlegal guardians have access to the medical record. If theperpetrator is the child’s father, he has access to the chartand could become aware of the caregiver-clinician IPVconversation, potentially increasing the risk of violence.In contrast, it may be legally beneficial (eg, in restrainingorders or child custody cases) to document IPV in thechild’s medical record. Although the risks of documen-tation are more theoretical than evidence-based, it isimportant to discuss the risks and benefits of documen-tation with the mother and abide by her wishes. A secondoption is to include in the chart a code known to otherclinicians in the practice that would not be readily under-standable to others, such as �MIPV (for positive mater-nal IPV).

Response to a Positive IPV ScreenAlthough there is a dearth of empiric data about effectiveinterventions in reducing IPV victimization or perpetra-tion or in improving health, the following responses to apositive IPV screen are recommended: 1) An empatheticstatement supporting the woman and emphasizing thatshe does not deserve to be treated that way; 2) Questionsabout escalation of violence, weapons, and comfort ingoing home to assess the woman and child’s currentsafety; 3) Detailed history and physical examinationlooking for possible child abuse; 4) Provision of socialwork assistance or national (1-800-799-SAFE) and localIPV resources; 5) Planning for safety (Table); and 6) Anassessment of whether the child would benefit frommental health intervention or from frequent visits withhis or her primary care clinician. If violence is escalatingor the family is in immediate danger, safe housing mustbe established, which IPV hotlines can help to facilitate.

Most IPV organizations offer shelter and counselingas well as legal and social advocacy for abused women.Therefore, pediatricians should familiarize themselveswith the services offered by local IPV organizations. Ifthe woman wishes, call the hotline from the pediatricoffice. The pediatric office also may be a safe place foran IPV advocate from a local organization to meet thewoman to provide assistance. Establishing a trustingcaregiver-clinician relationship is critical, and clinicians

should recognize that addressing IPV and helpingwomen make positive changes often takes time.

Clinicians should recognize that advising the womanto leave the relationship may not be the safest solutionfor her or her children. The risk of homicide for womenincreases by 75% around the time of leaving an abusiverelationship. Further, leaving an abuser may mean goinginto a shelter, removing children from their currentschools, and losing significant economic support.

Mandated ReportingWith any disclosure of IPV, a careful assessment of thechild’s well-being and safety is essential. Pediatric clini-cians are obligated to report to CPS any suspicion ofchild abuse or neglect or any concern that the child is inimminent danger. It can be helpful to have the motherfile the report with the clinician and to document, asappropriate, ways in which she has attempted to protectthe child. CPS should be made aware of the co-occurringIPV, and planning for safety is critical.

States vary regarding whether pediatric reporters aremandated reporters of childhood IPV exposure. Somestates’ laws explicitly say that pediatric clinicians aremandated reporters of childhood exposure to IPV. Inother states, it is not explicitly written. However, caselaw has interpreted IPV exposure as falling under emo-tional abuse, thereby making pediatric clinicians man-dated reporters. In some states, health-care clinicians alsomay be mandated reporters of IPV. Clinicians shouldknow their specific state’s reporting requirements prior

Table. Options for Safety PlanningEach recommendation should be discussed with the

woman to decide whether it is safe and feasible:● Advise women to collect the following items in

preparation for leaving: money, an extra set of houseand car keys, important documents (Social Securitycards, driver’s license, birth certificates, passports,marriage license, rent and utility receipts, carregistration, bank account numbers and check books,insurance policies and numbers, medical information)and a hidden bag with extra clothing, necessarymedications for mother and children, toiletries, andfood.

● Ask neighbors to call the police if they overhearviolence.

● Have a code with family or friends that communicatesthat the intimate partner violence is occurring or isabout to occur and establish what the family orfriends will do if the woman uses the code.

● Remove or disarm weapons in the home.




to screening. Websites such as www.endabuse.org orhttp://www.childwelfare.gov/ may help discern state-specific laws about mandated reporting.

If practicing in a state that has mandated reportinglaws for either IPV or childhood exposure to IPV, clini-cians should inform the mother of their responsibility toreport before screening. If practicing in a state that hasless specific reporting laws and there is no suspicion ofchild maltreatment, the pediatrician should considereach situation individually to determine what is best forthe child and mother.

Although proponents of mandated reporting forchildhood IPV exposure state that reporting is justifieddue to the negative health outcomes, others are con-cerned about such requirements. Apprehension aboutmandated reporting for IPV exposure stems from fearthat reporting is punitive to the victim, implying that shefailed to protect the child. Universal reporting also incor-rectly assumes that all children are affected similarly byIPV exposure and may increase demands on an alreadyoverburdened CPS. Mandated reporting decreasing

women’s likelihood of disclosing IPV and seeking helpalso is a concern.

References1. Identifying and Responding to Domestic Violence: Consensus Rec-ommendations for Child and Adolescent Health. San Francisco,Calif: The Family Violence Prevention Fund; 2004. Available at:http://www.endabuse.org/userfiles/file/HealthCare/pediatric.pdf2. Archer J. Sex differences in aggression between heterosexualpartners: a meta-analytic review. Psych Bull. 2000;126:651–6803. Anderson K. Perpetrator or victim? Relationships between in-timate partner violence and well-being. J Marr Fam. 2002;64:851–8534. McDonald R, Jouriles EN, Ramisetty-Mikler S, Caetano R,Green CE. Estimating the number of American children living inpartner-violent families. J Fam Psychol. 2006;20:137–1425. American Academy of Pediatrics. The role of the pediatrician inrecognizing and intervening on behalf of abused women. Pediat-rics. 1998;101:1091–10926. Appel A, Holden G. The co-occurrence of spouse and physi-cal child abuse: a review and appraisal. J Family Psych. 1998;12:578–5997. Jouriles EN, McDonald R, Smith Slep AM, Heyman RE,Garrido E. Child abuse in the context of domestic violence: preva-lence, explanations, and practice implications. Violence Vict. 2008;23:221–2358. Christian CW, Scribano P, Seidl T, Pinto-Martin JA. Pediatricinjury resulting from family violence. Pediatrics. 1997;99:e89. Kitzmann KM, Gaylord NK, Holt AR, Kenny ED. Child wit-nesses to domestic violence: a meta-analytic review. J Consult ClinPsych. 2003;71:339–35210. Bair-Merritt MH, Blackstone M, Feudtner C. Physical healthoutcomes of childhood exposure to intimate partner violence: asystematic review. Pediatrics. 2006;117:e278–e29011. Felitti VJ, Anda RF, Nordenberg D, et al. Relationship ofchildhood abuse and household dysfunction to many of the lead-ing causes of death in adults. The Adverse Childhood Experiences(ACE) study. Am J Prev Med. 1998;14:245–25812. MacMillan HL, Wathen CN, Jamieson E, et al. Screening forintimate partner violence in health care settings: a randomized trial.JAMA. 2009;302:493–50113. United States Preventive Services Task Force. Screening forfamily and intimate partner violence. Ann Intern Med. 2004;140:382–38614. Zink T, Jacobson J. Screening for intimate partner violencewhen children are present. J Interpers Viol. 2003;18:872–890

Summary• Based on strong research evidence, IPV and

childhood exposure to IPV are prevalent publichealth problems. (4)(5)(13)

• Based on strong research evidence, IPV and childmaltreatment commonly co-occur, particularly infamilies that have additional psychosocial riskfactors such as poverty, parental depression, andsubstance abuse. (6)(7)

• Based on strong research evidence, IPV negativelyaffects children’s social-emotional health, increasingthe risk for internalizing and externalizing disordersand aggression with peers. (9)

• Based on some research evidence, IPV may negativelyaffect children’s physical health and health-careuse. (10)

• Based on consensus and some research evidence,screening for IPV in the pediatric health-care settingmay help identify abused women and connect themto resources. (5)




PIR QuizQuiz also available online at http://pedsinreview.aappublications.org

6. Which of the following is true regarding children exposed to intimate partner violence?

A. They are less likely to be involved in violent relationships later in life.B. They are less likely to use the emergency department for care.C. They are more likely to be shy and withdrawn in their peer relationships.D. They are more likely to exhibit extreme separation anxiety in toddlerhood.E. They rarely are injured themselves during episodes of intimate partner violence.

7. You are considering initiating routine screening for intimate partner violence for parents in your pediatricclinic. Which of the following is the most appropriate action in this screening?

A. All mothers should be screened routinely by their child’s pediatrician because this has been shown todecrease violent episodes at home.

B. Children older than age 3 years should be asked questions routinely about the presence of violence athome.

C. Mothers of children who have frequent nonspecific symptoms, such as headache, should be queried withrespect to intimate partner violence.

D. Targeting mothers who are depressed and anxious or those who have children who have emotionalproblems likely will identify most cases of intimate partner violence.

E. Women should be asked general questions such as “Are you concerned about safety?” rather than morespecific questions related to intimate partner violence.

8. You have just identified a mother in your practice who is the victim of sporadic intimate partner violence.She states that the child’s father never has hit or spanked the child. The child’s physical examinationreveals no evidence of abuse. After additional questioning of the mother and examination of the child, youbelieve that she is the only direct victim of the violence. Which of the following is true regarding yourresponsibilities to this child and his mother?

A. The mother should be strongly encouraged to leave the father as soon as possible.B. You are legally obligated to report the abuse of the child only if physical evidence of abuse is present.C. You should order mandatory psychological counseling for the child.D. Your legal obligation to report the violence against the mother and the child’s exposure to it depends

on your state laws.E. Your staff should set up safe housing for both of them immediately.





Megan H. Bair-Merritt Intimate Partner Violence












Michael S. Leonard Patient Safety and Quality Improvement: Medical Errors and Adverse Events







Patient Safety and Quality Improvement:Medical Errors and Adverse EventsMichael S. Leonard, MD,

MS*

Author Disclosure

Dr Leonard has




commentary does not


of an unapproved/


commercial product/

device.


1. Define terms commonly used in patient safety discourse.2. Describe the scope of medical errors and adverse events, focusing on medication-

related issues.3. Identify barriers to improving patient safety.4. Address disclosure of medical errors and adverse events.5. Review principles and practices that can reduce the risk of harm to patients.

This is the first in a series of articles to review the topics of patient safety and quality im-provement in pediatrics.

IntroductionPatient safety is a subject that traverses all medical specialties and affects every health-careprofessional. The attention to medical errors and adverse events as well as the resultantliterature has grown exponentially over the past decade. A number of practicing physicians,however, remain unaware of the extent of the problem, the impact on patients, and theburden on the health-care system. Many also are unfamiliar with strategies to reduce therisk of harm.

TerminologyIt is important to note that definitions used in patient safety can vary across studies,between organizations, and over time. A medical error, as defined by the Institute ofMedicine (IOM), is “the failure to complete a planned action as intended or the use of awrong plan to achieve an aim.” (1)(2) It is a mistake in action or judgment. A medical errormust be distinguished from an adverse event, which is “an injury caused by medicalmanagement rather than by the underlying disease or condition of the patient.” An adverseevent results in harm to the patient. Not all medical errors lead to adverse events. In fact,most do not.

A medication error is the most common type of medical error (3)(4) and can occur atany point in the medication management process. (5) This chain of events includesordering, transcribing, preparing, delivering, and administering a drug. Medication errorscan be classified as errors of commission or omission. (6) An error of commission occurswhen an incorrect action is taken, such as prescribing ceftriaxone for a neonate who hashyperbilirubinemia. An error of omission results when a correct action is not taken, such asnot premedicating a patient who has a history of red man syndrome with an antihistaminebefore administering vancomycin.

An adverse drug event (ADE) is an injury due to a medication. ADEs are the mostcommon type of adverse events. Fortunately, less than 1% of medication errors result in anADE. (5) A preventable ADE is an ADE that, based on the medical information known atthe time, could have been avoided, such as a patient who has a known allergy to macrolidesbeing prescribed azithromycin and developing urticaria. This occurrence is in contrast to anonpreventable ADE, which could not have been foreseen based on the medical informa-tion known at the time. For example, a patient who has no known allergies is givenazithromycin and develops urticaria. An adverse drug reaction is synonymous with anonpreventable ADE. It is an event defined by the World Health Organization as “noxious

*Chief Quality & Safety Officer for Children’s Services, Associate Professor of Pediatrics, University of Rochester Medical Center,Rochester, NY.

Article patient safety



and unintended, and which occurs at doses used in manfor prophylaxis, diagnosis or therapy.” (7)(8)

A potential ADE is a medication error that places apatient at significant risk of injury but does not actuallyresult in harm. (7) Potential ADEs often are referred to as“near misses.” There are two subtypes of potentialADEs: intercepted and nonintercepted. An interceptedADE is an error that is identified and corrected before itreaches the patient. For example, a patient is prescribedan overdose of ranitidine that is detected by a pharmacist,who does not dispense the drug, but instead refers theprescription back to the prescriber for correction. A non-intercepted ADE is an error that reaches the patient but,by pure happenstance, does not cause harm to the pa-tient. For example, a patient is prescribed and given anoverdose of ranitidine without suffering any ill effects.The Figure shows the relationship among medicationerrors, ADEs, and harm.

A sentinel event is “an unexpected occurrence involv-ing death or serious physical or psychological injury, orthe risk thereof.” (9) An ADE or a potential ADE mayqualify as a sentinel event. Not all sentinel events are theresult of a medical error. A never event, as described bythe National Quality Forum, is one of 28 events resultingfrom an error in medical care that is serious, unambigu-ous, and usually preventable. (10) Never events, whichalso are referred to as serious adverse events, seriousreportable events, and serious reportable adverse events,never should occur within a health-care institution andgenerally signal a system safety failure.

General EpidemiologyPatient safety has become a national health-care focusover the past 10 years, but medical errors and adverseevents have been addressed for a far longer time. Theinvestigators for the Harvard Medical Practice Studyreviewed more than 30,000 records from patients dis-charged in 1984 from 51 hospitals across the state ofNew York. (11) Adverse events occurred in 3.7% of thesehospitalizations, most of which were preventable. (12) Ifgeneralized to all hospitals in the United States, thisincidence translates to more than 1 million people expe-riencing an adverse event and approximately 180,000patients dying from an adverse event every year. (7)Medication-related incidents were the most commontype of adverse event, at a rate of 0.7 ADEs per 100admissions. (5)(12)

A study of nearly 15,000 discharges across 28 hospi-tals in Utah and Colorado in 1992 identified 459 adverseevents. (13) Fifty-eight percent of these adverse eventswere found to be preventable. In a study of more than

4,000 admissions at two large Boston hospitals over a6-month period in 1993, investigators identified 247ADEs. (14) This translates to 6.5 ADEs per 100 admis-sions, a rate greater than nine times that found in theHarvard study. Twenty-eight percent of the ADEs wereconsidered preventable. These investigators also identi-fied 194 potential ADEs, of which 43% were interceptedbefore they reached the patient.

The IOM released its landmark report To Err isHuman at the end of 1999, heralding a new age for thefield of patient safety. (1) Extrapolating the results of theUtah-Colorado study and the Harvard study over the33.6 million admissions to United States hospitals in1997, the report estimated that 44,000 to 98,000 pa-tients die from medical errors annually. The IOM issuedrecommendations in four strategic areas to improve pa-tient safety: leadership and knowledge, identifying andlearning from errors, setting performance standards andexpectations for safety, and implementing safety systemsin health-care organizations.

The financial impact of medical errors and adverseevents is enormous. Based on the IOM report, the esti-mated cost of medical errors in the United States is$37.6 billion annually. Preventable adverse events com-prise an estimated $17 to $29 billion, more than 50% ofwhich represent direct costs to the health-care system.(1)(2) An ADE extends the length of hospital stay by1.74 to 2.2 days and increases costs by $2,000 to $3,200.A preventable ADE extends the hospital stay by 4.6 daysand increases costs by $5,800. (14)(15)(16) These fig-ures represent dollar values from the 1990s; currentestimates are likely even higher.

The financial burden of adverse events is shifting

Figure. The relationship between medication errors, adversedrug events, and harm. ADE�adverse drug event, ADR�adverse drug reaction.

patient safety adverse events



toward health-care institutions. Medicare no longer re-imburses the cost of treatment for some events occurringduring a hospital admission, and many state Medicaidagencies are considering similar financial disincentives.Private third-party payers may begin pursuing similarpractices.

Pediatric EpidemiologyThe prevalence of medical errors and adverse events inchildren is becoming clearer. Investigators analyzed asubset of data from the Utah-Colorado study represent-ing more than 3,700 pediatric hospitalizations, includingbirth admissions. (17) They found that adverse eventsoccurred in 1% of these hospitalizations, 60% of whichwere deemed preventable. Extrapolated to the entireUnited States population, approximately 70,000 hospi-talized children experience an adverse event each year, ofwhich 42,000 could be avoided.

Children have been recognized as a high-risk popula-tion for medication errors and ADEs. The risk for anADE is estimated to be three times higher in hospitalizedchildren than in adults. (4)(18) In a prospective study ofmedication orders in a children’s hospital, 23.7% con-tained errors. (19) In a study of more than 10,000medication orders representing 1,120 patients acrosstwo children’s hospitals, the investigators found 5.7% ofthe orders contained errors, with 1.1% representing po-tential ADEs. (18) They identified 2.3 ADEs per 100admissions, 19% of which were preventable. In a morerecent study across 12 children’s hospitals, investigatorsreported a rate of 11.1 ADEs per 100 patients, 15.7ADEs per 1,000 patient-days, and 1.23 ADEs per 1,000medication doses. (20) Rising ADE rates actually mayreflect an improved ability to detect such events over timerather than a true increase in incidence.

For many reasons, children are more vulnerable tomedication errors and ADEs. Unlike adults, for whomdosing tends to be a single or limited number of options,dosing for children usually is tailored to the patient basedon his or her weight. Proper dosing requires that theprescriber have an accurate weight for the child as well asthe proficiency to perform weight-based calculations.Many settings in which children are treated are predom-inantly adult-oriented, with pediatric patients compris-ing only a small fraction of admissions. In such venues,staff may not always be trained in safe pediatric medica-tion practices, and if they are, their skills may wane due tothe infrequency of providing care to these young pa-tients. Most drugs are packaged commercially for adultuse. Such preparations require compounding for pediat-ric usage, a task that requires a specific skill set. The need

for this additional step in the medication delivery processintroduces risk of error and, therefore, risk of harm. (4)

Children who experience extended lengths of stay,complex medication regimens, and higher severity ofillness are at increased risk of ADEs. (21) Neonates andyoung infants are at even greater risk due to their im-mature hepatic, renal, and immune systems. (4)(22) Insettings such as the neonatal intensive care unit, wherelengths of stays often are measured in months, patientscan have significant changes in weight over the course ofa hospitalization. This change requires vigilance to assurethat medication dosing regimens remain within safe andtherapeutic ranges.

