Article Text

other Versions

The child with a limp: a symptom and not a diagnosis
  1. Eve Smith1,
  2. Mark Anderson2,
  3. Helen Foster1,3
  1. 1Department of Paediatric Rheumatology, Great North Children's Hospital, Newcastle Upon Tyne, UK
  2. 2Department of Paediatric Medicine, Great North Children's Hospital, Newcastle Upon Tyne, UK
  3. 3Newcastle Musculoskeletal Research Group, Institute Cellular Medicine, The Medical School, Newcastle Upon Tyne, UK
  1. Correspondence to Dr Eve Smith, Department of Paediatric Rheumatology, Great North Children's Hospital, Richardson Road, Newcastle Upon Tyne NE1 4LP, UK; evemdsmith{at}

Statistics from


The limping child frequently poses a diagnostic challenge and clinical assessment may not be easy. Epidemiological studies are sparse; in one study1 children with an acute limp accounted for <2% of all paediatric emergency department (ED) attendances although the frequency may well be different in the primary care setting.

Trauma is the commonest cause of limping and many cases of atraumatic limp will resolve spontaneously. However, limp is not a diagnosis and it is important to assess limping children carefully as rarer, but serious, causes can be associated with significant morbidity, and even mortality, if there is a delay in diagnosis.

Children warranting urgent investigation are the very young (<3 years of age), the ill and febrile, the non-weight bearing and those with painful restricted hip movements. Teaching on the limping child correctly focuses on the hip, where significant pathology often occurs. However, limp may be due to extra-articular causes or joint problems other than those affecting the hip; these can be easily missed without careful assessment.

What is meant by the term ‘limping’?

In most cases, acute limping describes an antalgic (painful) gait, that is, minimising weight bearing on a sore limb, with a shortened stance phase and increased swing phase of the gait cycle. Acute refers to duration of 1–2 days in contrast to a chronic limp (>6 weeks) and subacute (2 days and up to 6 weeks). Subtle limping may be accentuated by asking the child to run: listening for an asymmetric cadence can be helpful. Limping is also used to describe other abnormal gait patterns, often due to a spectrum of causes that are not acute in origin (eg, cerebral palsy) and not covered in detail in this article.

The age of the child is helpful in establishing a differential diagnosis (table 1)2 ,3 which will be aided by careful initial assessment, judicious use and interpretation of blood tests, imaging and pattern recognition. A history of trauma is common in the young child and may be a ‘red herring’ coexisting with an alternative cause of limp. Conversely, the absence of witnessed trauma does not exclude it. Most importantly, the possibility of non-accidental injury must always be considered. Typical clinical presentations of the limping child (table 2) may help to refine the differential diagnosis.

Table 1

Common and significant causes of limp by age2

Table 2

Typical clinical presentations of the limping child

We present three case histories to highlight important diagnostic dilemmas and potential pitfalls when considering the acute limping child. We also summarise the evidence where available, and present a practical approach to investigations and initial management.

A case of acute limp with red flag features

A six-year-old boy presented to the ED with an 8 hour history of severe worsening left hip and thigh pain. He was previously fit and well apart from an upper respiratory tract infection 7 days previously. There was no history of preceding hip problems or trauma. He refused to weight bear and became very distressed with attempted examination of his hip. His temperature was 38.5°C and his mother had become worried when he was unable to sleep due to the pain. He was tachycardic, flushed and miserable. Systemic examination was unremarkable. Examination of his other joints was normal.


An acute non-weight bearing limp makes diagnoses such as infection, trauma and malignancy more likely. ‘Red flags’ for these conditions should be sought at initial presentation (table 3) with evidence from clinical examination and other sources as appropriate (eg, if non-accidental injury is suspected).

Table 3

‘Red flags’ for severe life-threatening conditions

Important key features of the history when eliciting red flag symptoms include; characteristics of the pain (eg, site, trigger factors, effect of weight bearing), the presence of systemic features (eg, fever, loss of appetite or weight), any recent history of travel (eg, to Lyme disease endemic areas although it must be noted that the typical history of a rash or tick bite may be absent4) and recent medication history (eg, recent antibiotic treatment may lead to partially treated septic arthritis or osteomyelitis).

