Pan African Clinical Trials Registry
South African Medical Research Council, South African Cochrane Centre
PO Box 19070, Tygerberg, 7505, South Africa
Telephone: +27 21 938 0835 or +27 21 938 0967
Email: pactradmin@mrc.ac.za Website: pactr.samrc.ac.za
Trial no.: PACTR202502517486411 Date of Registration: 24/02/2025
Trial Status: Registered in accordance with WHO and ICMJE standards
TRIAL DESCRIPTION
Public title Evaluating the Accuracy of AI-powered Paediatric TB Diagnostic Pathway: A two staged study (EVAL-PAEDTBAID)
Official scientific title Evaluating the Accuracy of AI-powered Paediatric TB Diagnostic Pathway: A two staged study (EVAL-PAEDTBAID)
Brief summary describing the background and objectives of the trial Background Pulmonary Tuberculosis (PTB) is a major health problem in children globally, with under-diagnosis leading to untreated disease, resulting in significant mortality and morbidity. Diagnosis is particularly challenging in young children due to paucibacillary disease, non-specific symptoms, and limited microbiological confirmation. Diagnosis of paediatric PTB is highly dependent on chest Xray (CXR) findings. There is limited expertise in interpreting CXRs in children, with significant inter- and intra-reader variability in identifying key signs of PTB. Data on the use of AI-assisted CXR interpretation for use in paediatric PTB remain scarce. Hypothesis: • AI-based CXR interpretation can accurately detect abnormalities of PTB and severity in children. • When aided by AI, clinicians will more accurately diagnose PTB in children on CXR compared to routine standard-of-care, unaided by AI. Objectives To prospectively enrol children up to 14 years presenting with suspected PTB to three study sites (2 in Pakistan, 1 in South Africa) into an observational diagnostics cohort. Analysis will be two staged: • STAGE I: Multi-site prospective diagnostic accuracy study using data from 300 confirmed and unconfirmed PTB cases and 250 non-TB LRTI cases using consecutive sampling. CXRs will be blindly analysed by 3 expert radiologists. • STAGE II: Multi-Reader Multi-Case study including 16 clinicians of varying expertise to investigate diagnosis of PTB on 350 CXRs when aided by AI compared to when unaided, using CXRs from stage 1.
Type of trial Observational
Acronym (If the trial has an acronym then please provide) EVAL PAEDTBAID
Disease(s) or condition(s) being studied Infections and Infestations,Paediatrics
Sub-Disease(s) or condition(s) being studied Tuberculosis
Purpose of the trial Diagnosis / Prognosis
Anticipated trial start date 01/03/2025
Actual trial start date 01/04/2025
Anticipated date of last follow up 31/10/2026
Actual Last follow-up date
Anticipated target sample size (number of participants) 550
Actual target sample size (number of participants)
Recruitment status Recruiting
Publication URL
Secondary Ids Issuing authority/Trial register
STUDY DESIGN
Intervention assignment Allocation to intervention If randomised, describe how the allocation sequence was generated Describe how the allocation sequence/code was concealed from the person allocating the participants to the intervention arms Masking If masking / blinding was used
Single Group Non-randomised Allocation Sequence/Code was not concealed Open-label(Masking Not Used)
INTERVENTIONS
Intervention type Intervention name Dose Duration Intervention description Group size Nature of control
Control Group Chest x ray classification Not applicable One time point The chest x-rays collected as part of the observational cohort will be examined by clinicians and classified in regards to the tuberculosis status with and without the use of a computer assisted platform. 550 Uncontrolled
ELIGIBILITY CRITERIA
List inclusion criteria List exclusion criteria Age Category Minimum age Maximum age Gender