Medication errors and ADEs occur in the communityas well, although this venue has been less researched.(6)(23) In a prospective study across six outpatient of-fices, 14% of pediatric patients experienced an ADE. (24)Twenty-three percent of these events were judged to bepreventable. Again, complex medication regimens in-creased the likelihood of preventable ADEs. Of note,parents whose English proficiency was limited were lesslikely to report an ADE, which may represent a sub-population of children at even greater risk of harm.

Not all outpatient errors and preventable adverseevents are iatrogenic; they also can be caused by well-meaning parents and nonmedical caregivers. Manyover-the-counter children’s medications are available in avariety of preparations and concentrations, which cancontribute to confusion and subsequent dosing errors.Outpatient medication errors and ADEs are not limitedto oral drugs. For example, in 2007, a 17-year-old athletedied of salicylate toxicity from excessive use of over-the-counter topical sports ointments. (25) A tragic eventsuch as this underscores the need for pediatricians to takeevery opportunity to stress to patients and families theimportance of using medications only as instructed andthe potential hazards of using seemingly benign drugsinappropriately.

Detection of Medical Errors andAdverse EventsDifficulty in identifying medical errors and adverse eventscreates a significant barrier to assessing risk reductionstrategies. Although the data presented in this article arestaggering, they are likely to represent considerable un-derestimates. In the absence of an accurate, reliablemethodology to measure errors and events, the ability toassess the impact of patient safety initiatives remainschallenging.

Historically, identification of medical errors and ad-verse events relied on incident reports, a methodology




that has many pitfalls. The reporting of incidents pre-sumes that an individual recognizes a medical error or anadverse event when it occurs. If an error or adverse eventis identified, staff may assume, mistakenly, that someoneelse will report it. Incident report forms generally requirea large amount of information, which makes them bothtime-consuming and cumbersome to complete. Person-nel who do take the time to report events often receive nofeedback. They may not see their efforts resulting innoticeable improvement, which can act as a disincentivefor reporting subsequent events. (26)

Staff may have valid apprehensions about reportingerrors and adverse events. (27) There is a natural hesi-tancy to point out one’s own mistakes for fear of beinglabeled incompetent and a reluctance to point out oth-ers’ mistakes for fear of being labeled a whistleblower.Although institutions across the country are transition-ing to a culture of blameless reporting, staff still mayworry about discipline by their institutions or licensingorganizations. Fear of litigation also remains a majordeterrent.

Chart reviews in search of medical errors and adverseevents are time- and labor-intensive. The cost often isprohibitive, and generally only a sampling of records canbe evaluated. Trigger methodology can streamline theprocess and help capture occurrences that otherwisemight go undetected. A trigger is an action or indicatorthat might signal the presence of an adverse event. Forexample, flumazenil usually is prescribed in response to abenzodiazepine-related ADE. An order for flumazenil,therefore, can serve as a trigger. Groups of triggers havebeen combined into trigger tools to increase the ability toidentify actual and potential ADEs. (28) The resultantfocused chart reviews save time and resources by improv-ing efficiency. Trigger methodology has been applied inchildren’s hospitals and has been shown to increase de-tection of ADEs. (20)

DisclosureThe general public is becoming increasingly aware ofissues of patient safety. A study reviewing newspaperarticles over a 10-year period found increasing coverageof pediatric medication safety. (29) The United Stateshad the highest absolute number of articles of the fivecountries included in the study. Nearly two thirds of thearticles covered specific adverse events. The good news isthat three quarters of the articles were written in a tonecharacterized as neutral, a positive step as health caretransitions from a culture of individual blame to one ofsystem-based improvement.

Patients and families often know when they have

experienced an adverse event. They may observe worsen-ing of clinical status, the need to perform additionaltesting, or adjustments made to treatments. They mayperceive a change in the behavior of the physician orother staff. They may overhear conversations betweenstaff members. Patients and families want to be told whenan error has occurred. They want to know what hap-pened, why it happened, how it will affect their health,and what is being done to prevent such an error fromrecurring. (30)

Informing patients and families when medical errorscause harm is the right thing to do, but despite physi-cians’ best intentions, full disclosure of such events isuncommon. Practitioners often fear that such disclosuremay result in litigation, loss of trust by the patient andfamily, or tarnishing of their professional reputation.Health-care professionals must overcome the instinct toignore, hide, or worst of all, deny when an adverse eventoccurs. Disclosure relieves the anxiety of not knowingand reaffirms an open, honest physician-patient-familyrelationship. Such transparency has been demonstratedto decrease litigation as well as the average settlementamount for claims that are filed. (30)(31)(32)

Patients and families also want an apology. An apol-ogy historically has been viewed as an admission of guilt,and for this reason, practitioners often ignore their in-stincts to say “I’m sorry.” However, this omission isperceived as cold, heartless, and impersonal by patientsand families. They correctly feel angry and distanced,which is toxic to the physician-patient-family relationshipand actually may increase the likelihood of litigation. Theimportance of an apology and the reluctance by physi-cians to provide one due to fear of litigation has resultedin legislative changes. As of January 2008, 35 states haveenacted statutes that prevent an apology from being usedagainst a physician in a law suit. (30) Other states havesimilar legislation in progress. The impact, effectiveness,and protection afforded by these recent statutes have yetto be established. Policies vary across organizations, andlegislation varies among states. Health-care professionalsshould consult risk management experts or legal counselfor guidance with disclosure of adverse events and saying“I’m sorry.”

Although patients are the primary and most visiblevictims of medical errors and adverse events, it is impor-tant to remember the “second victims”: the clinicians.(33) The emotional impact of medical errors on health-care personnel, especially those errors that cause harm topatients, should not be underestimated. In many cases,more than one clinician is involved and may feel respon-sible. Resultant feelings of guilt, anxiety, or incompe-




tence can be devastating, leading to inappropriate behav-iors such as lashing out at patients, families, andcolleagues and to unhealthy behaviors such as substanceabuse. Health-care professionals must be supportive andnonjudgmental of their colleagues when medical errorsoccur. Debriefing sessions can provide a therapeuticvenue to discuss not only the medical aspects of theevent, but also the emotional issues as they relate to theclinicians involved.

PreventionPatient safety initiatives can be framed within a publichealth model of prevention. (27) The disease we aretrying to eradicate is medical errors, and specifically forthis discussion, medication errors. It is logical for physi-cians’ efforts to focus on prescribing, the step in themedication management process at which errors mostcommonly occur. (7)(18)(34)

The goal of primary prevention is to reduce the inci-dence or risk of disease. For example, the purpose of theinfluenza vaccine is to prevent or decrease the likelihoodthat a patient will become ill from the virus. For medica-tion errors, there is no vaccine, but our armamentariumfor preventing errors includes a host of strategies. Thefollowing is a list of simple practices that can be employedby prescribers in any setting to help prevent errors andadverse drug events (Table):

1. Obtain a list of your patient’s current medications.This is part of a process termed medication reconcilia-tion, one of the National Patient Safety Goals. It helps toprevent inadvertent drug interactions and ensure thatpatients receive the intended therapy. (35) Include allprescription drugs, over-the-counter medications, vita-mins, and supplements.2. Obtain an accurate list of allergies and adverse reac-tions. Include all drugs, foods, and relevant substances.Document the specific symptom(s) the patient experi-ences when he or she is exposed.3. Print legibly. Few settings are completely paperless,and handwriting that is difficult to decipher can bemisinterpreted, propagating misinformation. This out-come is true not only for orders and prescriptions, but forother documents as well, such as progress notes, medi-cation lists, and even the self-stick message notes affixedto charts.4. Avoid the use of unsafe abbreviations. Many histori-cally used abbreviations have been identified as danger-ous, increasing the likelihood of medication errors andADEs. As part of the National Patient Safety Goals, the

Joint Commission has compiled a list of abbreviationsthat should not be used, (35), such as:

• QD, which easily can be mistaken for QID, result-ing in fourfold overdosing. Write “daily” instead.

• U, which can be misinterpreted as a zero (0), result-ing in 10-fold overdosing. Write out the word“units” instead.

A good maxim is: When in doubt, write it out.

5. Obtain an accurate patient weight. Most pediatricdosing is based on the child’s weight, at least until adultsize is reached. Weigh the patient in kilograms andinclude weight-based dosing information on the order orprescription. (4)(36) This practice allows a pharmacist ornurse to double-check your calculations.6. Mind your decimals. Dosing errors are the most com-mon type of prescribing errors; children, in particular, areat high risk for 10-fold errors due to misplaced decimals.(18)(34)(37)(38)(39)(40)(41) The adage “Always lead,never follow” can help avoid this type of error. Alwaysprecede a decimal with a zero (eg, 0.3 mg, not .3 mg).Never follow a decimal with a zero (eg, 7 mg, not7.0 mg). (35)7. Include the indication for therapy on orders andprescriptions. (36) This practice helps to prevent ortho-graphic, or look-alike, medication errors. Zyprexa� canbe mistaken easily for Zyrtec� on a poorly handwrittenprescription and has been. (42) Specifying the indication“allergic rhinitis” on a prescription for Zyrtec� drasticallyreduces the likelihood that Zyprexa� will be dispensed.

The indication for therapy also should be includedwith verbal orders to prevent phonetic, or sound-alike,medication errors. Diamox� and Trimox� may not bemistaken for each other on paper or on a computer

Table. Practices to Help PreventMedication Errors and AdverseDrug Events● Obtain a thorough list of the patient’s current

medications.● Obtain an accurate list of the patient’s allergies and

adverse reactions.● Print legibly.● Avoid the use of unsafe abbreviations.● Obtain an accurate patient weight.● Mind your decimals.● Include indication for therapy on orders and

prescriptions.● Educate patients and families.




screen, but a nurse working on a noisy unit receiving averbal order from a physician on a cell phone with only“one bar” easily could mistake them. Including the indi-cation for therapy decreases the risk that these two drugswill be interchanged.8. Educate patients and families. Review the dosing regi-men, indications for therapy, and potential adverse effectsfor every new prescription. Provide written instructions andinformation whenever possible. Encourage patients andfamilies to contact you with any questions or concerns.

Institution-based primary prevention strategies alsocan be implemented. Computerized physician (or pro-vider) order entry (CPOE) refers to a broad spectrum ofelectronic prescribing systems that have been shown todecrease medication errors and ADEs. (43)(44)(45)CPOE can ascertain that required information is in-cluded in an order or prescription using forced formatscreens and essentially can eliminate the issue of illegibil-ity. Weight-based calculators with pediatric guardrailscan ensure that dosing remains within safe parameters.Clinical decision support functionalities embedded inmany CPOE systems include checks for allergies anddrug interactions, integration of available laboratorydata, reminders to monitor serum drug concentrations,and promotion of standardized order sets.

However, CPOE is not a panacea. CPOE may helpsolve some problems, but it can introduce new sources oferror. For example, forced format screens can facilitateincorrect orders in some instances. (46) Implementationof CPOE requires an enormous investment in humanand financial resources. The cost of the technology re-mains prohibitive for many institutions. In a 2002 sur-vey, less than 10% of United States hospitals had CPOEsystems completely available. (47) Paper-based orderforms with forced formats can provide a temporizing,fiscally friendly solution for ensuring that all necessaryinformation is included in an order, (48) but they cannotresolve poor handwriting or provide the automated safe-guards of an electronic system.

Clinical pharmacist participation on inpatient roundshas been shown to be another effective means of primaryprevention in various settings. (22)(49)(50)(51) Recom-mendations, including choice of drug, dosing regimen,and pharmacokinetic monitoring, help reduce errors andpreventable ADEs. Unfortunately, the cost in humanand financial resources precludes this solution for manyorganizations.

Secondary prevention aims to detect and interveneearly in the course of disease. Antiviral drugs such asoseltamivir are prescribed upon early identification of

influenza to minimize symptom severity. Secondary pre-vention for medication safety is the rapid detection andremoval of errors introduced into the medication man-agement system, intercepting them before they have theopportunity to become preventable ADEs. Review ofmedication orders by pediatric pharmacists has beenshown to be an effective means of preventing potentialADEs from reaching the bedside. (37)(52) Pharmacistsand nurses should be empowered and encouraged todouble-check orders (53) and contact the prescriberwhen an error is identified. Physicians should be gratefulrather than annoyed by these calls; each representsaverted harm to a patient.

The goal of tertiary prevention is to minimize se-quelae associated with disease. Physical therapy assistspatients whose baseline status has been compromised bya severe bout of influenza. For patient safety, havingreversal agents readily available such as naloxone fornarcotics and having interventional systems in place suchas rapid response teams (54) may help reduce the short-and long-term morbidity associated with adverse eventswhen they do occur.

Primordial prevention strives to reduce environmen-tal, societal, cultural, and behavioral factors that increaserisk of disease. (55) For influenza, this is accomplishedthrough public health campaigns to promote immuniza-tion and measures to reduce disease spread. For patientsafety, primordial prevention entails increasing awarenessof the problem, an objective of this article; buildingsystems that minimize the risk of error and swiftly inter-cept them before they have an opportunity to causeharm; and establishing a culture in which safe practicesare the standard, error and event reporting is nonpuni-tive, and continual improvement is the paradigm.

References1. Kohn LT, Corrigan JM, Donaldson MS, eds. To Err is Human:Building a Safer Health System. Washington, DC: National Acad-emy Press; 1999

Summary• Strides are being made to keep patients safe despite

the enormity of the challenge and the inherentdifficulties in assessing improvement.

• Pediatricians need to be cognizant of general patientsafety principles as well as those specific to children.

• We must adopt safe practices and remain vigilant inour efforts to reduce the risk of harm to ourpatients. To err may be human, but to improve isdivine.




2. Medical Errors: The Scope of the Problem. Rockville, Md: Agencyfor Healthcare Research and Quality; 20003. Bates DW, Gawande AA. Error in medicine: what have welearned? Ann Intern Med. 2000;132:763–7674. Preventing pediatric medication errors. Sentinel Event Alert.Oakbrook Terrace, Ill: The Joint Commission; 20085. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L.Relationship between medication errors and adverse drug events.J Gen Intern Med. 1995;10:199–2056. Thomsen LA, Winterstein AG, Sondergaard B, Haugbolle LS,Melander A. Systematic review of the incidence and characteristicsof preventable adverse drug events in ambulatory care. Ann Phar-macother. 2007;41:1411–14267. Bates DW, Cullen DJ, Laird N, et al. Incidence of adverse drugevents and potential adverse drug events. Implications for preven-tion. ADE Prevention Study Group. JAMA. 1995;274:29–348. International drug monitoring. The role of the hospital. WHOTech Rep Ser. 1969;425:5–249. Sentinel Events. Oakbrook Terrace, Ill: The Joint Commission;2007. Accessed January 2010 at: http://www.jointcommission.org/SentinelEvents/10. Fact Sheet: CMS Improves Patient Safety for Medicare andMedicaid by Addressing Never Events. Washington, DC: Centers forMedicare & Medicaid Services; 200811. Brennan TA, Leape LL, Laird NM, Hebert L, et al. Incidenceof adverse events and negligence in hospitalized patients. Results ofthe Harvard Medical Practice Study I. N Engl J Medicine. 1991;324:370–37612. Leape LL, Brennan TA, Laird N, et al. The nature of adverseevents in hospitalized patients. Results of the Harvard MedicalPractice Study II. N Engl J Med. 1991;324:377–38413. Thomas E, Studdert D, Newhouse J, et al. Costs of medicalinjuries in Utah and Colorado. Inquiry. 1999;36:255–26414. Bates DW, Spell N, Cullen DJ, et al. The costs of adverse drugevents in hospitalized patients. Adverse Drug Events PreventionStudy Group. JAMA. 1997;277:307–31115. Evans R, Classen D, Stevens L, et al. Using a hospital informa-tion system to assess the effects of adverse drug events. Proc AnnuSymp Comput Appl Med Care. 1993:161–16516. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP.Adverse drug events in hospitalized patients. Excess length of stay,extra costs, and attributable mortality. JAMA. 1997;277:301–30617. Woods D, Thomas E, Holl J, Altman S, Brennan T. Adverseevents and preventable adverse events in children. Pediatrics. 2005;115:155–16018. Kaushal R, Bates DW, Landrigan C, et al. Medication errorsand adverse drug events in pediatric inpatients. JAMA. 2001;285:2114–212019. Marino BL, Reinhardt K, Eichelberger WJ, Steingard R. Prev-alence of errors in a pediatric hospital medication system: implica-tions for error proofing. Outcomes Manag Nurs Pract. 2000;4:129–13520. Takata GS, Mason W, Taketomo C, Logsdon T, Sharek PJ.Development, testing, and findings of a pediatric-focused triggertool to identify medication-related harm in US children’s hospitals.Pediatrics. 2008;121:e927–e93521. Holdsworth MT, Fichtl RE, Behta M, et al. Incidence andimpact of adverse drug events in pediatric inpatients. Arch PediatrAdolesc Med. 2003;157:60–6522. Chedoe I, Molendijk HA, Dittrich ST, et al. Incidence andnature of medication errors in neonatal intensive care with strategies

