R-TF-015-006 Clinical investigation report

Research Title

Project to enhance Dermatology E-Consultations in Primary Care Centres using Artificial Intelligence Tools.

Description

Study to validate the clinical utility of the device as a valid tool in managing patients with skin pathology by primary care practitioners and dermatologists at the Puerta de Hierro Majadahonda University Hospital involved in the project. In this study patients with suspected pathologies of the following in Primary Care: tumour pathology, inflammatory pathology and infectious pathology will be recruited. All of them will be treated at either Pozuelo Health Center or Majadahonda Health Center and referred to Puerta del Hierro Majadahonda University Hospital. For this study, 100 patients will be recruited.

Product Identification

	Information
Device name	Legit.Health Plus (hereinafter, the device)
Model and type	NA
Version	1.1.0.0
Basic UDI-DI	8437025550LegitCADx6X
Certificate number (if available)	MDR 000000 (Pending)
EMDN code(s)	Z12040192 (General medicine diagnosis and monitoring instruments - Medical device software)
GMDN code	65975
EU MDR 2017/745	Class IIb
EU MDR Classification rule	Rule 11
Novel product (True/False)	TRUE
Novel related clinical procedure (True/False)	TRUE
SRN	ES-MF-000025345

Sponsor Identification and Contact

	Manufacturer data
Legal manufacturer name	AI Labs Group S.L.
Address	Street Gran Vía 1, BAT Tower, 48001, Bilbao, Bizkaia (Spain)
SRN	ES-MF-000025345
Person responsible for regulatory compliance	Alfonso Medela, Saray Ugidos
E-mail	office@legit.health
Phone	+34 638127476
Trademark	Legit.Health
Authorized Representative	Not applicable (manufacturer is based in EU)

Identification of sponsors

• Sponsor: Puerta de Hierro Health Research Institute

Identification of the Clinical Investigation Plan (CIP)

	CIP
Title of the clinical investigation	Project to enhance Dermatology E-Consultations in Primary Care Centres using Artificial Intelligence Tools
Device under investigation	Legit.Health Plus
Protocol version	Version 1.0
Date	2022-06-29
Protocol code	LEGIT.HEALTH_DAO_Derivación_PH_2022
Sponsor	Instituto de Investigación Sanitaria Puerta de Hierro
Coordinating Investigator	Dr. Gastón Roustan Gullón
Principal Investigator(s)	Dr. Gastón Roustan Gullón
Investigational site(s)	Pozuelo and Majadahonda Health Centers and Puerta del Hierro Majadahonda University Hospital
Ethics Committee	Comité de Ética de la Investigación con Medicamentos del Hospital Universitario Puerta del Hierro de Majadahonda

Trial Registrations

• ClinicalTrials.gov (NCT): NCT07429123
• EMA RWD Catalogue (EUPAS): EUPAS108166

Public Access Database

The database used in this study is not publicly accessible due to privacy and confidentiality considerations.

Research Team

Principal Investigators

• Dr. Gastón Roustán Gullón (Hospital Universitario Puerta del Hierro Majadahonda).

Collaborating Investigators

• Centro de Salud Majadahonda
  • Dra. Esther Minguela
• Centro de Salud de Pozuelo
  • Dr. Fernando León
• Hospital Universitario Puerta del Hierro Majadahonda
  • Dr. Ángel Rosell Díaz

Investigational sites

• Hospital Universitario Puerta del Hierro de Majadahonda.
• Centro de Salud de Majadahonda.
• Centro de Salud de Pozuelo.

Compliance Statement

The clinical investigation will be conducted according to the Clinical Investigation Plan (CIP) and other applicable guidances and regulations. This includes compliance with:

• The ethical principles originating from the World Medical Association's Declaration of Helsinki
• Harmonized standard UNE-EN ISO 14155:2020
• Regulation (EU) 2017/745 on medical devices (MDR), including the applicable General Safety and Performance Requirements (GSPR) as outlined in Annex I, and the requirements of Annex XV (Chapter I and Chapter II, Section 3)
• Harmonized standard UNE-EN ISO 13485:2016
• MDCG 2024-3 for its structural and content expectations, MDCG 2021-8 concerning application requirements, and MDCG 2020-10/1 Rev 1 for safety reporting timelines and definitions
• Regulation (EU) 2016/679 (GDPR)
• Spanish Organic Law 3/2018 on the Protection of Personal Data and guarantee of digital rights.

All data processing within the device is carried out in accordance with the highest standards of data protection and privacy. Patient information is managed in an encrypted manner to ensure confidentiality and security.

The research team assumes the role of Data Controller, responsible for the collection and management of study data. Legit.Health acts as the Data Processor and is not involved in the processing of patient data.

The storage and transfer of data comply with European data protection regulations. At the conclusion of the study, all information stored in the device will be permanently and securely deleted.

The device employs robust technical and organizational security measures to safeguard personal data against unauthorized access, alteration, loss, or processing.

Report date

2024-01-10

Report author(s)

The full name, the ID and the signature for the authorship, as well as the approval process of this document, can be found in the verified commits at the repository. This information is saved alongside the digital signature, to ensure the integrity of the document.

Table of contents

Table of contents

• Research Title
• Description
• Product Identification
• Sponsor Identification and Contact
• Identification of sponsors
• Identification of the Clinical Investigation Plan (CIP)
  • Trial Registrations
• Public Access Database
• Research Team
  • Principal Investigators
  • Collaborating Investigators
  • Investigational sites
• Compliance Statement
• Report date
• Report author(s)
• Table of contents
• Abbreviations and definitions
• Summary
  • Title
  • Introduction
  • Objectives
    • Hypothesis
    • Primary objective
    • Secondary objectives
    • Endpoint
  • Population
  • Design and Methods
    • Study design
    • Number of subjects
    • Ethical considerations
    • Data quality assurance
    • Description of the clinical investigation methods
  • Results of the clinical investigation
  • Conclusion of the clinical investigation
• Introduction
• Investigational device and methods
  • Investigational device description
    • Device version under investigation and bridging to the CE-marked release
  • Clinical Investigation Plan
    • Objectives
    • Design
    • Ethical considerations
    • Data Quality Assurance
    • Subject Population
    • Treatment
    • Concomitant Medication/Treatment
    • Follow-Up Duration
    • Statistical Analysis
    • Derived metrics (for comparison)
• Results
  • Initiation and completion date
  • Subject and Investigational Product Management
  • Subject Demographics
  • Clinical Investigation Plan (CIP) Compliance
  • Analysis
    • Recruitment
    • Referral Appropriateness
    • Diagnosis
    • General Performance (malignancy detection)
    • Clinical Utility
    • Adverse Events and Adverse Reactions to the Product
    • Product Deficiencies
    • Subgroup Analysis for Special Populations
    • Protocol Deviations
• Discussion and Overall Conclusions
  • Clinical Performance, Efficacy, and Safety
    • Acceptance-criteria pass/fail summary
    • Primary-objective evaluability
    • Benefit-risk assessment against GSPR 1 and GSPR 8
    • Satisfaction, limitations and operational caveats
  • Clinical Relevance
  • Specific Benefit or Special Precaution
  • Implications for Future Research
  • Limitations of Clinical Research
  • Ethical Aspects of Clinical Research
    • Data quality assurance
• Investigators and Administrative Structure of Clinical Research
  • Brief Description
  • Investigators
    • Principal investigator
    • Collaborators
    • Center
  • External Organization
  • Sponsor and Monitor
• Report Annexes

