A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income Countries that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Abstract

Introduction

While ultrasound-guided breast biopsy (UGBB) performed by a radiologist is the standard of care in high-income countries for diagnosing breast cancer, blind or surgical biopsy has been the norm in low-and middle-income countries (LMIC) in part because LMIC radiologists lack the skill to perform UGBB. We present the study protocol of a competency-based UGBB training program for LMIC Nigerian radiologists that leverages mobile health technology.

Methods

This institutional review board-approved prospective multi-institutional single-arm clinical trial (ClinicalTrials.gov identifier: NCT04501419) involves 13 Nigerian radiologists from eight tertiary hospitals in South West and South East Nigeria. Our training program is unique because it uses a competency-based curriculum developed specifically for LMIC radiologists. The competency-based curriculum incorporates blended learning (e-learning and trainer-led), simulation (supervised and unsupervised), and patient biopsy (supervised and unsupervised) components. The study time frame is two years: 1 year for the trainees to complete active training and patient recruitment and another 1 year for patient follow-up. Primary outcome measures include trainees’ competency (measured using the Ottawa Surgical Competency Operating Room Evaluation (O-SCORE)), the radiology-pathology concordance rate, and the complication rate. Secondary outcome measures include the diagnostic interval and the positive predictive value of UGBB.

Conclusion

Building capacity for UGBB in Nigeria and other LMIC can potentially improve breast cancer outcomes through early diagnosis. This training program is part of an implementation multi-component strategy package in Nigeria to improve breast cancer outcomes. This training program can also be adapted for other image-guided procedures that could impact global cancer control through diagnosis, therapeutic intervention, and/or palliation.

Introduction

Breast cancer is a major global health problem that disproportionately affects low-and-middle-income countries (LMIC). ¹ Nigeria, the most populous country in Africa, ² has the highest breast cancer mortality rate in the world. ³ Delayed diagnosis is a significant contributor to this disparity. ⁴ “Delayed diagnosis” refers to the delay between when a patient presents with breast symptoms and when they receive a tissue diagnosis, which is essential for treatment. ⁵ In the context of tissue diagnosis, “early” refers to the speed at which a cancer is pathologically diagnosed after the first encounter for symptoms, recommended by the World Health Organization to be less than 60 days. ⁶

Ultrasound for breast cancer detection and biopsy is an evidence-based intervention that can significantly reduce the time to diagnosis, presenting one of the most promising opportunities for improving breast cancer survival outcomes in LMIC. ⁷ However, in Nigeria, the lack of ultrasound-guided breast biopsy (UGBB) is a major barrier to early diagnosis. To address this, capacity-building strategies for UGBB are urgently needed.^8,9 UGBB is as accurate as surgical biopsy but preferred because it is minimally invasive, inexpensive, and has less morbidity. ⁷ Yet, despite being the standard of care in high-income countries (HIC), UGBB is not performed widely by radiologists in LMIC, largely due to the lack of necessary skills, leading to the reliance on blind biopsy (biopsy guided by palpation without imaging) or surgical excision.

To our knowledge, no established UGBB training program exists for LMIC. Building capacity in UGBB in LMIC requires ultrasound devices as well as a training program for radiologists. While ultrasound is the mainstay of diagnostic imaging in LMIC, ¹⁰ ultrasound devices are often in constant clinical use, leaving little availability for training. Until recently, the high cost of ultrasound devices has limited their availability in LMIC. ¹¹ However, new technologies, particularly United States Food and Drug Administration (FDA)-approved mobile health (mHealth) ultrasound devices, offer a low-cost alternative. ¹² These handheld, battery-operated devices are ideal for point-of-care imaging ¹² and present a viable option for breast cancer capacity-building training projects.

Training radiologists to perform UGBB requires long-term investment and innovative training methods to overcome several existing barriers. Because of the scarcity of radiologists trained in UGBB in LMIC and the urgent need for skilled providers, accelerated training programs leveraging both remote and on-site training opportunities are needed. A time-based curriculum akin to that used in HIC is not feasible for practicing LMIC radiologists, who are in great demand and want to incorporate this skill into their practice. There is also a need to develop clear and transparent competencies as well as metrics to guide certification and maintenance of skills. New skills have different individual learning curves. HIC surgical literature supports the use of competency-based training in this setting.^13-16 These methods have also proven effective in LMIC. ¹⁷

