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

CASE 9-3

Four-Year-Old Girl



A 4-year-old Caucasian girl presented for evaluation of an eruption on her lower extremities. Four days prior to admission, she presented to her primary physician for evaluation of left knee pain. At that time she had mild swelling and tenderness of the joint without erythema. She was started on ibuprofen with some relief. One day later she developed abdominal pain, and was evaluated at a nearby hospital. Physical examination and abdominal radiograph with obstructive series were normal. Rectal examination was normal with negative stool guaiac. The patient was sent home, but the severe, intermittent abdominal pain persisted. On the day of admission the child’s mother noted the development of “small red bumps” all over both legs while giving the child a bath. The lesions became larger and continued to spread on her lower extremities throughout the day. There was no history of fever, vomiting, or diarrhea. The patient ate well between episodes of abdominal pain. She had some nasal congestion and rhinor-rhea 1 week prior to the onset of her symptoms. There was no history of tick bites or ingestions.


The child was a full-term infant with an uncomplicated birth history. There was no significant family history, travel, or exposures.


T 37.7°C; RR 22/min; HR 124 bpm; BP 98/65 mmHg

Height 50th percentile; Weight 50th percentile

Physical examination was remarkable for 2-10 mm nonblanching, erythematous papules on the lower extremities from the dorsum of her feet to waist bilaterally. A horizontal 4-cm wide purpuric plaque on the left leg at the sock line was also noted (Figure 9-3). Lung examination was normal. Cardiac examination revealed no murmurs, rubs, or gallops. There were active bowel sounds, a soft nondistended abdomen, and no hepatosplenomegaly. Rectal examination was normal with soft heme-negative stool in the vault. There was no prominent adenopathy. Joint and neurologic examinations were normal.


FIGURE 9-3. Rash on the lower extremities.


Laboratory analysis revealed 10 700 WBCs/mm3 with 50% segmented neutrophils, 40% lymphocytes, 4% eosinophils, and 6% monocytes. The hemoglobin was 10.9 g/dL and there were 436 000 platelets/mm3. Electrolytes, blood urea nitrogen, and creatinine were normal. Erythrocyte sedimentation rate was 22 mm/h. Urinalysis demonstrated a specific gravity of 1.010, with no protein, white blood cells, or red blood cells. Blood cultures were obtained.


The patient was admitted to the hospital where she received medication to manage her abdominal pain. The distribution and character of the rash suggested the diagnosis (Figure 9-3).



The differential diagnosis of a nonblanching eruption includes both endogenous and exogenous causes (Table 9-5). When isolated to the lower extremities, ecchymoses due to trauma in a healthy child must be considered. However, extent and bilaterally symmetric nature of the lesions, as well as the timing of the eruption in relation to her symptoms makes this less likely. Purpura can be seen in a variety of coagulopathies and in association with joint and gastrointestinal pain may reflect underlying hematologic malignancy (leukemia), idiopathic thrombocytopenic purpura, hemolytic uremic syndrome, or septicemia. Purpura due to meningococcemia or rickettsial disease should also be considered. Screening studies including normal complete blood count, blood urea nitrogen, and creatinine mitigate these diagnoses. Vasculitis in association with infection or underlying inflammatory or rheumatologic disease, such as lupus, must be considered. Although joint complaints can be seen in setting of juvenile idiopathic arthritis, the morphology of her skin lesions was not the evanescent pink lesions seen in this condition.

TABLE 9-5. Select causes of acute petechiae/purpura in children.



Rickettsial disease




  Idiopathic thrombocytopenic purpura (ITP)

  Hemolytic uremic syndrome (HUS)

  Thrombotic thrombocytopenic purpura (TTP)



  Vitamin K deficiency

  Hepatic dysfunction

  Disseminated intravascular coagulation

Prothrombotic state

  Protein C, S deficiency







  ANCA associated vasculitis


  Viral (e.g., Hepatitis, HIV)


  Henoch-Schölein Purpura (HSP)

   Acute hemorrhagic edema of infancy (AHEI)


Pigmented purpuric dermatoses


Traumatic purpura

Nonaccidental trauma

Self-inflicted trauma/factitial disorder


The characteristic palpable purpuric lesions on the lower extremities, especially along the sock line of the left leg (Figure 9-3) along with the constellation of symptoms in this patient led to the clinical diagnosis of Henoch-Schönlein purpura (HSP), also known as anaphylactoid purpura . One of the most common vasculitic syndromes affecting children, HSP is characterized by purpuric papules and plaques on the buttocks and lower extremities, polyarthritis, colicky abdominal pain, nephritis, or any combination of these symptoms. It is an IgA mediated leukocytoclastic vasculitis that involves inflammation of the precapillary, capillary, and postcapillary vessels in the skin, joints, gastrointestinal tract, and kidneys.