to improve safety: a review of the current literature. Drug Safety.2007;30:503–51323. Miller MR, Pronovost PJ, Burstin HR. Pediatric patient safetyin the ambulatory setting. Ambul Pediatr. 2004;4:47–5424. Zandieh SO, Goldmann DA, Keohane CA, Yoon C, BatesDW, Kaushal R. Risk factors in preventable adverse drug events inpediatric outpatients. J Pediatr. 2008;152:225–23125. Klatell JM. Track star’s death blamed on pain cream. CBSNews. New York, NY; 200726. Taylor JA, Brownstein D, Christakis DA, et al. Use of incidentreports by physicians and nurses to document medical errors inpediatric patients. Pediatrics. 2004;114:729–73527. Leonard MS. Application of a systems approach to medicationsafety. In: Caty MG, ed. Complications in Pediatric Surgery. NewYork, NY: Informa Healthcare; 2008:91–9828. Rozich JD, Haraden CR, Resar RK. Adverse drug event triggertool: a practical methodology for measuring medication relatedharm. Qual Saf Health Care. 2003;12:194–20029. Stebbing C, Kaushal R, Bates DW. Pediatric medication safetyand the media: what does the public see? Pediatrics. 2006;117:1907–191430. Pelt JL, Faldmo LP. Physician error and disclosure. Clin ObstetGynecol. 2008;51:700–70831. Laing LC. A legal perspective on errors in medicine. In: CatyMG, ed. Complications in Pediatric Surgery. New York, NY: In-forma Healthcare; 2008:25–6932. Matlow A, Stevens P, Harrison C, Laxer R. Achieving closurethrough disclosure: experience in a pediatric institution. J Pediatr.2004;144:559–56033. Wu AW. Medical error: the second victim. The doctor whomakes the mistake needs help too. BMJ. 2000;320:726–72734. Fortescue EB, Kaushal R, Landrigan CP, et al. Prioritizingstrategies for preventing medication errors and adverse drug eventsin pediatric inpatients. Pediatrics. 2003;111:722–72935. Accreditation Program: Ambulatory Health Care. OakbrookTerrace, Ill: The Joint Commission; 200836. Leonard MS, Cimino M, Shaha S, McDougal S, Pilliod J,Brodsky L. Risk reduction for adverse drug events through sequen-tial implementation of patient safety initiatives in a children’s hos-pital. Pediatrics. 2006;118:e1124–e112937. Folli HL, Poole RL, Benitz WE, Russo JC. Medication errorprevention by clinical pharmacists in two children’s hospitals. Pedi-atrics. 1987;79:718–72238. Ghaleb MA, Barber N, Franklin BD, Yeung VW, Khaki ZF,Wong IC. Systematic review of medication errors in pediatricpatients. Ann Pharmacother. 2006;40:1766–177639. Koren G, Barzilay Z, Greenwald M. Tenfold errors in admin-istration of drug doses: a neglected iatrogenic disease in pediatrics.Pediatrics. 1986;77:848–84940. Lesar TS. Errors in the use of medication dosage equations.Arch Pediatr Adolesc Med. 1998;152:340–34141. Lesar TS. Tenfold medication dose prescribing errors. AnnPharmacother. 2002;36:1833–183942. Bond GR, Thompson JD. Olanzapine pediatric overdose. AnnEmerg Med. 1999;34:292–29343. Bates DW, Teich JM, Lee J, et al. The impact of computerizedphysician order entry on medication error prevention. J Am MedInform Assoc. 1999;6:313–32144. Kuperman GJ, Gibson RF. Computer physician order entry:benefits, costs, and issues. Ann Intern Med. 2003;139:31–39




45. Kaushal R, Shojania KG, Bates DW. Effects of computerizedphysician order entry and clinical decision support systems on medica-tion safety: a systematic review. Arch Intern Med. 2003;163:1409–141646. Koppel R, Metlay JP, Cohen A, et al. Role of computerizedphysician order entry systems in facilitating medication errors.JAMA. 2005;293:1197–120347. Ash JS, Gorman PN, Seshadri V, Hersh WR. Computerizedphysician order entry in U.S. hospitals: results of a 2002 survey.J Am Med Inform Assoc. 2004;11:95–9948. Shaha S, Brodsky L, Leonard MS, et al. Establishing a cultureof patient safety through a low-tech approach to reducing medica-tion errors. In: Advances in Patient Safety. Rockville, Md: Agencyfor Healthcare Research and Quality; 200549. Kaboli PJ, Hoth AB, McClimon BJ, Schnipper JL. Clinicalpharmacists and inpatient medical care: a systematic review. ArchIntern Med. 2006;166:955–96450. Leape LL, Cullen DJ, Clapp MD, et al. Pharmacist participa-

tion on physician rounds and adverse drug events in the intensivecare unit. JAMA. 1999;282:267–27051. Kaushal R, Bates DW, Abramson EL, Soukup JR, GoldmannDA. Unit-based clinical pharmacists’ prevention of serious medica-tion errors in pediatric inpatients. Am J Health Syst Pharm. 2008;65:1254–126052. Krupicka MI, Bratton SL, Sonnenthal K, Goldstein B. Impactof a pediatric clinical pharmacist in the pediatric intensive care unit.Crit Care Med. 2002;30:919–92153. Walsh KE, Kaushal R, Chessare JB. How to avoid paediatricmedication errors: a user’s guide to the literature. Arch Dis Child.2005;90:698–70254. Mistry KP, Turi J, Hueckel R, Mericle JM, Meliones JN.Pediatric rapid response teams in the academic medical center. ClinPediatr Emerg Med. 2006;7:241–24755. Last J, ed. A Dictionary of Epidemiology. 4th ed. New York NY:Oxford University Press, Inc; 2001

ClarificationA perceptive reader notes a discrepancy between PIR’s policy not to discuss “unapprovedcommercial products” and comments on the use of the unapproved natriuretic peptidenesiritide by Madriago and Silberbach in their article “Heart Failure in Infants and Children”(January 2010, pp. 4–12). In their PIR-online reply (http://pedsinreview.aappublications.org/cgi/eletters/31/1/4), the authors lament that the majority of drugs routinelyemployed in pediatrics are neither studied nor approved by the United States Food andDrug Administration (FDA) and state that nesiritide may, in fact, be more effective thansome of the drugs that are approved by the FDA for use in pediatric heart failure.





Michael S. Leonard Patient Safety and Quality Improvement: Medical Errors and Adverse Events











DOI: 10.1542/pir.31-4-158 2010;31;158 Pediatr. Rev.

Clarification







45. Kaushal R, Shojania KG, Bates DW. Effects of computerizedphysician order entry and clinical decision support systems on medica-tion safety: a systematic review. Arch Intern Med. 2003;163:1409–141646. Koppel R, Metlay JP, Cohen A, et al. Role of computerizedphysician order entry systems in facilitating medication errors.JAMA. 2005;293:1197–120347. Ash JS, Gorman PN, Seshadri V, Hersh WR. Computerizedphysician order entry in U.S. hospitals: results of a 2002 survey.J Am Med Inform Assoc. 2004;11:95–9948. Shaha S, Brodsky L, Leonard MS, et al. Establishing a cultureof patient safety through a low-tech approach to reducing medica-tion errors. In: Advances in Patient Safety. Rockville, Md: Agencyfor Healthcare Research and Quality; 200549. Kaboli PJ, Hoth AB, McClimon BJ, Schnipper JL. Clinicalpharmacists and inpatient medical care: a systematic review. ArchIntern Med. 2006;166:955–96450. Leape LL, Cullen DJ, Clapp MD, et al. Pharmacist participa-

tion on physician rounds and adverse drug events in the intensivecare unit. JAMA. 1999;282:267–27051. Kaushal R, Bates DW, Abramson EL, Soukup JR, GoldmannDA. Unit-based clinical pharmacists’ prevention of serious medica-tion errors in pediatric inpatients. Am J Health Syst Pharm. 2008;65:1254–126052. Krupicka MI, Bratton SL, Sonnenthal K, Goldstein B. Impactof a pediatric clinical pharmacist in the pediatric intensive care unit.Crit Care Med. 2002;30:919–92153. Walsh KE, Kaushal R, Chessare JB. How to avoid paediatricmedication errors: a user’s guide to the literature. Arch Dis Child.2005;90:698–70254. Mistry KP, Turi J, Hueckel R, Mericle JM, Meliones JN.Pediatric rapid response teams in the academic medical center. ClinPediatr Emerg Med. 2006;7:241–24755. Last J, ed. A Dictionary of Epidemiology. 4th ed. New York NY:Oxford University Press, Inc; 2001

ClarificationA perceptive reader notes a discrepancy between PIR’s policy not to discuss “unapprovedcommercial products” and comments on the use of the unapproved natriuretic peptidenesiritide by Madriago and Silberbach in their article “Heart Failure in Infants and Children”(January 2010, pp. 4–12). In their PIR-online reply (http://pedsinreview.aappublications.org/cgi/eletters/31/1/4), the authors lament that the majority of drugs routinelyemployed in pediatrics are neither studied nor approved by the United States Food andDrug Administration (FDA) and state that nesiritide may, in fact, be more effective thansome of the drugs that are approved by the FDA for use in pediatric heart failure.




DOI: 10.1542/pir.31-4-158 2010;31;158 Pediatr. Rev.

Clarification












Beth Rezet, Benjamin D. Hoffman and Jeffrey Kaczorowski Training

Pediatrics in the Community: Integrating Community Pediatrics into Residency







Author Disclosure

Drs Rezet, Hoffman, Kaczorowski,

and Aligne have disclosed no

financial relationships relevant to

this article. This commentary does

not contain a discussion of an

unapproved/investigative use of a

commercial product/device.

Integrating Community Pediatricsinto Residency TrainingSince 1999, the American Academyof Pediatrics (AAP) has asserted thatpediatricians should “reaffirm theirrole as professionals in the commu-nity and prepare themselves for it,just as diligently as they prepare fortraditional clinical roles.” (1) Fur-thermore, as of 2001, the Accredita-tion Council for Graduate MedicalEducation (ACGME) Residency Re-view Committee for Pediatrics man-dated, ‘‘there must be structured ed-ucational experiences that prepareresidents for the role of advocate forthe health of children within thecommunity.” (2) As with clinical re-sponsibilities, the best approach topreparing for such a role is with activehands-on learning. Accomplishingthis task generally means participat-ing in a community-based project orother activity.

Fortunately, the skills gained dur-ing project participation translatesmoothly into ACGME competen-cies. The Table illustrates how thisteaching can be accomplished by us-ing, as an example, one of theprojects highlighted in this series.The Children’s Hospital of Philadel-phia pediatric residents have engagedin Ballroom Dance for L.I.F.E. toincrease physical activity for innercity fifth graders. (3) This endeavornot only allowed the residents toserve the community but also helpedthem to attain numerous proficien-cies. Faculty review of this project,with recording of feedback, progress,and outcome, fulfills the documenta-tion needs for ACGME competen-cies of medical knowledge, interper-sonal and communication skills,professionalism, and systems-basedpractice.

The AAP Community Pediatrics

Training Initiative (CPTI), in itsStructured Approach to CommunityHealth and Child Advocacy Train-ing: Integrating Goals, Activities,and Competencies, has identified ed-ucational goals and objectives to helpframe all community pediatrics expe-riences while also meeting require-ments of the ACGME. This docu-ment and additional resources areavailable on the CPTI web site:http://www.aap.org/commpeds/CPTI. This approach is based onthe work published in a supplementto Pediatrics in 2005 by the DysonInitiative curriculum committee.(4) The CPTI workgroup (5) iden-tified eight domains of communitypediatrics practice:

1. Culturally effective care2. Child advocacy3. Medical home4. Special populations5. Pediatrician as a consultant/

collaborative leader/partner6. Educational and child care set-

tings7. Public health and prevention8. Inquiry and applicationFor each domain, a goal and three

to six specific objectives are listed in atable with sample activities and doc-umentation, as well as correspondingACGME competencies. The Tablein this article presents one domain asan example. We believe that commu-nity pediatrics eventually will becomean integral part of pediatric training.(Beth Rezet, MD, Associate ClinicalProfessor of Pediatrics, Assistant Di-rector Pediatric Residency TrainingProgram, The Children’s Hospital ofPhiladelphia; Benjamin D. Hoffman,MD, Associate Professor of Pediatrics,Director, Pediatric Residency Pro-gram, University of New Mexico; Jef-

pediatrics in the community



frey Kaczorowski, MD, Director,Community Pediatrics Training Ini-tiative, American Academy of Pediat-rics, Associate Professor of Pediatrics,University of Rochester)

SECTION EDITOR’S NOTE. Dr Al-bert Schweitzer, the great medicalhumanitarian, said, “Example is notthe main thing in influencing others,it is the only thing.” I hope that thestories highlighted in this section todate have been inspirational. Thechallenge going forward will be tomultiply such examples systemati-cally so they become the norm ratherthan the exception. This articlepoints out that achieving this result

requires integrating community pe-diatrics into the routine processes ofresidency training. Documentingwhat residents learned by teachingchildren to waltz is not yet as simpleas 1, 2, 3, but it perhaps just becameeasier. (C. Andrew Aligne, MD, MPH)

References1. Rushton FE Jr. American Academy ofPediatrics Committee on CommunityHealth Services. The pediatrician’s role incommunity pediatrics. Pediatrics. 2005;115:1092–10942. Accreditation Council for GraduateMedical Education. Program Requirementsfor Residency Education in Pediatrics. Edu-cational Program: Community Experiences.

Accessed February 2010 at: http://www.acgme.org/acWebsite/downloads/RRC_progReq/320_pediatrics_07012007.pdf3. Fieldston E, Aligne CA. Pediatrics in thecommunity: Ballroom Dance for L.I.F.E.:mad hot community pediatrics. PediatrRev. 2007;28:276–2774. Rezet B, Risko W, Blaschke GS. Anne E.Dyson Community Pediatric Training Ini-tiative Curriculum Committee. Compe-tency in community pediatrics: consensusstatement of the Dyson Initiative Curricu-lum Committee. Pediatrics. 2005;115(suppl):1172–11835. CPTI Competency Workgroup: BenHoffman, MD, Chair; Lisa Chamberlain,MD, MPH; Susan Guralnick, MD; AmyJost-Starmer, MD; Jeffrey Kaczorowski,MD; Anda Kuo, MD, MPH; Gilbert Liu,MD; Beth Rezet, MD; Monique Evelyn,MA

Table. Matching Community Pediatrics Training With ACGME CompetenciesGoals and Objectives

Pediatrician as a consultant/collaborative leader/partner:Pediatricians must be child health consultants in their communities. Using collaborative skills, they must be able to work

with multidisciplinary teams, community members, educators, and representatives from community organizations andlegislative bodies.

Graduates are expected to:● Identify potential opportunities to serve as a health consultant in the community where he/she practices pediatrics and

demonstrate the ability to communicate effectively with a variety of audiences within that community.● Describe and discuss the essential qualities of community partnerships, including shared vision, the use of

complementary strengths, the willingness to collaborate, and the development of agreed-on boundaries.● Define and discuss principles of consensus building, including fostering inclusiveness, identifying mutual goals, setting

measurable outcomes, using effective problem-solving strategies, and negotiating toward consensus.

Sample Activity and Documentation of Resident Learning

Ballroom Dance For L.I.F.E. (Living Intelligently with Food & Exercise):A longitudinal senior advocacy project partnering with the community.Documentation of Resident Learning:1. The proposal is reviewed with faculty, and feedback and revision are documented. The documentation becomes part of

each resident’s portfolio.2. Residents are asked to reflect on the project and their involvement with the program coordinator.3. The presentation of health lessons are documented in handouts, posters, and slides.4. Attendance at meetings and other communication endeavors provide documentation of involvement.5. The project is presented at an end-of-year advocacy conference to be shared with faculty, nursing staff, and residents.6. The project is handed off to the next year’s class for ongoing, sustained involvement.

ACGME Competencies

Practice-based Learning and Improvement: Increasing physical activity in school-age children● Systems-based Practice: Partnering with a school and school district, obtaining grant funding● Interpersonal and Communication Skills: Providing leadership, sharing ideas and action, providing health curriculum● Professionalism: Uncompensated sharing of expertise and resources to benefit the health of the community● Medical Knowledge: Social determinants of health, barriers to medical care access

AGCME�Accreditation Council for Graduate Medical Education

pediatrics in the community




Beth Rezet, Benjamin D. Hoffman and Jeffrey Kaczorowski Training

Pediatrics in the Community: Integrating Community Pediatrics into Residency












Jacky M. Jennings and Erica Sibinga Research and Statistics: Understanding and Identifying Bias in Research Studies







Author Disclosure

Drs Jennings and Sibinga have

disclosed no financial relationships

relevant to this article. This

commentary does not contain a

discussion of an unapproved/

investigative use of a commercial

product/device.

Understanding and Identifying Bias inResearch StudiesJacky M. Jennings, PhD, MPH,* Erica Sibinga, MD*

Case StudyYou are seeing a healthy 18-year-old pa-tient who is interested in contraceptionand, specifically, the “Depo shot” or me-droxyprogesterone. The information youhave gathered from her during your visitand her medical history suggest that sheis a good candidate. You take a fewminutes out of the visit to review theliterature for issues of safety. Most of thearticles and reviews suggest that me-droxyprogesterone is a good choice forhealthy young women, but a recent studyshows an association between medroxy-progesterone and the development of hy-pertension. You look more closely at thestudy and discover that the youngwomen taking medroxyprogesterone hadtheir vital signs (including blood pres-sure) checked every 3 months when theycame in for the medroxyprogesterone in-jection. The other women included in thestudy for comparison only had their vitalsigns checked every year at annual healthsupervision visits. You are unclear aboutwhether the study design might have in-troduced bias and how the bias mightaffect your reading of the study results.

Bias DefinedBias is a major issue in epidemiologicresearch studies and can lead to infer-ences that systematically deviate fromtruth. Bias has been defined as “anysystematic error in the design, conductor analysis of a study that results in amistaken estimate of an exposure’s ef-fect on the risk of disease.” (1) Bias hasbeen written about extensively. In thisarticle, we review a few of the mostcommon types of bias encountered in

epidemiologic studies, discuss com-mon reasons for the biases, and pro-vide a guide for identifying bias whenreviewing or reading research studies.

