Improved survival in psoriatic arthritis with calendar time

Objective. To determine whether there has been a change in mortality rates over the last 3 decades in patients with psoriatic arthritis (PsA) whose cases were followed prospectively. Methods. Patients receiving followup care according to a standard protocol at the University of Toronto PsA Clinic between 1978 and 2004 were included. Information on patient deaths was collected prospectively. Mortality data for the general population of Ontario, Canada, stratified by 5-year age bands, sex, and calendar year from 1978 to 2004, were used to calculate the reference rates. Standardized mortality ratios (SMRs) were calculated through use of Poisson regression models for the number of observed deaths. Time trend analyses were performed through the use of 10-year "rolling-average" SMRs and followup period-specific SMRs stratified by the period of entry into clinic. Results. Of 680 patients with PsA, 106 (15.6%) (55 women and 51 men) have died. Major causes of death were disease of the circulatory system, neoplasms, diseases of the respiratory system, diseases of the gastrointestinal system, injuries/poisoning, and unknown. The overall SMR for the period 1978-2004 was 1.36 (95% confidence interval 1.12, 1.64). The estimated number of life-years lost by the PsA patient cohort overall was 2.99 years (95% confidence interval 1.14, 4.77). For patients who entered the cohort during the years 1978-1986, the SMRs were 1.89, 1.83, and 1.21 for followup periods 1978-1986, 1987-1995, and 1996-2004, respectively. For patients who entered the cohort during the years 19871995, the SMRs were 0.55 and 0.82, while the SMR for those who entered during 1996-2004 was 0.56. Conclusion. The drop in SMRs in this PsA clinic population suggests that the mortality risk has improved over time. This improved survival may reflect disease severity at presentation in the earlier cohort as well as earlier diagnosis and more aggressive treatment in the more recent followup period.