HSP can occur at any age, with peak incidence at 4-5 years; it is less common before age 1 or after age 10. Variable sex and ethnicity predominance has been reported. The incidence increases in the fall suggesting an infectious trigger and variable viral and bacterial infections have been described as inciting events.


The clinical picture of HSP is similar to this case: a previously healthy child acutely develops a distinctive eruption often preceded by arthritis or arthralgias, and/or abdominal pain. The rash is seen in all patients, and half the patients initially present with the skin rash. The eruption often occurs on the buttocks, lower extremities, and other areas of dependency. In severe cases it may include the face, trunk, and arms. The classic lesions are nonblanching erythematous macules and papules that last days with recurrences for up to 1-2 months. Nonpitting edema may be found in dependent areas. Arcuate purpura with a facial predominance may be a manifestation of acute hemorrhagic edema of infancy, a primarily cutaneous vasculitis that mimics HSP in younger children.

Arthralgia or arthritis are seen in 65%-85% of patients and may precede the eruption in 25% of the patients. The large joints are most commonly affected. These symptoms usually resolve within a few days.

Gastrointestinal manifestations can be seen in as many as three-fourths of patients. Abdominal pain is usually colicky in nature. Pain may be severe, mimicking appendicitis. Vomiting is common. Abdominal symptoms occasionally precede other manifestations by up to several weeks. While stools often contain occult blood, massive gastrointestinal bleeding occurs rarely. Other gastrointestinal complications include intussusception, bowel infarction, perforation, pancreatitis, and hydrops of the gallbladder. Intussusception, which occurs in 3% of cases, is usually ileo-ileal (rather than ileo-colic) and is seen in older children.

Renal manifestations range from transient hematuria or proteinuria to nephritis, nephrotic syndrome, and end-stage renal disease. While microscopic hematuria is a constant feature, up to 40% of patients have gross hematuria. Proteinuria occurs in conjunction with hematuria in two-thirds of patients, but proteinuria alone is rarely, if ever, a manifestation of HSP nephritis. Nephritis develops in 20%-50% of children with HSP. It usually appears days or weeks after the initial symptoms of HSP. Though it usually resolves, 2% may progress to renal failure. The histopathology of HSP renal disease demonstrates an IgA nephropathy. Those children who present with renal disease with nephritic and nephrotic features are at the highest risk to develop end stage renal disease.

Neurologic findings, including headaches and behavioral changes are seen in about one-third of the cases. Seizures, focal neurologic deficits, and peripheral neuropathies are rare. These symptoms may be seen in 2%-8% of those with the disease. Most neurologic findings are transient.

Orchitis develops in up to 10% of boys. Symptoms may mimic testicular torsion. Recurrences occur in a minority of patients and are manifested by appearance of symptoms after a period of resolution of 2 weeks or greater. Gastrointestinal and skin symptoms are most common.


The diagnosis is made clinically, as there are no specific laboratory tests to diagnose HSP. Laboratory studies may reveal suggestive features or complications of HSP but are more important to help differentiate HSP from other potential causes.

Complete blood count. Leukocytosis is seen in two-thirds of the children. Eosinophilia is occasionally present. Anemia is more commonly seen with Crohn disease and ulcerative colitis than HSP.

Serum IgA level. Serum IgA is elevated during the acute illness in approximately 50% of patients.

Complement levels. C3 and C4 levels are normal in HSP nephritis but may be low in lupus nephritis and poststreptococcal glomerulonephritis.

Urinalysis. Urinalysis should be performed on all the patients to look for findings suggestive of nephritis or nephrotic syndrome, including hematuria and proteinuria. Urinalysis should be repeated every 7 days while the disease is still active to monitor for the development of nephritis and renal function should be monitored for 3-6 months after disease onset given cases of late onset disease.