Common Types of BiasSurveillance bias may result from onepopulation being monitored moreclosely or more frequently than thegeneral population. In the study de-scribed, in which an association wasfound between medroxyprogesteroneand hypertension, the young womentaking medroxyprogesterone weremonitored at a greater frequency, ev-ery 3 months, than were the youngwomen who were not taking medroxy-progesterone, who were monitoredannually. The difference in the rate ofmonitoring may have introduced asurveillance bias. Because hypertensiontypically is diagnosed only after at leastthree blood pressure readings havebeen elevated above a certain level, theyoung women who were monitoredmore frequently had more of an op-portunity to have hypertension diag-nosed. The young women who weremonitored less frequently (annual vis-its only) may have had the same prev-alence of hypertension, but they mayhave had less of an opportunity to havehypertension diagnosed because of thelower frequency of visits. The system-atic difference in the monitoring of theyoung women taking medroxyproges-terone compared with the youngwomen not taking medroxyprogester-one may have resulted in a spuriousassociation.

Selection bias has two primary vari-eties. One arises from systematic differ-ences in the characteristics between in-dividuals selected for a study comparedwith those not selected for the study.

*Assistant Professor of Pediatrics, Division ofGeneral Pediatrics & Adolescent Medicine, JohnsHopkins University, Baltimore, Md.

research and statistics



(2) In the study described, for exam-ple, the study sample population con-sisted entirely of individuals attendingthe clinic, who may differ from thegeneral population. Therefore, associ-ations found are not necessarily gener-alizable to the larger population.

The second variety of selection biasarises from systematic differences in theselection of cases and controls or ex-posed and unexposed individuals.Such systematic differences can lead tofallacious associations that really do notexist but, rather, are a result of theselection bias. The medroxyprogester-one study described serves as an exam-ple of this type of selection bias.

Despite the flaws in the describedstudy, a follow-up study was designed.That study prospectively follows indi-viduals taking medroxyprogesterone(exposed) and those not taking me-droxyprogesterone (unexposed) to de-termine the incidence of hypertensionin each group and to test for differ-ences between the two groups. Theeligibility criteria for the two groupsinclude all females between the ages of12 and 18 years attending a teen healthclinic. Additional eligibility criteria forthe exposed group require that all fe-males currently are taking medroxy-progesterone. Do you have any con-cerns about the selection criteria?

One of our concerns is that in theprospective study, individuals selectedfor the study in the two groups maydiffer systematically from one another.For example, the young women takingmedroxyprogesterone presumably aresexually active, are aware of medroxy-progesterone as a contraceptive op-tion, and may be more likely to attendtheir clinic visits regularly. The youngwomen in the nonmedroxyproges-terone group, on the other hand, maynot be sexually active and may differ inother measurable and unmeasurablefactors (eg, health insurance status, re-liable transportation to and fromclinic). If the differences related to the

selection of the two groups also arerelated to the likelihood of hyperten-sion, the study may show findings thatare fallacious.

Misclassification bias results frommisclassifying individuals into diseasedor nondiseased groups or into exposedand unexposed groups. Individuals inthe medroxyprogesterone study, whowere followed less frequently (only forannual visits), may have had less op-portunity to have hypertension diag-nosed. This circumstance may have re-sulted in their misclassification into thenonhypertensive (nondiseased) group.This type of misclassification bias istermed differential misclassificationbecause the rate of misclassification dif-fers in the comparison study groups.(3) Differential misclassification canlead to identifying an associationwhere one does not exist or a lack of anassociation where one does, indeed,exist.

Misclassification bias also can occuras a result of nondifferential misclassi-fication. (3) This error occurs whencases and controls (exposed and unex-posed individuals) are misclassified atsimilar rates and the misclassification isnot related to case-control or exposurestatus. Nondifferential misclassifica-tion usually results in an attenuation ofa relative risk or odds ratio, resulting inless likelihood of an association appear-ing, although it may exist.

How to Identify BiasKnowing the more common types ofbias in epidemiologic research studiesmay not be enough to identify bias inresearch studies. Following are somebasic tips on how to assess a researchstudy for bias.

The most likely areas to contain bi-ases are in the study design and meth-ods of the study, so the reader shouldlook closely at the description of theseprocedures. Has the study design beenidentified clearly, including the generalapproach (descriptive or hypothesis

testing), the level of measurement (in-dividual or ecologic), and the specificdesign (cohort, retrospective cohort,cross-sectional, case-control, experi-mental, quasi-experimental)?

The next step is to examine how thestudy sample was selected. Have theauthors clearly stated where and fromwhat population the study sample wasselected? What were the eligibility andineligibility criteria for study selection?How may the study sample selectioncriteria have created a study populationthat is representative of the populationsampled? If the study population selec-tion criteria suggest that the studypopulation is no longer representativeof the larger population, it is likely thatsome biases have been introduced andthat the study will have limited gener-alizability.

The strategies of sampling for con-trols or other comparison group is im-portant. Were there any systematic dif-ferences in their selection? Try to assesswhether the controls were selectedsimilarly to the cases, such as similareligibility and ineligibility criteria, andwhether the controls are comparableto the cases. The comparison groupsshould be similar in all characteristicsexcept for one: the disease or exposureunder study. Finally, in considering theresults, try to determine whether theresults and any associations found havesufficient supporting evidence to sug-gest plausibility and, if available,whether the results are consistent withthose of other studies.

References1. Schlesselman JJ. Case-Control Studies:Design, Conduct, Analysis. New York, NY:Oxford University Press; 19822. Last JM. A Dictionary of Epidemiology.3rd ed. Oxford, United Kingdom: OxfordUniversity Press; 19953. Gordis L. Epidemiology. Philadelphia,Pa: WB Saunders Company; 1996

research and statistics




Jacky M. Jennings and Erica Sibinga Research and Statistics: Understanding and Identifying Bias in Research Studies












Renee D. Boss Practices

Ethics for the Pediatrician: Pediatric Research Ethics: Evolving Principles and







Author Disclosure

Dr Boss has disclosed no financial

relationships relevant to this article.

This commentary does not contain a



product/device.

Pediatric Research Ethics:Evolving Principles and PracticesRenee D. Boss, MD, MHS*

IntroductionEvidence-based medicine requires afoundation of excellent research. Yet,research in infants and children in-herently poses potential risks to thesevulnerable populations. The aca-demic pediatrician may experience anethical conflict in his or her dual rolesas a clinician, with a primary interestin benefiting the individual child,and as a researcher, with an interest inadvancing science and benefiting so-ciety. Understanding the evolutionof ethical principles for pediatric re-search permits insight into currentprinciples and the challenges of put-ting such principles into practice.This information is critical for boththe pediatrician designing human re-search protocols and the pediatricianwhose patients are enrolled in thoseprotocols.

Pediatric Research Ethics andRegulations in the 20thCenturyAlthough documented cases of re-search involving pediatric subjectsdate 2 and 3 centuries in the past,systematic examination and regula-tion of human research began in theearly 1900s. Initially, this oversightoccurred in just a few nations. As theethical principles of human researchbecame more widely recognized overtime, accountability at the nationalor international level historically hasbeen uncertain.

The ethical principles of consentand noncoercion were among thefirst to be applied to human research.

In their earliest manifestations, theseprinciples resulted in the exclusion ofchildren from medical research. In1900, the Prussian Ministry of Reli-gious, Educational, and Medical Af-fairs prohibited research involvingsubjects who were not fully compe-tent to provide consent, includingchildren. (1) Without legal bindingpower, it is unclear what impact thedirective had on medical research atthe time; research involving childrenlikely continued. In 1931, the Reichgovernment refined distinctions be-tween therapeutic and nontherapeu-tic research and continued to excludechildren from both.

Although German principles ofhuman research ethics were amongthe most advanced in the world atthat time, these principles clearly didnot prevent the practice of unethicalresearch. The Nuremburg Code of1947 was a response to medical ex-perimentation in Nazi Germany byphysicians in concentration camps.The Code defined ten principles oflegitimate medical research and againprohibited research involving sub-jects who lacked the legal capacity togive consent, including children. (2)

The principles outlined in theNuremberg Code had no legal bind-ing in the United States, and researchinvolving children continued herethroughout the mid-20th century.Multiple studies exposing infants andchildren to radiation and radioactiveisotopes were documented in the1950s and 1960s. Between 1943 and1973, investigators from the Massa-chusetts Institute of Technology andHarvard University conducted re-search in mentally retarded childrenliving in a state-run residential facil-

*Assistant Professor, Division of Neonatology,Department of Pediatrics, Johns Hopkins School ofMedicine, Baltimore, Md.

ethics for the pediatrician



ity. Radioactive iodine-131, radioac-tive calcium, and iron tracers wereadministered to the children. (3) Theresearch was not disclosed to parents,and parental consent was not ob-tained.

The first document of human re-search ethics authored by membersof the international medical commu-nity was the Declaration of Helsinki,developed by the World Medical As-sociation in 1964. (4) The Declara-tion is considered the basis for mostcontemporary guidelines for humanresearch, stipulating that the welfareof the research participant must takepriority over the interests of scienceand society. The Declaration in-cluded a more complex principle forresearch consent. It allowed for con-sent to be provided by legal guard-ians, thereby permitting research inchildren (Articles 23, 24). In 1983,the second revision acknowledged achild’s emerging autonomy by stat-ing that assent from minors shouldbe obtained whenever possible in pe-diatric research (Article 25).

Over the past 50 years, standardsfor the practical application of theethical principles for human researchhave become more defined. In theUnited States, the National Com-mission for the Protection of HumanSubjects of Biomedical and Behav-ioral Research was created in 1974.The Belmont Report summarizes thebasic ethical principles and guidelinesdeveloped by the Commission. (5) Itaffirmed that research involving chil-dren is permissible if consent is ob-tained from a legal guardian acting inthe child’s best interest and assent isobtained from the child, if possible.The Commission established the re-sponsibility of Institutional ReviewBoards (IRBs) to evaluate the risksand benefits involved in research pro-tocols.

Human research oversight effec-tively was tied to research funding in

the United States in 1991 by theDepartment of Health and HumanServices with the adoption of the“Common Rule,” a minimal stan-dard of protection for human partic-ipants. (6) Although institutions arebound to the Common Rule if theyreceive federal research funding,most academic institutions abide bythe principles. The Rule’s three basicprinciples of research ethics include:1) Respect for persons, which en-sures informed, voluntary consent;2) Beneficence, which maximizesbenefits while minimizing risks; and3) Justice, which assures that someindividuals are not exploited to ben-efit others. Subparts specifically ad-dress research in fetuses, neonates,pregnant women (Subpart B), andchildren (Subpart D). The guidelinesfor research involving children aremore specific than, but not substan-tially different from, those of the Bel-mont Report.

In the past decade, the pendulumhas continued to swing away from ahistorical prohibition against re-search in children toward active pur-suit of access to medical research forchildren. In response to concerns bythe United States House of Repre-sentatives and Senate that childrenwere denied fair access to medicaladvances as a result of their routineexclusion from research, the Na-tional Institutes of Health began re-quiring in 1998 that children be in-cluded in all human subjects researchprotocols unless the researchersdemonstrate a compelling reason toexclude them. (7)

Putting Pediatric ResearchEthics Into Practice:Continuing DilemmasThe fundamental ethical principlesrelevant to pediatric research are notdifferent from those relevant to adultresearch: respect for persons, justice,and beneficence. However, the prac-

tical application of these principles toresearch involving children and fam-ilies is complicated by children’s de-velopmental variability, a heightenedconcern about risk, and the complex-ities of family decision-making.

Assessing Risk Versus BenefitA primary function of IRBs is to as-sess the risks and benefits associatedwith any research protocol, particu-larly those that target vulnerablepopulations such as children. Untilrecently, however, IRBs were notcompelled to have members who hadpediatric expertise. In 2004, an Insti-tute of Medicine (IOM) study com-mittee examined the regulations cov-ering clinical research with children,with a focus on maximizing safety forpediatric research participants. A keycommittee recommendation wasthat IRBs reviewing pediatric re-search protocols have at least threemembers who have pediatric exper-tise. (8)

Contemporary IRBs often mea-sure risk and benefit according to thestandards of the Common Rule. TheCommon Rule states that healthychildren may participate only in re-search that involves no more thanminimal risk, defined as “The prob-ability and magnitude of harm or dis-comfort anticipated in the researchare not greater, in and of themselves,than those ordinarily encountered indaily life.” (Subpart D) (6) Healthychildren may participate in no othertype of research. Children who havedisorders or conditions may be eligi-ble for studies involving greater risk.Research entailing more than mini-mal risk may be approved if it poten-tially offers the child a direct benefit,as in a study of a novel cancer drugfor treating a child who has brain-stem glioma. Research involvingmore than minimal risk that has nodirect benefit to the child may beapproved if the study poses only a




minor increase over minimal risk andwill add vital information about thedisorder. There is agreement that thedefinition of “minimal risk” shouldbe the same for healthy and sick chil-dren and for children whose dailyexperiences include poverty and vio-lence and for those whose experi-ences do not.

The practical applications of“minimal risk” and “a minor in-crease over minimum risk” are thepurview of individual researchers andlocal IRBs. Despite attempts by theIOM to clarify these concepts and toprovide more practical measures, sig-nificant variability in risk assessmentby IRBs has been documented.

Consent and AssentInformed consent remains a complexissue in research with children. It en-tails proxy consent from a parent(who may have his or her personalreasons for wanting the child to par-ticipate in a study) and sometimesassent from the child (who will beexposed to the potential risk or ben-efit). For protocols involving signifi-cant potential risk, both parents maybe required to provide consent (whomay have different views on theirchild’s participation). Because of achild’s developmental variability,unique parent-child dynamics, andthe possibility of a life-limiting dis-ease, research consent must accom-modate the complexities of familydecision-making.

As articulated in the AmericanAcademy of Pediatrics Policy State-ment, the process of obtaining assentfrom a child requires: 1) confirma-tion that a child has a developmen-tally appropriate understanding ofher health, 2) a clear explanation ofthe potential outcomes of tests ortreatments, 3) assessment of thechild’s understanding and any poten-tial sources of coercion, and 4) con-firmation of the child’s agreement

with participation in the study. (9)The lower age limit for assent variessomewhat with the individual childand the complexity of the research.A child’s developmental limitationsin understanding potential out-comes, such as death from cancer,can limit frank discussions, even inolder children.

Many questions about consentand assent remain. Can a healthychild be compelled by his or her par-ents to participate in minimal riskresearch as a lesson in altruism? For asick child, how certain are we that thechild understands the difference be-tween medical care and medical re-search? Genetics research often en-tails family studies or “banking” ofbiologic samples; what level of coer-cion does a child experience whose

parent wishes to enroll in a familystudy? Can a child understand theramifications of diagnosing a geneticdisease years before symptoms de-velop? What potential harm does thatinformation do to a child? For longi-tudinal protocols, should a child’sassent at one developmental stage berevisited, and how often? Can a par-ent consent to future research involv-ing the child’s biologic samples? Ad-vances in research methodology willrequire careful adaptations of ethicsprinciples and practice.

References1. Vollmann J, Winau R. Informed consentin human experimentation before the Nurem-berg code. BMJ. 1996;313:1445–14492. Trials of War Criminals Before the Nurem-berg Military Tribunals Under Control Coun-cil Law. Vol 10. Washington, DC: US Gov-ernment Printing Office; 1949:181–1823. The Thyroid Studies: A Follow-up Reporton the Use of Radioactive Maternals in Hu-man Subject Research that Involved Resi-dents of State-operated Facilities within theCommonwealth of Massachusetts from 1943through 1973. Boston, Mass: The WorkingGroup on Human Subject Research; 19944. World Medical Organization. Declara-tion of Helsinki. 1964. Br Med J. 1996;313:1448–14495. The Belmont Report. Ethical Principlesand Guidelines for the Protection of HumanSubjects of Research. Washington, DC: De-partment of Health, Education and Welfare;19796. Code of Federal Regulations. Title 45Public Welfare. Part 46 Protection of Hu-man Subjects. Washington, DC: Depart-ment of Health and Human Services. Officefor Human Research Protections; 20057. Office of Extramural Research. Inclusionof Children Policy Implementation. Wash-ington, DC: United Sates Department ofHealth & Human Services. 1998. http://grants.nih.gov/grants/funding/children/children.htm8. Institute of Medicine of the NationalAcademies. The Ethical Conduct of ClinicalResearch Involving Children. Washington,DC: The National Academies Press; 20049. American Academy of Pediatrics. Com-mittee on Bioethics. Informed consent, pa-rental permission, and assent in pediatricpractice. Pediatrics. 1995;95:314–317

Summary• In contrast to a historical

prohibition against researchinvolving children, there is acurrent emphasis on enhancingchildren’s access to research toprevent and treat pediatricdisease.

• Healthy children may notparticipate in research protocolsentailing more than minimalrisk. Sick children may bepermitted to participate inprotocols involving slightlymore risk, if they may receivedirect benefit or if vitalinformation about theirdisorder can be gathered.

• As the number of pediatricresearch participants increases,there must be a commitment toproviding the local, national,and international oversight andresources necessary to maximizesafety.

• Pediatricians, researchers, IRBs,and parents must work togetherto refine and apply theprinciples of respect forpersons, justice, andbeneficence to children.





Renee D. Boss Practices

Ethics for the Pediatrician: Pediatric Research Ethics: Evolving Principles and












Index of Suspicion







The reader is encouraged to writepossible diagnoses for each case beforeturning to the discussion. We invitereaders to contribute case presentationsand discussions. Please inquire first bycontacting Dr. Deepak Kamat [email protected].

Author Disclosure

Drs Zadeh, Bernstein, Stiasny, Callaghan,

Flores, Tytko, and Mannarino and Mr

Moore have disclosed no financial

relationships relevant to these cases. This

commentary does not contain a

discussion of an unapproved/investigative

use of a commercial product/device.

Case 1 PresentationA 7-week-old girl presents to the EDwith respiratory distress. She has hadincreasingly labored breathing overthe past several hours without ac-companying fever, cough, or rhinor-rhea. During evaluation, she devel-ops apnea and requires intubation.Chest radiography demonstratesdiffuse bilateral opacities. Echocardi-ography reveals dilated cardiomyop-athy. During a 2-week hospitaliza-tion, her clinical status and cardiacfunction gradually improve on med-ical therapy.

Her cardiac function normalizeswithin months of discharge. Labora-tory evaluation and endomyocardialbiopsy do not identify a cause for hercardiomyopathy. Viral titers are neg-ative. During this interval, she isnoted to have nystagmus. Dilatedophthalmologic examination andMRI of the head yield normal results.

