Commentary

The study by Wilson et al is a helpful addition to the clinical literature on psoriatic arthritis and may be the best available source at this time. The Rochester Epidemiology Project provided a template that is somewhat unique and allowed this project to proceed.

It is unclear from the authors’ description of their methods whether a standard set of forms was used to ensure consistency and completeness of the observations recorded at all clinical encounters. Although patients were not systematically questioned about their rheumatologic symptoms, it is likely that those who sought medical care for arthritis would have been captured. Indication of the number of clinical observations per patient, with confidence intervals, could help to estimate the likelihood of capturing arthritis manifestations in this population.

The study population was large by rheumatology standards, and use of the CASPAR criteria provided a definition of psoriatic arthritis. It is notable that 56% of patients were still under observation at 10 years after presentation and 22% at 20 years. With regard to the study’s findings, it is apparent that the incidence of psoriatic arthritis is at the lower end of previously reported observations.1 The association between nail changes and psoriatic arthritis is in keeping with standard clinical teaching. The observed association between scalp psoriasis and psoriatic arthritis is new and somewhat surprising in that the hazard ratio was higher than that for nail involvement and raises the question of possible ascertainment bias. The methods indicate that in cases where classification of psoriasis was not specified, patients were assigned to the chronic plaque psoriasis group, suggesting that observations were not complete. The number of patients classified this way was not reported, and thus the impact on the results cannot be estimated. It is difficult to ascertain the level of association between scalp, intergluteal/perianal, and nail involvement.