1. Children with signs and symptoms suggestive of PTB aged 0-14 years presenting to study sites; AND 2. Children who have undergone a digital PA/AP CXR examination as part of diagnostic workup; AND 3. Informed consent and assent (in children over 7 yrs) 1. Children who are diagnosed positive for extrapulmonary TB (e.g. tuberculous meningitis) without any evidence of PTB, OR 2. Children who are currently on treatment for TB or those who have taken any treatment for TB in the 6-months prior to study site visit. In exceptions, children who have received less than three daily doses given since treatment initiation before the study site visit can be considered eligible. Child: 6 Year-12 Year,Infant: 1 Month(s)-12 Month(s),Infant: 13 Month(s)-24 Month(s),New born: 0 Day-1 Month,Preschool Child: 2 Year-5 Year 0 Month(s) 14 Year(s) Both
ETHICS APPROVAL
Has the study received appropriate ethics committee approval Date the study will be submitted for approval Date of approval Name of the ethics committee
Yes 06/12/2024 University of Cape Town Human Research Ethics Committee
Ethics Committee Address
Street address City Postal code Country
Faculty of Health Sciences Old Main Building, Groote Schuur Hospital Floor E53, Room 46 Rondebosch Cape Town 7700 South Africa
Has the study received appropriate ethics committee approval Date the study will be submitted for approval Date of approval Name of the ethics committee
Yes 06/01/2025 The Childrens Hospital University of Health Sciences
Ethics Committee Address
Street address City Postal code Country
Ferozepur Road Lahore 54000 Pakistan
Has the study received appropriate ethics committee approval Date the study will be submitted for approval Date of approval Name of the ethics committee
Yes 18/11/2024 Al Aleem Medical College
Ethics Committee Address
Street address City Postal code Country
Ferozepur Road Lahore 54000 Pakistan
OUTCOMES
Type of outcome Outcome Timepoint(s) at which outcome measured
Primary Outcome Sensitivity and specificity of AI in detecting confirmed and unconfirmed paediatric PTB cases (composite reference standard)* Following the period of statistical analysis after Stage I is completed
Secondary Outcome •Sensitivity and specificity of AI in in correctly identifying PTB cases using microbiological reference standard. •AUROC of AI using microbiological and composite reference standard. •Sensitivity, Specificity and AUROC of AI stratified by subgroups - age (0-9 years, 10-14 years), sex, severity of PTB (non-severe versus severe), and by study site. •Sensitivity, specificity and AUROC of AI versus radiological reference standard; overall and by subgroups. •Agreement statistics# between radiological reference standard and AI The end of Stage I of the study
Primary Outcome Difference in AUROC of readers with and without AI in classifying a CXR as consistent with PTB or not using radiological reference standard as reference standard. After Stage II of the study is completed
Primary Outcome Sensitivity and specificity of AI in detecting confirmed and unconfirmed paediatric PTB cases (composite reference standard)* Following the period of statistical analysis after Stage I is completed
Secondary Outcome •Sensitivity and specificity of AI in in correctly identifying PTB cases using microbiological reference standard. •AUROC of AI using microbiological and composite reference standard. •Sensitivity, Specificity and AUROC of AI stratified by subgroups - age (0-9 years, 10-14 years), sex, severity of PTB (non-severe versus severe), and by study site. •Sensitivity, specificity and AUROC of AI versus radiological reference standard; overall and by subgroups. •Agreement statistics# between radiological reference standard and AI The end of Stage I of the study
Secondary Outcome •Difference in sensitivity and specificity of readers with and without AI in correctly classifying CXRs into those which are consistent with PTB or not radiological reference standard as the reference standard. •Difference in accuracy of readers with and without AI based on WHO-TDA elements. •AUROC, sensitivity and specificity of readers in detecting radiological findings relevant for paediatric PTB . •Accuracy, AUROC, sensitivity and specificity stratified by subgroups, including type and experience of readers, and severity of disease (severe/non-severe), confirmed/unconfirmed PTB. The end of Stage II of the study
RECRUITMENT CENTRES
Name of recruitment centre Street address City Postal code Country