  • ▸ The clinical assessment (table 4) needs to be comprehensive as the history may be relatively scant and young children frequently experience non-specific pain (eg, ‘my leg is hurting’).7 The hip will often be the initial focus of the examination, since acute unexplained limp is frequently caused by hip pathology. However, referred pain must not be forgotten, and in the case of the hip, examination must include the spine, abdomen, pelvis and testes as appropriate. For other lower limb joints, a minimum of the joint above and below the affected joint must be examined. Septic arthritis tends to involve one joint but can (rarely) affect multiple joints; conversely, the involvement of multiple joints raises suspicion of a more systemic process (including malignancy) (tables 2 and 3).

  • ▸ Clearly, gait will be difficult to assess if the child is non-weight bearing, and in the severely ill child the approach to musculoskeletal examination will focus on passive rather than active movements but the key point is that all joints should be screened and paediatric gait, arms, legs and spine (pGALS) examination may be helpful (see table 4 and box 1).

Table 4

Clinical assessment of the acute limping child

Box 1 The paediatric gait, arms, legs and spine (pGALS) musculoskeletal screening examination6

  • Screening questions

  •  Do you (or does your child) have any pain or stiffness in your joints, muscles or back?

  •  Do you (or does your child) have any difficulty getting yourself dressed without any help?

  •  Do you (or does your child) have any difficulty going up and down stairs?

  • Gait

  •  Observe the child walking and turning.

  •  ‘Walk on your tip-toes/walk on your heels’

  • Arms

  •  ‘Put your hands out in front of you’

  •  ‘Turn your hands over and make a fist’

  •  ‘Pinch your index finger and thumb together’

  •  ‘Touch the tips of your fingers with your thumb’

  • Squeeze metacarpophalangeal joints

  •  ‘Put your hands together/put your hands back to back’

  •  ‘Reach up and touch the sky’

  •  ‘Look up at the ceiling’

  •  ‘Put your hands behind your neck’

  • Legs

  •  Feel for effusion at the knee

  •  ‘Bend and then straighten your knee’ (active movement of knees and examiner feels for crepitus)

  •  Passive flexion (90°) with internal rotation of the hip

  • Spine

  •  ‘Open your mouth and put three of your fingers in your mouth’

  •  Lateral flexion of the spine: ‘Try and touch your shoulder with your ear’

  •  Observe spine from behind

  •  ‘Can you bend and touch your toes’: observe curve of spine from side and behind.

  • ▸ Further details are available with a video demonstration of pGALS performed on a normal child (

The doctor who assessed this child initially suspected septic arthritis or possibly a reactive arthritis. Blood results showed a white cell count of 11.5×109/l (90% neutrophils), C reactive protein (CRP) of 30 mg/l and erythrocyte sedimentation rate (ESR) of 15 mm/h. Blood film was normal. Plain radiography of the hip was normal. An urgent hip ultrasound scan (USS) confirmed a significant effusion.


  • ▸ During the assessment of a limping child who appears acutely unwell, essential investigations include full blood count (FBC), acute phase reactants (CRP, ESR), blood cultures, blood film with other tests depending on the clinical presentation (table 5). The diagnosis of septic arthritis or osteomyelitis can be problematic as even patients with culture positive septic arthritis can have normal inflammatory markers and be apyrexial initially.8

  • ▸ Radiographs (including ‘frog leg’ lateral views) and urgent hip USS are required if clinical examination reveals the hip to be the suspected site of pathology. It is important to note that hip pain can be referred to the knee.

  • ▸ Transient synovitis and septic arthritis can both result in significant effusions on USS and distinguishing between septic arthritis and transient synovitis is a matter of clinical judgement. Kocher et al9 have proposed a ‘clinical prediction rule’ which helps to differentiate septic arthritis from transient synovitis in the presence of a confirmed hip effusion. The risk of septic arthritis increases with the number of factors present (box 2). The presence of elevated CRP levels (>10 ml/l) further increases the risk of sepsis.11 ,18

  • ▸ If hip USS is normal but clinical concerns about septic arthritis or osteomyelitis remain, then bone scan or MRI are indicated to rule out osteomyelitis, psoas abscess or other potential septic ‘hot spots’. The role of imaging and what tests to do and when remains controversial (table 5). The choice of imaging modality is influenced by local access, clinical judgement and experience.

A presumptive diagnosis of septic arthritis was made based on the clinical presentation and investigations. Joint aspiration and wash-out were undertaken in theatre under general anaesthetic and synovial fluid was sent for microscopy and culture. Gram stain was negative but >50 000/mm3 white cells were seen on microscopy (mainly neutrophils). Intravenous antibiotics were commenced urgently. This child was treated with intravenous antibiotics for 2 weeks, followed by 4 weeks of oral antibiotics. His symptoms improved rapidly and he made a full recovery.

