The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers
Trial registered on ANZCTR


Registration number
Help: Registration number
ACTRN12626000968370
Ethics application status
Help: Ethics application status
Approved
Date submitted
Help: Date submitted
27/03/2026
Date registered
Help: Date registered
6/08/2026
Date last updated
Help: Date last updated
6/08/2026
Date data sharing statement initially provided
Help: Data sharing statement date
6/08/2026
Type of registration
Help: Type of registration
Prospectively registered

Titles & IDs
Public title
Assessment of acute limp or joint pain in children presenting to emergency departments.
Scientific title
Paediatric Acute Limp Study (PALS) - A prospective observational cohort study to derive and validate a new paediatric limp / lower limb pain clinical decision rule for optimal evaluation of acute non-traumatic lower limb pain / limp in children presenting to the emergency department.
Additional trial identifiers [1] 316892 0
NHMRC GNT2043758
Universal Trial Number (UTN)
Trial acronym
PALS
Related trial records

Health condition
Health condition(s) studied:
Acute limp 341293 0
Acute lower limb pain 341296 0
Condition category
Condition code
Emergency medicine 337370 337370 0 0
Other emergency care
Musculoskeletal 337372 337372 0 0
Other muscular and skeletal disorders

Intervention/exposure
Study type
Observational
Patient registry
False
Target follow-up duration
Target follow-up type
Description of the intervention or exposure
This study aims to derive and validate a paediatric ED Clinical Decision Rule (PAL CDR) that (i) accurately detects serious pathology bone and joint infection, malignancy, slipped upper femoral epiphysis, ARF) and (ii) risk-stratifies children with suspected serious pathology* into those who do and do not require blood tests or imaging. The CDR will not be used to guide patient care during this study. Instead, the study aims to derive and validate the rule and evaluate how it could potentially influence future decisions about blood tests and imaging.

Data will be collected prospectively at the time of ED presentation and retrospectively (30-60 days and 6 months) after ED presentation.

• Prospective audit, recording routinely collected clinical assessment data:

At the time of ED presentation the treating clinician will collect predictor variables;
-Patient History: duration of pain, temperature, medications, history of trauma or illness in previous 2-4 weeks, previous bone/joint problems.
- Examination findings
- Clinical impression
- POCUS findings (if performed)

•Retrospective review of the medical record:

At 30-60 days, and 6 months after ED presentation, the research team will collect the following variables via chart review;
- Basic demographic/ED presentation data
- Vital signs and pain scores
- Clinical notes: location of pain, systemic symptoms, past medical history, ED disposition, mobility at discharge, outpatient follow up
- Radiology results
- Pathology results
- Treatment in ED
- Diagnosis

Participant data such as quality of life, ongoing symptoms, health service use, diagnoses, representations to hospital, treatment, procedures and test results will be collected via questionnaire;
- On enrolment
- 30-60 days after ED visit
- 6 months after ED visit
Intervention code [1] 333599 0
Not applicable
Description of the comparator or control
No control group - as an observational study, derivation and validation and evaluation of performance accuracy (sensitivity, specificity, negative predictive value [NPV], positive predictive value [PPV]) in identifying any serious condition (i.e. septic arthritis, osteomyelitis, haematological or bone malignancy, slipped upper femoral epiphysis, ARF) of a new CDR (PAL CDR) and existing CDRs (Kocher's rule, Caird criteria etc) via analysis of observational data collected.

We are planning to compare performance accuracy between existing clinical decision rules (Kocher’s rule and Caird’s rule) and (a) the newly derived CDR and (b) the developed machine learning model.
Kocher’s rule is our primary comparator, which will be compared against (a) the newly derived CDR and (b) the developed machine learning model. When validating against outcomes derived from existing clinical care, we will use the treating clinician’s assessment and resulting clinical outcomes as the comparator.
Control group
Active