Given Nigeria’s high breast cancer mortality rate ³ and large population, ² it is an ideal setting for the initial implementation of a competency-based UGBB training program designed for LMIC. In 2013, the African Research Group for Oncology (ARGO), a United States National Cancer Institute (NCI)-recognized consortium, was established. ARGO is dedicated to collaborative research and has adopted a multidisciplinary approach to breast cancer diagnosis and treatment, with full access to core services, including ultrasound and pathology services. ¹⁸ ARGO radiologists have institutional and multidisciplinary (i.e., surgeons, oncologists, and pathologists) support for the clinical implementation of UGBB once trained. Notably, none of the ARGO radiologists could perform an UGBB as of early 2017. ARGO radiologists indicated that they had no prior training in UGBB, that a training program is the most important need, and that local barriers to a training program were time, access to ultrasound devices for training, and the lack of skilled trainers. ⁸ With the availability of ultrasound and pathology services, the main barrier to capacity building in UGBB for ARGO radiologists was no longer ultrasound or pathology services but the lack of an established UGBB training program for LMIC radiologists.

To address the urgent need for a UGBB training program in Nigeria and other LMIC to allow early diagnosis, which would potentially improve breast cancer survival outcomes, an initial feasibility study was conducted with one Nigerian radiologist (ADO) and one trainer (EJS). From 2017-2018, the Nigerian radiologist successfully completed competency-based UGBB training involving blended learning, as well as supervised and unsupervised simulation and patient biopsies. Insights from the feasibility study have been adapted to address context-specific obstacles and unforeseen clinical challenges in the design of a competency-based training program that leverages mobile health to build capacity for UGBB in LMIC. This manuscript presents the protocol of our study to train LMIC Nigerian radiologists in UGBB using our competency-based UGBB training curriculum that leverages mHealth technology. This is a 2-year study that includes our training curriculum and patient follow-up. The training curriculum has been completed while patient follow-up is ongoing.

We hypothesize that this study will build capacity for UGBB among Nigerian radiologists, which may improve breast cancer outcomes by decreasing the diagnostic interval (time from presentation with symptom to pathology diagnosis).

Methods

Aims

The goal of this study is to establish a training program to train Nigerian radiologists in UGBB to address the problem of late breast cancer diagnosis in LMIC. Specifically, the study aims to develop a scalable competency-based training curriculum to train LMIC radiologists in UGBB by leveraging mobile health technology. The reporting of this study conforms to SPIRIT guidelines. ¹⁹

Ethical Considerations and Approvals

This study is Health Insurance Portability and Accountability Act (HIPAA)-compliant. Memorial Sloan Kettering Cancer Center (MSK) has provided IRB approval (approval number 18-114 [File S1] and 21-325 [File S2]). The approved study protocols (initially as Appendix E of an umbrella 18-114 protocol [File S3] and subsequently as an independent 21-325 protocol [File S4]) allow the use of MSK Research Electronic Data Capture (REDCap) to capture the data collected during the unsupervised patient-based biopsies. Appendix E of the umbrella 18-114 study protocol has also been approved by the institutional review board at Obafemi Awolowo University Teaching Hospitals Complex (OAUTHC), where the study is domiciled (File S5). In addition, the seven additional hospitals of the participating radiologists have obtained IRB approval for the unsupervised patient-based biopsies. Any important protocol modifications will be communicated to the relevant parties via monthly study calls or via email/phone if sooner notification is warranted. The study is registered at clinicaltrials.org (clinicaltrials.gov identifier: NCT04501419) (registration at clinicaltrials.org was completed after the enrollment of participants began because of unanticipated administrative delays). All research for all enrolled participants is conducted under full ethical and administrative institutional approvals. The authors confirm that all ongoing and related trials for this intervention are registered.

Study Design

This prospective multi-institutional single-arm non-randomized trial will involve the training of thirteen board-certified radiologists at eight ARGO tertiary-care teaching hospitals in South West and South East Nigeria (File S6). A SPIRIT schedule of enrollment, interventions, and assessments is presented in Figure S1.

Inclusion Criteria

Trainees

Board-certified Nigerian radiologists with a full-time appointment in one of the eight selected government tertiary care hospitals in Nigeria.