  • ▸ With such a patient, it is best to err on the side of caution and adopt a careful approach to management, rather than miss a septic joint. The hip joint should be drained, irrigated and synovial fluid sent for urgent microscopy, gram stain and culture. If there are concerns about atypical infection such as in the immunocompromised child, it is important to discuss with microbiology and paediatric infectious disease colleagues to ensure that the appropriate tests are undertaken.

  • ▸ Differentiating between septic arthritis and aseptic/inflammatory arthritis based on synovial fluid findings may be difficult as gram stain and culture have been reported to be negative in 50%–80% of septic arthritis, and children with inflammatory and septic arthritis can have similar synovial fluid white cell counts.8 ,20

  • ▸ Empirical antibiotics must be started urgently in suspected septic arthritis with the choice of antibiotic altered if a causative organism is identified (table 6). Staphylococcus aureus is the most common causative organism. Septic arthritis is an orthopaedic emergency and outcomes can be dramatically worse if antibiotic treatment is delayed.21

  • ▸ The response to antibiotic treatment is monitored clinically (temperature, pain, spontaneous movement of the joint) as well as by serial monitoring of inflammatory markers with normalisation of the CRP being the earliest laboratory parameter to indicate improvement.17 Controversy exists regarding the length of treatment and when to switch to oral antibiotics but the regimes outlined in table 6 are commonly used in the first instance.

  • ▸ In children under the age of 2 years, the blood supply to the joint is via the metaphysis, which may be intra-articular at certain joints (namely the hip, ankle, shoulder and elbow) and explains why septic arthritis and osteomyelitis frequently coexist at these sites. Diagnosing concomitant osteomyelitis is important as it commonly requires a longer duration of antibiotics as compared with septic arthritis alone, and long-term sequelae are more likely.

  • ▸ The outcome of septic arthritis is variable with worse prognosis occurring with hip involvement, associated proximal femur osteomyelitis, S aureus infection, onset in a child <6 months of age and where there has been a delay in diagnosis of ≥4 days. Potential sequelae following septic arthritis include avascular necrosis and premature degenerative joint disease.16 ,22

  • ▸ Mycobacterial joint infection should be considered in high-risk patients (certain ethnicities immunosuppressed, ethnicity or family contacts) or if there is a history of increasing pain, night sweats and weight loss (with or without associated cough). Atypical mycobacteria and other unusual organisms (eg, fungal infection) also need to be considered in immunocompromised children.

Table 5

Investigations in the acute limping child

Table 6

Common causative organisms for septic arthritis and suggested antibiotic choices

A case of subacute limp and irritable hip

A six-year-old boy presented to the ED with left thigh pain and limp, having been reluctant to walk for 4 days. He continued to be playful if his toys were around him and he could remain seated. There was no history of previous joint problems, antecedent trauma or any other medical history of note. Although he had not had a temperature since the onset of the limp he had a sore throat 7 days before.

On examination he was apyrexial and alert. He improved following arrival in the ED, had a mildly antalgic gait but was happy to weight bear. When encouraged to run his limp was more pronounced. His left hip had a reduced range of movement compared with the right. A comprehensive musculoskeletal, neurological and systemic examination was undertaken and found to be unremarkable. Blood tests (FBC, ESR, CRP) and hip radiograph were all normal. The patient was allowed home with advice regarding analgesia and his parents were given written instructions of when to return to hospital (ie, if he became unwell, developed a high temperature, had increasing leg pain or night pain, was unable to weight bear, developed involvement of other joints or if the limp persisted beyond 2 weeks).


  • ▸ The patient had no red flag features. He was not excessively distressed, was able to weight bear with some movement of his hip. The initial acute investigations of such children is controversial with a wide variation in clinical practice, from ‘watchful waiting’ to blood tests, plain x-ray and/or hip USS at initial presentation. Safety-net advice for parents/carers is vital and children need review if symptoms do not settle.

  • ▸ Given his age and the clinical scenario, the most likely diagnosis was transient synovitis (or early Perthes’ disease). In the older child (>11-years-old), slipped upper femoral epiphysis (SUFE) would need to be considered and frog leg lateral hip x-ray undertaken.

 Two weeks after the initial presentation, the child continued to have a mild limp and so his parents contacted the ED and he was reviewed in the orthopaedic department. On examination, he still had some limitation of hip movement and pain on abduction and internal rotation. Plain hip x-ray was repeated and showed early signs of Perthes’ disease.