Abbreviations and definitions

• AE: Adverse Event
• AEMPS: Spanish Agency of Medicines and Medical Devices
• AEP: Adverse Reaction to Product
• AUC: Area Under the ROC Curve
• CAD: Computer-Aided Diagnosis
• CMD: Data Monitoring Committee
• CIP: Clinical Investigation Plan
• CUS: Clinical Utility Questionnaire
• DLQI: Dermatology Quality of Life Index
• GCP: Standards of Good Clinical Practice
• ICH: International Conference of Harmonization
• IFU: Instructions For Use
• IRB: Institutional Review Board
• N/A: Not Applicable
• NCA: National Competent Authority
• PI: Principal Investigator
• PPV: Positive Predictive Value
• NPV: Negative Predictive Value
• SAE: Serious Adverse Events
• SAEP: Serious Adverse Event to Product
• SUAEP: Serious and Unexpected Adverse Event to the Product
• SUS: System Usability Scale

Summary

This is a prospective analytical observational study of a series of clinical cases designed to validate the clinical utility of the device in the management of patients with skin pathology by primary care practitioners and dermatologists at the Puerta de Hierro Majadahonda University Hospital involved in the project. This study, 100 patients were recruited from primary care centres who could potentially suffer from tumoral, inflammatory or infectious diseases. The primary objective is to validate that the information provided by the device increases the true accuracy of healthcare professionals (HCPs) for the diagnosis of multiple dermatological conditions.

Title

Project to enhance Dermatology E-Consultations in Primary Care Centres using Artificial Intelligence Tools.

Introduction

Skin diseases represent about 5% of primary care (PC) visits, primarily among the active population, consuming significant healthcare resources. There is a notable diagnostic discrepancy between PC doctors and dermatologists, with agreement rates around 57–65%, partly due to limited dermatological training among PC doctors, leading to misdiagnoses and unnecessary referrals. Given the scarcity of dermatologists, especially in rural areas, PC doctors often perform preliminary assessments of skin conditions, though their diagnostic capacity is limited. To improve dermatology access and efficiency, teledermatology initiatives like teleDERMADRID have been launched in Madrid, allowing PC centres to send clinical images and information to dermatologists, reducing in-person consultations, travel, and delays in diagnosis and treatment. In parallel, AI-based decision-support tools have been associated with improved PCP referral adequacy and patient-triage accuracy in published primary-care evaluations [10][11].

This study focuses on evaluating whether the information provided by the device increases the accuracy of primary care practitioners in the diagnosis of different skin conditions. With a diverse cohort of 100 patients, rigorous ethical standards, and adherence to regulatory guidelines, this investigation evaluates whether the device can improve the diagnostic accuracy of primary care practitioners for dermatological conditions.

Objectives

Hypothesis

The device significantly improves the appropriateness of dermatology referrals. This is owing to the introduction of significant changes in the diagnostic process, such as greater sensitivity and specificity than a primary care practitioner in diagnosing skin conditions, especially in differentiating between malignant and benign lesions, and it also provides the reassurance of a second medical opinion, which has been clinically validated.

Primary objective

To validate that the information provided by the device increases the true accuracy of healthcare professionals (HCPs) for the diagnosis of multiple dermatological conditions.

Secondary objectives

• Reduce and correct the referral of patients with skin pathologies from primary care to dermatology.
• Individualize and improve the ongoing training of primary care practitioners in dermatology.
• Offer healthcare adapted to technological innovations.
• Measure the satisfaction of primary care practitioners with the device.
• Measure the satisfaction of dermatologists with the device.

Endpoint

• AUC (area under the ROC curve) equal to or greater than 80.00% detecting malignancy.(User Group: Primary care practitioners)
• increase in the adequacy of referrals equal to or greater than 15.00%.(User Group: Primary care practitioners)
• Expert consensus equal to or greater than 70.00%.(User Group: Primary care practitioners)

Population

Adult patients (≥ 18 years) with skin pathologies seen at their Health Centres (Majadahona or Pozuelo Health Centers) and whose reference hospital is Puerta del Hierro Majadahonda University Hospital. These patients should be diagnosed at their health centre and diagnosed with tumoral, inflammatory or infectious diseases.

Design and Methods

Study design

This is a prospective analytical observational study of a series of clinical cases.

Number of subjects

At least 15 primary care practitioners made at least one diagnostic report using the device. In total, 180 diagnostic reports were recorded for 131 patients. In the review carried out in June 2023, the number of patients was 92, and as of the date of this report, the target of 100 has been exceeded.

Ethical considerations

Ethics Committee Approval

Approved by CEIm of Puerta de Hierro University Hospital on 2022-06-24, reference number 47/395984.9/22.

This study adhered to international Good Clinical Practice (GCP) guidelines, the Declaration of Helsinki in its latest amendment, and applicable international and national regulations. As applicable, approval from the relevant Ethics Committee was obtained prior to the initiation of the study. When applicable, modifications to the protocol were reviewed and approved by the Principal Investigator (PI) and subsequently evaluated by the Ethics Committee before subjects were enrolled under a modified protocol.

This study was conducted in compliance with European Regulation 2016/679, of 27 April, concerning the protection of natural persons with regard to the processing of personal data and the free movement of such data (General Data Protection Regulation, GDPR), and Organic Law 3/2018, of 5 December, on the Protection of Personal Data and the guarantee of digital rights. In accordance with these regulations, no data enabling the personal identification of participants was collected, and all information was managed securely in an encrypted format.

Participants were informed both orally and in writing about all relevant aspects of the study, with the information being tailored to their level of understanding. They were provided with a copy of the informed consent form and the accompanying patient information sheet. Adequate time was given to patients to ask questions and fully comprehend the details of the study before providing their consent.

The PI was responsible for the preparation of the informed consent form, ensuring it included all elements required by the International Conference on Harmonisation (ICH), adhered to current regulatory guidelines, and complied with the ethical principles of GCP and the Declaration of Helsinki.

The original signed informed consent forms were securely stored in a restricted access area under the custody of the PI. These documents remained at the research site at all times. Participants were provided with a copy of their signed consent form for their records.

Data confidentiality

Current legislation will be complied with in terms of data confidentiality protection (European Regulation 2016/679, of 27 April, on the protection of natural persons with regard to the processing of personal data and the free movement of such data and Organic Law 3/2018, of 5 December, on Personal Data Protection and guarantee of digital rights). For this purpose, when applicable, each participant will receive an alphanumeric identification code in the study that will not include any data allowing personal identification (coded CRD). The Principal Investigator will have an independent list that will allow the connection of the identification codes of the patients participating in the study with their clinical and personal data. This document will be filed in a secure area with restricted access, under the custody of the Principal Investigator and will never leave the centre.

Once the paper CRDs are completed and closed by the Principal Investigator, the data will be transferred to a database.

As in the CRDs, the Database will comply with current legislation in terms of data confidentiality protection (European Regulation 2016/679, of 27 April, on the protection of natural persons about the processing of personal data and the free movement of such data and Organic Law 3/2018, of 5 December, on the Protection of Personal Data and guarantee of digital rights) in which no data allowing personal identification of patients will be included.

