Abstract
OBJECTIVE: Early detection of renal involvement in lupus prevents poor outcomes. Although published guidelines recommend urine dipstick as an appropriate screening test and evidence suggests a majority of American rheumatologists use dipstick to screen for proteinuria, the performance of this diagnostic approach in lupus has not been reported. We examined the validity of qualitative urine dipstick versus quantitative 24-hour measurement to accurately detect proteinuria, including low-level proteinuria. METHODS: We performed a diagnostic accuracy study using paired samples from the Johns Hopkins University School of Medicine and the Ohio State University School of Medicine lupus cohorts. All qualitative urine dipstick values were obtained within 1 day of a 24-hour urine collection. RESULTS: We analyzed the performance of 3 urine dipstick assays to detect proteinuria compared to 24-hour protein/creatinine ratios, using 2224 dipstick measures from 296 patients. The sensitivity of a > or = 1+ dipstick result to detect quantitative proteinuria (> or = 0.50 g protein/g creatinine) was 82.7% for the Clinitek, 97.7% for the Atlas, and 85.5% for the Bayer assay. The corresponding sensitivity to detect low-level proteinuria, (0.50-0.99 g protein/g creatinine) was 63.1%, 96.4%, and 80.7%, respectively. The specificity to correctly exclude proteinuria (< 0.50 g protein/g creatinine) with negative/trace results was 86.1%, 62.2%, and 59.4%. There was considerable variability in the range of protein/creatinine ratios detected at each dipstick level of proteinuria. CONCLUSION: Urine dipsticks demonstrate substantial variability and often poor validity to accurately detect proteinuria at quantitative levels; this warrants further diagnostic evaluation. Clinicians should consider quantified proteinuria assays as a more accurate screening tool in the diagnostic evaluation of lupus nephritis.