Patients

Women aged 18 years and older with a breast ultrasound demonstrating a solid mass that is suspicious for cancer (Breast Imaging Reporting and Data System (BI-RADS) 4 and 5 masses), ²⁰ which would typically undergo either a blind biopsy or surgical excision at the Nigerian hospital where the patient is seeking diagnosis.

Exclusion Criteria

Trainees

Board-certified Nigerian radiologists without a full-time appointment in one of the eight selected government tertiary hospitals in Nigeria. Non-radiologist physicians will be excluded from the trainee group.

Patients

Women under the age of 18 years and/or medical reason precluding study participation.

Consent

Written informed consent will be obtained for all study participants, whether radiologist trainees (by the principal investigators—EJS and ADO) or patients (by the radiologist trainees) after reading and understanding the participant information sheet attached to the consent form (File S7). All questions and clarifications raised by the radiologist trainees will be addressed by the principal investigators (EJS and ADO) while those raised by the patients will be addressed by the radiologist trainees.

Sample Size

The sample size of radiologists will be determined by the number of Nigerian ARGO sites interested in building capacity for UGBB and the availability of interested radiologists at those sites; no power calculation will be performed. Eight Nigerian ARGO sites have shown interest in building capacity for UGBB, and thirteen eligible radiologists from these eight sites have consented to be participate as trainees in this study.

For the training, each radiologist trainee will be required to enroll a minimum of 25 patients each with a suspicious breast mass on diagnostic breast ultrasound. The investigators agreed on the 25 as the minimum required by each trainee based on insights from the feasibility study and considering their busy work schedules with their full-time appointments in their respective hospitals. Using the ARGO platform, the investigators will collaborate with breast surgeons in each study site to facilitate patient recruitment for the patient biopsies and patient retention for the patient follow-up.

The choice of 13 radiologists and 25 patients per radiologist was purely based on the feasibility of recruiting trainee radiologists and the maximum anticipated availability of suspicious cases, respectively. No statistical sample size calculations were performed.

Study Activities and Approach

This is a 2-year study that includes a competency-based training curriculum that leverages mobile health technology and patient follow-up.

Trainees

Thirteen board-certified Nigerian radiologists with a full-time academic appointment at eight tertiary hospitals in South West and South East Nigeria will be consented to be trainees. Like in the initial feasibility study, trainees agree to be future trainers if successfully trained.

Trainers

The training program will be led by one board certified fellowship-trained breast radiologist from MSK (EJS) and one board-certified Nigerian radiologist (ADO) who successfully completed training as part of the initial feasibility study, leveraging train-the-trainer techniques. The training program will be additionally supported by five board-certified fellowship-trained breast radiologists from MSK.

Design of the Competency-Based Ultrasound-Guided Breast Biopsy Training Program that Leverages Mobile Health Technology

The curriculum of the training program will be tailored to Nigerian hospitals and defined based on consensus opinion after reviewing American College of Radiology (ACR) practice parameters, ²¹ the ACR Breast Imaging Reporting and Data System (BI-RADS) atlas, ²⁰ and the MSK Breast Imaging Fellowship Curriculum. Curriculum development will be facilitated in collaboration with content experts and the MSK Patient-Reported Outcomes, Community-Engagement and Language (PRO-CEL) core facility to address the specific barriers and targeted needs of LMIC.

The training program will be organized into three components: (1) blended learning: e-learning and trainer-led, (2) simulation: supervised and unsupervised, and (3) patient biopsy: supervised and unsupervised (Figure S2).

Supervised training is when the trainer is physically present “elbow-to-elbow” with the trainee and unsupervised is when they are not. Each component has 2 subcomponents, making a total of six steps to be completed by each trainee in the training curriculum.

Blended learning

Blended learning will involve a combination of e-learning and trainer-led modules, which cover knowledge, skills, and attitudes required for competent practice. Course content includes (a) ultrasound physics and quality assurance, (b) breast anatomy, (c) ultrasound characterization of breast masses, which includes the BI-RADS standard lexicon ²⁰ (d) patient positioning, (e) universal precautions, (f) management of sharps and biohazards, (g) biopsy complications, (h) radiology–pathology concordance, (i) management recommendations, (j) communicating bad news to patients, and (k) mHealth ultrasound training. The e-learning module will be created on a premium, affordable, and accessible e-learning management application called talentlms (Epignosis 2018), ²² which has over 1 million users. Talentlms is cloud-based, user-friendly, scalable, and allows trainers to remotely supervise progress of trainees. Talentlms delivers across any device including Samsung Galaxy smartphones and can be downloaded for free. E-learning content consists of lectures, videos, and 3-D animations to guide self-directed simulation. We aim to provide foundational knowledge and practical skills through an interactive and engaging platform that will supplement the lack of UGBB-trained radiologists in LMIC. As part of the blended learning component, trainees will also be required to observe a minimum of three complete UGBB procedures performed by one of the trainers.

Simulation

Because of the lack of UGBB-trained radiologists in LMIC, we will incorporate simulation-based training whereby trainees will practice UGBB on a breast phantom containing multiple masses with graded interventional difficulty. Simulation-based training is linked to better academic performance. It provides a method for consolidating learning and developing clinical reasoning. Simulation provides the opportunity to develop skills, i.e., competencies, in performing procedures, and builds confidence and clinical decision-making. Simulation is less costly, time efficient, requires fewer personnel, allows for individual performance tracking, and is associated with fewer medical errors and improved patient safety compared with training with actual patients. ²³ Simulation biopsy is where the trainee practices UGBB skills on a breast phantom with different masses providing varying levels of difficulty.

Patient biopsy

Following successful completion of simulation biopsy, trainees will perform UGBB in actual patients. During the supervised patient biopsy stage, narrative-based supervision will be employed to address complex clinical scenarios and allow trainees to ask questions, fostering a learner-centered environment. ²⁴ During the unsupervised patient biopsy stage, which will take places at the trainees’ hospitals, trainees will enroll a minimum of 25 patients each with a suspicious breast mass on diagnostic breast ultrasound. Trainees will obtain informed consent (File S7), collect clinicopathologic variables into a case report form (File S8), perform UGBB, document any complications, determine radiology-pathology concordance, communicate pathology results to patients, and navigate patients to surgeons if the biopsy result is malignant. Trainees will also document study outcome metrics for breast cancer including clinical stage, treatment, and overall survival. Repeat biopsies and/or surgical excision will be done for discordant benign (imaging features suspicious for malignancy but benign pathology) lesions. Follow-up with trainees will occur at three-month intervals for one year as they clinically implement UGBB. Trainees will also be engaged in a semi-structured interview over video conference to elicit feedback on the course structure and barriers and facilitators to clinical implementation.

Of note, during the training program, trainees will attend two separate on-site training components (trainer-led/supervised simulation and supervised patient biopsy) for a combined 10 days of travel to OAUTHC, which is where the program is domiciled and where the Nigerian IRB of record is located. The other training components (e-learning, unsupervised simulation, and unsupervised patient biopsy) will be completed at the trainees’ hospitals. The instructor-led and e-learning modules are expected to be completed in four weeks, the supervised and unsupervised simulation-based training in seven weeks, the supervised patient biopsy in one week, and the unsupervised patient biopsies in 36 weeks. Trainees are expected to complete the training program in 12 months (Figure S1), and this will be the first year of the study period.

Training Program Competency Assessment

Competency assessments will be conducted during the supervised simulation, supervised patient biopsy, and unsupervised patient biopsy stages in this consecutive order. Each competency step must be completed, and a passing score obtained before progressing on to the next. Because there is no reference standard to assess simulation and patient biopsy, the Ottawa Surgical Competency Operating Room Evaluation (O-SCORE), a validated tool, has been selected and adapted for the assessment of trainee competency (File S9). UGBB is a skill analogous to a minor operation; thus, it was felt to be the best means of assessing competencies. There are 8 variables rated on a scale of 1-5 ²⁵ and we will use a passing criterion of 80%. The simulation biopsy competency assesses whether trainees are competent to perform supervised patient biopsies. The supervised patient biopsy competency assesses whether trainees are competent to perform unsupervised patient biopsies at their hospital, as part of the prospective single-arm clinical trial described above. The simulation and patient biopsy examinations will be run by two raters to account for inter-rater reliability. As part of the unsupervised patient biopsy competency assessment, the trainees will be required to document complications and radiology-pathology concordance.