An attempt is made to take her offangiotensin-converting enzyme in-hibitors at 1 month of age. However,after the withdrawal of medications,her cardiac function decreases butnormalizes within 4 weeks of reinitia-tion of the medication.

At 3 years of age, the patient hasslightly delayed developmental mile-stones and photophobia. Her heightand weight are at the 95th percentilefor age, with a body mass index(BMI) of 20 kg/cm2 (�97th percen-tile). She has acanthosis nigricans atthe posterior neck and truncal obe-sity. Follow-up ophthalmology eval-uation performed at age 3 yearsreveals decreased visual acuity; pho-tophobia; high-frequency, small-amplitude pendular nystagmus; andhypoplastic optic nerves with poorfoveal reflex, consistent with conedysfunction. The patient does nothave electroretinography (ERG) per-formed. Review of the family historyidentifies remote consanguinity.

Case 2 PresentationA 12-year-old African American boypresents with increasing bilateralneck swelling and left-sided abdomi-nal pain of 3 weeks’ duration. Hedenies fever, throat pain, weight loss,or night sweats. Recently, he had aviral infection diagnosed. When theneck swelling continued to worsen,he came to the ED.

On physical examination, hisweight is 58.1 kg, temperature is37.0°C, heart rate is 80 beats/min,blood pressure is 130/58 mm Hg,respiratory rate is 20 breaths/min,and oxygen saturation is 100% onroom air. He has bilateral anteriorcervical and supraclavicular lymphad-enopathy as well as a spleen tip pal-pable below the left costal margin.He has no hepatomegaly or abnor-mal findings on the cardiovascular,respiratory, or neurologic examina-tions.

Laboratory results in the ED showa WBC count of 42.9�103/mcL(42.9�109/L), with 58% neutro-phils, 11% lymphocytes, 13% mono-cytes, 5% bands, 5% myelocytes, and8% blasts. His Hgb is 14.0 g/dL(140 g/L), platelet count is123�103/mcL (123�109/L), uricacid concentration is 10.6 mg/dL(630.5 mcmol/L), and lactate dehy-drogenase (LDH) is 2,244 U/L.

Leukemia is suspected, based onhis high WBC count, markedly ele-vated uric acid and LDH values, andpresence of blasts on peripheralsmear. He is hospitalized after receiv-ing two doses of recombinant urateoxidase for hyperuricemia in the ED.Two and one-half hours after admis-sion, his oxygen saturation is 80% onroom air by transcutaneous pulseoximetry. He is transferred to thepediatric intensive care unit, whereadditional laboratory evaluationleads to the diagnosis.

Frequently Used Abbreviations

ALT: alanine aminotransferaseAST: aspartate aminotransferaseBUN: blood urea nitrogenCBC: complete blood countCNS: central nervous systemCSF: cerebrospinal fluidCT: computed tomographyECG: electrocardiographyED: emergency departmentEEG: electroencephalographyESR: erythrocyte sedimentation

rateGI: gastrointestinalGU: genitourinaryHct: hematocritHgb: hemoglobinMRI: magnetic resonance imagingWBC: white blood cell

index of suspicion



Case 3 PresentationAn 11-year-old boy presents to thesports medicine clinic with a5-month history of worsening painand swelling of both ankles. He hasbeen referred by a pediatric rheuma-tologist who evaluated him recently.His pain is located at the anterioraspect of his ankles and is worse withhigh-impact activities such as basket-ball. He cannot recall an acute injury.His mother reports that sometimeshe walks “like an old man” because ofthe pain. He has had no fever, weightloss, rash, night pain, or swelling ofany other joints. Family history is notcontributory.

Examination of his ankles reveals amild effusion and pes planus withvalgus orientation of both ankles(Fig. 1). Passive range of motion ofboth ankles is normal and does notelicit any pain or crepitus. The anklesare not warm, and he is neurovascu-larly intact. There is generalized ten-derness along the anterior aspects ofhis ankles in the location of his talar-tibial joint. His gait appears normal.

CBC, comprehensive metabolicprofile, urinalysis, C-reactive protein,and HLA-B27 results are normal.Radiographs of the ankles reveal anirregularity along the medial aspects

of his talar domes (Fig. 2), and adiagnosis is suspected.

He is advised to stop high-impactactivities while his evaluation is beingcompleted. A subsequent imagingstudy helps to delineate the diagno-sis.

Case 1 DiscussionMultiorgan disease in infancy shouldprompt consideration of genetic dis-orders. In this case, the family historyof consanguinity as well as co-occurrence of disease in more thanone organ system (ie, cardiac andophthalmologic) greatly increasessuspicion for a genetic disorder.

Differential DiagnosisA variety of infectious and noninfec-tious conditions should be consid-ered in the differential diagnosis ofrespiratory distress in an infant. Non-infectious causes include structuralairway anomalies, disorders of thechest wall and diaphragm, cardiacdisease, and neurologic disorders.Anemia and metabolic disorders alsoare considerations.

Cardiomyopathy during infancyand toddlerhood may be a presentingsign of multiple syndromic and non-syndromic genetic disorders. Theseconditions include familial dilatedand hypertrophic cardiomyopathies,lysosomal storage disease, disordersof fatty acid oxidation and carnitinetransport, mitochondrial disorders,connective tissue disorders, Noonanand related genetic syndromes, andAlstrom syndrome. The differentialdiagnosis for nongenetic causes ofearly-onset cardiomyopathy includesinfectious myocarditis, sequelae ofcongenital heart disease, and ar-rhythmias.

The findings of nystagmus and di-lated cardiomyopathy in this patientsupported mitochondrial disease andAlstrom syndrome. The patient’s in-creased BMI favored Alstrom syn-drome rather than mitochondrialdisease. Normalization of cardiacfunction in a fairly short period oftime after initial onset has been ob-served in Alstrom syndrome and mayreassure the clinician falsely and delayidentification of an already diagnos-tically challenging condition.

Figure 1. Pes planus with valgus orien-tation of both hind feet.

Figure 2. Radiographs of the ankles reveal an irregularity along the medial aspects ofthe talar domes.

index of suspicion



Alstrom syndrome was diagnosedclinically on the basis of the patient’shistory of dilated cardiomyopathy,photophobia, nystagmus, obesity,and acanthosis nigricans. The diag-nosis was confirmed by moleculartesting for the ALMS1 gene. Thesensitivity of molecular diagnosis islimited and is estimated to be in therange of 25% to 40%. Thus, the phys-ical examination and associated clin-ical findings often are essential inmaking this diagnosis.

The ConditionAlstrom syndrome is an autosomalrecessive condition resulting frommutations of the ALMS1 gene. Clin-ical diagnosis is based on recognitionof a distinctive constellation of find-ings. As observed in this case, dilatedcardiomyopathy, cone-rod dystro-phy, truncal obesity, and insulin re-sistance are common in the first de-cade of life. ERG is considered thegold standard for detecting cone-roddystrophy because retinal examina-tion may yield normal results duringinfancy. Later retinal findings mayinclude narrowing of retinal vessels,patchy atrophy of retinal pigment ep-ithelium, increased visibility of cho-roidal vessels, macular sheen, andpale optic discs. Hyperlipidemia andsensorineural hearing loss are fre-quent findings in infancy. By the sec-ond decade, vision loss is prominent.Cardiomyopathy usually developsprior to the onset of visual distur-bances but may occur later in life aswell. A small proportion of patientsdevelop renal disease due to intersti-tial fibrosis.

ManagementThere is no specific therapy for Al-strom syndrome. Management issupportive and aimed at minimizingcomplications of the condition. In-terdisciplinary treatment is recom-mended. Contributing specialists

should include audiology, cardiol-ogy, endocrinology, genetics, ne-phrology, gastroenterology, andophthalmology.

Early recognition of Alstrom syn-drome is important for optimal man-agement as well as counseling re-garding recurrence risk andprognosis. Affected families shouldbe advised that blindness is to beexpected and that early training fornonvisual skills is recommended. Es-tablishing a healthy diet and regularexercise at an early age may help tominimize obesity. Periodic screeningfor insulin resistance also is indicated.

Lessons for the Clinician● Unexplained severe illness in an in-

fant should raise concern for a ge-netic disorder, especially whenmultiple organ systems are af-fected.

● Identifying a specific genetic disor-der allows for accurate counselingabout recurrence risk, better antic-ipation of expected complications,and proper medical management,with the aim of maximizing patientand family quality of life.

(Neda Zadeh, MD, Jonathan A.Bernstein, MD, PhD, Stanford Uni-versity, Stanford, Calif.)

Case 2 DiscussionBlood gas analysis revealed 15.3%methemoglobin. The boy was givena dose of methylene blue as treat-ment for methemoglobinemia, buthe continued to demonstrate low ox-ygen saturation and worsening of re-spiratory distress and, therefore, wasstarted on 100% oxygen via nonre-breather mask, with which heshowed moderate improvement. Thepatient was found to have lowglucose-6-phosphate (G6PD) activ-ity. The cause of his methemoglobin-emia was the administration of the

urate oxidase and G6PD deficiency.Urate oxidase was discontinued, andhe was started on allopurinol for hy-peruricemia. In addition, the diagno-sis of T-cell acute lymphoblastic leu-kemia was confirmed by bonemarrow studies.

The ConditionMethemoglobinemia is a disorderencountered in a number of settingsin children (Table). Methemoglobinis the result of heme-bound ironlocked in the Fe3� (ferric) form, andthis compound gives the Hgb pig-ment a bluish-gray hue. Methemo-globin is unable to deliver oxygen totissues. It is important to identify thisdiagnosis early because the cause andtreatment are distinct from othercauses of cyanosis.

Congenital causes of methemo-globinemia, including Hgb M vari-ants, usually are diagnosed in theperinatal period. Newborns have arelative deficiency of methemoglobinreductase, the enzyme responsiblefor conversion of methemoglobin tonative Hgb. Thus, methemoglobin-emia sometimes is encountered inthe newborn intensive care unit. Thispathway requires reductive capacityin the form of reduced glutathione,which is provided by the pentosephosphate shunt.

The most common cause of met-hemoglobinemia is diarrheal illnessassociated with metabolic acidosis ininfancy. This condition occurs fre-quently in children 2 to 6 months ofage and is believed to result fromincreased intestinal permeability tonitrites, increased concentrations offetal Hgb (which is more susceptibleto oxidation), and age-related dimin-ished capacity to reduce methemo-globin. Another acquired cause ofmethemoglobinemia is exposure tooxidizing medications (Table), suchas, in this case, recombinant urateoxidase. Recombinant urate oxidase

index of suspicion



is an enzyme that catalyzes the con-version of uric acid into allantoin toenable its excretion through the uri-nary tract (allantoin is five times moresoluble than uric acid) and, thus, candecrease serum uric acid concentra-tions rapidly and protect the kidneysin patients who have tumor lysis syn-drome.

Signs and symptoms of methemo-globinemia include a cyanotic orgrayish appearance, which usually isnoted at methemoglobin concentra-tions in excess of 16%, and respira-tory distress. The treatments for met-hemoglobinemia are removal of thesource, supplemental oxygen, meth-ylene blue, and when severe, ex-change transfusion.

G6PD deficiency is the most com-mon red blood cell (RBC) enzy-mopathy, with approximately200 million people affected. ThisX-linked recessive disorder can causeepisodic hemolytic anemia from in-fections, medications (such as sulfadrugs, antimalarials, or aspirin), orfava beans (especially common in thediets of Middle Eastern cultures).Chronic hemolysis occurs rarely.Males are affected commonly, al-though extreme lyonization of theunaffected X chromosome can causesome women to be affected. Muta-tions can occur on multiple sites inthe G6PD gene, causing a wide rangeof manifestations. There are two ma-jor mutation categories: the A vari-ant, usually seen in individuals of Af-rican descent and associated with amilder form of the disease, and the Bvariant, usually seen in individuals ofMediterranean or Middle Easterndescent.

G6PD is an essential enzyme inthe pentose phosphate pathway thatnormally oxidizes glucose-6-phosphate to 6-phosphogluconicacid through reduction of nicotin-amide adenine dinucleotide phos-phate (NADP) to nicotinamideadenine dinucleotide phosphate hy-drogen (NADPH) (Fig. 3). NADPHsubsequently is reoxidized to NADPthrough the reduction of oxidizedglutathione to reduced glutathione.However, enzyme deficiency causes adeficiency of NADPH, leading to de-creased concentrations of reducedglutathione and impaired response to

oxidative stress. Exposure to oxi-dants results in accumulation of freeradicals, which affect red blood cellRBC membranes, resulting in hemo-lysis. Symptoms usually develop 1 to2 days after ingestion of the sub-stance causing the oxidative stress.Laboratory studies demonstrate a de-cline in Hgb and Hct as well as serumhaptoglobin. Hgb often is present inthe urine. Peripheral smear reveals“basket” or “bucket handle” RBCs,and Heinz bodies are seen on supra-vital stains.

The patient can be determined tobe deficient by testing for enzymeactivity in RBCs, although testsyielding negative results should berepeated several weeks after an acuteepisode because concentrations canbe elevated in reticulocytes, whichordinarily are increased during anacute episode. Acute treatment mayrequire RBC transfusion as well asremoval of the substance causing theoxidative stress.

When oxidized glutathione con-centrations within RBCs are low inpatients who have G6PD deficiency,free radicals and other oxidized sub-stances can induce oxidation of ironin Hgb from Fe2� to Fe3�, inducingmethemoglobinemia. Patients bornwith G6PD deficiency who receiveurate oxidase can develop methemo-globinemia by this process. The reac-tion catalyzed by urate oxidase causesoxidation of uric acid to allantoin andreduction of oxygen to H2O2-, whichthen can participate additionally inreactions that cause oxidation ofFe2� in the Hgb molecule to its ferricform through reduction of H2O2 towater. Without resolution of storesof reduced glutathione or removalof the oxidative stress, a vicious cy-cle of methemoglobinemia and he-molytic anemia ensues in affectedpatients.

Table. Causes ofMethemoglobinemiain ChildrenCongenital

● Hemoglobin M variants● Methemoglobin reductase

deficiency

Acquired

● Diarrheal illness of infancy● Oxidant drugs

–Urate oxidase–Nitroprusside–Nitroglycerin–Nitric oxide–Silver nitrate–Lidocaine–Prilocaine–Benzocaine–Nitrous oxides–Dapsone–Phenazopyridine–Primaquine–Chloroquine–Acetaminophen–Cyclophosphamide–Ifosfamide–Celecoxib

● Inorganic agents–Nitrates–Chlorates–Copper sulfate

● Organic nitrites/nitrates–Amyl nitrite–Isobutyl nitrite–Sodium nitrite–Nitrogen dioxide–Trinitrotoluene

● Industrial/household agents–Aniline dyes–Nitrobenzene–Naphthalene (moth balls)–Aminophenol–Nitroethane

index of suspicion



Treatment and PrognosisMethylene blue (1%) in an intrave-nous dose of 1 to 2 mg/kg is thetreatment of choice for methemoglo-binemia not associated with G6PDdeficiency. Methylene blue helps toreduce iron to its Fe2� state by ac-cepting electrons from NADPH,producing NADP in the process viathe enzyme NADPH methemoglo-bin reductase. This mechanism is notfeasible when G6PD is deficient.

This patient recovered well fromhis episode of methemoglobinemiaas well as from the early onset oftumor lysis syndrome. He respondedwell to the induction phase of hischemotherapeutic protocol for T-cellacute lymphocytic leukemia and con-tinues to undergo treatment for hismalignancy. He has not had any re-current episodes of methemoglobin-emia.

Lessons for the Clinician● Methemoglobinemia is a condition

encountered in a number of cir-cumstances in pediatrics.

● Clinicians should be aware of thepresentation, causes, and treat-ment of this disorder. Specifically,this report highlights the associa-tion of G6PD deficiency and met-hemoglobinemia and the contrain-dication to the use of methyleneblue in G6PD deficiency patients.

● Physicians should take great care inconsidering the possible adverse ef-fects and contraindications ofnewer medications used in clinicalpractice.

(David Stiasny, MD, Michael U. Cal-laghan, MD, Children’s Hospital ofMichigan, Detroit, Mich.)

Case 3 DiscussionDue to his ankle effusion and theabnormality noted on his talar domeson plain radiographs, MRI of theboy’s ankles was obtained, which re-vealed two medial osteochondral le-sions in the right talus and one me-dial osteochondral lesion in the lefttalus. These findings confirmed thediagnosis of bilateral osteochondritis

dissecans (OCD) of the talus, ex-plaining his chronic ankle pain andeffusion.

The ConditionThe term OCD is a misnomer andoriginated in 1888 when Konig triedto describe the pathologic processthat led to atraumatic loose bodies inthe knee and hip joints. He believedthat an underlying inflammatory re-action of bone and cartilage was theprimary component of this process,but inflammatory cells in histologicspecimens of removed osteochondralloose bodies never have been identi-fied. Thus, the more accepted termi-nology today is “osteochondral le-sion.” Nevertheless, use of the termOCD has persisted. We use the moreappropriate term of osteochondrallesion of the talus (OLT).

The articular cartilage and sub-chondral bone are involved in anosteochondral lesion. It is character-ized by separation of an osteochon-dral fragment from the underlyingbone. It occurs in the knee 75% ofthe time, the elbow 6% of the time,and the ankle 4% of the time. OLT israre in children and occurs more of-ten in males than in females, usuallypresenting at 15 to 35 years of age.

The cause for OLT is unknown,although trauma, ischemia, and he-reditary propensity have been pro-posed. Most experts believe that thedevelopment of osteochondral le-sions is multifactorial.

Clinically, the patient who hastraumatic or atraumatic OLT pre-sents with a few or all of the ensuingsigns and symptoms: 1) unresolvingankle pain after an inversion injury orchronic ankle pain without a specificinjury; 2) stiffness, swelling, or re-duced range of motion; 3) ecchymo-sis (rare); and 4) catching, clicking,locking, or giving way at the anklejoint. In children, the predominant

Figure 3. Mechanism by which urate oxidase causes methemoglobinemia in patientswho have G-6-PD deficiency.

index of suspicion



presenting complaints are pain andswelling.