Outcomes
Primary outcome [1] 344896 0
Accuracy in identifying the presence of any serious condition (i.e. septic arthritis, osteomyelitis, haematological or bone malignancy, slipped upper femoral epiphysis, ARF) causing paediatric acute limp/lower limb pain using the reference standard diagnosis (or follow up for those who did not undergo tests or imaging). The reference standard diagnoses for the following serious conditions are; Septic arthritis - joint effusion, synovial fluid analysis (fluid culture and cell count). Osteomyelitis - imaging (MRI +/- xray), blood cultures & laboratory findings (CPR/ESR, WCC). Haematological or bone malignancy - imaging, histopathology and bone marrow aspirate. Slipped upper femoral epiphysis - plain xray (AP and frog leg lateral view). ARF - imaging (US), clinical findings, evidence of GAS infection.
Timepoint [1] 344896 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Primary outcome [2] 345911 0
Presence of septic arthritis
Timepoint [2] 345911 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Primary outcome [3] 345912 0
Presence of osteomyelitis
Timepoint [3] 345912 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [1] 459663 0
Secondary outcome - performance accuracy of the Paediatric Acute Limp / Lower Limb Pain (PAL) CDR and currently published CDRs for identifying any serious condition (defined as malignancy, osteomyelitis, septic arthritis, acute rheumatic fever (ARF) and slipped upper femoral epiphysis) will be assessed as a composite secondary outcome.
Timepoint [1] 459663 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [2] 459665 0
Rates and results of investigation such as joint aspiration, blood tests.
Timepoint [2] 459665 0
Between 30-60 days, and 6 months post ED visit.
Secondary outcome [3] 459669 0
Inter-observer reliability of the newly developed CDR rule.
Timepoint [3] 459669 0
Within 30 mins of treating clinician seeing patient in ED
Secondary outcome [4] 459670 0
Cost effectiveness of implementing each CDR
Timepoint [4] 459670 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [5] 459688 0
Test characteristics of Point of Care Ultrasound (POCUS) for acute limp (composite outcome) including sensitivity, specificity, negative predictive value [NPV] and positive predictive value [PPV] for identifying any serious condition.
Timepoint [5] 459688 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [6] 459689 0
Agreement between ED and admitting clinician radiology interpretation and formal radiology reports in suspected bone and joint infection
Timepoint [6] 459689 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [7] 463752 0
Primary Outcome: Presence of haematological or bone malignancy.
Timepoint [7] 463752 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [8] 463753 0
Primary Outcome: Presence of slipped upper femoral epiphysis.
Timepoint [8] 463753 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [9] 463754 0
Primary Outcome: Presence of Acute Rheumatic Fever (ARF).
Timepoint [9] 463754 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [10] 463755 0
Secondary Outcome: performance accuracy of the new Paediatric Acute Limp / Lower Limb Pain CDR and currently published CDRs (Kocher’s rule, Caird’s rule) for detecting septic arthritis.
Timepoint [10] 463755 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [11] 463756 0
Secondary Outcome: performance accuracy of the new Paediatric Acute Limp / Lower Limb Pain CDR and currently published CDR (Novel 4-Point Risk Score for Acute Hematogenous Osteomyelitis) for detecting osteomyelitis.
Timepoint [11] 463756 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [12] 463757 0
Secondary Outcome: performance accuracy of the new Paediatric Acute Limp / Lower Limb Pain CDR and currently published CDRs (TRIM Score and Bone-RADS) for detecting haematological or bone malignancy.
Timepoint [12] 463757 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [13] 463758 0
Secondary Outcome: performance accuracy of the new Paediatric Acute Limp / Lower Limb Pain CDR and currently published CDRs (Age-Weight Test and Age-Height Test, Southwick Slip Angle) for detecting slipped upper femoral epiphysis.
Timepoint [13] 463758 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [14] 463881 0
Number of participants who under went joint aspiration procedure, pathology testing and ECG investigations (assessed as individual outcomes).
Timepoint [14] 463881 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [15] 463882 0
