Kocher Criteria for Septic Arthritis

Learn and apply the Kocher clinical prediction rule for pediatric acute hip pain: fever, non–weight-bearing, ESR ≥40 mm/h, and WBC >12k. Understand score interpretation, limitations, and how it supports—but does not replace—clinical judgment.

Kocher Criteria for Septic Arthritis

Select each criterion that is met. The score is the number of predictors present (0–4). Intended for children with acute hip pain; external validation varies—use with clinical judgment.

Clinical and laboratory predictors

Fever

Temperature > 38.5°C (101.3°F)

+1 predictor

Non–weight-bearing on affected lower extremity

Unable to bear weight or clinical equivalent (e.g., refusal to walk)

+1 predictor

ESR ≥ 40 mm/h

Erythrocyte sedimentation rate at or above 40 mm/h

+1 predictor

WBC > 12,000 / mm³

Peripheral white blood cell count above 12,000 cells/mm³

+1 predictor

Disclaimer: This tool summarizes the Kocher prediction rule for educational purposes. It does not establish a diagnosis, does not replace specialist evaluation, and may not perform equally in all settings or joints. Use institutional pathways for aspiration, antibiotics, and surgery.

Kocher Criteria — Scoring Guide

Purpose

The Kocher criteria are a clinical prediction rule developed to help distinguish acute septic arthritis of the hip from transient synovitis in children presenting with a painful hip. Each criterion is worth one point; the total number of predictors correlates with the probability of septic arthritis in the original derivation cohort.

Confirm cutoffs and definitions with the original publication and current institutional guidelines. Performance varies in external validation; the rule is most specific to the pediatric hip presentation for which it was derived.

The four criteria (1 point each)

1. Fever

History of measured fever: body temperature > 38.5°C (101.3°F).

2. Non–weight-bearing

Inability to bear weight on the affected lower extremity (including refusal to walk or limp that represents inability to bear weight as judged clinically).

3. Elevated ESR

Erythrocyte sedimentation rate ≥ 40 mm/h.

4. Elevated WBC

Peripheral white blood cell count > 12,000 cells/mm³.

Interpretation (derivation cohort — Kocher et al., 1999)

Predictors present	Approximate probability of septic arthritis
0	~0.2%
1	~3%
2	~37–40%
3	~62%
4	~93%

These estimates reflect the original study population and should not replace bedside judgment, orthopaedic or infectious disease consultation, or local pathways for joint aspiration and imaging.

Reference

Kocher MS, Zurakowski D, Kasser JR. 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):1662-1670.

Clinical background

Acute hip pain in children creates a high-stakes diagnostic problem. Septic arthritis is a surgical emergency: bacterial infection within the joint space can destroy cartilage within hours to days, and delay in drainage is associated with worse outcomes. Transient synovitis—a benign, self-limited inflammatory condition of the hip—presents with overlapping features: limp, pain, irritability, and sometimes fever. Because the stakes of missing septic arthritis are so high, clinicians have long sought structured ways to estimate risk and to decide how aggressively to pursue imaging, aspiration, and empiric therapy.

The Kocher criteria (often called the Kocher score or Kocher prediction rule) summarize four readily available bedside and laboratory findings into a simple integer from 0 to 4. The rule was derived specifically in a cohort of children evaluated for acute irritable hip, with the goal of estimating the probability that the child had septic arthritis rather than transient synovitis. It is not a stand-alone diagnostic test; it is a clinical decision support framework meant to complement examination, imaging, specialist input, and institutional pathways.

Why a prediction rule helps

Children with hip pain may appear relatively well early in the disease course, or may be difficult to examine because of pain and anxiety. Plain radiographs are often normal or nonspecific early on. Ultrasound can demonstrate an effusion but cannot reliably distinguish sterile effusion from infection. The gold standard for diagnosing septic arthritis remains joint fluid analysis—cell count, Gram stain, culture, and crystal evaluation when indicated—often obtained operatively in clinically suspicious cases. Because aspiration and drainage carry procedural and anesthetic risks, and because overtreatment has its own harms, a transparent scoring system can help teams communicate risk, document reasoning, and trigger escalation in a consistent way.

The Kocher approach is attractive because it uses four simple binary predictors that clinicians already collect in most emergency and inpatient evaluations: fever history, weight-bearing status, erythrocyte sedimentation rate (ESR), and peripheral white blood cell (WBC) count. Each predictor is counted once, and the total number of positive predictors maps to approximate probabilities reported in the original derivation cohort.

The four Kocher predictors (one point each)

1. Fever: temperature greater than 38.5 °C (101.3 °F)

Fever reflects systemic inflammatory or infectious response. In the Kocher model, a history of measured fever above the threshold counts as a positive predictor. In practice, clinicians should confirm how temperature was taken (oral, rectal, axillary, temporal) and whether antipyretics were given, because treatment can mask fever. The criterion is framed as a history of fever in the original context of clinical evaluation; exact documentation practices vary by site. When fever is absent but the child appears toxic, clinicians should not anchor solely on this single element—overall appearance, pain, and inability to move the hip remain critical.

2. Non–weight-bearing on the affected lower extremity

This predictor captures severe functional limitation: the child is unable to bear weight on the affected side, or refuses to walk, or demonstrates a pattern that clinicians judge equivalent to inability to bear weight. In a painful hip, guarding and refusal to ambulate can occur in both septic arthritis and transient synovitis, but inability to bear weight has been emphasized as a marker of more severe intra-articular pathology. Interpretation requires age-appropriate assessment; preverbal children may not cooperate with formal gait testing, and clinical judgment substitutes for formal “walking” in some encounters. Compare with the unaffected side when possible and document exam limitations clearly.