Scope/ Definition of Terms

For this study, complications will be defined as the number of (1) hematomas requiring intervention, (2) infections requiring treatment, and (3) other complications requiring intervention, occurring within 14 days of the procedure. Radiology-pathology concordance will be defined as the agreement between ultrasound and pathologic findings, such that the pathology satisfactorily explains the imaging findings.^26,27

mHealth Ultrasound Device

For this study, the HIPAA-compliant Phillips Lumify L12-4 Hz broadband linear array transducer (Phillips 2018) will serve as the mHealth ultrasound device. The Phillips Lumify L12-4 Hz broadband linear array transducer is a high-frequency probe with a scan depth of up to 12 cm and is well suited for high-resolution breast imaging (Figure S3). It is compatible with encryption and data security systems and can comply with enterprise data security policies to ensure HIPAA compliance. The ultrasound application will be downloaded onto a 10.5-inch Samsung Galaxy Tablet (Samsung 2018); the tablet will be connected by USB-C to the Phillips Lumify L12-4 Hz broadband linear array transducer when ready to scan.

Ultrasound-Guided Breast Biopsy Procedure

UGBB will be performed as follows: The Phillips Lumify L12-4 Hz broadband linear array transducer will be placed on the skin of the breast to identify the breast mass, which will be displayed on the tablet. Once the breast mass is identified, an antiseptic solution will used to clean the skin overlying the mass area and both superficial and deep injection of local anesthetic will be given to the area. Once the area is numb, the radiologist will make a small incision (surgical cut) in the breast and then insert the ultrasound core biopsy device (12- or 14-needle gauge) into the breast mass under ultrasound guidance to obtain sequential tissue samples. To obtain a representative sampling of the breast mass, a minimum of three core biopsy specimens will be required per breast mass. Core biopsy specimens will be placed in a container filled with formalin, sealed, and taken to the pathology department.

Pathology Assessment

All core biopsy specimens will undergo standard-of-care processing as well as staining at the pathology department to reach a definitive diagnosis. We will recommend to all trainees that all pathology samples demonstrating breast cancer undergo immunohistochemistry to evaluate for hormone receptors (estrogen and progesterone) and HER2 overexpression.

Patient Follow-Up

Patient follow-up will take place in year two of the study for 12 months. Using the ARGO platform, the investigators will collaborate with breast surgeons in each study site to facilitate patient retention and complete patient follow-up. Additionally, the study research assistants in each of the participating site will track these patients via phone call and their clinic visits. Outcome data will be collected for all patients. Patients who are diagnosed with a benign mass will undergo follow-up at 12 months to ensure the biopsy was a true negative. Patients who are diagnosed with a benign high-risk lesion will undergo surgical excision and pathology will be documented. Outcome variables will be collected for patients that are diagnosed with breast cancer including breast cancer stage, treatment received (surgery, neoadjuvant/adjuvant chemotherapy, radiation therapy), recurrence status, and vital status.

Invariably, there will be challenges with patient follow-up due to various factors including relocation, change in telephone numbers, or a patient opting for alternative medicine. To mitigate these challenges, the following strategies will be put into place: (1) research training and capacity-building including a quality assurance and audit program as well as regular study meetings, (2) updates to the electronic data recording system that will utilize branching logic and hard stops as well as internal team trackers to support timely follow-up on data gaps, and (3) enhanced data collection at time of consent. Patients will be asked their preferred method of contact for follow-up and will also be asked to provide an alternative contact method or telephone numbers to help reach them for follow-up purposes.

When data points are missing, we have developed a missing data codes standard operating procedure, whereby we will follow a formal procedure to handle and track missing data.

Program Evaluation and Outcome Measures

The study time frame is two years: one year for the trainees to complete active training and patient recruitment and another one year for patient follow-up. The following outcomes have been defined a priori and will be used as metrics to assess the impact of the training program.

Primary Outcome Measures

1. Trainees’ competency: This will be assessed using the O-SCORE during the supervised simulation and during supervised patient and unsupervised patient biopsies. An O-SCORE of 80% and above in each of the above-mentioned phases will be taken as the passing score for competency.

2. Radiology-pathology concordance rate: The radiology-pathology concordance rate will be assessed by the percentage of lesions with imaging features determined by trainees to be concordant with pathology out of the total lesions biopsied at the end of the study period.

3. Complication rates: Complication rates will be assessed by the percentage of biopsies with complications requiring interventions within 14 days out of the total number of biopsies done during the study period.

Secondary Outcome Measures

1. Diagnostic interval: This will be defined by the median time from the presentation of the patients to the time of pathologic diagnosis.