Data quality assurance

The Principal Investigator is responsible for reviewing and approving the protocol, signing the Principal Investigator commitment, guaranteeing that the persons involved in the centre will respect the confidentiality of patient information and protect personal data, and reviewing and approving the final study report together with the sponsor. All the clinical members of the research team assess the eligibility of the patients in the study, inform and request written informed consent, collect the source data of the study in the clinical record and transfer them to the Data Collection Notebook (DCN) or Data Collection Forms (CRF).

Description of the clinical investigation methods

Primary Care Practitioner Recruitment

The Principal Investigator and/or the collaborating investigators assigned to this task will explain to the primary care practitioners the purpose of their participation in the study, and practitioners will be able to ask any questions they consider necessary to clarify their doubts about the study. Additionally, all practitioners participating in the study will receive specific information on the use of the artificial intelligence tool, provided by an expert in the field.

If a doctor wishes to participate in the study, they will sign the Informed Consent form, and a study code will be assigned to them. Once the informed consent is signed, the data collection process begins. The Principal Investigator and/or the collaborating investigators assigned to this task will collect demographic data (age, centre, years of practice) and explain to the doctor the steps they must follow as part of the study.

Patient Identification

The primary objective of the study is to assess the perceived clinical utility of the doctors, so the clinical variables of the patients are not the focus of the study. Therefore, patients will only be identified in the tool to enable its use, and no further analysis or data processing of the patient's information will take place.

The primary care doctor will explain to the patient what their participation in the study would entail through the Patient Information Sheet. The patient, in turn, can ask any questions they deem appropriate to clarify their doubts about the study.

If the patient wishes to participate in the study, they will sign the Informed Consent form and be assigned a study code. Once the informed consent is signed, the data collection process begins. The Principal Investigator and/or the collaborating investigators assigned to this task will enter demographic data (age, sex) and information related to the diagnosis, characteristics, and treatment of the condition into the device.

Procedures to be Performed by Primary Care Practitioners

Consultation

Primary care practitioners must take photographs showing the areas affected by the condition. These photographs will be taken with their own smartphones or a mobile dermatoscope if their use is clinically relevant. Primary care practitioners will upload the photographs to the device.

They will assess the patient's condition through direct observation and guided by the platform's results. They will enter the diagnosis in the device by selecting the pathology if it is among the top five options or through a dropdown menu displaying a wide range of conditions.

Besides the probable diagnoses, practitioners will have access to referral criteria, a clinical referral questionnaire, and basic treatment recommendations. When deemed necessary, they will refer the patient in their Selene system, as they normally would, not through the device.

Completion of Questionnaires

Primary care practitioners will complete the Clinical Utility and Satisfaction Questionnaire twice during the study: once at 2 months and again at 4 months after the start of the study.

Procedures to be Performed by Specialist Doctors

Consultation

The Principal Investigator will see their patients while maintaining their standard clinical routine. They will note the appropriateness of the patient referral in the Case Report Form (CRF). If they consider the referral to dermatology inappropriate, this will also be recorded. The initial diagnosis provided by the primary care doctor will be documented, as well as the specialist's diagnosis. If relevant, the pathology diagnosis will be included additionally.

Completion of questionnaires

Dermatologists will complete the Clinical Utility and Satisfaction Questionnaire twice during the study: once at 2 months and again at 4 months after the start of the study.

Results of the clinical investigation

15 primary care practitioners were included in this study. In total, 180 diagnostic reports were recorded for 131 patients. In diagnosing specific conditions such as hidradenitis suppurativa (HS), the device helped identify two cases initially undiagnosed by primary care practitioners (PCPs), with one confirmed later by a dermatologist. Additionally, HS appeared as a suggested diagnosis in four other reports. For urticaria, the tool guided one confirmed diagnosis, with no further suggestions. For psoriasis, eight cases were confirmed, with the tool suggesting psoriasis in 23 instances, resulting in 10 diagnoses by PCPs.

In skin cancer cases, specifically melanoma, five cases were confirmed through pathology, with dermatologists suspecting melanoma in 10 cases. Using the device, specialists achieved a sensitivity of 60% and specificity of 91% for melanoma detection. Overall, the tool's AUC for malignancy detection (including melanoma and carcinomas) was 0.84. Clinicians completed a survey on the clinical utility of the device, with responses from 8 primary care practitioners and 2 dermatologists.

Conclusion of the clinical investigation

Three pre-specified secondary acceptance criteria were evaluated, and each was met against its pre-specified threshold: malignancy-detection AUC reached 0.842 (95% CI [0.82, 0.86]) against a ≥ 0.80 threshold; the relative change in the adequacy of referrals (proportion of cases correctly assigned to primary-care management) rose from 26.66% to 33.33%, a +25.03% relative change against a ≥ +15% relative threshold; and expert-consensus clinical utility reached 0.80 against a ≥ 0.70 threshold. The CIP primary objective — whether the device increases PCPs' own true diagnostic accuracy — was not evaluable due to a major protocol deviation on baseline capture of unaided PCP diagnoses, which produced missing baseline data and broke the pairing integrity of the primary analysis; confirmatory evaluation of the primary endpoint is deferred to the Post-Market Clinical Follow-up Plan (R-TF-007-002) and the post-market observational study R-TF-015-012. Supporting descriptive data on melanoma-specific sensitivity and specificity are reported with wide exact confidence intervals on small denominators and do not substantiate a melanoma-specific performance claim on their own. No adverse events and no product deficiencies were observed under the surveillance mechanisms described in §Adverse Events and §Product Deficiencies. PCPs and dermatologists reported positive scores on the Clinical Utility and Satisfaction Questionnaire, with response rate and non-response caveats disclosed. This investigation supplies Pillar 3 real-patient prospective clinical-performance evidence at Rank 2–4 under MDCG 2020-6 Appendix III for the three secondary acceptance criteria above; the CIP primary endpoint remains a PMCF deliverable.

Introduction

Skin-related diseases are a common reason for primary care (PC) consultations, estimated to account for about 5% of all visits, mostly among the working population. This significant use of resources highlights the need for an efficient approach to optimize primary care operations.

Numerous studies reveal discrepancies between PC doctors and dermatologists, with diagnostic agreement rates ranging between 57% and 65.5%, depending on the study. In general, PC doctors demonstrate limited expertise in skin disease diagnosis and treatment. This knowledge gap impacts the time and effort required to assess the severity and stage of a patient's condition, often resulting in low protocol adherence and inadequate referrals. Given the shortage of medical professionals, especially dermatologists—averaging just three per 100,000 inhabitants—access to specialists is particularly challenging in smaller communities, leaving PC doctors to screen dermatological conditions, despite an increased risk of misdiagnosis. Studies show a diagnostic discordance rate between 55% and 65% between PC doctors and specialists, with frequent misdiagnoses of common dermatological diseases by non-dermatologists, such as drug-induced rashes and fungal infections.

Additionally, patient self-reporting introduces potential bias, especially when treatment relies on patient-reported information, and healthcare teams often lack the resources to verify the accuracy of these reports. Since 2020, healthcare information systems have included an asynchronous e-consultation for dermatology, allowing clinical information and images from health centres to be sent to the dermatology department at Hospital Puerta de Hierro Majadahonda, integrated within the hospital's Selene system.

The gradual introduction of teledermatology as part of the strategic plan in the Madrid region culminated in the teleDERMADRID project, led by Pilar Sánchez Pobre Bejarano and Manuel Grandal Martín (GAOAIO), in coordination with DGASA. This initiative offers a new, non-face-to-face consultation mode, adapted to the current social and technological landscape. It includes an Image Acquisition Management (IAM) tool, accessible via QR code on mobile devices, linked to the integrated request system (SIPE), facilitating fast, generally satisfactory responses to health issues, reducing diagnostic delays, unnecessary travel, and expediting treatment access.

