Symptom-Based Diagnosis in Pediatrics (CHOP Morning Report) 1st Ed.

CASE 17-2

Two-Year-Old Boy



The patient is a 2-year-old boy, who had been previously well until 3 days prior to presentation to the hospital with watery diarrhea and decreased appetite. The next day, he also developed vomiting at which point he was seen by his primary physician who treated him with trimethobenzamide hydrochloride (Tigan) suppositories which provided no relief. The symptoms progressed to 20 episodes of diarrhea, now with blood and mucus and abdominal cramping on the day of presentation. The patient was admitted to an outside hospital for presumed bacterial gastroenteritis. There were no known ill contacts, no known ingestion of undercooked foods, no recent travel, and no recent exposure to antibiotics.


The boy’s history was significant only for one hospital admission for an asthma exacerbation. There was no surgical history, no medications. His mother has a history of irritable bowel syndrome.


T 37.5°C; RR 30/min; HR 150 bpm; BP 105/50 mmHg

Weight 50th percentile

The patient was alert, but quiet and ill appearing. His eyes were mildly sunken. He had dry mucous membranes. There was no lymphadenopathy. The remainder of his head and neck examination was unremarkable. His cardiac examination revealed tachycardia, but no murmur or gallop. There was tachypnea, but no rales or wheezing. His neurologic examination was nonfocal. His skin turgor was diminished.


Complete blood count revealed a WBC count of 16 600 cells/mm3 with a differential of 62% segmented neutrophils, 24% band forms, 10% lymphocytes, 3% monocytes, and 1% atypical lymphocytes. Hemoglobin was 14.2 g/dL and platelet count was 381 000 cells/mm3. White blood cells were present on Gram stain of the stool. Routine bacterial and viral stool cultures were sent which eventually returned negative. Serum electrolytes were significant for a chloride of 100 mmol/L and a bicarbonate of 19 mmol/L. Blood urea nitrogen was 73 mg/dL and creatinine was 1.2 mg/dL.


The patient was admitted to the hospital for intravenous fluid rehydration and given nothing by mouth after failing of repeated oral clear liquid challenges while continuing to have diarrhea. A Foley catheter was placed to monitor urine output (Figure 17-2). On the second day of hospitalization the patient developed a fever to 39°C with increased abdominal pain, particularly in the periumbilical region and right lower quadrant. His abdominal examination now revealed hypoactive bowel sounds and an abdominal radiograph demonstrated ileus. An abdominal ultrasound showed a small amount of ascites. Ampicillin, gentamicin, and clindamycin were started intravenously. A sigmoidoscopy was performed to 50 cm, demonstrating a normal appearing colon. A biopsy was obtained which showed signs of chronic inflammation. Later that day, the patient was taken for an exploratory laparotomy due to worsening abdominal pain and signs of an acute abdomen. Surgery revealed a leathery thickening of the descending colon, possibly consistent with chronic inflammation, without involvement of the transverse colon or distal sigmoid. A large amount (500 mL) of clear yellow ascitic fluid was also removed. A central venous catheter was placed, given the anticipated need for prolonged intravenous fluids and medications.


FIGURE 17-2. Photo of the patient’s leg (note the poor perfusion) that also includes a portion of the urinary catheter drainage.



The key to this diagnosis is having the appropriate level of suspicion in the right clinical context. During the initial prodrome of bloody diarrhea, any of the other bacterial causes of enteroinvasive diarrhea would be on the list of differential diagnoses. SalmonellaShigella, and Campylobacter species would be among some of the etiologies. In fact, HUS has been associated with Shigella infection, as a result of elaboration of Shiga toxin. All of these bacterial causes would be routinely screened for in most stool cultures. It is important to remember that E. coli O157:H7 may not be routinely screened for, although it may now be the most common cause of bloody diarrhea in the United States. Clostridium difficile colitis could also present in this manner and testing for toxins A and B would be important in distinguishing this entity, as would a prior history of recent antibiotic use.

Inflammatory bowel disease could also present in this fashion, especially as an acute flare with signs of systemic toxicity and abdominal symptoms severe enough sometimes to warrant surgical exploration.


The Foley catheter revealed bloody urine (Figure 17-2). The patient’s lower extremity, also included in the photograph, revealed evidence of poor perfusion. Postoperatively the patient was noted to be edematous while on maintenance intravenous fluids. Repeat laboratory studies revealed a serum sodium of 133 mmol/L, chloride 114 mmol/L, bicarbonate 11 mmol/L, blood urea nitrogen 24 mg/dL, and creatinine 3.0 mg/dL. The hemoglobin decreased to 11.3 g/dL and the platelet count was 118 000 cells/mm3. The peripheral smear showed the presence of schistocytes, suggesting a hemolytic process. Liver function tests were significant for an aspartate aminotransferase of 532 U/L, alanine aminotransferase 287 U/L, lactate dehydrogenase 4290 U/L, and albumin 1.7 g/dL. The partial thromboplastin time was slightly elevated at 37.4 seconds. During the next 3 hours, the patient’s urine output stopped completely. He was given 25% albumin intravenously with intravenous fluids followed by furosemide which resulted in only 10 cc of urine output. At this point, the patient was transferred to our institution for continuous arteriovenous hemoperfusion. A stool culture sent specifically for detection of E. coli O157:H7 returned positive. This culture confirmed the diagnosis of hemolytic uremic syndrome (HUS) secondary to infection with E. coliO157:H7, otherwise known as D(+) HUS, in that it is associated with diarrhea. During his 4-week hospitalization, his blood urea nitrogen level peaked at 101 mg/dL and creatinine at 9.9 mg/dL. His hemoglobin nadir was 4.7 g/dL and platelet count nadir was 103 000/mm3. He required continuous arteriovenous hemoperfusion for most of the remainder of his hospitalization. Upon discharge, his blood urea nitrogen was 74 mg/dL and his creatinine was 5.5 mg/dL.


HUS is one of the most common causes of acute renal failure in children, and in turn, E. coli O157:H7 infection is the most common cause of HUS. As a member of the enterohemorrhagic group of E. coli, it causes a hemorrhagic colitis and elaborates a verotoxin that is similar to Shiga toxin, produced by Shigella dysenteriae type 1. In the United States, it appears to be one of the most common causes of bloody diarrhea. Cases can occur on a sporadic as well as epidemic basis, which has appeared to be increasing in the last decade. There is a seasonal pattern to infection, in that cases are more common in the summer months, although cases do sometimes occur in the winter. This may also be associated with the fact that cattle and their undercooked meat products or unpasteurized dairy products are the most significant factors in transmission and are more likely to be consumed during the summertime.

The highest attack rates occur at the extremes of the age spectrum—the very old and the very young appear to be at greatest risk. In particular, children younger than 4 years of age are at great risk of contracting E. coli O157:H7 infection and are also at greater risk for developing HUS as a result. HUS occurs in approximately 10% of patients who acquire sporadic infection with E. coli O157:H7 and appears to occur at a higher attack rate of up to 40% of patients infected in an outbreak. The mortality rate is approximately 5% and another 5% of patients suffer from severe neurologic sequelae or end-stage renal failure. Extremely poor prognostic factors include a high leukocyte count, a severe gastrointestinal prodrome, age younger than 2 years, and early onset of anuria.


Infection with E. coli O157:H7 begins as a non-bloody diarrhea that progresses to bloody diarrhea, often a few days after the onset of illness. Vomiting occurs in half of the patients and a fever in approximately one-third. Fecal leukocytes are often present. The hemorrhagic colitis is otherwise nonspecific and may appear indistinguishable clinically from many other colitides. Abdominal symptoms may become severe enough to mimic an acute abdomen, prompting surgical intervention. Complications include intussusception, perforation, or stricture. The classic triad of microangiopathic hemolytic anemia, thrombocytopenia, and acute renal failure follows the gastrointestinal prodrome. Leukocytosis develops, as can cerebral edema with seizure activity. Irritability progressing to stupor and coma may also develop. Cortical blindness and stroke may occur as well.

