A Study Using Artificial Intelligence to Identify Adults With Complex Perianal Fistulas Associated With Crohn's Disease

Use of Natural Language Processing (NLP) and Machine Learning (ML) for the Identification of Patients With Crohn's Disease (CD) and Complex Perianal Fistulas (CPF) and Their Characterization in Terms of Clinical and Demographic Characteristics. A Multicentre, Retrospective, NLP Based Study

Status

Completed

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT04844593

Acronym

INTUITION-CPF

Enrollment

Registered

2021-04-14

Start date

2022-03-08

Completion date

2024-04-29

Last updated

2024-05-10

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Crohn Disease, Rectal Fistula

Keywords

Drug Therapy

Brief summary

Natural Language Processing and machine learning are examples of artificial intelligence tools. This study will check if these tools correctly identify people with Crohn's disease with complex perianal fistulas from their medical records.

Detailed description

This is a non-interventional, retrospective study of participants with CD and CPF in a clinical practice setting. The study will enroll approximately 100 participants. The study will have a retrospective data collection to select and analyze information from EMRs processed by an AI based analytics framework that uses machine learning and NLP methodologies. All participants will be enrolled in one observational group. • Participants with CD This multi-center trial will be conducted in Spain. The overall duration of the study is approximately 36 months.

Interventions

None listed

Study design

Observational model

COHORT

Time perspective

RETROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to No maximum

Healthy volunteers

Inclusion criteria

1\. CD participant diagnosed or not with CPF between January 1st 2015 and December 31st 2021.

Exclusion criteria

Not applicable.

Design outcomes

Primary

Measure	Time frame	Description
Percentage of Participants With CD and CPF Accurately Identified With the use of NLP and Medical Language (MEL)	Up to Month 36	Percentage of participants will be measured in terms of accuracy and precision (sensitivity and specificity) of the algorithm used to identify participants with CPF associated with CD. Data obtained through the artificial intelligence (AI) technology will be compared with data obtained through traditional electronic data capture (EDC) and source data verification methods.

Secondary

Measure	Time frame	Description
Number of Participants With CD and CPF Characterized Using NLP and Machine Learning Techniques	Up to Month 36	The following information at the moment of CPF diagnosis will be extracted from the electronical medical records (EMRs): age, gender, date of diagnosis of CPF, smoking status, date of diagnosis of CD, luminal disease characteristics (localization, behaviour and activity) at diagnosis, treatments (medical and surgical) established for luminal disease in the study period, treatments (medical and surgical) established for CPF since first occurrence, fistula characteristics at diagnosis: type of fistula (following American Gastroenterological Association \[AGA\] classification) number of fistula internal and external openings, fistula activity.

Countries

Spain

Outcome results