MRC Unit on Child and Adolescent Health Red Cross Childrens Hospital Rondebosch Cape Town 7700 South Africa
Gulab Devi Teaching Hospital Ferozepur Road Lahore Pakistan
University of Child Health Sciences Childrens Hospital Ferozepur Rd, Nishtar Town, Lahore Lahore Pakistan
FUNDING SOURCES
Name of source Street address City Postal code Country
Thrasher Research Fund 68 S. Main Street, Suite 400 Salt Lake City 84101 United States of America
SPONSORS
Sponsor level Name Street address City Postal code Country Nature of sponsor
Primary Sponsor Thrasher Research Fund 68 S. Main Street, Suite 400 Salt Late City 84101 United States of America Charities/Societies/Foundation
COLLABORATORS
Name Street address City Postal code Country
CONTACT PEOPLE
Role Name Email Phone Street address
Principal Investigator Heather Zar heather.zar@uct.ac.za +27216585319 Red Cross Childrens Hospital
City Postal code Country Position/Affiliation
Rondebosch Cape Town 7700 South Africa Principal Investigator
Role Name Email Phone Street address
Public Enquiries Cynthia Baard cynthia.baard@uct.ac.za +27216585518 Red Cross Childrens Hospital, REACH UNIT B11
City Postal code Country Position/Affiliation
Rondebosch Cape Town 7700 South Africa Project Manager
Role Name Email Phone Street address
Scientific Enquiries Cynthia Baard cynthia.baard@uct.ac.za +27216585518 Red Cross Childrens Hospital, REACH UNIT B11
City Postal code Country Position/Affiliation
Rondebosch Cape Town 7700 South Africa Project Manager
Role Name Email Phone Street address
Scientific Enquiries Brekhna Aurangzeb brekhna.aurangzeb@gmail.com +41774055343 15 Rue Gardiol Grand Sacconex
City Postal code Country Position/Affiliation
Geneva 1218 Switzerland Project Manager Pakistan Site
Role Name Email Phone Street address
Public Enquiries Atiqa Ambareen atiqaambren@gmail.coim 00923214536690 Ferozpur Road
City Postal code Country Position/Affiliation
Lahore Pakistan Professor of Microbiology
REPORTING
Share IPD Description Additional Document Types Sharing Time Frame Key Access Criteria
Yes Informed consent will be obtained from a parent/legal guardian and assent from a child from 7 years of age. A qualified staff member will conduct the informed consent and assent process with the parent/legal guardian of the child and the child. The child will complete a separate assent form, when applicable. The informed consent and assent process will be conducted before enrolment in their preferred language (South Africa: isiXhosa, Afrikaans, or English; Pakistan: Urdu) by trained study staff according to a standardised SOP. The consent and assent process will detail the purpose of the study, study procedures, and the risks and benefits to children and caregivers during the study. Consent will include permissions to abstract clinical data prospectively as well as for data sharing. Signed and witnessed documentation of informed consent and assent will be done. Following IRB guidelines in this setting, parents/legal guardians and children who are illiterate and unable to provide their signature will be asked to provide a thumbprint, which will be independently witnessed. Data and confidentiality All participants' information will be kept strictly confidential. All clinical and radiological data will be de-linked from identifying information and encrypted to ensure confidentiality of participants. All personnel involved in data collection and management undergo specific training for the study in confidentiality and related patient protection issues. Anonymous participant identification numbers are used on all study documents and all electronic records are kept in password-protected files to reduce access by anyone outside of the study. Informed Consent Form,Study Protocol Within 12 months of the study completion date of 30 January 2028. Dissemination will take place through the publication of study results in peer-reviewed journals. The study consortium (study team) reserves the rights (controlled access) to data collected as part of the study. Only anonymised data will be considered for sharing. The reference standard database will be open-source, to potentially facilitate faster evaluations and/or development of other paediatric TB AI models.
