Saliva, Urine, and Anal Swabs as Viable Alternatives to Lesion Swabs for Molecular Diagnosis of Mpox

Background Uganda experienced an outbreak of Mpox, declared on August 2, 2024, by the Ministry of Health following the confirmation of two cases in Kasese District. We evaluated the diagnostic performance of alternative specimen types in comparison with conventional lesion swabs for the molecular diagnosis of Mpox. Methods We conducted a cross-sectional diagnostic evaluation study comparing saliva, anal swabs, and urine samples with lesion swabs from 120 suspected Mpox cases at the Entebbe Isolation Unit in Uganda during the 2024/2025 outbreak. All specimens were tested for Mpox virus DNA using the Bioperfectus Real-Time PCR Kit. Diagnostic performance (sensitivity, specificity, PPV, NPV) with 95% CIs was calculated for each alternative specimen, along with agreement metrics (Cohen’s kappa, Gwet’s AC1). Discordance was assessed using exact McNemar tests with Holm correction and odds ratios; Ct differences were compared using Wilcoxon signed-rank tests with effect sizes and Bland–Altman plots. Equivalence was tested via TOST with ± 10% bounds. Generalized estimating equations were used to account for within-participant correlation. Descriptive analyses were stratified by sex and case category. Findings: Among 120 participants (52.5% female; median age 28), PCR positivity was highest in lesion swabs (83.3%), followed by saliva (81.7%), anal swabs (78.3%), and urine (75.8%). Ct values were generally non-normal except for anal swabs; median values ranged from 26.8 (urine) to 28.3 (saliva). Paired comparisons showed no significant differences (all p > 0.58; trivial effect sizes). Compared with lesion swabs, all alternative specimens demonstrated high sensitivity (≥ 89%). Urine had the highest specificity (95%) and PPV (99%), while anal and saliva had moderate specificity (75%). Significant discordance was observed for urine (Holm-adjusted p = 0.035; OR ≈ 10), but not for anal or saliva. Equivalence testing showed that saliva was diagnostically equivalent within ± 0.10 bounds; anal and urine were not. Regression analyses revealed lower detection odds for urine (OR = 0.40; p = 0.0035), whereas anal and saliva were not statistically significant. No clinically meaningful Ct differences were observed across specimen types. Conclusion Saliva demonstrated diagnostic equivalence to lesion swabs for Mpox PCR, anal swabs showed high sensitivity, and urine was highly specific but suboptimal as a standalone specimen. Together, these findings support the integration of multi-site sampling into Mpox diagnostic algorithms to reduce missed diagnoses and identify saliva as a practical, non-invasive alternative when lesion sampling is infeasible or clinical presentations are atypical.