2. Positive predictive value of UGBB: This will be assessed by the number of lesions with imaging features determined by trainees to be positive for malignancy that have malignant pathology (true positives) divided by the number of lesions with imaging features suspicious for malignancy (true positives plus false positives).

Data Collection and Management

According to local research regulations, anonymized patient data obtained during the unsupervised patient biopsy stage will be recorded on paper forms and stored in a secure, locked area to ensure data security and patient privacy initially. Subsequently, a unique identifier will be assigned to each patient record and anonymized patient data will be transferred and stored in a REDCap database by trained research assistants at the various study sites. Research assistants will be provided with REDCap project server login details, allowing them to submit data to the REDCap database securely. The REDCap server will be managed by MSK. Only the principal investigator and research collaborators will have access to the final study dataset.

Statistical Analysis

Data interpretation will be performed in concert with the MSK Department of Epidemiology and Biostatistics, with the type I error rate (α) for all statistical tests set to 0.05.

Trainee characteristics including age, number of years in practice, prior UGBB experience, and hospital infrastructure will be summarized using frequencies and percentages. We will derive a descriptive summary (medians, interquartile ranges) of the trainee scores. The performance of the trainees will be evaluated for the simulation and patient biopsy components using a passing criterion of 80% on the adapted O-SCORE ¹³ for each component. The choice of 13 radiologist trainees from 8 hospitals was chosen based on the feasibility of recruitment. No statistical sample size calculations were performed, and the results of this study will be used to inform future power analysis. Complication rates will be compared informally with complication rates reported in literature. For example, the usual rate of hematomas is around 10%, <1% for infections, and <1% for other complications. ²⁰ Radiology-pathology concordance will be listed as a percentage of lesions that are pathologically concordant. 95% confidence intervals (CIs) for concordance will be evaluated using the bootstrapping technique to account for intra-patient correlation which is inherent in this analysis where there are multiple readings (i.e., multiple biopsies) per patient.

The positive predictive value (PPV) of mHealth UGBB will be estimated at the biopsy level, using the CLUSTER statement in SAS 9.4, which accounts for patient-level random effects. A true-positive biopsy will be defined as a suspicious breast mass identified on mHealth ultrasound that underwent UGBB and yields malignant concordant pathology. In the case of a benign discordant result (where ultrasound features are suspicious for malignancy and pathology is benign and does not explain the ultrasound findings) and a repeat biopsy is recommended, the measure of agreement will be based on the first discordant biopsy. A false-positive biopsy will be defined as a suspicious breast mass identified on mHealth ultrasound that underwent UGBB and yields benign discordant pathology, and then excision of the entire lesion gives a benign pathology or remains unchanged on either clinical or imaging follow-up at 12 months or more. Data from the patients lost to follow-up will be excluded from PPV analyses.

Patient stage at diagnosis and treatment among patients who were biopsied will be summarized using frequencies and percentages, while time between first medical presentation and diagnosis will be summarized using medians and interquartile ranges.

Clinical variables such as patient age, family history of breast cancer, breast symptom(s) at presentation, duration of symptoms, and size of breast mass will be compared between benign and malignant pathology (gold standard) diagnosis at the patient level using the Wilcoxon rank sum test for continuous variables (i.e., patient age, duration of symptoms, and size of breast mass) and Fisher’s or Chi-square test for categorical variables (i.e., family history of breast cancer, and presence or absence of breast symptoms at presentation).

Protocol and Data Monitoring Team

The principal investigators (EJS and ADO) will be primarily responsible for the clinical trial. EJS and ADO are part of ARGO, a consortium recognized by the United States NCI. ARGO’s mission includes generating data to inform regional evidence-based management recommendations, investigating prevention and early detection strategies, increasing access to cancer care, and improving cancer care training in rural and underserved communities. OAUTHC serves as the major Nigerian hospital in this consortium, while MSK is the principal center in the United States. MSK’s research activities within ARGO are coordinated through the MSK Global Cancer Disparities Initiatives (GCDI).

The protocol monitoring team (PMT) will oversee protocol and data monitoring for this study. The PMT comprises a clinical research manager, a research project manager, and clinical research coordinators from MSK who are dedicated to GCDI efforts, as well as a program manager from OAUTHC who is dedicated to ARGO efforts. The PMT’s responsibilities will span a wide range, including overseeing project compliance, data collection, abstraction and entry, data reporting, regulatory monitoring, and problem resolution and prioritization. Moreover, the PMT will monitor the study for safety and quality. Given that the study uses an FDA-approved mHealth tablet ultrasound device, an independent data monitoring committee will not be formally constituted.