URL Results Available Results Summary Result Posting Date First Journal Publication Date
No
Result Upload 1: Result Upload 2: Result Upload 3: Result Upload 4: Result Upload 5:
Result URL Hyperlinks Link To Protocol
Result URL Hyperlinks
Changes to trial information
Section Name Field Name Date Reason Old Value Updated Value
Trial Information Actual trial start date 21/05/2025 The trial has started recruitment 01 Apr 2025
Section Name Field Name Date Reason Old Value Updated Value
Trial Information Anticipated date of last follow up 21/05/2025 The end of trial date has been updated to include a 3-month visit which is essential for classifying the TB status of participants. One more month has been added on to be inclusive of the full 30 day window period for the 3-month visit. 31 May 2026 31 Oct 2026
Section Name Field Name Date Reason Old Value Updated Value
Trial Information Recruitment status 21/05/2025 The trial has now begun recruitment. Not yet recruiting Recruiting
Section Name Field Name Date Reason Old Value Updated Value
Eligibility Inclusion criteria 21/05/2025 We updated the inclusion criteria to reflect that children must be exhibiting the signs and symptoms suggestive of pulmonary TB, rather than a diagnosis itself. We have also included details in terms of the ages that are required to sign assent in the participating countries. • Presumptive PTB children aged 0-14 years presenting to study sites who have undergone diagnostic workup for determining the TB disease status AND • Children who have undergone a digital PA/AP (posterior-anterior/anterior-posterior) view CXR examination as part of diagnostic workup AND • Children with PTB diagnosis information available (confirmed TB, unconfirmed TB, unlikely TB). AND • Signed written consent/assent or witnessed oral consent/assent in case of illiteracy. 1. Children with signs and symptoms suggestive of PTB aged 0-14 years presenting to study sites; AND 2. Children who have undergone a digital PA/AP CXR examination as part of diagnostic workup; AND 3. Informed consent and assent (in children over 7 yrs)
Section Name Field Name Date Reason Old Value Updated Value
Eligibility Exclusion criteria 21/05/2025 As having a CXR is an inclusion criteria, not having it has been removed from the list of exclusion criteria to remove redundancy. • Children who do not undergo CXR examination as part of the diagnostic workup OR • Children who are diagnosed positive for extrapulmonary TB (e.g. tuberculous meningitis) without any evidence for PTB OR • Children who are currently on treatment for TB, including tuberculosis preventive treatment (TPT), or those who have taken any treatment for tuberculosis in the 6-months prior to study site visit. In exceptions, children who have received up to three daily doses given since treatment initiation before the study site visit can be considered eligible. 1. Children who are diagnosed positive for extrapulmonary TB (e.g. tuberculous meningitis) without any evidence of PTB, OR 2. Children who are currently on treatment for TB or those who have taken any treatment for TB in the 6-months prior to study site visit. In exceptions, children who have received less than three daily doses given since treatment initiation before the study site visit can be considered eligible.
Section Name Field Name Date Reason Old Value Updated Value
Eligibility Age group 21/05/2025 PACTR Admin New born: 0 Day-1 Month, Infant: 1 Month-23 Month, Infant: 0 Month-23 Month, Preschool Child: 2 Year-5 Year, Child: 6 Year-12 Year New born: 0 Day-1 Month, Infant: 1 Month-23 Month, Infant: 0 Month-23 Month, Preschool Child: 2 Year-5 Year, Child: 6 Year-12 Year, Adolescent: 13 Year-18 Year
Section Name Field Name Date Reason Old Value Updated Value
Eligibility Age group 02/06/2025 PACTR Admin New born: 0 Day-1 Month, Infant: 13 Month(s)-24 Month(s), Infant: 0 Month(s)-12 Month(s), Preschool Child: 2 Year-5 Year, Child: 6 Year-12 Year New born: 0 Day-1 Month, Infant: 13 Month(s)-24 Month(s), Infant: 0 Month(s)-12 Month(s), Preschool Child: 2 Year-5 Year, Child: 6 Year-12 Year, Adolescent: 13 Year-18 Year
Section Name Field Name Date Reason Old Value Updated Value
Intervention Intervention List 21/02/2025 More details included Control Group, Unassisted chest xray classification , Not applicable, One time point, The chest x-rays collected as part of the observational cohort will be examined by clinicians and classified in regards to the tuberculosis status. , 550, Uncontrolled
Section Name Field Name Date Reason Old Value Updated Value
