Design: Nationwide prospective register cohort (BADBIR), doubly robust Poisson regression
Population: 18,635 adults with psoriasis on systemic therapy; 118,018 person-years
Primary outcome: Tuberculosis, meningitis, candidiasis and other fungal infections
Effect: TB 0.19 and meningitis 0.25 per 1,000 PY; candidiasis IRR 2.67-4.65 with IL-17 inhibitors

This nationwide cohort used the British Association of Dermatologists Biologic and Immunomodulators Register (BADBIR), 2007-2024: 40,930 treatment episodes in 18,635 adults with psoriasis over 118,018 person-years, with entropy balancing and doubly robust models.

Tuberculosis (22 cases; 0.19 per 1,000 person-years) and meningitis (29 cases; 0.25, mostly viral) were rare, and most followed TNF-alpha inhibitors, particularly adalimumab. Fungal infections were more common (7.53 per 1,000 person-years; 450 candidiasis). IL-17 inhibitors had 2.67 times the candidiasis rate of apremilast (95% CI 1.33-5.36) and 4.65 times that of IL-12/23 inhibition (3.46-6.24), with no difference between individual IL-17 inhibitors. Other fungal infections did not differ between classes.

These are real-world comparative rates including the newer IL-17 and IL-23 classes, which trials were too small to provide. They let the choice of biologic reflect the patient's own infection history.

In India, where latent tuberculosis is common, the concentration of TB cases on TNF inhibitors reinforces screening and favours IL-17 or IL-23 inhibitors where TB risk is high, while IL-17 inhibitors need a candidiasis conversation.

Screen for latent TB before any biologic, and repeat where exposure risk continues.
In patients with high TB risk, prefer IL-23 or IL-17 inhibitors over TNF inhibitors where suitable.
Ask about recurrent oral, genital or skin candidiasis before choosing an IL-17 inhibitor.
Warn patients starting IL-17 inhibitors about candidiasis symptoms; switching within the class does not reduce risk.
Consider IL-23 inhibition for patients who develop recurrent candidiasis on IL-17 blockade.

Why it matters

Real-world rates across all biologic classes let infection history, not habit, decide between TNF, IL-17 and IL-23 inhibition.

The statistics, in plain English

Incidence per 1,000 person-years counts events per 1,000 patients treated for a year. With only 22 TB cases, class comparisons for TB are imprecise, which is why the paper describes where cases occurred rather than giving firm ratios. The candidiasis ratios rest on 450 events and are far more reliable.