Frequency and Impact of Serious Infections in Hospitalised Patients with Sjögren’s Disease: A Longitudinal Cohort Study

INTRODUCTION: As little is known about the risk, type and severity of infections in patients with Sjögren's disease (SjD), we compared the frequency and impact of serious infections (SI) between hospitalised patients with SjD and controls (Co). METHODS: Population-level observational study of patients with SjD (n = 692, 87.9% female, age 67 years) identified in the Hospital Morbidity Data Collection (HDMC) in Western Australia (WA) between 1980 and 2015 by diagnostic codes (ICD-9 CM 710.2, ICD10-AM M35.0) and age- and gender-matched controls free of rheumatic disease (n = 3737, 87.5% female, age 67 years). SI were defined by a validated algorithm and primary SjD, as the absence of other autoimmune rheumatic disease. Odds ratios (OR), incidence rates (IR) per 100 person-years (PY) and IR ratios (IRR) with 95% CI for SI were estimated. HDMC data were linked to WA Death Registry to evaluate the impact of SI on mortality rates (MR) per 100 PY and Kaplan-Meier survival estimates. RESULTS: Serious infections occurred more in patients with SjD (OR 1.87, CI 1.59, 2.21, p < 0.01) with a fourfold higher IR for SI (IRR 4.28, CI 4.02-4.56) and increased IRR for pneumonia (4.01, CI 3.50-4.60), urogenital (3.06, CI 2.73-3.43), skin/soft tissue (5.22, CI 4.60-5.90), opportunistic infections (6.93, CI 5.77-8.32) and sepsis (6.15, CI 4.87-7.74) (all p < 0.01). The MR for patients with SjD with SI was almost twice that for Co with SI (MR ratio 1.89, CI 1.65-2.16). Patients with associated SjD (n = 270, 39%) had increased OR (2.05, CI 1.49-2.82) and IRR (1.61, CI 1.44-1.79) for SI compared to patients with primary SjD (n = 422, 61%). Few differences in SI existed between patients with SjD associated with rheumatoid arthritis (n = 197) or SLE (n = 79). CONCLUSIONS: SI occurred at a higher rate in all patients with SjD and was associated with increased long-term mortality. These results emphasize the increased susceptibility to and impact of SI for both patients with primary and secondary SjD.