3. Erythrocyte sedimentation rate (ESR) ≥ 40 mm/h

The ESR is a nonspecific marker of inflammation. It typically rises over hours to days and may remain elevated even as clinical improvement begins. An ESR at or above 40 mm/h counts as a positive predictor in the Kocher rule. Clinicians should remember confounders: recent surgery, other infections, anemia, autoimmune conditions, and malignancy can influence ESR. Conversely, early presentation may precede laboratory elevation. If the ESR is below threshold but suspicion remains high, repeating inflammatory markers and integrating imaging may still be appropriate depending on trajectory and examination findings.

4. Peripheral white blood cell count > 12,000 cells/mm³

Leukocytosis suggests systemic immune activation. A WBC above 12,000/mm³ counts as positive in the Kocher model. As with ESR, WBC is nonspecific: viral illness, stress response, other bacterial infections, corticosteroid exposure, and hematologic conditions can shift counts. The differential count and bandemia may add context but are not part of the classic Kocher score. Serial trends can be informative when the presentation is evolving.

How the score is calculated

Each predictor is binary: present or absent. The Kocher score equals the number of predictors present, ranging from 0 to 4. There are no fractional points and no weighting beyond inclusion or exclusion of each criterion.

Operationally, teams should agree locally on how “present” is determined for each element—for example, whether fever must be documented in the medical record versus caregiver report, and whether “non–weight-bearing” requires a formal orthopedic examination versus emergency physician assessment. Consistency improves inter-provider communication and reduces silent drift in how the score is applied.

Interpreting the score in the derivation cohort

The probabilities most commonly cited with the Kocher rule come from the original derivation dataset and describe the estimated risk of septic arthritis given the number of predictors present. Commonly taught ranges are approximately:

0 predictors: very low estimated probability (on the order of a fraction of a percent in the derivation cohort).
1 predictor: low estimated probability (a few percent).
2 predictors: intermediate probability (often cited around the high thirties to forty percent range).
3 predictors: high probability (often cited around sixty percent).
4 predictors: very high probability (often cited around ninety percent).

These figures are useful for teaching and shared decision-making, but they are not individualized post-test probabilities for every patient in every setting. Prevalence of disease, referral patterns, timing of labs, and prior antibiotic exposure change the meaning of the same score. The score should be understood as a structured summary of concerning features rather than a replacement for clinician judgment.

How the rule fits into modern evaluation

In many centers, children with suspected septic hip undergo a layered evaluation: bloodwork, imaging (often ultrasound), orthopaedic consultation, and planning for operative joint drainage when clinical and laboratory data align with high risk. The Kocher score can help stratify urgency and clarify why a team is pursuing aspiration or observation, but it does not define absolute thresholds for surgery or antibiotics by itself.

When scores are intermediate (often around two predictors), pathways frequently emphasize repeat examination, repeat inflammatory markers, ultrasound assessment of effusion, and early specialist involvement because missing septic arthritis in this band remains consequential. When scores are high (three or four predictors), many institutions treat the situation as an emergency and expedite operative management and culture-directed therapy, pending local protocols.

Antibiotic timing involves balancing culture yield with clinical stability. Institutional infectious disease and orthopaedic standards typically guide whether antibiotics can be briefly deferred for operative sampling in stable patients, versus immediate empiric therapy in toxic-appearing children. The Kocher score informs risk discussion but does not resolve these protocol-dependent decisions alone.

Limitations, validation, and common pitfalls

The Kocher criteria were developed in a specific population and era of practice. External validation studies have reported variable sensitivity and specificity, and some datasets show lower positive predictive value than the original estimates, particularly when disease prevalence differs or when care pathways selectively test higher-risk patients. Applying the rule outside its intended context—such as to adults, other joints, or chronic presentations—may be misleading unless separately validated.

Common pitfalls include the following:

Treating the score as diagnostic. A low score does not exclude septic arthritis in an ill-appearing child; a high score does not replace synovial fluid diagnosis.
Ignoring evolution. Early labs may be deceptively normal; repeat assessment matters.
Anchor bias after partial treatment. Antipyretics, analgesia, and antibiotics can change temperature, WBC, and functional status.
Overreliance on ESR/WBC delay. If labs are pending, clinical escalation should not be deferred solely because a computed score is not yet available.
Forgetting alternative diagnoses. Osteomyelitis, pyomyositis, Lyme arthritis, inflammatory arthritis, leukemia, and slipped capital femoral epiphysis (in the appropriate age group) can mimic septic hip presentations.

Modified Kocher algorithms and additional biomarkers (including C-reactive protein in many modern pathways) appear in subsequent literature and institutional guidelines. When your hospital uses an augmented rule, document which version you are applying so consultants can interpret the risk estimate consistently.

Documentation and medicolegal awareness

Clear documentation of hip examination (range of motion, pain with internal rotation, neurovascular status), fever pattern, weight-bearing status, laboratory values with timestamps, imaging results, consultant recommendations, and return precautions supports safe care regardless of the prediction score. If a clinician chooses not to aspirate or operate despite an elevated Kocher score, the chart should reflect the reasoning (for example, rapid clinical improvement, alternative diagnosis supported by objective data, or shared decision-making with guardians when clinically appropriate).

Using this site’s calculator responsibly

The accompanying calculator tabulates the Kocher score when you indicate which predictors are present. It is intended for education and workflow support. Always interpret results in the full clinical context, follow local pediatric emergency medicine and orthopaedic protocols, and escalate care when the patient appears septic, is hemodynamically unstable, or has severe uncontrolled pain or concern for compartment-equivalent joint compromise.