Reducing the pressure on in-person hospital consultations lowers conventional healthcare costs while upholding patient-centred care as a priority to improve health-related quality of life. Alongside the image tool, a clinical form is used to complement image submissions with necessary items.

The development of Information and Communication Technologies (ICTs) has enabled advancements in medicine through AI-based decision-support tools, including automation of patient-triage support to prioritise severe cases and refine primary-care referrals for skin conditions. Published evaluations of such tools in primary care report measurable improvements in PCP referral adequacy and triage accuracy [10][11].

Thus, this study aims to clinically validate an innovative AI tool to enhance the appropriateness of primary care referrals to dermatology.

Investigational device and methods

Investigational device description

This section contains a short summary of the device. A complete description of the intended purpose, including device description, can be found in the record Legit.Health Plus description and specifications.

Product description

The device is a computational software-only medical device leveraging computer vision algorithms to process images of the epidermis, the dermis and its appendages, among other skin structures. Its principal function is to provide a wide range of clinical data from the analyzed images to assist healthcare practitioners in their clinical evaluations and allow healthcare provider organisations to gather data and improve their workflows.

The generated data is intended to aid healthcare practitioners and organizations in their clinical decision-making process, thus enhancing the efficiency and accuracy of care delivery.

The device should never be used to confirm a clinical diagnosis. On the contrary, its result is one element of the overall clinical assessment. Indeed, the device is designed to be used when a healthcare practitioner chooses to obtain additional information to consider a decision.

Intended purpose

The device is a computational software-only medical device intended to support health care providers in the assessment of skin structures, enhancing efficiency and accuracy of care delivery, by providing:

• quantification of intensity, count, extent of visible clinical signs
• interpretative distribution representation of possible International Classification of Diseases (ICD) categories.

Intended previous uses

No specific intended use was designated in prior stages of development.

Product changes during clinical research

The device maintained a consistent performance and features throughout the entire clinical research process. No alterations or modifications were made during this period.

Device version under investigation and bridging to the CE-marked release

The version of the device used throughout this clinical investigation is v1.1.0.0. This is the sole CE-marked release placed on the market, and the version under investigation is identical to the marketed release; an identity bridge is therefore applicable and no clinical-relevance assessment of version changes is required. The Person Responsible for Regulatory Compliance (PRRC) confirms this identity position. Throughout this document, references to "the device" refer to the investigational product identified above.

Clinical Investigation Plan

Objectives

This study aims to validate that the information provided by the device increases the true accuracy of healthcare professionals (HCPs) for the diagnosis of multiple dermatological conditions.

Design

This is a prospective observational analytical study of a clinical case series. The study does not involve an active or control group, as it is focused on the evaluation of the device in a real-world clinical setting. The group of primary care practitioners will act as their control group, first without using the device and making their diagnosis and afterwards using the device to confirm or change their diagnosis. The assessment criteria include the completion of questionnaires, photograph submission, and patient-reported outcomes through the device. The study employs a variety of methods, including data collection through questionnaires, photograph analysis, and patient satisfaction surveys.

Ethical considerations

This study adhered to international Good Clinical Practice (GCP) guidelines, the Declaration of Helsinki in its latest amendment, and applicable international and national regulations. As applicable, approval from the relevant Ethics Committee was obtained prior to the initiation of the study. When applicable, modifications to the protocol were reviewed and approved by the Principal Investigator (PI) and subsequently evaluated by the Ethics Committee before subjects were enrolled under a modified protocol.

This study was conducted in compliance with European Regulation 2016/679, of 27 April, concerning the protection of natural persons with regard to the processing of personal data and the free movement of such data (General Data Protection Regulation, GDPR), and Organic Law 3/2018, of 5 December, on the Protection of Personal Data and the guarantee of digital rights. In accordance with these regulations, no data enabling the personal identification of participants was collected, and all information was managed securely in an encrypted format.

Participants were informed both orally and in writing about all relevant aspects of the study, with the information being tailored to their level of understanding. They were provided with a copy of the informed consent form and the accompanying patient information sheet. Adequate time was given to patients to ask questions and fully comprehend the details of the study before providing their consent.

The PI was responsible for the preparation of the informed consent form, ensuring it included all elements required by the International Conference on Harmonisation (ICH), adhered to current regulatory guidelines, and complied with the ethical principles of GCP and the Declaration of Helsinki.

The original signed informed consent forms were securely stored in a restricted access area under the custody of the PI. These documents remained at the research site at all times. Participants were provided with a copy of their signed consent form for their records.

Data Quality Assurance

The Principal Investigator is responsible for reviewing and approving the protocol, signing the Principal Investigator commitment, guaranteeing that the persons involved in the centre will respect the confidentiality of patient information and protect personal data, and reviewing and approving the final study report together with the sponsor. All the clinical members of the research team assess the eligibility of the patients in the study, inform and request written informed consent, collect the source data of the study in the clinical record and transfer them to the Data Collection Notebook (DCN) or Data Collection Forms (CRF).

Subject Population

Adult patients (>18 years) who come to their primary care centre with tumour pathology, inflammatory pathology or infectious pathology. They can be treated in the Health Center of Pozuelo and the Health Center of Majadahonda. It is important to note that some patients presented with more than one issue, contributing to the total count.

Inclusion Criteria

Patients suspected of having the following conditions in Primary Care:

• Tumor pathology:
  • Benign:
    • Histiocytoma
    • Seborrheic keratosis
    • Angiomas
  • Precancerous:
    • Actinic keratosis
  • Suspected malignancy:
    • Basal cell carcinoma
    • Squamous cell carcinoma
  • Pigmented lesions:
    • Melanocytic nevus
    • Malignant melanoma
• Inflammatory pathology:
  • Psoriasis
  • Atopic dermatitis
  • Urticaria
  • Hidradenitis suppurativa
  • Lichen planus
• Infectious pathology:
  • Viral warts
  • Molluscs
  • Herpes simplex
• Patients aged 18 years or older.
• Patients who have signed the informed consent for the study.

Exclusion Criteria

• Patients under 18 years of age.
• Pregnant patients.
• Patients who, in the opinion of the researcher, will not comply with the study procedures.

Sample size

The goal of this study is to evaluate whether the use of the device by HCPs can improve the diagnostic accuracy of skin pathologies by at least 10%. The 10% improvement refers to an increase in the proportion of correct diagnoses compared to baseline performance. This level of improvement would provide a strong justification for a meaningful change in clinical practice. Consequently, 15 HCPs will participate in this study, which provides a strong basis and ensures sufficient statistical power. For this reason, assuming the diagnostic accuracy is measured as a proportion (e.g., correct diagnoses), a sample size of 100 patients can provide a 95% confidence level with a margin of error of around 9-10% for the accuracy rate and 80% of power.

The sample size is sufficient for evaluating qualitative assessments, such as evaluating usability, satisfaction, and clinical utility, using validated scales and descriptive statistics. Each HCP will interact with 6–7 patients on average (100 patients distributed across 15 doctors), ensuring a manageable workload and thorough evaluation of each case. However, subgroup analysis may be limited by the small number of participants per group. The design provides a solid foundation for detecting meaningful improvements in diagnostic accuracy, HCP satisfaction and exploring the usability of the device in clinical settings.