Results of participant investigations - joint aspiration, pathology testing, ECG investigations (assessed as individual outcomes).
Timepoint [15] 463882 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [16] 463883 0
Patient Outcome: Disposition
Timepoint [16] 463883 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [17] 463884 0
Patient Outcome: Surgical or interventional radiology procedures (composite outcome)
Timepoint [17] 463884 0
Between 30-60 days, 6 months post ED visit.
Secondary outcome [18] 463885 0
Patient Outcome: Re-presentations within 72 hours
Timepoint [18] 463885 0
Between 30-60 days post ED visit
Secondary outcome [19] 463888 0
Patient Outcome: Re-presentations within 7 days
Timepoint [19] 463888 0
Between 30-60 days post ED visit
Secondary outcome [20] 463889 0
Patient Outcome: 30-day outcome – quality of life
Timepoint [20] 463889 0
30 days post ED visit
Secondary outcome [21] 463890 0
Patient Outcome: 30-day outcome – days of school missed
Timepoint [21] 463890 0
30 days post ED visit
Secondary outcome [22] 463891 0
Patient Outcome: 30-day outcome – days of work missed
Timepoint [22] 463891 0
30 days post ED visit
Secondary outcome [23] 463892 0
Patient Outcome: 6 month outcome – quality of life
Timepoint [23] 463892 0
6 months post ED visit
Secondary outcome [24] 463893 0
Patient Outcome: 6 month outcome – days of school missed
Timepoint [24] 463893 0
6 months post ED visit
Secondary outcome [25] 463894 0
Patient Outcome: 6 month outcome – days of work missed
Timepoint [25] 463894 0
6 Months post ED visit
Secondary outcome [26] 463895 0
Inter-observer reliability of Kocher’s rule
Timepoint [26] 463895 0
Within 30 mins of treating clinician seeing patient in ED
Secondary outcome [27] 463896 0
Inter-observer reliability of Caird criteria
Timepoint [27] 463896 0
Within 30 mins of treating clinician seeing patient in ED
Secondary outcome [28] 463897 0
Inter-observer reliability of Novel 4-Point Risk Score for Acute Hematogenous Osteomyelitis
Timepoint [28] 463897 0
Within 30 mins of treating clinician seeing patient in ED
Secondary outcome [29] 463898 0
Variation in pathology testing for children with suspected bone and joint infection.
Timepoint [29] 463898 0
Between 30-60 days post ED visit, 6 months post ED visit.
Secondary outcome [30] 463899 0
Variation in imaging undertaken for children with suspected bone and joint infection.
Timepoint [30] 463899 0
Between 30-60 days post ED visit, 6 months post ED visit.
Secondary outcome [31] 463900 0
Variation in operative approach for children with suspected bone and joint infection.
Timepoint [31] 463900 0
Between 30-60 days post ED visit, 6 months post ED visit.
Secondary outcome [32] 463901 0
Variation in treatment approach for children with suspected bone and joint infection.
Timepoint [32] 463901 0
Between 30-60 days post ED visit, 6 months post ED visit.
Secondary outcome [33] 463902 0
External validation of the Copley severity of illness score for acute haematogenous osteomyelitis scoring system for bone and joint infection in children.
Timepoint [33] 463902 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [34] 463905 0
External validation of Caird criteria for bone and joint infection in children.
Timepoint [34] 463905 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [35] 463906 0
External validation of Novel 4-Point Risk Score for Acute Hematogenous Osteomyelitis scoring system for bone and joint infection in children.
Timepoint [35] 463906 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [36] 463907 0
External validation of Modified Osteomyelitis Severity of Illness scoring system for bone and joint infection in children.
Timepoint [36] 463907 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [37] 463908 0
External validation of the A-SCORE scoring system to predict acute complications for children bone and joint infection.
Timepoint [37] 463908 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [38] 463909 0
External validation of the C-SCORE scoring system to predict chronic morbidity in children bone and joint infection.
Timepoint [38] 463909 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.
Secondary outcome [39] 463910 0
External validation of the modified Copley severity of illness score for acute haematogenous osteomyelitis scoring system for bone and joint infection in children.
Timepoint [39] 463910 0
Baseline (ED visit), between 30-60 days, 6 months post ED visit.

