Background Urinary tract infections (UTIs) are considered to be the most common bacterial infection. UTIs elevate the risk of pyelonephritis, premature delivery, and fetal mortality among pregnant women, and are associated with impaired renal function and end-stage renal disease. UTIs are also very common in patients with rheumatic disease.
Objectives To analyse the spectrum of pathogenic agents (PA) of UTIs in patients with Rheumatoid Arthritis (RA), Systemic Lupus Erythematodes (SLE) and controls patients with no rheumatic disease and to relate these findings to gender, comorbidity and immunosuppressive therapy (IT).
Methods A retrospective study was performed and urine cultures in period of 9 years (1995-2004) from all inpatients of the Department of Internal Medicine were collected. All patients had at least one UTI. The following data were analysed: demographics, diagnosis of RA or SLE if present, PA of UTIs, comorbidity. In a subsequent step of data analysis we also took into account the current IT. Descriptive statistics were calculated with respect to age, gender distribution, the most common rheumatic disease and the five most common PAs. Fisher’s exact tests were done for comparing frequencies of the five most common PAs between RA-, SLE-patients and controls. SPSS 18 was used for all descriptive and inferential two-sided chi-square tests.
Results 621 patients with UTI were included in the study. Mean age (±SD) was 59.30 (±18.38), ranging from 16 to 96. 80.9% were female whereas 56.8% of all patients received IT. 64.9% had no rheumatic disease, 35.1 had either RA (22.5%) or SLE (12.6%), 4.4% of them with renal involvement. The most common comorbidity was diabetes mellitus (19.5%) followed by chronic renal failure (14.7%) and coronary heart disease (14.3%). Patients of both groups (RA/SLE vs. controls) showed almost the same distribution of the five most common PAs. RA/SLE-patients showed significantly higher frequency distribution of E. coli (p=0.033) according to Fisher’s exact test and no significance in frequency distribution by other four most common PAs [(Enterococcus faecalis (15.4% vs. 17.5%); Staphylococcus coagulase negative (SCN) (15.4% vs. 14.4%); Enterococcus species (5.4% vs. 4.5%); Pseudomonas aeruginosa (5.2% vs. 6.8)]. Furthermore, we found a significantly higher prevalence of E. coli in RA-patients when compared to SLE–patients (χ2=5.89; p=0.015) and significantly elevated occurrence of SCN in SLE-patients when compared to RA-patients (χ2=10.30; p=0.001). The occurrence of other common PAs did not reach significance in the analysis of these subgroups.
Conclusions We found significantly higher frequency of E. coli in RA/SLE-patients vs. controls as compared to other common PAs causing UTIs. Although, only 79% of RA/SLE-patients received IT the spectrum of PAs causing UTIs is not different from inpatients with no rheumatic disease. However, the observation of a more frequent occurrence of E. coli in RA and SCN in SLE needs to be addressed in the future.
Disclosure of Interest None Declared