Spondyloarthritis & psoriatic arthritis
Natural language processing to identify and characterize spondyloarthritis in clinical practice
Authors' conclusion
The application of NLP technology facilitated the characterisation of the SpA patient profile, including demographics, clinical features, comorbidities and treatments.
Study context
- Population
- All patients with spondyloarthritis (SpA) (including psoriatic arthritis (PsA)) at Hospital Universitario La Paz, 2020–2022of 639,474 hospital patients
- Participants
- 4,337 patients
- Outcome
- Clinical profile of SpA patients (demographics, SpA subtypes, comorbidities, treatments) extracted by NLP
Methodological detailsobservational, single-centre, natural language processing (NLP)-based extraction from unstructured electronic health records (EHRs)
- Observational
- Single-centre
- NLP-based extraction from unstructured EHRs
Main finding
As reported in the abstract
SpA clinical profile
- SpA diagnosis in hospital population
- 0.7%
- Men
- 55.3%
- Mean age (years)
- 50.9
- Peripheral SpA (incl. PsA)
- 31.6%
- Axial SpA
- 20.9%
- Both axial and peripheral SpA
- 3.7%
- SpA subtype not reported
- 43.7%
- Hypertension
- 25.0%
- Dyslipidaemia
- 22.2%
- Diabetes mellitus
- 15.5%
- MTX (most used csDMARD)
- 25.3%
- Adalimumab (most used bDMARD)
- 10.6%
Observational data: associations, not effects.
Further findings
NLP performance in detecting SpA clinical entities (precision, recall, F-1 score)
- All assessed F-1 score values over 0.80; high precision and recall
Original publication
Natural language processing to identify and characterize spondyloarthritis in clinical practice. RMD Open. Published 24 May 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →