Approach to the Pediatric Limp


A 4 year old otherwise healthy boy presents to the ED with left leg pain and limp with decreased ability to bear weight. The symptoms started yesterday after watching a movie, with moderate improvement after Motrin. Mom denies any trauma or fall. Notably, he has been afebrile, but has had a mild coryzal illness over the past 3-5 days.

On presentation to the ED, his vital signs are recorded as: T 36.9 C, pulse 120, BP 100/60, RR 22, SpO2 100%. He is alert, active, and non-toxic in appearance. He is lying on the bed, with his hip flexed, abducted, and in external rotation. There is no overlying erythema of the hip, knee, or ankle. The left hip has mild restriction in abduction in comparison the right hip. With the left leg fully extended, the patient exhibits no involuntary muscle guarding or obvious discomfort with log-rolling of the left lower extremity. He additionally has a nontender lumbar spine and ipsilateral knee, ankle, and tibia/fibula. The patient is able to bear weight, but has an antalgic gait without exhibiting toe walking or circumduction. Plain radiographs of the pelvis and left hip are obtained which reveal medial joint space widening in comparison to the right hip.

What is the most likely cause of his symptoms?

A.   Toddler’s fracture

B.   Septic arthritis

C.   Transient synovitis

D.   Osteomyelitis

E.   Limb length discrepancy

C: Transient synovitis


Limp accounts for approximately 4/1000 visits to the Emergency Department, with the location of the pain not always reflecting the location of pathology. Problems in the hip can cause knee pain, and similarly, back problems can refer pain to the lateral thigh or posterior leg.


Minor trauma is the most common etiology for a limp, with the median age being 4 years, boys outnumbering girls 2:1, and localization typically to the hip. In those that deny history of trauma, the most common diagnosis is transient synovitis, with 77% having a benign cause without requiring surgical intervention or hospital admission.

PEM Playbook has a great mnemonic: STOP LIMPING

S: Septic arthritis (hip>knee)

T: Toddler’s fracture (1-3 years, minor fall with rotational component)

O: Osteomyelitis (2% of those children presenting with limp)

P: Perthes disease (Legg-Calve-Perthes disease, an idiopathic AVN, affecting children 3-12 years old)

L: Limb length discrepancy

I: Inflammatory (transient synovitis, 3-6 years of age after viral illness)

M: Malignancy

P: Pyomyositis (possible viral cause such as influenza, often with tender calves)

I: Iliopsoas abscess

N: Neurologic (stroke, will often have underlying pathology such as cardiac lesion, sickle cell disease, metabolic history; ataxia can present with a reported ‘limp’)

G: Gastrointestinal (appendicitis), genitourinary (testicular/ovarian torsion)

One of the ‘can’t-misses’ is the pediatric septic hip, which typically results from three sources: hematogenous spread, local spread (osteomyelitis), direct inoculation (trauma, surgery). S. aureus is the most common causative organism in all age groups, with Salmonella considered in sick cell disease patients, and N. gonorrhea in sexually active patients.


CBC, ESR, CRP are indicated in an acutely limping child in whom infectious etiology is a possible cause. Also consider Lyme if living in an endemic area, and there are no abnormalities on plain radiography.

If suspicion for septic arthritis remains high despite negative plain radiographs, recent studies have show utility in ultrasonography to identify effusions. Unfortunately, US cannot differentiate between septic arthritis and toxic/transient synovitis as both will result in effusions and mild widening of the joint space, as seen in this patient. MRI is favored over radionuclide scanning for osteomyelitis, stress fracture, and early avascular necrosis. MRI is also indicated if spinal pathology is suggested. Computed tomography (CT) is rarely useful in the patient with a limp, but can diagnose appendicitis, deep soft tissue infections of the paraspinal and retroperitoneal regions.