Population

100 patients were intended to be recruited in this study. Finally, 180 diagnostic reports of 131 patients were recorded, exceeding the sample size planned at the beginning. This study is proposed as a pilot proof-of-concept study in which the sample size has been estimated based on the number of primary care practitioners who treat patients diagnosed with skin diseases who can be treated in the primary care service of the following centres:

• Majadahonda Health Center
• Pozuelo Health Center

During the recruitment period of the study, all patients diagnosed with skin diseases who met the selection criteria were included.

Treatment

Patients in this study did not receive any specific treatment as part of the research protocol.

Concomitant Medication/Treatment

Patients continued their regular prescribed medications and treatments as directed by their primary healthcare providers. No additional medications or treatments were administered as part of this study.

Follow-Up Duration

In this study, there will be no follow-up period since it is a cross-sectional study. The ethics committee will be notified of the start of the study. Annual monitoring reports will be sent subsequently. After obtaining the conclusions of the study, a final report will be prepared and presented to the ethics committee.

The PI or investigators completed the CUS (Appendix I of the protocol) as required by the study protocol.

Statistical Analysis

To estimate mean responses and their variability, we calculated the mean and standard deviation for all questions. Additionally, to evaluate the secondary objective concerning the CUS value, we conducted a descriptive analysis so as to know the opinion of the HCPs regarding the Clinical Utility of the device.

Derived metrics (for comparison)

The following table summarises the metrics derived from the manufacturer's performance-claims dataset for this investigation, for side-by-side comparison with the acceptance criteria declared in the CIP.

Error: Could not determine study code. Provide studyCode or folderSlug prop, or ensure this component is used within an Investigation document with a registered folder slug.

Results

Initiation and completion date

• Initiation: 2022-06-24
• Completion: 2024-01-10

Subject and Investigational Product Management

The CIP pre-specified 100 patients as the intended sample. Actual enrolment reached 131 patients and 180 diagnostic reports; the extension beyond the planned 100 patients was not covered by a formal CIP amendment and is declared as a post-hoc analytical-unit decision under §Protocol Deviations. The analysis unit per endpoint is stated explicitly in §Clinical Performance, Efficacy, and Safety: the malignancy-detection AUC and expert-consensus endpoints use the report-level and respondent-level units respectively, and the referral-adequacy endpoint uses the 120 diagnostic reports for which both the unaided PCP triage response and the dermatologist adjudication were available.

Investigational product management for this software medical device consisted of controlled web-based access to the device through individual authenticated user accounts issued by the manufacturer to each PCP. Access provisioning and revocation events are recorded in the manufacturer's access-control log held within the Trial Master File. No physical product accountability is applicable.

Subject Demographics

The study did not place specific emphasis on gender, age, or nationality as primary factors of investigation. Instead, it encompassed a diverse patient cohort with a range of chronic dermatologic pathologies, providing a comprehensive representation of the population affected by these conditions. Along with this, it includes the feedback received by the HCPs regarding the Clinical Utility of the device.

Fitzpatrick Phototype	Count	Percentage
Phototype I	63	48.3%
Phototype II	48	36.7%
Phototype III	16	12.2%
Phototype IV	3	2.2%
Phototype V	1	0.6%
Phototype VI	0	0.0%

The cohort includes only one Fitzpatrick V participant and no Fitzpatrick VI participants. Accordingly, this investigation does not, on its own, support performance claims on Fitzpatrick V or VI skin. Phototype coverage for darker skin tones is addressed at Clinical Evaluation level (R-TF-015-003) and by the Post-Market Clinical Follow-up Plan (R-TF-007-002); the IFU is aligned with the dedicated phototype-bridging evidence.

Clinical Investigation Plan (CIP) Compliance

At the end of the study, the pre-specified secondary acceptance criteria were all evaluated, and the associated performance claims are reported against their pre-specified thresholds in §Clinical Performance, Efficacy, and Safety. A major protocol deviation, however, prevented direct evaluation of the CIP primary objective: assessing whether the device increases PCPs' own diagnostic accuracy. Many PCPs did not record their initial unaided diagnosis, instead relying on the device output from the start, which produced missing baseline data for the paired without-device vs with-device comparison and broke the pairing integrity of the primary analysis. Confirmatory evaluation of the PCP-accuracy primary endpoint is therefore deferred to the Post-Market Clinical Follow-up Plan (R-TF-007-002) and the post-market observational study (R-TF-015-012), which will enforce software-level blinding of the device suggestion until the unaided diagnosis is submitted. The present investigation remains a Pillar 3 real-patient prospective clinical-performance investigation at Rank 2–4 under MDCG 2020-6 Appendix III for the three secondary acceptance criteria that were evaluable (malignancy AUC, relative change in the adequacy of referrals, and expert-consensus satisfaction).

Analysis

Recruitment

At least 15 primary care practitioners provided at least one diagnostic report using the device. A total of 180 diagnostic reports were registered for 131 patients. In the review conducted in June 2023, the patient count was 92, and as of this report, the target of 100 patients has been exceeded.

Referral Appropriateness

Three potential patient pathways were defined:

1. Primary care management,
2. Dermatology consultation,
3. Surgical management.

Once referral criteria were established, diagnostic reports were reviewed to assign each case to its appropriate pathway. For each diagnostic report, a referral decision was classified as avoidable (i.e., the case could have been managed in primary care) when the primary care practitioner's response matched that of the dermatologist and the case was assigned to primary-care management by the dermatologist adjudication; all other referrals were classified as necessary. Under this classification, a higher avoidable-referral percentage corresponds to a greater proportion of cases correctly kept in primary care — i.e., a positive change in the adequacy of referrals. Results on the 120 diagnostic reports for which both the unaided PCP response and the dermatologist adjudication were available are summarised below:

Question #	Avoidable Referral	Necessary Referral
Without using the device	32 (26.66%)	88 (73.33%)
Using the device	40 (33.33%)	80 (66.66%)

The proportion of cases correctly identified as manageable within primary care rose from 26.66% (32/120) without the device to 33.33% (40/120) with the device, corresponding to a +25.03% relative change in the adequacy of referrals. The pre-specified acceptance threshold for this secondary endpoint was a ≥ 15% relative increase; the observed point estimate therefore meets the threshold. The 95% confidence interval for the relative change, estimated by the Katz log method, is approximately [−15%, +85%]; the interval is wide and includes the null, reflecting the exploratory nature of this secondary endpoint and the impact of the missing-baseline deviation on pairing integrity. Confirmatory inferential testing is committed as a Post-Market Clinical Follow-up deliverable under R-TF-007-002.

Diagnosis

In diagnosing specific conditions such as hidradenitis suppurativa (HS), the device identified two cases that primary care practitioners had not diagnosed without the tool. Of these, one case was later confirmed by a dermatologist. Additionally, while not all were confirmed as HS, the tool suggested this condition as a possible diagnosis in four reports, the first option in three instances and the second in one.

For urticaria, only one patient was diagnosed with the condition. Although the diagnosis without the tool is not available, the guided diagnosis by the device was urticaria, subsequently confirmed by a dermatologist. No further urticaria suggestions were made by the tool.

Regarding psoriasis, eight diagnoses were confirmed. The tool suggested psoriasis or its variants 23 times, leading to 10 psoriasis diagnoses by the primary care practitioners.

In skin-cancer cases, five melanoma cases were confirmed by histopathology and a further cases were dermatologist-suspected (total ten suspected). On the histologically-confirmed subset (n = 5) the dermatologist's melanoma-detection sensitivity when using the device was 3/5 (60%; exact 95% CI [14.66%, 94.73%] by Clopper–Pearson); the corresponding specificity on the dermatologist-suspected non-melanoma contrast subset was 91% (with a wide exact confidence interval driven by the small denominator). Because many PCP unaided responses were missing (see §Protocol Deviations), a direct PCP-with-device-vs-PCP-alone comparison for melanoma could not be performed on this investigation. These melanoma-specific sensitivity and specificity values are reported as supporting descriptive data with wide confidence intervals; no melanoma-specific performance claim is made on the basis of this investigation alone, and confirmatory evaluation at a pre-specified operating threshold is a PMCF deliverable under R-TF-007-002.

One dermatologist-undiagnosed case presented a device-reported malignancy risk of 30%, which exceeded the Top-5 differential referral-recommendation threshold, and basal-cell carcinoma and melanoma appeared among the top-three algorithm suggestions.

General Performance (malignancy detection)

On the malignancy-detection secondary endpoint, the device achieved an AUC of 0.842 (95% CI [0.82, 0.86]) against a pre-specified acceptance threshold of AUC ≥ 0.80 and an external state-of-the-art value of 0.778 (95% CI [0.74, 0.80]). The acceptance criterion is met. The observed point estimate exceeds both the pre-specified threshold and the external state-of-the-art value.

Clinical Utility

Primary care practitioners and dermatologists completed a survey designed to assess the clinical utility of the medical device. Eight primary care practitioners and two dermatologists completed the questionnaire, with results presented in the following table.

Aspect Evaluated #	Result (Score / % Affirmation)
Performance	90%
Ease of Use	7.9/10
Usefulness of Information	7.1/10
Triage	7.1/10
Speed	7.9/10
Diagnostic Support	7.4/10
Patient Status Information	80%
Overall Satisfaction	7.6/10
Recommendation Level	7.7/10

The device received positive feedback from primary care practitioners. It received a high score especially for performance (90%) and knowledge of patient status information (80%). The lowest scores — still positive — were "Usefulness of Information" (7.1/10) and "Triage" (7.1/10). Overall Satisfaction with the device was 7.6/10 and the Recommendation Level was 7.7/10. Response rate for this questionnaire was 8 of 15 (53%) primary care practitioners and 2 of 2 dermatologists; these descriptive data are reported without confirmatory inference and the questionnaire analysis is treated as exploratory.

The full text of each questionnaire item is reproduced in Appendix I of the study protocol.

Adverse Events and Adverse Reactions to the Product

Adverse events (AEs) and adverse events to the product (AEPs) were actively surveilled throughout the investigation by the investigator at each consultation, with a dedicated CRF section for AE/AEP capture and a standing reporting path from investigator to sponsor within 24 hours of awareness. The sponsor reviewed AE/AEP reports on a rolling basis and maintained an AE/AEP register held within the Trial Master File. Over the full investigation period, no adverse events or adverse reactions related to the investigational product were observed or reported. Adverse-event surveillance after completion of this investigation is continued under the general post-market surveillance procedure (GP-009) and the Post-Market Clinical Follow-up Plan (R-TF-007-002).

Product Deficiencies

Product deficiencies were actively surveilled via a CRF field for investigator-reported deficiencies and via platform-side error logging and session monitoring, with reports routed to the sponsor within 24 hours of awareness. The sponsor reviewed deficiency reports on a rolling basis and maintained a product-deficiency register held within the Trial Master File. Over the full investigation period, no product deficiencies were observed or reported. Deficiency surveillance after completion of this investigation is continued under the vigilance procedure and the Post-Market Surveillance system (GP-009).

Subgroup Analysis for Special Populations

In the context of the analysed pathologies, no special-population subgroups were pre-specified for this investigation. Phototype coverage for darker skin tones (Fitzpatrick V and VI) is addressed at Clinical Evaluation level (R-TF-015-003) and in the PMCF Plan (R-TF-007-002), as disclosed under §Subject Demographics.

Protocol Deviations

The following protocol deviations were identified during the conduct of the investigation and are declared here in numbered form for traceability to CAPA and PMCF commitments.

1. Missing unaided PCP diagnoses (major deviation on primary endpoint) — For several PCPs across the investigation, the unaided diagnosis was not captured in the CRF before the PCP consulted the device output, producing missing baseline data and breaking the pairing integrity of the primary paired analysis. Impact: the CIP primary objective (PCP accuracy improvement) is not evaluable in this investigation. Mitigation: confirmatory evaluation of the PCP-accuracy primary endpoint is a PMCF deliverable under R-TF-007-002, and the post-market observational study (R-TF-015-012) will enforce software-level blinding of the device suggestion until the unaided diagnosis is submitted.

2. Enrolment exceeded planned sample (minor deviation on sample-size pre-specification) — The CIP pre-specified 100 patients; actual enrolment reached 131 patients and 180 diagnostic reports. The extension was not covered by a formal CIP amendment. Impact: none on the evaluable secondary endpoints, as each endpoint is reported on its specific analysis unit (see §Subject and Investigational Product Management); inference on the secondary endpoints remains descriptive/exploratory as pre-specified.

3. Analysis unit not pre-specified per endpoint (minor deviation on statistical-plan pre-specification) — The CIP §Statistical analysis did not pre-specify the analysis unit for each endpoint (patient vs report vs HCP) or the primary analysis population (ITT vs per-protocol) or the missing-data handling rule or the multiplicity strategy for secondary endpoints. These pre-specification gaps are declared here and the investigation is treated as supporting Pillar 3 evidence on the three evaluable secondary acceptance criteria rather than as confirmatory evidence on the CIP primary endpoint. No post-hoc acceptance-threshold changes were made.

4. Operational logging gaps (minor deviation on CRF completeness) — Challenges in logging all cases due to constraints in hospital information systems resulted in incomplete CRF capture on a subset of cases. Impact: contributes to the pairing-integrity issue under deviation 1. Mitigation: PMCF Plan (R-TF-007-002) pre-specifies software-level CRF enforcement and standalone eCRF independent of hospital information-system availability.

5. Questionnaire response rate (minor deviation on secondary-endpoint data completeness) — The Clinical Utility and Satisfaction Questionnaire was completed by 8 of 15 PCPs (53%) and 2 of 2 dermatologists. Impact: the satisfaction analysis is descriptive and exploratory; no confirmatory inference is drawn. Mitigation: PMCF monitoring of satisfaction under R-TF-007-002 with pre-specified response-rate targets.

No critical, serious or unexpected device-related safety deviations were identified.

Discussion and Overall Conclusions

Clinical Performance, Efficacy, and Safety

Acceptance-criteria pass/fail summary

Error: Could not determine study code. Provide studyCode or folderSlug prop, or ensure this component is used within an Investigation document with a registered folder slug.

The three pre-specified acceptance criteria that this investigation supports in the manufacturer's performance-claims dataset — (i) malignancy detection AUC, (ii) relative change in the adequacy of referrals, and (iii) expert-consensus clinical utility — are each reported against their pre-specified thresholds, together with the derivation of each result from the §Results analyses.