Radiologic studies. When gastrointestinal signs are present, contrast radiologic studies of the gastrointestinal tract demonstrate small bowel involvement with thickened folds, hypomotility and “thumb-printing,” characteristic of submucosal hemorrhage. Ultrasound may help diagnose intussusception, as intussusception proximal to the ileo-colic location may be missed by air or liquid contrast enemas.

Other studies. The erythrocyte sedimentation rate (ESR) may be elevated but is rarely helpful in diagnosing HSP. The ESR is always elevated with Crohn disease and ulcerative colitis. The prothrombin time and partial thromboplastin time are normal in children with HSP but are frequently elevated in children with meningococcemia, sepsis, and other coagulation disorders presenting with purpura. Blood cultures should be performed in ill-appearing children when HSP cannot be readily distinguished from meningococcemia or sepsis. Renal biopsy should be considered in children with nephritis complicated by nephrotic syndrome, hypertension, or substantial renal insufficiency to predict prognosis more precisely. Skin biopsy for routine histology and immunofluorescence in severe or atypical cases is warranted: unlike other small vessel vasculitic processes, HSP classically manifests with cutaneous IgA and C3 deposition on immunofluorescence.


Specific treatment is not usually required and most children can be managed in the outpatient setting. Hospital admission should be considered in children who develop gastrointestinal hemorrhage, colicky abdominal pain suggestive of intussusception, significant renal disease, or changes in mental status. The duration of illness is typically 3-4 weeks. Relapses are uncommon but have occurred in up to 15% of children in some studies.

The use of prednisone (1-2 mg/kg/day) has been suggested to hasten improvement of abdominal pain as well as improve outcomes in hospitalized patients with HSP. This may also reduce the risk of intussusception. Some reports suggest that corticosteroids alone or combined with other immunosuppressive agents, such as azathioprine or cyclophosphamide, reduce the risk of renal failure in children with HSP nephritis, although prospective trials are still needed.

The prognosis for children who develop HSP is generally good. Approximately 95% of children have complete recovery. However, of those children who required renal biopsy to evaluate persistent renal disease, 18% ultimately developed chronic renal failure. The risk factors for renal failure include age at onset greater than 7 years, severe abdominal symptoms, and persistent purpura.


1. Al-Sheyyab M, El-Shanti H, Ajlouni S, Sawalha D, Daoud AS. The clinical spectrum of Henoch-Schönlein purpura in infants and young children. Eur J Pediatr. 1995;154:969-972.

2. Hyams JS. Corticosteroids in the treatment of gastrointestinal disease. Curr Opinion Pediatr. 2000;12:451-455.

3. Lanzkowsky S, Lanzkowsky L, Lanzkowsky P. HenochSchönlein purpura. Pediatr Rev. 1992;13:130-137.

4. Saulsbury FT. Henoch-Schönlein purpura in children: report of 100 patients and review of the literature. Medicine. 1999;78:395-409.

5. Walker WA, Higuchi L. Henoch-Schönlein syndrome. In: McMillan JA, DeAngelis CD, Feigin RD, Warshaw JB, eds. Oski’s Pediatrics: Principles and Practice. 3rd ed. Philadelphia: Lippincott Williams & Wilkins; 1999:2176-2179.

6. Sano H, Izumida M, Shimizo H, et al. Risk factors of renal involvement and significant proteinuria in Henoch–Schönlein purpura. Eur J Pediatr. 2002;161:196-201.

7. Shah D, Goraya JS, Poddar B, et al. Acute infantile hemorrhagic edema and Henoch–Schönlein purpura overlap in a child. Pediatr Dermatol. 2002;19:92-93.

8. Praid D, Amir J, Nussinovitch M. Recurrent Henoch–Schonlein purpura in children. J Clin Rheumatol. 2007;13:25-28.

9. Mills JA, Michel BA, Bloch DA, et al. The American College of Rheumatology 1990 criteria for the classification of Henoch–Schönlein purpura. Arthritis Rheum. 1990;33:1114-1121.

10. Weiss PF, Klink AJ, Localio R, et al. Corticosteroids may improve clinical outcomes during hospitalization for Henoch-Schönlein purpura. Pediatrics. 2010 Oct;126(4):674-81. Epub 2010 Sep 20.