One classification catalogs theOLT into four different radiographicstages: stage I involves a small area ofcompression of subchondral bone,stage II represents a partially de-tached osteochondral fragment,stage III is characterized as a com-pletely detached osteochondral frag-ment remaining in the crater bed,and stage IV is a displaced (loose)osteochondral fragment.

EvaluationThe history of chronic pain without aspecific injury combined with thephysical findings of an effusionshould raise suspicion for an osteo-chondral lesion or rheumatologicdisease. Absence of constitutionalsymptoms in this patient made aninfectious process less likely. In addi-tion, when any child has musculo-skeletal pain, it is vital to ask aboutpain at night; an underlying onco-logic process or osteoid osteoma canbe associated with night pain.

Physical examination of any pa-tient who has ankle problems shouldbegin with inspection of his or herstance and gait. The physical findingin this patient of pes planus with val-gus alignment suggested the possi-bility of an abnormal distribution offorces in the ankle joint. Althoughthis type of malalignment could con-tribute to this type of pain, it wouldnot explain the ankle effusions.

If ankle effusion or tenderness isdetected, plain radiographs with an-teroposterior, lateral, and mortiseviews of the ankle are needed to look

for underlying joint pathology. Inmost cases, plain radiographs havebeen sufficient for diagnosing OLT.If the radiographs appear normal andthe symptoms persist, MRI can beused to identify injuries of the sub-chondral bone and cartilage that maynot be detected on radiographs. MRIalso can be used to assist with thediagnosis of OLT identified via plainradiograph, to aid in classification,and to evaluate for loose fragments inthe joint that usually require surgicalremoval.

TreatmentPatients diagnosed with or suspectedof having OLT should be referred toan orthopedic surgeon for additionalevaluation. Symptomatic lesions inchildren or skeletally immature pa-tients who do not have loose frag-ments on radiographic examinationcan be treated initially with a trial ofconservative therapy. This regimenincludes rest from high-impact activ-ity, with an initial period of immobi-lization of the limb in a cast, Cam-Walker boot, or brace.

This patient was provided withankle braces and advised to rest untilhe was pain-free and his effusion hadresolved. At his 4-week follow-upvisit, he had no pain or effusion andwas permitted to resume full activi-ties. He returns to the clinic every 6months for monitoring of his condi-tion.

Lessons for the Clinician● OLT is a rare disorder in children

that should be considered in thedifferential diagnosis of any child

suffering persistent ankle pain andeffusion with or without a previousinjury.

● It is important to inquire about aspecific injury in any child who hasankle pain because such knowledgecan help differentiate between anacute ankle sprain or fracture andan underlying disorder such asOLT.

● Plain radiographs should be ob-tained for any child suspected ofhaving OLT, followed by MRI toassist in diagnosis, classification ofthe lesion, and detection of loosefragments.

● If OLT is diagnosed, referral to anorthopedic surgeon is warrantedfor additional evaluation, treat-ment, and possible surgical inter-vention.

● For asymptomatic lesions diag-nosed incidentally after an acuteankle injury, MRI may not be nec-essary; instead, treating the acuteinjury, monitoring for symptomsover time, and following with ra-diographs every 6 months to eval-uate for loose fragments is suffi-cient.

(Charles E. Flores, MD, PediatricsDay and Night, Hamilton, NJ; JamesM. Tytko, MD, Frank P. Mannarino,MD, Jim Moore, ATC, KetteringSports Medicine Center, Kettering,Ohio)

To view Suggested Reading lists forthese cases, visit http://pedsinreview.aappublications.org and click on In-dex of Suspicion.

index of suspicion




Index of Suspicion



Supplementary Material

/DC1http://pedsinreview.aappublications.org/cgi/content/full/31/4/167Supplementary material can be found at:






http://pedsinreview.aappublications.org/cgi/content/full/31/4/167/DC1





Nasreen Bhumbra and Mary Elizabeth Wroblewski Adenovirus







In BriefAdenovirusNasreen Bhumbra, MDMary Elizabeth Wroblewski, MDUniversity of ToledoToledo, Ohio

Author Disclosure

Drs Bhumbra, Wroblewski, and

Serwint have disclosed no financial

relationships relevant to this In Brief.




product/device.

Adenoviruses. Cherry JD, Chen TK. In:Feigin RD, Cherry JD, Demmler-Harrison GJ, Kaplan SL, eds. Feiginand Cherry’s Textbook of PediatricInfectious Diseases. 6th ed. Philadel-phia, Pa: Elsevier Saunders; 2009:1949–1972

Adenovirus Infections. American Acad-emy of Pediatrics. In: Pickering LK,Baker CJ, Kimberlin DW, Long SS,eds. Red Book: 2009 Report of theCommittee on Infectious Diseases.28th ed. Elk Grove Village, Ill: Ameri-can Academy of Pediatrics; 2009:204–206

Laboratory Approaches to the Diagnosisof Adenovirus Infection Dependingon Clinical Manifestations.Terletskala-Ladwig E, Leinmuller M,Schneider F, Meier S, Enders M.Infection. 2007;35:438–443

Adenoviruses. Langley JM. Pediatr Rev.2005; 26:244–248

Adenoviruses, so named because theywere discovered initially in adenoidaltissue, are DNA viruses that can cause amultitude of clinically significant hu-man disease syndromes. More than 50different strains of the virus have been

discovered. Most children become in-fected with at least one strain withinthe first 5 years after birth, and infec-tion is more likely in child care andovercrowded conditions. Communityoutbreaks of adenoviral disease havebeen recognized worldwide. Infectioncan be asymptomatic, and reinfectionalso is possible.

Adenovirus is transmitted from per-son to person through contact withrespiratory secretions, through fecal-oral transmission, and via fomites. Ex-posure to contaminated swimming poolwater and lakes has caused outbreaks.Nosocomial transmission in health-caresettings from personnel and improperlycleaned equipment also can occur. Ad-enoviral infection can take place at anytime throughout the year, but out-breaks usually are concentrated in win-ter, spring, and early summer. Althoughthe virus can survive on surfaces fordays, the incubation period varies from2 to 14 days for respiratory infectionsand 3 to 10 days for gastrointestinaldisease.

The respiratory and gastrointestinalsystems are affected most commonly.Symptoms of respiratory tract infectionin children include those associatedwith a nonspecific febrile illness, upperrespiratory tract infection, otitis media,pharyngitis, exudative tonsillitis, andpneumonia. Pharyngoconjunctival feveris characterized by fever, tonsillitis(sometimes suppurative), follicular con-junctivitis, coryza, and diarrhea. Cervi-cal and preauricular lymphadenopathyis common. The presence of a general-ized rash in association with fever,conjunctivitis, and pharyngitis can bemistaken for Kawasaki disease. Otherclinical manifestations of adenoviral

disease include a pertussislike illnessand bronchiolitis obliterans in younginfants. Occasionally, especially in im-munocompromised children, severe in-fection can lead to meningitis, myocar-ditis, pericarditis, hemorrhagic cystitis,or hepatitis. Fulminant infections withmultiorgan failure can occur in neo-nates. Gastroenteritis follows infectionwith enteric adenovirus serotypes 40or 41 and less often with serotype 31.More recently, severe and fatal infec-tions due to serotype 14 have beenreported in all age groups.

Adenoviral infections are highlycontagious, but spread can be reducedby using proper hand hygiene, surfacecleaning with bleach, and contact anddroplet precautions when treating hos-pitalized patients known to have ad-enoviral disease. Pharyngoconjunctivalfever prevention requires proper chlori-nation of swimming pools. Infection ismost communicable during the initialillness, but virus shedding can persistlong after symptoms have resolved.

Several diagnostic techniques canidentify an adenoviral infection. Themost readily available tests are directantigen detection and viral isolation inclinical specimens from various bodysites and fluids. A rapid antigen test foradenovirus has a high sensitivity andspecificity compared with viral culture.Traditional viral culture detection takeslonger, with cytopathic effects com-monly noted within 1 week. A modifi-cation of viral culture known as therapid shell vial technique can detectvirus growth 1 to 2 days after specimeninoculation. Enteric serotypes 40 and41 are identified best by antigen detec-tion in stool specimens.

Other identification methods in-clude restriction endonuclease analysis,

in brief



polymerase chain reaction, serology,and electron microscopy. These tech-niques are not universally available,despite both restriction endonucleaseanalysis and polymerase chain reactionbeing even more sensitive and specificthan antigen detection by immunoflu-orescence. Suitable clinical specimensinclude nasopharyngeal aspirate orswab, throat swab or wash, rectal swab,conjunctival swab or scrapings, stool,urine, cerebrospinal fluid, and tissue.Specimens should be obtained as earlyas possible after the onset of disease toincrease the chances of virus detection.

Although adenoviral infections gen-erally are self-limited and require nomore than supportive care, more severedisease can occur in immunocompro-mised patients, and antiviral agents

such as ribavirin and cidofovir havebeen used with inconsistent results.

Comment: The discovery of the clin-ical manifestations of adenovirus spansthe past century. Although viral strainsinitially were isolated in 1953, epidem-ics of keratoconjunctivitis were de-scribed in Austria as early as 1889,followed by descriptions of pharyngo-conjunctival fever in the 1920s andreports of adenovirus strains thatcaused enteritis in 1975. With such awide array of presentations, cliniciansneed to consider adenoviral infectionswhen evaluating all of these symptoms.Although immunocompetent childrenmay be asymptomatic or have self-limited disease, specific diagnosis ofadenovirus can be important in certain

circumstances. Because the signs andsymptoms of adenovirus may mimicKawasaki disease, differentiation be-tween Kawasaki and adenovirus is es-sential because of the time sensitivityof treatment for Kawasaki disease. Pre-cise diagnosis in immunocompromisedpatients, especially pediatric transplantpatients, can be equally important, as isdiagnosis during outbreaks or for hos-pital “cohorting” procedures. When di-agnosis is important, polymerase chainreaction and shell viral culture haveenhanced sensitivity over rapid immu-nofluorescence tests; practitioners needto be aware of the types of testsavailable in the laboratories they use.

Janet R. Serwint, MDConsulting Editor

In BriefToxic PlantsKevin Carter, MDDaniel R. Neuspiel, MD, MPHDepartment of PediatricsLevine Children’s Hospital of Carolinas

Medical CenterCharlotte, NC

Author Disclosure

Drs Carter, Neuspiel, and Serwint

have disclosed no financial





product/device.

2007 Annual Report of the AmericanAssociation of Poison Control Cen-ters’ National Poison Data System(NPDS): 25th Annual Report. Bron-

stein A, Spyker D, Cantilena L Jr,Green J, Rumack B, Heard S. ClinToxicol. 2008;46:927–1057

Toxic Plant Ingestions. Graeme KA. In:Wilderness Medicine. 5th ed. Phila-delphia, Pa: Mosby; 2007

Plants. Palmer M, Betz J. In: Goldfrank’sToxicologic Emergencies. 8th ed.New York, NY: McGraw-Hill; 2006

Toxic Mushroom Ingestions. SchneiderS, Donnelly M. In: Wilderness Medi-cine. 5th ed. Philadelphia, Pa: Mosby;2007

More than 60,000 calls are made an-nually to poison control centers (PCCs)for cases of suspected plant toxicity.Children younger than age 6 yearscomprise two thirds of cases, due totheir natural curiosity and limited judg-ment. Most of these exposures are be-nign; fewer than 10% result in treat-ment by a health professional. The PCC

is a valuable source of information toassist in management.

Most ingestions of plant material byyoung children are of small quantity,and symptoms, if present, typically areshort-lived and self-limited. Gastroin-testinal effects are common and maybe a clue to seek other, more subtlesigns of poisoning. Plant ingestions inolder children and adolescents gener-ally are intentional and of larger quan-tity, the result of either substance ex-perimentation or attempted self-harm.

Autonomic toxidromes can be seenin many plant poisonings. Deadly night-shade (Atropa belladonna) and Jimsonweed (Datura stramonium) produce at-ropine, scopolamine, and hyoscyamine,all anticholinergic toxins. Victims canpresent with classic symptoms of flush-ing, hyperthermia, blurred vision, drymouth, and hallucinations. Common

in brief




Nasreen Bhumbra and Mary Elizabeth Wroblewski Adenovirus












Kevin Carter and Daniel R. Neuspiel Toxic Plants







polymerase chain reaction, serology,and electron microscopy. These tech-niques are not universally available,despite both restriction endonucleaseanalysis and polymerase chain reactionbeing even more sensitive and specificthan antigen detection by immunoflu-orescence. Suitable clinical specimensinclude nasopharyngeal aspirate orswab, throat swab or wash, rectal swab,conjunctival swab or scrapings, stool,urine, cerebrospinal fluid, and tissue.Specimens should be obtained as earlyas possible after the onset of disease toincrease the chances of virus detection.

Although adenoviral infections gen-erally are self-limited and require nomore than supportive care, more severedisease can occur in immunocompro-mised patients, and antiviral agents

such as ribavirin and cidofovir havebeen used with inconsistent results.

Comment: The discovery of the clin-ical manifestations of adenovirus spansthe past century. Although viral strainsinitially were isolated in 1953, epidem-ics of keratoconjunctivitis were de-scribed in Austria as early as 1889,followed by descriptions of pharyngo-conjunctival fever in the 1920s andreports of adenovirus strains thatcaused enteritis in 1975. With such awide array of presentations, cliniciansneed to consider adenoviral infectionswhen evaluating all of these symptoms.Although immunocompetent childrenmay be asymptomatic or have self-limited disease, specific diagnosis ofadenovirus can be important in certain

circumstances. Because the signs andsymptoms of adenovirus may mimicKawasaki disease, differentiation be-tween Kawasaki and adenovirus is es-sential because of the time sensitivityof treatment for Kawasaki disease. Pre-cise diagnosis in immunocompromisedpatients, especially pediatric transplantpatients, can be equally important, as isdiagnosis during outbreaks or for hos-pital “cohorting” procedures. When di-agnosis is important, polymerase chainreaction and shell viral culture haveenhanced sensitivity over rapid immu-nofluorescence tests; practitioners needto be aware of the types of testsavailable in the laboratories they use.

Janet R. Serwint, MDConsulting Editor

In BriefToxic PlantsKevin Carter, MDDaniel R. Neuspiel, MD, MPHDepartment of PediatricsLevine Children’s Hospital of Carolinas

Medical CenterCharlotte, NC

Author Disclosure

Drs Carter, Neuspiel, and Serwint

have disclosed no financial





product/device.

2007 Annual Report of the AmericanAssociation of Poison Control Cen-ters’ National Poison Data System(NPDS): 25th Annual Report. Bron-

stein A, Spyker D, Cantilena L Jr,Green J, Rumack B, Heard S. ClinToxicol. 2008;46:927–1057

Toxic Plant Ingestions. Graeme KA. In:Wilderness Medicine. 5th ed. Phila-delphia, Pa: Mosby; 2007

Plants. Palmer M, Betz J. In: Goldfrank’sToxicologic Emergencies. 8th ed.New York, NY: McGraw-Hill; 2006

Toxic Mushroom Ingestions. SchneiderS, Donnelly M. In: Wilderness Medi-cine. 5th ed. Philadelphia, Pa: Mosby;2007

More than 60,000 calls are made an-nually to poison control centers (PCCs)for cases of suspected plant toxicity.Children younger than age 6 yearscomprise two thirds of cases, due totheir natural curiosity and limited judg-ment. Most of these exposures are be-nign; fewer than 10% result in treat-ment by a health professional. The PCC

is a valuable source of information toassist in management.

Most ingestions of plant material byyoung children are of small quantity,and symptoms, if present, typically areshort-lived and self-limited. Gastroin-testinal effects are common and maybe a clue to seek other, more subtlesigns of poisoning. Plant ingestions inolder children and adolescents gener-ally are intentional and of larger quan-tity, the result of either substance ex-perimentation or attempted self-harm.

Autonomic toxidromes can be seenin many plant poisonings. Deadly night-shade (Atropa belladonna) and Jimsonweed (Datura stramonium) produce at-ropine, scopolamine, and hyoscyamine,all anticholinergic toxins. Victims canpresent with classic symptoms of flush-ing, hyperthermia, blurred vision, drymouth, and hallucinations. Common

in brief



garden vegetables in the Solanum ge-nus, including tomatoes, potatoes, andeggplants, also can cause anticholin-ergic symptoms when blossoms or un-ripe buds are ingested. Physostigminemay be indicated to treat severe orpersistent symptoms.

A variety of central nervous system(CNS) responses follow plant ingestion.Hallucinations are common with mari-juana ingestion by children and withingestion of nutmeg or morning gloryseeds by teenagers. Tobacco plant ex-posure results in parasympatheticsymptoms (miosis, bronchorrhea, gas-trointestinal distress) as well as neuro-muscular derangement due to un-checked nicotinic receptor response.

Cardioactive glycosides are pro-duced by foxglove (Digitalis), but theyalso are found in lily of the valley(Convallaria) and oleander (Nerium andThevetia). Symptoms cannot be distin-guished from those of digoxin toxicityand include hyperkalemia, CNS depres-sion, and cardiac conduction abnormal-ities. Treatment with digoxin-specificantibody fragments can be lifesaving.

Potentially dangerous toxins canshow up in unexpected sources. Berriesof the holly and mistletoe plants, com-mon in holiday decorations, carry a riskof significant gastrointestinal distress.Amygdalin, contained in seeds and pitsof Prunus species fruits (cherries, apri-cots, peaches, apples, plums), generatescyanide when metabolized. The result-ing inhibition of cellular respiration canbe lethal.

Ingestion of mushrooms also mayhave fatal consequences. Species thatharbor amatoxins (Amanita) and re-lated compounds typically cause de-layed onset (6 hours) of nausea, vom-iting, and diarrhea. A second latentperiod is followed by acute and possiblyfulminant hepatitis beginning 48 to72 hours after ingestion. Effective de-contamination and therapies directedat the toxins generally are ineffective,and supportive care, including livertransplant if necessary, is the mainstayof therapy. Other species of mushroomscan cause hallucinations, muscarinictoxicity, or general gastrointestinal ir-ritation. Although most mushroom spe-cies are nontoxic, caretakers of a childwho has eaten or who is suspected ofeating any wild mushroom should callthe PCC for guidance.