HUS may present in a D(–) form which is not associated with diarrhea. Cases of D(–) HUS, otherwise known as atypical HUS, have been associated with medications, familial patterns of inheritance, recurrent episodes of HUS, and bacterial infections, such as Streptococcus pneumoniae. In general, patients with D(–) HUS have a higher incidence of neurologic sequelae in addition to the renal sequelae and overall have worse outcomes than patients with D(+)HUS.

HUS is occasionally associated with S. pneumoniae. Children with pneumococcal HUS typically present with pneumonia complicated by empyema. Blood cultures are typically positive. In one multicenter case series of children with pneumococcal HUS, 4 of 37 patients have concomitant pneumococcal meningitis. Up to one-third of children with pneumococcal HUS have residual renal dysfunction, including elevated creatinine, proteinuria, and hypertension.


Clinical and laboratory findings. The key to diagnosis is recognition of the clinical syndrome in the appropriate context. An abnormal complete blood count with thrombocytopenia and anemia with schistocytes and red blood cell fragments on the peripheral smear in the context of fluid overload and oliguria or anuria with a rising serum creatinine confirm the diagnosis of this syndrome.

Microbiology. Stool culture is important to rule out other infectious causes of colitis, but specific biochemical tests are necessary to identify the specific serotype, O157:H7, and in some laboratories, may not be routinely performed and must be specifically requested.

Enzyme immunoassays. New, rapid methods to detect E. coli O157:H7 lipopolysaccharide and shiga toxin are very useful in diagnosing infection in a timely manner.


Current therapy remains supportive in nature; however, prevention is essential by ensuring complete cooking of all beef products, particularly ground beef, thorough hand-washing when interacting with animals which may be carriers, and avoidance of unpasteurized dairy or other products. Patients who develop HUS require significant volume-support with careful fluid and electrolyte management. Transfusion may be necessary due to significant gastrointestinal blood loss and microangiopathic hemolytic anemia. Hypertension and renal failure must be anticipated in the clinical management and patients often require dialysis. Neurologic complications, such as seizures, may require antiepileptic therapy. There is no evidence that antibiotics are helpful and they may be potentially harmful, by causing increased release of toxin. Antimotility agents are also contraindicated because of the increased absorption of toxin.


1. Loirat C, Fremeaux-Bacchi V. Atypical hemolytic uremic syndrome. Orphanet J Rare Dis. 2011;6:60.

2. Zoja C, Buelli S, Morigi M. Shiga toxin-associated hemolytic uremic syndrome: pathophysiology of endothelial dysfunction. Pediatr Nephrol. 2010;2(11):2231-2240.

3. Bitzan M, Schaefer F, Reymond D. Treatment of typical (enteropathic) hemolytic uremic syndrome. Semin Thromb Hemost. 2010;36(6):594-610.

4. Malina M, Orumenina LT, Seeman T, et al. Genetics of hemolytic uremic syndromes. Presse Med. 2012;41(3Pt2): e10-e14.

5. Scheiring J, Rosales A, Zimmerhackl LB. Clinical practice: today’s understanding of haemolytic uremic syndrome. Eur J Pediatr. 2010;169(1):7-13.

6. Palerma MS, Exeni RA, Fernandez GC. Hemolytic uremic syndrome: pathogenesis and update of interventions. Expert Rev Anti Infect Ther. 2009;7(6): 697-707.

7. Copelovitch L, Kaplan BS. Streptococcus pneumoniae-associated hemolytic uremic syndrome. Pediatr Nephrol. 2008;23(11):191-196.

8. Iijima K, Kamioka I, Nosu K. Management of diarrhea-associated hemolytic uremic syndrome. Clin Exp Nephrol. 2008;12(1):16-19.

9. Banerjee R, Hersh AL, Newland J, et al. Streptococcus pneumoniae-associated hemolytic uremic syndrome among children in North America. Pediatr Infect Dis J. 2011;30:736-739.

10. Waters AM, Kerecuk L, Luk D, et al. Hemolytic uremic syndrome associated with invasive pneumococcal disease: the United Kingdom experience. J Pediatr. 2007;151:140-144.