Patient-level data will be managed using REDCap, a secure data management software system. REDCap allows the creation of customized, secure data management systems, including web-based data entry forms, reporting tools, and a comprehensive set of security features. These features, such as user- and group-based privileges and a complete audit trail of data manipulation and export procedures, are designed to protect the integrity and confidentiality of the data. The MSK REDCap administrative team will ensure the highest level of data security. At MSK, the REDCap system is maintained on locally owned servers in a locked server room with unique environmental controls, supported by an uninterrupted power supply, and backed up nightly with some backup tapes stored off-site. All connections to REDCap are encrypted (SSL-based). REDCap will only be used to house survey and patient-level data, and its use has been approved by the MSK webmaster. Source documentation will be available to support the electronic record.

Quality Assurance

The PMT will generate weekly registration reports to monitor patient accrual and ensure the completeness of registration data. Routine data quality reports will be created to identify missing data and inconsistencies. Patient accrual rates, evaluation accuracy, and follow-up completeness will be reviewed periodically throughout the study period. Any potential issues will be promptly reported to the PMT for discussion and corrective action. Additionally, the PMT will conduct random-sample data quality and protocol compliance audits at least twice a year or more frequently if necessary.

Data Availability

While patients are being accrued and prior to any publication reporting the results of the study, study data will not be made publicly available. Data requests can be made to the corresponding author to access anonymized patient data underlying each publication, between 9 months and 36 months after each publication, and will be granted based on scientific merit. After approval of a data request, data will be shared via a secure online platform after establishing a data sharing agreement. Authorship eligibility will be commensurate with involvement in study design, data analysis, and manuscript development. There is no intended use of professional writers.

Project Status and Timeline

The project was originally designed to take place over a 2-year period (Figure S1). However, the COVID-19 pandemic significantly altered the project status and timeline. The recruitment period for the 13 radiologists began on September 1, 2019, and ended on November 16, 2019, which was the start of the training program in Ile-Ife, Nigeria. All 13 radiologists are still considered on study until patient follow-up is complete. Patient enrollment began on September 7, 2020, and concluded on August 25, 2022. All enrolled patients are still on study for follow-up and data quality assurance by the PMT, which is ongoing and expected to conclude on November 10, 2024.

Discussion

Training LMIC radiologists to perform UGBB requires long-term investment and innovative methods to overcome several existing structural barriers. Given the scarcity of trained radiologists in LMIC and the urgent need for skilled providers, we aimed to develop a competency-based approach tailored to Nigerian hospitals whereby LMIC radiologists from Nigeria can be trained successfully to perform and clinically implement UGBB by leveraging mHealth technology.

Our training program is unique because it uses a competency-based curriculum developed specifically for LMIC radiologists. Considering the need for sustainability and scalability, the training program leverages mHealth ultrasound devices. mHealth ultrasound devices are affordable, scalable, enable independent learning, and can be used for both training and clinical purposes. ¹² The battery-operated mHealth ultrasound devices offer 4.5 hours of continuous scan time, and although electricity in Nigeria is sometimes interrupted, it is generally reliable for recharging batteries. ²⁸ Additionally, in our initial technology acceptability and usage survey administered to Nigerian radiologists from all six geopolitical zones, 94% of participants reported that their hospitals had backup generators in case of power outages. The survey also indicated that Nigerian radiologists are “very comfortable” or “extremely comfortable” in using mobile applications and smart devices, which supports anticipated success with the e-learning component of our training program. Furthermore, most of the training program does not depend on an internet connection. Nevertheless, the majority (75%) of ARGO sites have reported having broad-based internet solutions/Wi-Fi access “the entire day” or “most of the day.” ²⁸

The training program also allows for a self-propagating “train-the-trainer” model, which is essential for long-term scalability. Nigerian radiologists who are trained have agreed to become trainers in subsequent studies. One of our long-term goals is to validate the train-the-trainer model of our training program.

