JMIR Publications

ABSTRACT

Background:

Psoriasis severity is commonly assessed using the Psoriasis Area and Severity Index (PASI), but PASI scores are frequently unavailable in real-world data due to inconsistent documentation and is often affected by interobserver variability. Medication-based proxies may offer a pragmatic alternative for severity stratification in registry-based research.

Objective:

This study aimed to develop a medication-based severity scale for psoriasis and evaluate its association with PASI scores in Danish real-world registry and electronic health record (EHR) data.

Methods:

We conducted a retrospective registry-based study including individuals with an ICD-10 diagnosis of psoriasis in Eastern Denmark between 2006 and 2016. Data were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and a population-wide EHR repository. PASI scores were extracted from unstructured clinical notes using regular expressions. Anti-psoriatic medications were categorized into mild, moderate, and severe groups based on Danish treatment guidelines. Correlations between PASI scores and the medication severity scale were analyzed.

Results:

Among 19218 individuals with psoriasis, 18848 had relevant prescriptions and 2884 had PASI scores. The automated PASI extraction algorithm demonstrated high performance (F1 score 0.98). A weak but statistically significant correlation was observed between PASI scores and the medication severity scale (r = 0.057, P = .001), indicating partial overlap between clinical severity and prescriptions. Moderate and severe therapies were generally associated with higher PASI values, though exceptions reflected treatment history and healthcare system structure.

Conclusions:

Automated extraction of PASI scores from EHR data is feasible and reliable in large-scale registry research. The modest association between PASI and medication-based severity highlights differences between clinical severity measures and real-world prescribing practices. Medication-based stratification may therefore serve as a complementary proxy for disease severity in registry-based studies when PASI scores are unavailable.