Malignancy detection (AUC)

The pre-specified threshold was AUC ≥ 0.80. The observed malignancy-detection AUC was 0.842 (95% CI [0.82, 0.86]), exceeding both the pre-specified threshold and the external state-of-the-art value (0.778, 95% CI [0.74, 0.80]). The acceptance criterion is met. Sensitivity (0.60) and specificity (0.91) observed on the histologically-confirmed melanoma subset (n = 5 biopsied, n = 10 dermatologist-suspected) are reported as supporting secondary descriptive data; the small denominator yields wide exact 95% confidence intervals (Clopper–Pearson: sensitivity ≈ [0.15, 0.95]; specificity based on 10 suspected cases ≈ [0.59, 1.00]), and no malignancy-specific sensitivity or specificity claim is made on the basis of this investigation alone. Confirmatory sensitivity/specificity evaluation at a pre-specified operating threshold is a Post-Market Clinical Follow-up deliverable and is addressed under R-TF-007-002 and by the post-market observational study R-TF-015-012.

Relative change in the adequacy of referrals

The pre-specified threshold was a relative increase of at least 15% in the proportion of cases correctly identified as manageable within primary care (cases whose PCP triage response matched the dermatologist adjudication and the case was assigned to primary-care management). The proportion of cases so classified rose from 26.66% (32/120) without the device to 33.33% (40/120) with the device, corresponding to a +25.03% relative change. The acceptance criterion is met at the point estimate. The 95% confidence interval for the relative change, estimated by the Katz log method on the available data, is approximately [−15%, +85%], which is wide and includes the null, reflecting the exploratory nature of this secondary endpoint and the impact of the missing-baseline deviation on pairing integrity. Confirmatory inferential testing is committed as a PMCF deliverable.

Expert-consensus clinical utility

The pre-specified threshold was ≥ 0.70 (absolute value). The observed expert-consensus score was 0.80. The acceptance criterion is met. The questionnaire response rate was 8 of 15 primary care practitioners (53%) and 2 of 2 dermatologists; the analysis is descriptive and exploratory, and no confirmatory inference is drawn from it.

Primary-objective evaluability

The CIP primary objective — whether the information provided by the device increases PCPs' own true diagnostic accuracy — was not evaluable in this investigation. As documented in §Clinical Investigation Plan (CIP) Compliance and in §Protocol Deviations, the without-device baseline diagnosis was not captured for many diagnostic reports because several PCPs used the device from the outset, without first recording an unaided diagnosis. This produced missing baseline data and broke the pairing integrity of the primary paired analysis. No conclusion on the primary objective is drawn from the present data. Confirmatory evaluation of the PCP-accuracy primary endpoint is addressed under the Post-Market Clinical Follow-up Plan (R-TF-007-002) and the post-market observational study R-TF-015-012, and is supported at a lower rank under MDCG 2020-6 Appendix III by the manufacturer's MRMC simulated-use reader investigation (LEGIT.HEALTH_PH_2024, Rank 11 supporting Pillar 3 §4.4 evidence).

Benefit-risk assessment against GSPR 1 and GSPR 8

GSPR 1 (benefits outweigh risks): on the pre-specified secondary acceptance criteria that this investigation supports, the device meets or exceeds the pre-specified thresholds (malignancy AUC 0.842 ≥ 0.80; relative change in the adequacy of referrals +25.03% ≥ +15%; expert-consensus clinical utility 0.80 ≥ 0.70). The principal residual benefit–risk tension sits with the melanoma secondary descriptive data (sensitivity 0.60 on five histologically-confirmed cases): this is acknowledged, its wide confidence interval is disclosed, and the residual risk of a false-negative malignancy classification is mitigated by the device's Top-5 prioritised differential view, the standing protocol that the final decision remains with the clinician, and the PMCF confirmatory commitment.

GSPR 8 (use-error risk): zero adverse events and zero product deficiencies were observed under the surveillance mechanisms described in §Adverse Events and §Product Deficiencies. The primary residual use-error risk is misinterpretation of the device output as a diagnostic determination rather than as decision support. This is mitigated by the IFU decision-support framing, by the device's integration requirements which mandate display of the Top-5 prioritised differential, the malignancy gauge and the referral recommendation, and by the standing instruction that the clinician retains the final diagnostic and referral decision.

Satisfaction, limitations and operational caveats

Primary care practitioners and dermatologists reported positive scores on the Clinical Utility and Satisfaction Questionnaire across performance, ease of use, usefulness of information, diagnostic support, triage and speed, with response rate and non-response caveats as noted in §Expert-consensus clinical utility. Operational limitations included challenges in logging all cases due to constraints in hospital information systems; these are captured in §Limitations of Clinical Research and are addressed in the PMCF Plan (R-TF-007-002) by pre-specifying software-level enforcement of the unaided-diagnosis step.

Clinical Relevance

The device applies deep-learning-based image analysis to support dermatological assessment, consistent with the current state of the art in AI-based decision-support tools for skin disease [1][2][3][4]. This is aligned with published research on the integration of artificial intelligence and machine learning in dermatological diagnostics [5][6].

Machine-learning algorithms applied to dermatological image analysis have been reported for a range of pathologies including basal cell carcinoma, melanoma, nevi and psoriasis [7]. The device's suitability for remote assessment is relevant to the clinical context of teledermatology and primary-care triage [10][11]. The malignancy-detection AUC observed in this investigation (0.842, 95% CI [0.82, 0.86]) supports the device's potential use in primary-care triage where the clinical question is prioritisation of patients with suspected malignancy.

Regarding the clinical utility perceived by healthcare professionals, both primary care practitioners and dermatologists provided positive scores on the Clinical Utility and Satisfaction Questionnaire, with response-rate and non-response caveats as disclosed above. Prior studies indicate that AI decision-support tools in teledermatology can reduce referred patients and increase the proportion of cases managed in primary care [8], contributing to reduced dermatology waiting lists and healthcare-resource optimisation [9].

The emphasis on patient-centred triage and reduced consultation time aligns with the broader trend in healthcare towards patient-centric and efficient care delivery. The absence of adverse events and product deficiencies observed under the surveillance mechanisms described in §Adverse Events and §Product Deficiencies is consistent with the device's favourable safety profile in this investigation.

References [1], [2], [3] and [4] include co-authors who are personnel of the manufacturer. This conflict of interest is declared in accordance with MEDDEV 2.7/1 Rev 4 §8 and ISO 14155:2020 §6.6.

References:

[1] Mac Carthy, Taig, et al. "Automatic Urticaria Activity Score (AUAS): Deep Learning-based Automatic Hive Counting for Urticaria Severity Assessment." JID Innovations (2023): 100218.

[2] Hernández Montilla, Ignacio, et al. "Automatic International Hidradenitis Suppurativa Severity Score System (AIHS4): A novel tool to assess the severity of hidradenitis suppurativa using artificial intelligence." Skin Research and Technology 29.6 (2023): e13357.

[3] Montilla, Ignacio Hernández, et al. "Dermatology Image Quality Assessment (DIQA): Artificial intelligence to ensure the clinical utility of images for remote consultations and clinical trials." Journal of the American Academy of Dermatology 88.4 (2023): 927–928.

[4] Medela, Alfonso, Taig Mac Carthy, S. Andy Aguilar Robles, Carlos M. Chiesa-Estomba, and Ramon Grimalt. "Automatic SCOring of atopic dermatitis using deep learning: a pilot study." JID Innovations 2, no. 3 (2022): 100107.

[5] Esteva, Andre, et al. "Dermatologist-level classification of skin cancer with deep neural networks." Nature 542.7639 (2017): 115–118.