As noted earlier, the training program uses a competency-based curriculum incorporating blended learning (e-learning and trainer-led), simulation (supervised and unsupervised), and patient biopsy (supervised and unsupervised) components. The traditional time-based curriculum used in HIC is not feasible for practicing LMIC radiologists, who are limited in number and time but seek to incorporate this skill into their practice. HIC radiologists are trained to perform UGBB using a time-based approach of a minimum of several months in residency and/or fellowship programs where they work “elbow-to-elbow” with their trainer. However, this is not feasible for the few LMIC radiologists. Thus, accelerated training programs leveraging both remote and on-site training opportunities are needed. HIC surgical literature supports the use of accelerated competency-based training.^13-16 These methods have also proven effective in LMIC. ¹⁷ Simulation-based training particularly enables trainees to acquire vital skills through adequate training sessions. Within the competency-based curriculum, simulation-based training allows radiologists to develop skills in performing procedures independently on a breast phantom. This ensures expertise before training on patients, which reduces the rates of complications that might occur in actual patients during training.

There are several limitations to this study. Firstly, the single-arm, non-randomized design limits the ability to draw causal inferences. The study would be strengthened by having a control group, such as radiologists receiving the reference standard training program. Unfortunately, to our knowledge, no UGBB training program exists. Radiologists in mostly HIC are trained to perform UGBB in residency and/or fellowship training programs where time-based practice guidelines are followed with no established curriculum. Secondly, no statistical sample size calculations were performed for either radiologist trainees or patients as the numbers were chosen based upon feasibility. We hope the results of this pilot study will inform a larger appropriately powered study. Thirdly, we adapted the O-SCORE, which is a clinically validated tool; however, the study would be strengthened if we had validated it for use in the context of UGBB and LMIC conditions. Finally, the training program may not be generalizable to all LMIC as key program requirements include radiologists, Wi-Fi access, and pathology services.

While the sample size of trainees limits the study, the successful training of 13 Nigerian radiologists is a promising start with the hope that this study can demonstrate that the different components of the training program can all work together, showcasing the effectiveness of a comprehensive training program for radiologists in Nigeria. This training program, designed for board-certified radiologists with expertise in ultrasound and breast imaging, can serve as a model for similar training programs in other LMICs. To make this program more generalizable, we must acknowledge that not all LMIC have enough board-certified radiologists and therefore we could adapt our program for broader applicability by offering it to non-radiologist clinicians or other health workers. Since non-radiologist clinicians or other health workers do not have baseline ultrasound skills, our program would need to incorporate a competency-based approach to the foundational proficiency in ultrasound, which is essential for the success of such a program.

Relevant bodies like the Breast Imaging Society of Nigeria (BISON) were involved in the project design from the outset to facilitate the implementation and integration of the project’s results into practice. The study team is committed to rapidly sharing research results with investigators, the cancer research community, and relevant stakeholders. The results of our study will be disseminated to the scientific community through publication in international peer-reviewed journals. Findings from this project will also be disseminated to all stakeholders; we will share our results with BISON and the Association of Radiologists in Nigeria (ARIN) to promote broader implementation across Nigeria. The African Organization for Research and Training in Cancer (AORTIC) provides a valuable platform for disseminating research relevant to LMICs. We aim to present our findings at the biennial AORTIC conference. We will also submit our results for presentation at other local and international conferences of relevant bodies, such as the National Postgraduate Medical College of Nigeria, West African College of Surgeons, BISON and the American Society of Clinical Oncology. This project will utilize the existing ARGO platform, which currently includes 31 Nigerian and 4 North American hospitals, to disseminate and implement the project’s results.

Conclusion

We anticipate that this training program will demonstrate that LMIC radiologists can be effectively trained to perform and clinically implement UGBB. This competency-based approach, utilizing innovative remote and on-site training methods and mobile health technology, may offer a promising solution for building capacity for UGBB to enhance early diagnosis and improve the survival outcomes of breast cancer patients in LMIC.

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control

Supplemental Material

Supplemental Material - A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol

Supplemental Material for A Competency-Based Ultrasound-Guided Breast Biopsy Training Program for Radiologists From Low-and-Middle-Income that Leverages Mobile Health Technology (NCT04501419): A Study Protocol by Adeleye Dorcas Omisore, Adedeji Ayoola Egberongbe, Lydia Eleanor Pace, Sughra Raza, Rachael Adeyanju Akinola, Millicent Olubunmi Obajimi, Varadan Sevilimedu, Yolanda Bryce, Victoria Lee Mango, Olusegun Isaac Alatise, T. Peter Kingham, Elizabeth Anne Morris, and Elizabeth Jane Sutton in Cancer Control