The PCC can be an important aid inmedical decision making, particularlywith symptomatic patients for whomthe identity of the plant is unknown.Electronic transmission of digital im-ages may allow the PCC and expertbotanists to identify the offendingplant quickly and confidently and pro-vide data on managing the exposure.

Management of a potentially lethalexposure always should include com-munication with a toxicologist. In thesesituations, establishing control of air-way, breathing, and circulation shouldbe priorities. Aggressive decontamina-tion, with gastric emptying, activatedcharcoal, and possibly whole bowel ir-rigation, may be warranted.

Parents should be aware of thetypes of plants kept inside the homeas well as in any landscaping in theyard or neighborhood. Unknown plantsor shrubs, especially those that havebright colors or other features thatmight seem inviting to the curiouschild, can be identified with the help ofa local nursery.

Comment: Although pediatriciansmay consider drug ingestion readily inthe differential diagnosis for certainsigns and symptoms, I dare say we donot consider toxic plant exposures asoften as we should, especially inyounger children. Plant exposures inchildren 6 years of age and youngeraccounted for 4.6% of calls to PCCs,and this figure only represents in-stances when exposure was considered.The prevalence, therefore, is probablyunderestimated. Another important as-pect to consider is the overlap of pre-sentations between herbal remediesand plant exposures, another areawhere we need to expand our question-ing. Consideration of toxic plant expo-sures in our patients reinforces theimportance and value to parents andpediatricians of PCCs and the vastknowledge of diagnosis and treatmenttheir staff impart. The website http://www.aapcc.org contains useful infor-mation about poison centers.

Janet R. Serwint, MDConsulting Editor, In Brief

in brief




Kevin Carter and Daniel R. Neuspiel Toxic Plants











DOI: 10.1542/pir.31-4-e28 2010;31;e28-e35 Pediatr. Rev.

Ashlesha Kaushik, Carol Pineda and Helen Kest Diagnosis and Management of Dengue Fever in Children

http://pedsinreview.aappublications.org/cgi/content/full/31/4/e28located on the World Wide Web at:




http://pedsinreview.aappublications.org/cgi/content/full/31/4/e28


Diagnosis and Management of Dengue Feverin ChildrenAshlesha Kaushik, MD,*

Carol Pineda, MD,*

Helen Kest, MD, MPH,

CPH†

Author Disclosure

Drs Kaushik, Pineda,

and Kest have




commentary does not


of an unapproved/


commercial product/

device.


1. Describe the epidemiology and clinical spectrum of dengue viral infections.2. Recognize when to consider dengue in the differential diagnosis of acute fever.3. Discuss the diagnosis and management of this common tropical illness.4. Identify other diseases that can mimic dengue viral infections.

Case 1 PresentationA 17-year-old Hispanic girl presents with a 5-day history of temperature of 39.4°C to 40.5°Cand a 4-day history of severe bifrontal and intermittent headaches. She also has a 3-day historyof malaise, generalized body aches, and mild epigastric pain. On the day of admission, shedevelops a dark reddish-purple, nonpruritic, and nonblanching rash over her arms and thighsand is brought to the emergency department. There is no cough, sore throat, vomiting, ordiarrhea. She denies illicit drug use, tick exposure, sexual activity, or allergies.

On physical examination, the girl appears alert, oriented, and in no acute distress. Hertemperature is 38.5°C, heart rate is 119 beats/min, respiratory rate is 18 breaths/min, andblood pressure is 130/68 mm Hg (90th to 95th percentile). Capillary refill time is less than2 seconds. A petechial rash is present over her arms and anterior thighs. She has mild epigastrictenderness, with no rebound tenderness, guarding, hepatosplenomegaly, or masses. Tourniquettest is positive. Kernig and Brudzinski signs are negative. The remainder of the physicalexamination findings are normal.

Her white blood cell count is 3.5�103/mcL (3.5�109/L) with 59% neutrophils, 33%lymphocytes, 6% monocytes, and 1% eosinophils; hemoglobin is 13.4 g/dL (134 g/L); hematocritis 39.6% (0.396); platelet count is 126�103/mcL (126�109/L); and erythrocyte sedimen-tation rate is 13 mm/hour. The electrolytes, urinalysis, and coagulation profile are normal.The stool is negative for blood. Liver function test results include a protein concentration of6.9 g/dL (69 g/L), albumin of 4 g/dL (40 g/L), aspartate aminotransferase of 39 U/L,alanine aminotransferase of 18 U/L, alkaline phosphatase of 103 U/L, total bilirubin of0.9 mg/dL (15.4 mcmol/L), and direct bilirubin of 0.3 mg/dL (5.1 mcmol/L). Thick andthin smears are negative for malarial parasites.

The patient is admitted for monitoring, and intravenous hydration is started. On addi-tional questioning, she states that she had returned from the Dominican Republic yesterday.Additional serum testing reveals the diagnosis of classic dengue fever. Viral serology revealsa dengue virus immunoglobin M (IgM) enzyme-linked immunosorbent assay (ELISA) titerof 4.93 (negative, �1.11) and dengue virus IgG ELISA titer of 9.69 (negative, �1.11).Platelet counts and hematocrit values are monitored every day. On the second hospital day, herplatelet count declines to 96�103/mcL (96�109/L) with no evidence of bleeding. The countsincrease to 109�103/mcL (109�109/L) on the third hospital day and 121�103/mcL(121�109/L) on the following day. Her hematocrit value remains stable at 40% (0.40). Threedays after hospitalization, the fever resolves, the patient has stable vital signs, and she isdischarged.

Case 2 PresentationA 15-year-old boy who has a 5-day history of a temperature of 39.0°C to 39.5°C with chills andsevere pain in both legs is admitted to the hospital for evaluation. Three days ago, he developed

*Department of Pediatrics, St Joseph’s Children’s Hospital, Patterson, NJ.†Division of Pediatric Infectious Disease, Department of Pediatrics, St. Joseph’s Children’s Hospital, Patterson, NJ.

Article infectious diseases

e28 Pediatrics in Review Vol.31 No.4 April 2010 by Enrique Mendoza-Lopez on April 2, 2010 http://pedsinreview.aappublications.orgDownloaded from


nausea, headache, myalgias, and severe fatigue. He deniesany vomiting, somnolence, abdominal pain, or neck pain.Findings on past and family histories are unremarkable.

Physical examination reveals a tired-looking adolescentwho has a temperature of 39.4°C, heart rate of 95 beats/min, respiratory rate of 16 breaths/min, and blood pressureof 120/70 mm Hg (75th to 90th percentile). No neckstiffness, rash, or lesions are apparent. After the bloodpressure cuff is removed, he develops petechiae all over hisantecubital fossa. He has shotty cervical lymphadenopathy.All other physical findings are normal.

Initial laboratory results reveal a white blood cell countof 3.6�103/mcL (3.6�109/L) with 55% neutrophils, 29%bands, 14% lymphocytes, and 10% monocytes; hemoglobin of15.7 g/dL (157 g/L); hematocrit of 42.2% (0.422); andplatelet count of 87�103/mcl (87�109/L). Electrolytesand fibrinogen values are normal. Prothrombin time is15.2 seconds, international normalized ratio (INR) is 1.2,and activated partial thromboplastin time is 41.2 seconds.D-dimers are 3.3 mcg/mL (normal, �0.50 mcg/mL).Thick and thin smears for malarial parasites are negative.Liver function tests include a total protein of 6.2 g/dL(62 g/L), albumin of 3.4 g/dL (34 g/L), aspartate ami-notransferase of 587 U/L, alanine aminotransferase of412 U/L, alkaline phosphatase of 157 U/L, total bilirubinof 1 mg/dL (17.1 mcmol/L), and direct bilirubin of0.2 mg/dL (3.4 mcmol/L).

Additional questioning reveals that the boy had been inthe Dominican Republic for 2 weeks and returned the dayhis fever started. He is admitted with a diagnosis of sus-pected dengue hemorrhagic fever.

Viral studies sent on admission reveal an acute-phasedengue virus IgM ELISA titer of 1.8 (negative, �1.11)and a dengue virus IgG ELISA titer of 8.8 (negative,�1.11). On the second hospital day, he has one episode ofvomiting containing blood. Laboratory results reveal aplatelet count of 37�103/mcL (37�109/L), prothrombintime of 16.2 seconds, INR of 1.2, and partial thromboplas-tin time of 42.9 seconds.

On the third hospital day, due to another episode ofhematemesis and a low platelet count of 21�103/mcL(21�109/L), he receives a platelet transfusion. On thesame day, he develops diarrhea and mild lower abdominaltenderness. He has a hemoglobin of 18.3 g/dL (183 g/L)and hematocrit of 52.6% (0.53), which is a 20% increasefrom the baseline. Liver function tests yield normal resultsexcept for an albumin of 3.1 g/dL (31 g/L). Ultrasonog-raphy of the abdomen shows normal results, and stool ex-amination for parasites and culture produces negativeresults. Serum amylase and lipase values are normal. Onthe fifth hospital day, the diarrhea subsides and the patient

is afebrile. By the sixth hospital day, the child is feeling betterand has stable vital signs. His hematocrit and plateletcounts have improved, and he is discharged the followingday.

Case 3 PresentationAn 11-year-old Hispanic boy develops a temperature of39.1°C with diarrhea. Two days later, the diarrhea sub-sides, but he develops severe abdominal pain and vomiting,becomes increasingly lethargic, and is brought to the emer-gency department. His parents state that he has been com-plaining of severe back pain and headache. They deny anymedication use, previous hospitalization, or tick bites.

Physical examination reveals a sick-looking boy who issomnolent but arousable. His temperature is 40.0°C, heartrate is 140 beats/min, respiratory rate is 26 breaths/min,blood pressure is 80/50 mm Hg (�5th percentile), andGlasgow Coma Scale score is 12. He has a weak pulse, coldextremities, and a capillary refill time of 6 to 8 seconds. Hisabdomen is diffusely tender, with the liver enlarged 4 cmbelow the right costal margin. A petechial rash is presentover his chest and trunk. Meningeal signs are absent. Allother physical findings are unremarkable.

Initial laboratory results reveal a white blood cell countof 2.5�103/mcL (2.5�109/L) with 58% neutrophils, 8%bands, 10% lymphocytes, 18% atypical lymphocytes, and4% monocytes; hemoglobin of 14 g/dL (140 g/L); hemato-crit of 42% (0.42); and platelet count of 27�103/mcL(27�109/L). Electrolyte values are normal, and urinalysisshows trace blood. Coagulation profile reveals a prothrom-bin time of 17.8 seconds, activated partial thromboplastintime of 44 seconds, D-dimers of 4.4 mcg/mL (normal,�0.50 mcg/mL), and fibrinogen value of 200 mg% (nor-mal, 183 to 503 mg%). Liver function tests reveal a totalprotein of 5.6 g/dL (56 g/L), albumin of 3 g/dL (30 g/L),aspartate aminotransferase of 455 U/L, alanine amino-transferase of 324 U/L, alkaline phosphatase of 140 U/L,total bilirubin of 1.2 mg/dL (20.5 mcmol/L), and directbilirubin of 0.1 mg/dL (1.7 mcmol/L). Smears for malar-ial parasites are negative.

The parents share their concern about multiple relativeswho had self-limiting fevers and body aches in the Carib-bean, from where they had returned 4 days ago. The childhas been to the Caribbean twice within the past 2 years.

The clinical and laboratory picture suggest dengue shocksyndrome. The boy is admitted to the intensive care unitand started on intravenous hydration. On the second hos-pital day, he has two episodes of vomiting containing bloodand develops frank hematuria. Laboratory results show aplatelet count of 14�103/mcL (14�109/L), prothrombintime of 19.5 seconds, INR of 1.2, and partial thromboplas-

infectious diseases dengue fever

Pediatrics in Review Vol.31 No.4 April 2010 e29 by Enrique Mendoza-Lopez on April 2, 2010 http://pedsinreview.aappublications.orgDownloaded from


tin time of 52 seconds. His hematocrit is 33% (0.33), a 22%decrease from the initial value. He receives platelet andfresh frozen plasma transfusions. There is no recurrence ofbleeding, and by the fourth hospital day, he is hemodynam-ically stable and has improvement in hematocrit and plate-let counts (34% [0.34]and 90�103/mcL [90�109/L],respectively). On day 5, he continues to be afebrile, is feelingbetter, and is discharged. Viral studies sent on the thirdhospital day are strongly positive for dengue virus infec-tion, with a dengue virus IgM ELISA titer of 5.4 (nega-tive, �1.11) and dengue virus IgG ELISA titer of 12(negative, �1.11).

IntroductionAccording to the World Health Organization, about50 million dengue infections and 25,000 deaths occurworldwide annually, making dengue one of the mostimportant arthropod-borne viral diseases in humans. Allcontinents are endemic for dengue except Europe. Anestimated 2.5 billion of the world’s population live inareas at risk for epidemic dengue transmission, and itremains a leading cause of morbidity and mortalityamong children in some Asian countries. Most of thesevere cases and deaths occur in children younger than15 years of age. (1)(2)

The first reported epidemic occurred in the FrenchWest Indies in the 17th century, (3) but it was theSoutheast Asia pandemic created by the ecological dis-ruption that followed World War II that is credited for itsworldwide spread. Over the past several decades, thegradually increasing incidence has been attributed tomultiple factors, including global demographic changeswith associated uncontrolled urbanization and popula-tion growth, overcrowding with inappropriate sanita-tion, infrastructural problems, lack of preventive pro-grams for epidemic transmission, and poor mosquitocontrol efforts. (4)(5) Most cases in the United States areimported from other countries. Currently, dengue feveris the most common cause of fever in travelers returningfrom certain high-risk areas that include the Caribbean,Central America, and South Central Asia.

Areas bordering Mexico and southeastern states serveas niches for imported and locally acquired cases ofdengue due to population migration. According to theCenters for Disease Control and Prevention, Aedes ae-gypti and Aedes albopictus are the established vectors inthese areas and are a potential threat for dengue trans-mission throughout the United States.

Dengue fever is caused by the dengue virus and istransmitted by the bite of an infective female Aedesmosquito. The diagnosis is based on history, physical

findings, and laboratory markers. There are four majorclinical syndromes: 1) undifferentiated fever, 2) denguefever, 3) dengue hemorrhagic fever (DHF), and 4) den-gue shock syndrome (DSS). Most cases are mild. How-ever, DHF case fatality rates can reach 20% if not treatedappropriately or in a timely manner. It is highly likelythat dengue cases are unreported in the United Statesbecause physicians often do not include it in the differ-ential diagnosis of travelers returning from endemicareas. (6)(7)

PathogenesisDengue virus is an arbovirus of the flavivirus family thathas four different serotypes (DEN-1, -2, -3, and -4). Itsclassification is based on biologic and immunologic char-acteristics. (8) Because there is no cross-protection be-tween the different serotypes, lifetime immunity is ob-tained only after infection by each type. Therefore,persons living in endemic areas may be infected morethan once with different serotypes. Genetic variationwithin each serotype confers distinct virulence capacityand epidemic potential that may result in epidemics bythe same serotype in different years and locations. (5)After repeated infections, the chance of developing DHFand DSS increases. (9)

Mosquito CycleA aegypti is the primary vector responsible for transmis-sion; other vectors include A albopictus, A polynesiensis,and A niveus. A aegypti is primarily a daytime feeder. Itbreeds mainly in artificial water collections created bypoor sanitation or infrastructure such as jars, plates,flowerpots, glass containers, drainpipes, and cupboards.Although transmission is year round, the rainy seasoncreates ideal larval habitats and ecologically suitableniches for mosquito breeding and subsequent endemic-ity. (10)

The life cycle begins when an uninfected female mos-quito takes blood from an infected person during theviremic phase of illness. Within the mosquito’s digestivesystem, the virus replicates for 8 to 12 days (extrinsicincubation period). When this infective mosquito bitesagain, it transmits the virus to another person by inject-ing its salivary fluid. Once the virus is in the body, itreplicates in target organs and is released into the blood(intrinsic incubation period). Symptoms appear 3 to14 days after inoculation and may last up to 7 days ormore. Dengue should not be considered in the differen-tial diagnosis of a patient who develops fever more than2 weeks after leaving a dengue endemic area. (2)




Clinical Presentation of Dengue InfectionInfection with dengue viruses in children can have variedpresentations, ranging from asymptomatic to severeshock and death. Table 1 lists definitions of probable andconfirmed dengue syndromes. (2)(11)

Undifferentiated FeverPatients are mildly symptomatic, with nonspecific flulikesymptoms. This pattern usually occurs during a primaryinfection with dengue viruses and may be the mostcommon manifestation.

Dengue FeverClassic dengue fever is characterized by abrupt onset ofhigh-grade fever (temperature of 38.9°C to 40.6°C)associated with headache (especially retroorbital painthat worsens with eye movement), severe myalgia, ar-thralgia, nausea/vomiting, altered taste sensation (oftendescribed as metallic), and sometimes a rash. (2)(12)(13)The constellation of symptoms of severe and incapacitat-ing body ache, back pain, and arthralgia often is called“break bone fever.” Fever may last from 2 days to 1 week

and occasionally is described as having two peaks orbeing “saddle-backed,” that is, the initial 2 to 5 days offever are followed by 1 to 2 days of defervescence, afterwhich the temperature may rise again. (8)(14)(15)

Dengue fever rash may be erythematous, macular, ormaculopapular, and lymphadenopathy may be present.Infants and young children usually present with nonspe-cific symptoms such as fever, runny nose, rash, anddiarrhea; older children and adults have the classic breakbone fever, as described previously. Dengue fever canhave hemorrhagic manifestations without including theentire constellation of DHF. The hemorrhagic manifes-tations associated with dengue fever include a positivetourniquet test (Figure), petechiae/purpura, mucosalbleeding, and gastrointestinal bleeding. (8)(14) Thechild in the first case had petechiae and a positive tour-niquet test.

Most patients who have dengue fever recover un-eventfully. Rarely, patients may present with uncommonmanifestations such as seizures, paresis, meningitis, andmental status changes that can include lethargy, somno-lence, and coma.

Table 1. WHO and CDC Definitions of Dengue Clinical Syndromes (2)(11)

Type Clinical Case Definition

Dengue Fever Probable dengue fever: Fever of 2 to 7 days’ duration, with two or more of thefollowing:Headache, retroorbital pain, myalgia, arthralgia, rash, hemorrhagic manifestations,

leukopenia, and supportive serology or occurrence at the same location and time asother confirmed cases of dengue.