Eligibility
Key inclusion criteria
Children aged <16 years presenting to the ED with acute onset (within the last 7 days) of non-traumatic lower limb pain and/or limp.
Minimum age
0 Years
Maximum age
15 Years
Sex
Both males and females
Healthy volunteers
No
Key exclusion criteria
• Children who present with a history of trauma where the symptoms were maximal at the time of the trauma (i.e. traumatic limb pain). This will be operationalised as “Acute trauma (e.g. clear injury with immediate pain such as fall / direct blow) directly leading to ED visit.”
• Known chronic joint problem affecting the same limb (e.g. previously diagnosed chronic arthritis, Perthes disease, slipped upper femoral epiphysis)
• Unable to obtain an accurate history (e.g. parent / guardian unavailable, language other than English and no interpreter available)
• Child is in out of home care and staff are unable to obtain consent from appropriate carer / guardian


Study design
Purpose
Screening
Duration
Longitudinal
Selection
Defined population
Timing
Prospective
Statistical methods / analysis
A new PAL CDR is planned to be developed.
The enrolled study population will be divided into derivation and validation cohorts. Enrolment into the validation cohort will continue until it is at least two-thirds the size of the planned derivation cohort. We have described derivation and validation of a CDR for septic arthritis below. The same approach will be used for a CDR for any serious pathology.
The CDR will not be applied in this study. We aim to derive and validate it only. As such, we do not anticipate any changes to clinical decisions, treatment, or any other aspects of patient care.

Study population (predictor variables and outcome variables) variables will be dichotomised (i.e. present or absent) or analysed as discrete categories and summarised with counts and percentages with 95% confidence intervals. Associations between candidate risk factors, modifiers, confounders, and confirmed septic arthritis will be analysed by Pearson chi-squared test or Fisher’s exact test (as appropriate) and presented as bivariable risk ratios (risk difference and relative risk) with 95% CIs.

The prediction rule will be constructed using data from the derivation cohort by first evaluating the bivariable association of individual variables with septic arthritis and retaining those with p values of less than 0·2 for subsequent modelling. A clinically sensible variable subset will be individually selected to identify children whose risk of septic arthritis exceeds 10%. The strength of association of these high-risk variables will be determined with logistic regression or bivariable Poisson regression with cluster robust error estimates as appropriate. These variables will then be used to categorise participants to a high-risk group; these participants will then be removed from further analysis. The remaining participants in the derivation cohort will be evaluated by binary recursive partitioning using machine learning (classification and regression tree (CART) analysis) to identify non-negligible risk factors for septic arthritis. Ten-fold cross-validation will be used to tune tree parameters for a more generalisable rule. In the CART analysis, we applied a misclassification cost of 250 to one for failure to identify a patient with septic arthritis versus incorrect classification of a patient without septic arthritis.

Rule performance measures and associated 95% CIs will be calculated by applying the derived prediction rule to the derivation and validation cohorts. Sensitivity, specificity, positive predictive value, negative predictive value, positive and negative likelihood ratios will be calculated.

We will perform a medical record review of participants in the derivation and validation cohorts who had been misclassified by the derived prediction rule. To evaluate the effect of potential bias in data collection and inclusion criteria, we evaluated the test characteristics of the prediction rule when applied to specific subsets of children. These subsets excluded children with suspected child abuse (for whom the time of injury and injury mechanism are unknown), children transferred to the study site from other EDs, children whose data collection occurred after limb imaging results were known, and children aged 9–15 years who might have findings more like those of adults.

Using the prediction rule factors, we will develop a clinical care algorithm for determining which children warrant pathology (blood) testing for further risk stratification, and for those who require imaging (formal ultrasound). We will construct the clinical algorithm to recommend imaging and blood tests for children presenting with high-risk factors and blood tests alone for children presenting with non-negligible risk factors identified in the CART analysis. In order to quantify the algorithm's potential effect on clinical decisions, we will group participants into three mutually exclusive groups: no testing (clinically cleared), blood tests obtained, and blood tests and imaging obtained. We will then calculate the blood testing and imaging rates observed in the study population and the imaging rates associated with applying the algorithm to the study population. We will use R and Python to conduct all analyses.

We will complement traditional analyses with machine learning approaches to identify the most predictive models for our data. Candidate methods will include boosting, decision trees, k-nearest neighbours, neural networks, random forests, support vector machines, and logistic regression. For feature selection, we will apply least absolute shrinkage and selection operator (LASSO) and elastic net (EN) techniques. Model hyperparameters will be tuned using cross-validation, and predictive performance will be evaluated by splitting the dataset into training (80%) and test (20%) sets to reduce overfitting. The best-performing model will then be retrained on the full dataset to estimate variable importance. Standardized variable importance measures will be presented using bar plots, and permutation tests will be conducted to assess their stability. Analyses will be conducted using R version 4.5.1 (https://cran.r-project.org/) within RStudio version 2025.05.1 Build 513 (https://posit.co/download/rstudio-desktop/), JASP (v0.95.0.0) (JASP Team, 2025), or Python 3.13 (http://www.python.org). As an alternative ensemble approach, we will also employ Super Learner to optimize predictive performance across candidate algorithms.

For prediction models, we will split data into training (80%) and test (20%) data set. The models will be fitted on the trained data and their performance will be assessed on the test data to avoid overfitting. Finally, the best model will be trained again on whole data to estimate the variable importance.

Recruitment
Recruitment status
Not yet recruiting
Date of first participant enrolment
Anticipated
Actual
Date of last participant enrolment
Anticipated
Actual
Date of last data collection
Anticipated
Actual
Sample size
Target
Accrual to date
Final
Recruitment in Australia
Recruitment state(s)
NSW,NT,QLD,SA,TAS,WA,VIC
Recruitment outside Australia
Country [1] 27986 0
New Zealand
State/province [1] 27986 0

Funding & Sponsors
Funding source category [1] 321488 0
Government body
Name [1] 321488 0
NHMRC
Country [1] 321488 0
Australia
Primary sponsor type
University
Name
Monash University
Address
Country
Australia
Secondary sponsor category [1] 324251 0
None
Name [1] 324251 0
Address [1] 324251 0
Country [1] 324251 0

Ethics approval
Ethics application status
Approved
Ethics committee name [1] 320021 0
Monash Health Human Research Ethics Committee A
Ethics committee address [1] 320021 0
Ethics committee country [1] 320021 0
Australia
Date submitted for ethics approval [1] 320021 0
17/11/2025
Approval date [1] 320021 0
13/03/2026
Ethics approval number [1] 320021 0

Summary
Brief summary
Trial website
Trial related presentations / publications
Public notes

Contacts
Principal investigator
Name 149170 0
Prof Simon Craig
Address 149170 0
Monash University Department of Paediatrics, Level 5, Monash Children's Hospital 246 Clayton Rd Clayton VIC 3168
Country 149170 0
Australia
Phone 149170 0
+61 03 9594 2707
Fax 149170 0
Email 149170 0
Contact person for public queries
Name 149171 0
Naomi Loftus
Address 149171 0
Monash University Department of Paediatrics, Level 5, Monash Children's Hospital 246 Clayton Rd Clayton VIC 3168
Country 149171 0
Australia
Phone 149171 0
+61 3 8572 3961
Fax 149171 0
Email 149171 0
Contact person for scientific queries
Name 149172 0
Simon Craig
Address 149172 0
Monash University Department of Paediatrics, Level 5, Monash Children's Hospital 246 Clayton Rd Clayton VIC 3168
Country 149172 0
Australia
Phone 149172 0
+61 03 9594 2707
Fax 149172 0
Email 149172 0

Data sharing statement
Will the study consider sharing individual participant data?
Yes
Will there be any conditions when requesting access to individual participant data?
Persons/groups eligible to request access:
• Individual researchers and/or collaborative research groups with ethical approval.

Conditions for requesting access:
• Yes, conditions apply:
• Requires review on a case-by-case basis by the trial custodian, sponsor or data sharing committee
• Requires a scientifically sound proposal or protocol
• Requires approval by an ethics committee
• Requires a data sharing agreement between data requester and trial custodian or sponsor
What individual participant data might be shared?
• De-identified individual participant data:
• Data dictionary

What types of analyses could be done with individual participant data?
• Systematic reviews and meta-analyses
When can requests for individual participant data be made (start and end dates)?
From:
12 months after publication of results.

To:
5 years after publication

Where can requests to access individual participant data be made, or data be obtained directly?
• The study team will review the request and consider the scientific merit of the proposed use of the data, and the legal, regulatory and ethical issues pertinent to the request. Presuming all constraints are addressed, the data will be shared using a secure file transfer platform.

For all data sharing enquiries, please contact Prof Simon Craig via email - [email protected]


Are there extra considerations when requesting access to individual participant data?
No


What supporting documents are/will be available?

No Supporting Document Provided


Results publications and other study-related documents

Documents added manually
No documents have been uploaded by study researchers.

Documents added automatically
No additional documents have been identified.