RESUMEN
BACKGROUND: The diagnostic criteria for Parkinson's disease (PD) remain complex, which is especially problematic for nonmovement disorder experts. A test is required to establish a diagnosis of PD with improved accuracy and reproducibility. OBJECTIVE: The study aimed to investigate the sensitivity and specificity of tests using sniffer dogs to diagnose PD. METHODS: A prospective, diagnostic case-control study was conducted in four tertiary medical centers in China to evaluate the accuracy of sniffer dogs to distinguish between 109 clinically established medicated patients with PD, 654 subjects without PD, 37 drug-naïve patients with PD, and 185 non-PD controls. The primary outcomes were sensitivity and specificity of sniffer dog's identification. RESULTS: In the study with patients who were medicated, when two or all three sniffer dogs yielded positive detection results in a sample tested, the index test sensitivity, specificity, and positive and negative likelihood ratios were 91% (95% CI: 84%-96%), 95% (95% CI: 93%-97%), and 19.16 (95% CI: 13.52-27.16) and 0.10 (95% CI: 0.05-0.17), respectively. The corresponding sensitivity, specificity, and positive and negative likelihood ratios in patients who were drug-naïve were 89% (95% CI: 75%-96%), 86% (95% CI: 81%-91%), and 6.6 (95% CI: 4.51-9.66) and 0.13 (95% CI: 0.05-0.32), respectively. CONCLUSIONS: Tests using sniffer dogs may be a useful, noninvasive, fast, and cost-effective method to identify patients with PD in community screening and health prevention checkups as well as in neurological practice. © 2022 International Parkinson and Movement Disorder Society.