Confirmed dengue fever: Confirmed by laboratory criteria (isolation of the denguevirus, demonstration of the dengue virus antigen, serology, or genomic sequence).

Dengue Hemorrhagic Fever All of the following criteria must be fulfilled:(DHF) 1. Fever or history of acute fever, lasting 2 to 7 days, occasionally biphasic

2. Hemorrhagic manifestations in the form of at least one of the following:- A positive tourniquet test- Petechiae, ecchymosis, or purpura- Bleeding from the mucosa or injection sites- Hematemesis, melena, hematochezia, hematuria, increased menstrual flow

3. Thrombocytopenia (<100�103/mcL [100�109/L])4. Objective evidence of plasma leakage caused by increased vascular permeability, as

evidenced by one or more of the following:- A rise in the hematocrit (defined as >20% over baseline)- A drop in hematocrit following volume replacement treatment <20% of baseline- Low albumin or- Pleural effusion, ascites, or other effusions

Dengue Shock Syndrome DHF plus evidence of circulatory failure manifested by shock or all of the following:(DSS) - Rapid and weak pulse

- Narrow pulse pressure (<20 mm Hg) or hypotension for age (systolic pressure<80 mm Hg for children younger than 5 years of age or <90 mm Hg forchildren 5 years of age and older)

- Cold, clammy skin and altered mental status

CDC�Centers for Disease Control and Prevention, WHO�World Health Organization




Dengue Hemorrhagic FeverDHF is a potentially fatal illness marked by high fever,hemorrhagic manifestations, and evidence of plasma leak-age (Table 1). DHF begins with the sudden onset of a hightemperature that lasts 2 to 7 days, with accompanying chills,flulike constitutional symptoms, and a flushed face. As thefever subsides, patients may recover or progress to a state ofplasma leakage. Features of plasma leakage include ascites,pleural effusion (right-sided in most cases), and rarely,pericardial effusion associated with a high mortality. (8)(16)If untreated, the condition may deteriorate rapidly to pro-found shock and death within hours. (12) The hemorrhagicmanifestations of DHF include skin hemorrhages such aspetechiae, purpura, and ecchymoses; bleeding from mucousmembranes (epistaxis, gingival bleeding); and bleeding fromthe gastrointestinal, vaginal, and urinary tracts. These manifes-tations usually occur after the fever subsides, with the gastro-intestinal tract being the most common site of bleeding.

DHF is classified into four grades according to severity(Table 2). Laboratory abnormalities associated with DHFinclude thrombocytopenia (�100�103/mcL [100�109/L]), leukopenia, prolonged prothrombin time and acti-vated partial thromboplastin time, elevated fibrin degrada-tion products, low serum albumin, and elevated liverenzymes, as in the patient in case 2. Atypical lymphocytosis(�15%) and electrolyte abnormalities also may be seen.

Recovery from DHF usually is uneventful, marked by areturn of appetite and often a recovery rash (erythematouspetechial rash with islands of clearing). (2)(8) In a fewpatients, symptoms such as weakness and malaise may per-sist for several weeks after the acute illness has subsided.

Dengue Shock SyndromeMost patients who have DHF do not develop DSS. DSSoccurs during defervescence 3 to 6 days after the onset of

symptoms and has a high mortality rate of 10% to 47%.(2)(17) Warning signs of DSS include severe abdominalpain; persistent vomiting (with or without blood);abrupt change of temperature from fever to hypother-mia; and altered mental status, including irritability,somnolence, or obtundation. In DSS, capillary leakageand loss of intravascular volume result in shock ratherthan in hemorrhage (Table 1). (11)

Complications of Dengue InfectionsSevere dengue complications include liver dysfunction, en-cephalitis, cardiomyopathy (usually reversible), pancreatitis,acalculous cholecystitis, peripheral neuropathy, and acuterenal failure. (8)(13)(18) Liver involvement is one of themost important gastrointestinal manifestations, especiallywith infection by DEN-3 and DEN-4 serotypes, (19) and itcan present as acute hepatitis with elevated liver enzymevalues (aspartate aminotransferase being more significantlyelevated than alanine aminotransferase), jaundice, alteredmental status, seizures, and severe hypoglycemia. Transam-inase values are highest on day 9 of illness and normalizewithin 3 weeks. (18) Central nervous system (CNS) diseaseis attributed to various factors, including direct viral inva-sion of the CNS, liver failure, electrolyte disturbances, andcerebral edema. Some case series have reported a highmortality rate (up to 20%) in children who develop enceph-alopathy. (20) (21)

DiagnosisDengue infection can be diagnosed via serologic meth-ods, virus isolation, or molecular methods (Table 3).Table 4 shows the characteristics of similar infections thatshould be in the differential diagnosis of dengue fever.

Table 2. Grades of DengueHemorrhagic Fever (11)

Grades Definitions

Grade I Fever and nonspecific constitutionalsymptoms, with a positivetourniquet test being the onlyhemorrhagic manifestation

Grade II Grade I manifestations plusspontaneous bleeding

Grade III* Circulatory failure manifested asrapid /weak pulse, with cold skin,restlessness, and narrow pulsepressure or hypotension

Grade IV* Profound shock with nondetectablepulse or blood pressure

* Grades III and IV constitute dengue shock syndrome.Figure. Positive tourniquet test.




Table 3. Methods of Laboratory Diagnosis of Dengue InfectionDiagnostic Method Comments

Serology:● ELISA The IgM ELISA is the most common test for serologic diagnosis. Sensitivity is● Hemagglutination inhibition test 83.9% to 98.4% and specificity is 100%. IgM antibodies remain detectable● Complement fixation tests from day 5 to 4 to 5 weeks of illness. (2)(22)● Antigen capture enzyme

immunosorbent assayVirus Isolation: Virus isolation methods are employed to determine the serotype of the

● Mosquito cell cultures infecting virus. Because this procedure takes 2 weeks and is costly, it is● Mosquito inoculation:

Toxorhynchites amboinensisor Aedes albopictus areused commonly for inoculation

used primarily for research purposes. Mosquito inoculation technique ismore sensitive than cell cultures and is the preferred method of virusisolation. (23)

Molecular Methods:● RT-PCR for viral RNA

RT-PCR is a rapid method of diagnosis (allowing detection within 24 hours).It is more sensitive than virus isolation and useful in the early phase ofillness when antibodies are not circulating. However, this method is costlyand needs expertise. (2)(24)

ELISA�enzyme-linked immunosorbent assay, IgM�immunoglobulin M, RT-PCR�reverse transcription polymerase chain reaction

Table 4. Differential Diagnosis of Dengue FeverDisease Classic Signs and Symptoms Differentiating Features of Dengue

Influenza Fever, headache, myalgias, malaise, respiratory tractsymptoms. (25)

● Similar to undifferentiated fever form● Dengue should be included in differential

diagnosis when there is a history of travelto endemic area

Malaria High fever, chills, rigor, sweats that may beparoxysmal. Vomiting, diarrhea, cough, arthralgias,abdominal and back pain, hepatosplenomegaly,anemia, and thrombocytopenia are common. (26)

● Fever can have a biphasic pattern but isnot cyclic

● Thrombocytopenia may be present inboth; diagnostic studies are needed

Typhoid fever Fever, headache, malaise, anorexia, abdominal pain,hepatosplenomegaly, rose spots, altered mentalstatus. (27)

● Typical break bone features are absent● Severe disease may be difficult to

differentiate from DSS, and diagnosticstudies may be needed

Leptospirosis Initial phase: Fever, chills, headache, vomiting, transientrash, myalgias of calf and lumbar regions, andconjunctival discharge (nonpurulent). Second phase:meningitis, liver disease, and renal failure. (28)

● Biphasic presentation in leptospirosis andeye findings

● Usually exposure to animals/farms● Travel history to high-risk areas may be

absent in leptospirosisMeningococcemia Fever, chills, malaise, prostration, rash (macular,

maculopapular, petechial). Can progress to fulminantwith disseminated intravascular coagulation, purpura,shock, and death. (29)

● DHF or shock due to dengue may beundistinguishable from meningococcemia;it is reasonable to manage asmeningococcemia, and dengue fevershould be considered as a possibility inreturned travelers

Chikungunya Fever, rash (petechial or maculopapular) of trunk orlimbs, arthralgias, arthritis. Other: Headache,conjunctivitis, and photophobia.

Travel history is usually positive for Africa or Asia. (30)

● Febrile illness in dengue does not last aslong and is not followed by arthralgicdisease, which is prolonged inchikungunya

Rubella Rash, posterior auricular or suboccipitallymphadenopathy, headache, conjunctivitis,polyarthritis. (31)

● Rash does not have the cephalocaudaldistribution of rubella; similar toundifferentiated fever form

DHF�dengue hemorrhagic fever, DSS�dengue shock syndrome




TreatmentTreatment is supportive. Fever is controlled with acetamin-ophen. Nonsteroidal anti-inflammatory agents should beavoided due to their anticoagulant properties and risk ofReye syndrome in children. Most cases of dengue feverare mild and occur as undifferentiated fever or classicdengue fever.

Early recognition and treatment decreases morbidityand mortality. Home therapy with adequate fluid intakeand bed rest should be reinforced. Patients do not haveto be admitted to the hospital or receive intravenousfluids unless they present with severe vomiting, dehydra-tion, bleeding, altered mental status, clinical deteriora-tion, or evidence of DHF or DSS. Patients who haveDHF and can be managed as outpatients include thosewho have platelet counts of at least 50�103/mcL(50�109/L), no active bleeding (besides the petechiae),and a hematocrit that is not elevated. Patients who haveDHF and those who are in shock should be treated inan intensive care setting. Hematologic, cardiovascular,and fluid and electrolyte status should be observed andsupported. The platelet transfusion threshold in DHFis controversial, and transfusion is required only in pa-tients who have severe thrombocytopenia or hemor-rhagic manifestations. (8) When dengue is considered ina differential diagnosis, acute-phase (0 to 5 days) andconvalescent-phase samples (14 to 21 days) should becollected and sent for viral isolation and serology.

Prevention/Infection ControlPretravel counseling for all patients visiting endemicregions should emphasize measures to prevent human-vector contact. Using repellants containing N,N-diethyl-3-methylbenzamide (DEET), wearing protec-tive clothing (long-sleeved shirts and pants) during themosquito-biting period (morning and afternoon), andusing bed nets can minimize mosquito bites. Insectrepellants can be used safely in children older than 2months of age. Travelers also can reduce their risk bystaying in screened or air-conditioned areas when possi-ble and avoiding potential mosquito breeding sites.Eliminating mosquito breeding by covering water con-tainers and eliminating standing water can prevent thetransmission of dengue virus.

Worldwide epidemic control involves aggressive initi-atives targeting prevention of dengue transmission andincludes educating the medical community and improv-ing public health infrastructure that provides integratedA aegypti control and an active, laboratory-based surveil-lance system that has a rapid response contingency planfor epidemic prevention. However, factors such as air

travel, global migration, and climate change continue toaffect this global reemergence. (32) Dengue nevershould be overlooked in countries whose prevalence islow, such as the United States. Dengue fever continuesto appear, and the possibility of epidemics exists evenafter several years of sporadic outbreak.

References1. World Health Organization. Dengue and Dengue HemorrhagicFever. Publication No. 117. Geneva, Switzerland: World HealthOrganization; 20082. Division of Vector Borne Infectious Disease. Dengue Fever.Atlanta, Ga: Centers for Disease Control and Prevention; 20083. Howe GM. World Geography of Human Diseases. New York,NY: Academic Press; 19774. Guzman MG, Kouri G. Dengue: an update. Lancet Infect Dis.2002;2:33–425. World Health Organization. Initiative of vaccine research:vector-borne viral infections. The World Health Report. 20036. Centers for Disease Control and Prevention. Travel-associateddengue—United States, 2005. MMWR Morb Mort Wkly Rep. 2006;55:700–7027. Centers for Disease Control and Prevention. Travel-associateddengue infections, 2001–2004. MMWR Morb Mort Wkly Rep.2005;54:556–5588. Malavige GN, Fernando S, Fernando DJ, Seneviratne SL. Den-gue viral infections. Postgrad Med J. 2004;80:588–6019. Halstead SB. Dengue haemorrhagic fever—a public healthproblem and a field for research. Bull WHO. 1980;58:1–2110. Promprou S, Jaroensutasinee M, Jaroensutasinee K. Climaticfactors affecting dengue hemorrhagic fever: incidence in southernThailand. Dengue Bulletin. 2005;2911. World Health Organization. Dengue in the Context of Inte-

Summary• Today, dengue is considered among the most

important arthropod-borne viral diseases in humans.It is transmitted by the bite of an infective femaleAedes mosquito. (2)(6)

• According to the World Health Organization andCenters for Disease Control and Prevention, theincidence of dengue in the United States isincreasing and may be underreported due toinadequate disease recognition and low index ofsuspicion. (1)(7)

• Children younger than 15 years of age are at highestrisk for severe disease and death. (2)(25)

• The spectrum of dengue viral infections includesfour categories: undifferentiated fever, dengue fever,DHF, and DSS. (1)(11)

• DHF and shock forms should be managedaggressively in an intensive care setting to preventmorbidity and mortality. (2)(12)

• Dengue should be considered in the differentialdiagnosis of any child who develops fever within2 weeks of travel to endemic areas. (2)




grated Management of Childhood Illness. Geneva, Switzerland:World Health Organization; 200512. World Health Organization. Dengue Haemorrhagic Fever:Diagnosis, Treatment, Prevention and Control. 2nd ed. Geneva,Switzerland: World Health Organization; 199713. World Health Organization. Regional Guidelines of Dengue:DHF Prevention and Control. Geneva, Switzerland: World HealthOrganization; 199914. Ahmed FU, Mahmood CB, Sharma JD, et al. Dengue feverand dengue haemorrhagic fever in children: the 2000 outbreak inChittagong, Bangladesh. Dengue Bulletin. 2001;25:33–3915. Narayanan M, Aravind MA, Thilothammal N, et al. Denguefever epidemic in Chennai–a study of clinical profile and outcome.Indian Pediatr. 2002;39:1027–103316. Kalayanarooj S, Chansiriwongs V, Nimmannitya S. Denguepatients at the Children’s Hospital, Bangkok: 1995–1999. DengueBulletin. 2002;26:33–4317. Kabra SK, Jain Y, Pandey RM, et al. Dengue haemorrhagicfever in children in the 1996 Delhi epidemic. Trans R Soc Trop MedHyg. 1999;93:294–29818. Gulati S, Maheshwari A. Atypical manifestations of dengue.Trop Med Int Health. 2007;12:1087–109519. Kalayanarooj S, Nimmannitya S. Clinical and laboratory pre-sentations of dengue patients with different serotypes. DengueBulletin. 2000;24:53–5920. Cam BV, Fonsmark L, Hue NB, et al. Prospective case-controlstudy of encephalopathy in children with dengue hemorrhagicfever. Am J Trop Med Hyg. 2001;65:848–85121. Lum LC, Lam SK, Choy YS, et al. Dengue encephalitis: a trueentity? Am J Trop Med Hyg. 1996;54:256–25922. Guzman MG, Kouri G, Soler M. Dengue 2 virus enhancementin asthmatic and non-asthmatic individuals. Mem Inst OswaldoCruz. 1992;87:559–56423. Guzman MG, Kouri G. Advances in dengue diagnosis. ClinDiagn Lab Immunol. 1996;3:621–62724. De Paula SO, Pires Neto RJ, Correa JA, et al. The use of reverse

transcription polymerase chain reaction (RT-PCR) for the rapiddetection and identification of dengue virus in an endemic region: avalidation study. Trans R Soc Trop Med Hyg. 2002;96:266–29925. American Academy of Pediatrics. Influenza. In: Pickering LK,Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009 Report ofthe Committee on Infectious Diseases. 28th ed. Elk Grove Village, Ill:American Academy of Pediatrics; 2009:400–41226. American Academy of Pediatrics. Malaria. In: Pickering LK,Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009 Report ofthe Committee on Infectious Diseases. 28th ed. Elk Grove Village, Ill:American Academy of Pediatrics; 2009:438–44427. American Academy of Pediatrics. Epidemic typhus. In: Pick-ering LK, Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009Report of the Committee on Infectious Diseases. 28th ed. Elk GroveVillage, Ill: American Academy of Pediatrics; 2009:711–71228. American Academy of Pediatrics. Leptospirosis. In: PickeringLK, Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009 Reportof the Committee on Infectious Diseases. 28th ed. Elk Grove Village,Ill: American Academy of Pediatrics; 2009:427–42829. American Academy of Pediatrics. Meningococcal infections.In: Pickering LK, Baker CJ, Kimberlin DW, Long SS, eds. RedBook: 2009 Report of the Committee on Infectious Diseases. 28th ed.Elk Grove Village, Ill: American Academy of Pediatrics; 2009:455–46330. American Academy of Pediatrics. Arboviruses. In: PickeringLK, Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009 Reportof the Committee on Infectious Diseases. 28th ed. Elk Grove Village,Ill: American Academy of Pediatrics; 2009:214–22031. American Academy of Pediatrics. Rubella. In: Pickering LK,Baker CJ, Kimberlin DW, Long SS, eds. Red Book: 2009 Report ofthe Committee on Infectious Diseases. 28th ed. Elk Grove Village, Ill:American Academy of Pediatrics; 2009:579–58432. Patz JA, Martens WJM, Focks DA, Jetten TH. Denguefever epidemic potential as projected by general circulationmodels of global climate change. Environ Health Perspect. 1998;106:147–153




DOI: 10.1542/pir.31-4-e28 2010;31;e28-e35 Pediatr. Rev.

Ashlesha Kaushik, Carol Pineda and Helen Kest Diagnosis and Management of Dengue Fever in Children


http://pedsinreview.aappublications.org/cgi/content/full/31/4/e28including high-resolution figures, can be found at:









PIR_abril_2010

Documents

location of pain

pain emanating

type of pain

somatic abdominal pain

medical abdominal pain

source of abdominal

visceral pain receptors

somatoparietal pain