Abstract
BACKGROUND: Kidney disease affects millions globally, especially in low and middle-income countries where access to diagnostic testing is limited. Saliva urea nitrogen (SUN) has been proposed as a simple, non-invasive alternative to traditional serum-based diagnostics. OBJECTIVE: This study aimed to evaluate the diagnostic accuracy of SUN for detecting kidney disease in adults through a systematic review and meta-analysis. METHODS: This review adhered to the PRISMA-DTA guidelines. A comprehensive search of five databases was conducted without language or date restrictions. Study quality was assessed using the QUADAS-2 tool. STATA version 17 was used for analysis. A random-effects model was used to estimate pooled sensitivity, specificity, and diagnostic odds ratios (DOR). Subgroup analysis was conducted based on the reference test used (serum creatinine or blood urea nitrogen). Heterogeneity was assessed using the I² statistic, and meta-regression explored sources of heterogeneity. RESULTS: Seven studies (n = 1,933) met the inclusion criteria. In the serum creatinine (sCr) subgroup (2 studies), SUN showed pooled sensitivity of 0.44 (95% CI: 0.38-0.49), specificity 0.96 (95% CI: 0.95-0.98), DOR 18.89 (95% CI: 15.19-23.57), and AUC ~ 0.90. In the blood urea nitrogen (BUN) subgroup (5 studies), sensitivity was 0.83 (95% CI: 0.69-0.91), specificity 0.88 (95% CI: 0.78-0.94), DOR 37 (95% CI: 15-91), and AUC 0.93. Heterogeneity was moderate in the BUN subgroup (bivariate I² = 51%), with 42% of variability attributed to threshold effects. Meta-regression identified study country (p = 0.01), and reference test used (p = 0.02) as contributors to heterogeneity in sensitivity, while comorbidity (p = 0.001) significantly affected specificity. CONCLUSION: SUN shows high diagnostic specificity and a good overall accuracy, particularly when compared to BUN, and may serve as a practical non-invasive screening tool in low- resource settings. While heterogeneity was present, SUN remains a promising diagnostic alternative and warrants further validation in diverse clinical populations.