Extended testing across, not within, tasks raises diagnostic accuracy of smell testing in Parkinson's disease.

The aim of this study was to determine whether extended olfactory testing within a single olfactory task and/or across olfactory tasks increases diagnostic accuracy of olfactory testing in Parkinson's disease (PD). Olfactory function was assessed using an extended version of the "Sniffin' Sticks", comprising 32-item odor identification and discrimination tasks, and a detection threshold task in 52 PD patients and 50 controls, all aged between 49 and 78 years. ROC curves based on sensitivity and specificity estimates were used to compare the diagnostic accuracy of extended and combined olfactory testing. There was no significant difference in diagnostic accuracy between the 16-item and the 32-item versions of the odor identification or discrimination test. The single olfactory test that was best in discriminating between PD patients and controls was a 16-item odor identification test. A combination of the 16-item identification test and the detection threshold task had a significantly higher area under the curve than the 16-item odor identification test alone. In conclusion, extended testing across, and not within, olfactory tasks increases diagnostic accuracy of olfactory testing in PD. A combination of an odor detection threshold task and a 16-item odor identification test had the highest sensitivity and specificity in distinguishing between PD patients and controls.