  • ▸ Most cases of transient synovitis respond quickly to analgesia and rest. Review is necessary when the limp persists to exclude evolving Perthes’ disease (table 2). If repeat radiographs are normal, then bone scan or MRI may be indicated.

  • ▸ The aim of treatment of Perthes’ disease is to prevent deformity of the femoral head, which could lead to early osteoarthritis of the hip. Prognosis is variable but best with early detection, in younger children (<6 years) and where the femoral head is minimally involved.23 ,24

A case of chronic intermittent limp

A previously fit and well 4-year-old girl, presented to her general practitioner with a 7 week history of limp associated with intermittent right knee swelling observed by her parents. There was no history of preceding trauma or systemic upset. She was reluctant to weight bear in the mornings and more ‘grumpy’ than usual but otherwise was well in herself. She was reluctant to sit cross-legged on the floor when playing, and kept her right leg outstretched. On examination, she was afebrile, looked well and general examination was normal. Her right knee was warm and slightly swollen but not red or tender. She was reluctant to fully extend or flex her knee. Initial investigations revealed normal FBC, blood film, ESR, CRP and knee X-ray. A presumptive diagnosis of reactive arthritis was made and the patient was discharged after 48 h of observation on the ward.


  • ▸ The patient's history was of several weeks duration and she was systemically well. There were no red flags in the initial presentation and the working differential diagnosis should include conditions that are associated with chronic or persistent limp (tables 1 and 2). By 6 weeks, an episode of reactive arthritis should have resolved.

  • ▸ The absence of a definitive diagnosis and persistent symptoms always warrants planned review.

Two weeks later, the patient was referred to the paediatric department by her general practitioner. She had become increasingly reluctant to walk. Both knees were swollen and displayed early flexion deformities. Examination of her joints also revealed swelling of the right ankle. Systemic examination was otherwise unremarkable. Blood tests (FBC, blood film and acute phase reactants) were normal and auto-antibodies were negative. The orthopaedic team arranged an MRI of both lower limbs under general anaesthetic; an effusion in both knees and the right ankle was confirmed.


  • ▸ The chronicity and periodicity of the patient's symptoms, gelling and stiffness after rest, and involvement of more than one joint suggest an inflammatory arthritis. In the UK, juvenile idiopathic arthritis (JIA) is the most likely cause of chronic arthritis (ie, lasting >6 weeks). JIA is a group of disorders characterised by arthritis of >6 weeks duration, presenting before the age of 16 following exclusion of other conditions and with a spectrum of presentations and clinical courses.25 Optimal outcome rests on early diagnosis and prompt referral to paediatric rheumatology specialist teams as emphasised in the Standards of Care for JIA (

Early diagnosis of JIA rests on suspicion and careful clinical assessment:

  • ▸ In JIA, pain may not be verbalised especially in the very young, but may be suggested by a change in the child's mood, sleep pattern, change in activities (play, sport and where appropriate school work) and the effect of analgesics or non-steroidal anti-inflammatory drugs. The child with JIA may avoid or adapt certain activities that are uncomfortable or may be noted to be ‘clumsy’ or have regression in achieved motor milestones (such as walking).

  • ▸ The pGALS musculoskeletal examination (box 1) helps identify abnormal joints which can then be examined further using a more detailed approach to joint examination such as the paediatric regional examination of the musculoskeletal system (table 4). pGALS has been validated for the school aged child,6 but can be used in younger children in a ‘copy me’ style and has been shown to be effective in acute paediatric practice.27 pGALS may help detect abnormal joints that are not apparent from the history alone7 especially when the symptoms are vague and illocalised. Joint swelling can be difficult to detect clinically. Ankle swelling as part of pGALS is best observed from behind, with associated calf wasting suggesting chronicity.

  • ▸ It is advisable to refer to paediatric rheumatology when JIA is suspected, and especially prior to contemplating invasive procedures (eg, arthroscopy/synovial biopsy/MRI), which are invariably not necessary to confirm the diagnosis; delay in access to such tests may incur further delay to referral and starting treatment. If MRI is required, then gadolinium should be given, as enhancement is helpful to detect synovitis. Ultrasound is increasingly used to assess for synovitis in children as it does not require sedation or general anaesthesia.

  • ▸ Eye screening (using slit lamp examination) is essential when JIA is suspected. Visual involvement with uveitis is potentially blinding and is invariably asymptomatic in the early stages.

  • ▸ There is no diagnostic test for JIA. Investigations (FBC, ESR and CRP) may be normal although more severe subtypes have raised acute phase reactants and anaemia. A positive antinuclear antibody (ANA) is not diagnostic and can be found in up to 33% of normal healthy children.28 When present in children with JIA, ANA indicates a higher risk of chronic anterior uveitis. Rheumatoid factor is invariably negative but in polyarticular JIA indicates a more guarded prognosis.

  • ▸ Growing pains are a common label used to describe children with aches and pains of unclear cause.29 Persistent limp and daytime symptoms are exclusion criteria for growing pains and in such children further assessment is needed.30

  • ▸ It is noteworthy that in a child with a persistent limp it is important to also consider inflammatory muscle disease and careful assessment for skin rash and proximal muscle weakness is necessary (inability to jump in a school aged child or an abnormal Gower's test suggests proximal muscle weakness). Measurement of muscle enzymes is warranted.

In this case the child had a diagnosis of oligoarticular JIA (ie, arthritis of ≤4 joints during the first 6 months of disease). This is the most common JIA subtype with an excellent prognosis albeit high risk of uveitis. She had flexion contractures, which can be avoided by early joint injection with steroids and physiotherapy. Regular eye screening is mandatory. More than a third of children with oligoarticular JIA will develop the so-called extended oligoarticular JIA with a guarded prognosis and invariably treated with methotrexate.24

Limping child guidelines

  • ▸ There is currently no agreed evidence or consensus based clinical guidelines for the limping child.

  • ▸ Limping child guidelines should be locally agreed among A&E, paediatric and orthopaedic departments to help exclude serious life-threatening pathology and facilitate early detection of potentially disabling conditions such as Perthes’ disease, slipped upper femoral epiphysis and JIA.

  • ▸ Limping child guidelines need to contain discharge criteria, indications for review and referral for children who fail to improve.

  • ▸ Parent information leaflets need to include advice on analgesia, and when to return the child for review if symptoms do not settle or worsen.

  • ▸ The management of persistent limp and referral to other subspecialties such as paediatric rheumatology needs to avoid any undue delay.

  • ▸ Common pitfalls to be avoided in the limping child are shown in box 3.

Box 2 Kocher's criteria to differentiate between septic arthritis and transient synovitis in the presence of confirmed hip effusion9


  • Fever >38°C

  • Unable to weight bear

  • Erythrocyte sedimentation rate >40 mm/h

  • Serum white cell count >12×106/l

Probability of septic arthritis

  • No factors present <0.2%

  • Two factors present 40%

  • Three factors present 93%

  • Four factors present >99%

Box 3 Common pitfalls to be avoided

  • Ascribing limp to trauma and overlooking features that suggest other causes. Referred pain (eg, from the abdomen (and testes in boys), back or chest and hip pathology manifesting as knee pain).

  • Think beyond the hip (!) and examine the child comprehensively.

  • Classical clinical features of sepsis may be masked in the immunosuppressed child.

  • Mycobacterial infection can be easily missed.

  • Synovial fluid may be sterile in partially treated septic arthritis.

  • Labelling children with daytime symptoms as having ‘growing pains’.

  • Medically unexplained limp or physical symptoms warrant specific management and referral (ie, discharge without a diagnosis and follow-up plan is not advised).

  • The blood film may be normal in children with malignancy.

  • Radiographs are often normal in children with early sepsis or arthritis.

  • Acute phase reactants may be normal in children with arthritis.

  • Rheumatoid factor is usually negative in children with arthritis.

  • Antinuclear antibody and rheumatoid factors may be false positives in children without inflammatory joint or muscle disease.

Clinical messages

  • The ‘limping child’ is a common presentation: careful clinical assessment, knowledge and judicious use of often simple investigations will often facilitate a correct diagnosis.

  • The hip is a common site of pathology but it is important to exclude pathology elsewhere.

  • Kocher's clinical prediction rule is the most useful tool to date to help distinguish between septic arthritis and transient synovitis; however, it requires further validation in a large prospective study.

  • Ultrasound scan is more sensitive than plain x-ray for the detection of hip effusions.

  • Limping is not a diagnosis: all children need clear follow-up plans and a parent information leaflet to indicate when and how parents can seek further medical advice.

  • If a limp persists (>3 weeks), then the likelihood of juvenile idiopathic arthritis is high and referral to paediatric rheumatology is recommended and before contemplating invasive investigations which may be unnecessary.


View Abstract


  • Funding None.

  • Competing interests None.

  • Patient consent Obtained from the parent/guardian.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.