Intervention Intervention List 21/02/2025 Added more details Control Group, Unassisted chest xray classification , Not applicable, One time point, The chest x-rays collected as part of the observational cohort will be examined by clinicians and classified in regards to the tuberculosis status. , 550, Uncontrolled Control Group, Chest x ray classification , Not applicable, One time point, The chest x-rays collected as part of the observational cohort will be examined by clinicians and classified in regards to the tuberculosis status with and without the use of a computer assisted platform. , 550, Uncontrolled
Section Name Field Name Date Reason Old Value Updated Value
Outcome OutCome List 20/05/2025 Additional outcomes for Stage II of the study are included here Primary Outcome, Difference in AUROC of readers with and without AI in classifying a CXR as consistent with PTB or not using radiological reference standard as reference standard., After Stage II of the study is completed
Section Name Field Name Date Reason Old Value Updated Value
Outcome OutCome List 20/05/2025 The time point is defined as being following Stage I of the trial Primary Outcome, 1) Sensitivity (Se) and specificity (Sp) of AI in detecting confirmed and unconfirmed paediatric TB cases. , Following the period of statistical analysis after the trial completion. Primary Outcome, Sensitivity and specificity of AI in detecting confirmed and unconfirmed paediatric PTB cases (composite reference standard)* , Following the period of statistical analysis after Stage I is completed
Section Name Field Name Date Reason Old Value Updated Value
Outcome OutCome List 20/05/2025 The timepoint for the secondary outcomes has been updated Secondary Outcome, 1) Se and Sp of AI in in correctly identifying TB cases stratified by confirmed and unconfirmed TB. 2) AUROC of AI in detecting paediatric confirmed and unconfirmed paediatric TB cases (diseased cases) and also stratified by confirmed and unconfirmed TB. 3) Positive and negative predictive value (PPV and NPV) of AI in detecting paediatric TB. 4) Se, Sp and AUROC of AI stratified by subgroups - age (0-9 years, 10-14 years), gender, nutritional status, severity of TB (non-severe versus severe) and by study sites. 5) Se, Sp and AUROC of AI versus radiological reference standard; overall and by subgroups. 6) Agreement statistics between radiological reference standard and AI for the presence/absence of each finding - cavity, opacities, lymphadenopathy, pleural effusion and miliary pattern in the CXR., Following the analysis period at the end of the study 30 January 2028 Secondary Outcome, •Sensitivity and specificity of AI in in correctly identifying PTB cases using microbiological reference standard. •AUROC of AI using microbiological and composite reference standard. •Sensitivity, Specificity and AUROC of AI stratified by subgroups - age (0-9 years, 10-14 years), sex, severity of PTB (non-severe versus severe), and by study site. •Sensitivity, specificity and AUROC of AI versus radiological reference standard; overall and by subgroups. •Agreement statistics# between radiological reference standard and AI, The end of Stage I of the study
Section Name Field Name Date Reason Old Value Updated Value
Outcome OutCome List 20/05/2025 A new secondary endpoint is entered for Stage II of the study which was previously not defined. Secondary Outcome, •Difference in sensitivity and specificity of readers with and without AI in correctly classifying CXRs into those which are consistent with PTB or not radiological reference standard as the reference standard. •Difference in accuracy of readers with and without AI based on WHO-TDA elements. •AUROC, sensitivity and specificity of readers in detecting radiological findings relevant for paediatric PTB . •Accuracy, AUROC, sensitivity and specificity stratified by subgroups, including type and experience of readers, and severity of disease (severe/non-severe), confirmed/unconfirmed PTB., The end of Stage II of the study
Section Name Field Name Date Reason Old Value Updated Value
Contact People Contacs List 21/02/2025 Additional Contact Scientific Enquiries, Brekhna, Aurangzeb, Dr., brekhna.aurangzeb@gmail.com, brekhna.aurangzeb@student.uib.no, +41774055343, 15 Rue Gardiol Grand Sacconex, Geneva, 1218, Switzerland, Project Manager Pakistan Site
Section Name Field Name Date Reason Old Value Updated Value
Contact People Contacs List 21/02/2025 Additional Contact Public Enquiries, Atiqa, Ambareen, Prof., atiqaambren@gmail.coim, , 00923214536690, Ferozpur Road, Lahore, , Pakistan, Professor of Microbiology