American College of Radiology Appropriateness Criteria

Traumatic – XR

Atraumatic, no signs of infection – XR, if negative then US hip

Atraumatic, signs of infection – US hip, if negative consider XR, if negative and still concerned for septic arthritis consider MRI


Patients with high concern for bone or joint infection require orthopedic consultation, emergent bone or joint aspiration, and early initiation of antibiotic therapy. A child with an oncologic process requires admission for staging workup and initiation of treatment. Most children have a benign etiology. Afebrile children with normal radiographs are suitable to followup with PCP, after discharging with NSAIDs. Ambulatory febrile children, with normal radiography and blood studies can also be followed up as an outpatient. If the patient remains febrile and unable to bear weight, have a low threshold to pursue joint aspiration. If the patient is afebrile, but unable to bear weight despite an adequate dose of analgesia, consider observation for MRI if early osteomyelitis, AVN, or spinal pathology is high on your differential.

Any unstable patient requires treatment as presumed sepsis, with fluid resuscitation and initiation of empiric antibiotics, orthopedics consultation for source control, and consideration of ultrasonography or MRI. One antibiotic regimen includes nafcillin 50 mg/kg, ceftriaxone 50 mg/kg; consider adding vancomycin 10 mg/kg if concerned for MRSA or sepsis.

Kocher’s criteria:

Our orthopedic colleagues utilize the Kocher criteria to determine the probability of whether the pediatric patient has a septic arthritis. The four elements include:

  • Erythrocyte sedimentation rate >40
  • WBC >12
  • Non-weight bearing on the affected joint
  • Fever >38.5 C

If elements are present, the probability of septic arthritis was determined to be:

  • 0/4 = 0%
  • 1/4 = 3%
  • 2/4 = 40%
  • 3/4 = 93%
  • 4/4 = >99%

HIGH pretest probability? – Kocher's criteria predictive value is HIGH

LOW pretest probability? – Kocher's criteria predictive value is LOW

A 2011 ACEP News Release confirms that Kocher’s criteria remains the best method for EM providers to differentiate transient synovitis and septic arthritis. If 2 or more criteria are present, talk with your orthopedic colleagues.


  • Assume that any child with a fever who presents with refusal to walk has septic arthritis or osteomyelitis until proven otherwise. Transient synovitis is a diagnosis of exclusion!
  • 2 or more Kocher criteria should prompt orthopedic consultation for consideration of joint aspiration.
  • Discuss care with PCP for prompt re-evaluation if patient’s ED examination is improving and they are stable for discharge.


Brady M. The child with a limp. J Pediatr Health Care. 1993:7:226.

Singer J. The cause of gait disturbance in 425 pediatric patients. Pediatr Emerg Care. 1985;1:7.

Chung S. Identifying the cause of acute limp in childhood. Clin Pediatr. 1974;13:769.

Fischer S, Beattie T. The limping child: epidemiology, assessment and outcome. J Bone Joint Surg Br. 1999;81:1029.

Pediatric Emergency Playbook. Please, just stop limping! Available at: Accessed April 3, 2017.

Leet A, Skaggs D. Evaluation of the acutely limping child. Am Fam Physician. 2000;61:1011.

Huttenlocher A, Newman T. Evaluation of the erythrocyte sedimentation rate in children presenting with limp, fever, or abdominal pain. Clin Pediatr. 1997;36:339.

American College of Radiology ACR Appropriateness Criteria. Limping child – ages 0-5 years. Available at: Accessed April 3, 2017.

ACEP News. Kocher criteria still the best way to ID septic arthritis in children. Available at: Accessed April 3, 2017.

Kocher M, Zurakoski D, Kasser J. Differentiating between septic arthritis and transient synovitis of the hip in children: an evidence-based clinical prediction algorithm. J Bone Joint Surg Am. 1999;81(12):1662.

Kocher M, Mandiga R, Zurakoski D, et al. Validation of a clinical prediction rule for the differentiation between septic arthritis and transient synovitis of the hip in children. J Bone Joint Surg Am. 2004;86:1629.