[6] Haenssle, Holger A., et al. "Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists." Annals of Oncology 29.8 (2018): 1836–1842.

[7] Han, Seung Seog, et al. "Classification of the clinical images for benign and malignant cutaneous tumors using a deep learning algorithm." Journal of Investigative Dermatology 138.7 (2018): 1529–1538.

[8] Giavina-Bianchi, M., et al. "Teledermatology reduces dermatology referrals and improves access to specialists." eClinicalMedicine 29–30 (2020).

[9] Eminovic, Nina, et al. "Teledermatologic consultation and reduction in referrals to dermatologists: a cluster randomized controlled trial." Archives of Dermatology 145(5) (2009).

[10] Papachristou, Panagiotis, et al. "Evaluation of an artificial intelligence-based decision support for the detection of cutaneous melanoma in primary care: a prospective real-life clinical trial." British Journal of Dermatology 191.1 (2024): 125–133.

[11] Marsden, Helen, et al. "Accuracy of an Artificial Intelligence as a medical device as part of a UK-based skin cancer teledermatology service." Frontiers in Medicine 11:1302363 (2024).

Specific Benefit or Special Precaution

One specific benefit of the device is its ability to streamline the diagnostic process for dermatologists. By automating image analysis and populating measurement scales, the tool reduces the cognitive load on clinicians, allowing them to focus on critical decision-making. This not only leads to more efficient consultations but also minimizes the risk of errors in assessment.

However, it is important to note that while the tool offers valuable clinical support, it should not replace the expertise of the dermatologist. It is designed to augment, not replace, the clinician's judgment and experience. Patients should always receive a comprehensive evaluation that takes into account both clinical data and any additional contextual information.

Additionally, clinicians should exercise caution in cases where the tool's image analysis capabilities may be limited, such as in instances of poor image quality. In such scenarios, it is imperative for dermatologists to rely on their clinical judgment and consider seeking additional diagnostic methods.

Implications for Future Research

The positive outcomes of this study pave the way for several avenues of future research. Firstly, expanding the scope of the device to encompass a broader range of dermatologic pathologies and conditions would enhance its clinical utility.

Furthermore, exploring the integration of artificial intelligence and machine learning techniques to refine the tool's diagnostic capabilities warrants attention. This could lead to even more accurate and reliable assessments, potentially revolutionizing the field of dermatology.

Additionally, conducting long-term studies to evaluate the impact of the device on patient outcomes, including treatment adherence and quality of life, would provide a comprehensive understanding of its broader clinical implications.

Regarding specifically this investigation, the deviation on baseline capture of PCPs' unaided diagnoses means the CIP primary objective — whether the device increases PCPs' own diagnostic accuracy in real-world primary care — cannot be confirmed from the present data. Confirmatory evaluation of the PCP-accuracy primary endpoint is addressed under the Post-Market Clinical Follow-up Plan (R-TF-007-002) and the post-market observational study (R-TF-015-012). The manufacturer's MRMC simulated-use reader study (LEGIT.HEALTH_PH_2024) provides Rank 11 supporting Pillar 3 §4.4 evidence on the same PCP-triage hypothesis in a controlled, blinded, within-subject design and is documented separately.

Limitations of Clinical Research

The main limitation of machine learning is the quantity and quality of the images collected. Variability in lighting, colour, shape, size and focus are key factors, as well as the number of images per patient. This means that high variability within the same patient and an insufficient number of images to reflect this variability can result in lower than expected accuracy.

Along with this, the main limitation of this study was the large amount of missing data of diagnosis made by primary care without using the device, which does not allow us to assess the main objective of this study and validate the device as a useful tool in diagnosis support.

Ethical Aspects of Clinical Research

This study adhered to international Good Clinical Practice (GCP) guidelines, the Declaration of Helsinki in its latest amendment, and applicable international and national regulations. As applicable, approval from the relevant Ethics Committee was obtained prior to the initiation of the study. When applicable, modifications to the protocol were reviewed and approved by the Principal Investigator (PI) and subsequently evaluated by the Ethics Committee before subjects were enrolled under a modified protocol.

This study was conducted in compliance with European Regulation 2016/679, of 27 April, concerning the protection of natural persons with regard to the processing of personal data and the free movement of such data (General Data Protection Regulation, GDPR), and Organic Law 3/2018, of 5 December, on the Protection of Personal Data and the guarantee of digital rights. In accordance with these regulations, no data enabling the personal identification of participants was collected, and all information was managed securely in an encrypted format.

Participants were informed both orally and in writing about all relevant aspects of the study, with the information being tailored to their level of understanding. They were provided with a copy of the informed consent form and the accompanying patient information sheet. Adequate time was given to patients to ask questions and fully comprehend the details of the study before providing their consent.

The PI was responsible for the preparation of the informed consent form, ensuring it included all elements required by the International Conference on Harmonisation (ICH), adhered to current regulatory guidelines, and complied with the ethical principles of GCP and the Declaration of Helsinki.

The original signed informed consent forms were securely stored in a restricted access area under the custody of the PI. These documents remained at the research site at all times. Participants were provided with a copy of their signed consent form for their records.

Data quality assurance

The Principal Investigator is responsible for reviewing and approving the protocol, signing the Principal Investigator commitment, guaranteeing that the persons involved in the centre will respect the confidentiality of patient information and protect personal data, and reviewing and approving the final study report together with the sponsor. All the clinical members of the research team assess the eligibility of the patients in the study, inform and request written informed consent, collect the source data of the study in the clinical record and transfer them to the Data Collection Notebook (DCN) or Data Collection Forms (CRF).

Investigators and Administrative Structure of Clinical Research

Brief Description

This clinical investigation was conducted jointly by the Dermatology Department of Hospital Universitario Puerta del Hierro de Majadahonda and the manufacturer.

Investigators

Principal investigator

• Dr. Gastón Roustán Gullón (Dermatology Department)

Collaborators

• Hospital Universitario Puerta del Hierro de Majadahonda
  • Dr. Ángel Rosell Díaz
• Centro de salud de Majadahonda
  • Dr. Esther Minguela
• Centro de salud de Pozuelo
  • Dr. Fernando León

Center

• Hospital Universitario Puerta del Hierro de Majadahonda
• Centro de salud de Majadahonda
• Centro de salud de Pozuelo

External Organization

No additional organizations, beyond those previously mentioned, contributed to the clinical research. The study was conducted with the collaboration and resources of the specified entities.

Sponsor and Monitor

Sponsor: Puerta de Hierro Health Research Institute. Monitor: the manufacturer (see §Sponsor Identification and Contact).

Report Annexes

• The Clinical Investigation Plan is provided as R-TF-015-004 for this investigation.
• The Ethics Committee favourable opinion is provided as Annex II to the CIP (CEIm Puerta de Hierro, reference 47/395984.9/22, 2022-06-24).
• The Instructions For Use (IFU v1.1.0.0) for the investigational device are provided to the investigators and are cited in the Device Description and Specifications record held within the technical file.
• Questionnaires are reproduced in Appendix I of the study protocol (Annex I to the CIP).

Signature meaning

The signatures for the approval process of this document can be found in the verified commits at the repository for the QMS. As a reference, the team members who are expected to participate in this document and their roles in the approval process, as defined in Annex I Responsibility Matrix of the GP-001, are:

• Author: Team members involved
• Reviewer: JD-018 Clinical Research Coordinator
• Approver: JD-022 Medical Manager