Skip to main content
. 2019 May 3;266(8):1897–1906. doi: 10.1007/s00415-019-09340-x

Table 3.

Screening performance for detecting hyposmia defined by the UPSIT in PD patients using “winning” smell subsets which contain at least 2 of the 10 most discriminating smell and current abbreviated smell tests

Total no. of smells No. of discriminatory smells Hyposmia cut-off score Abbreviated smell test Sensitivity (%) Specificity (%)
4 4  ≤ 3 Menthol, Clove, Gingerbread, Orange 100 87.5
5 4  ≤ 4 Menthol, Clove, Gingerbread, Orange, Onion 100 87.5
12 5  ≤ 9 BSIT-B (includes, Clove, Coconut, Lemon, Wintergreen, Banana) 100 62.5
7 5  ≤ 5 Menthol, Cherry, Clove, Gingerbread, Orange, Pine, Pizza 100 37.5
6 5  ≤ 4 Menthol, Cherry, Clove, Gingerbread,orange, Root beer 100 12.5
7 2  ≤ 6 Menthol, Clove, Leather, Lilac, Watermelon, Smoke, Rose 96.9 37.5
6 3  ≤ 5 Menthol, Clove, Gingerbread, Lilac, Watermelon, Smoke 96.9 25
3 3  ≤ 2 Menthol, Clove, Coconut 96.9 25
5 4  ≤ 3 Menthol, Cherry, Clove, Coconut, Root beer 96.9 12.5
7 3  ≤ 4 Cherry, Clove, Gingerbread, Fruit punch, Root beer, Pine, Mint 96.9 12.5
12 2  ≤ 9 BSIT-A (includes Lemon, Banana) 90.6 25
4 0  ≤ 3 PST-A (none) 81.3 37.5
2 2  ≤ 1 Clove, Coconut 81.3 25
4 3  ≤ 2 Cherry, Clove, Coconut, Root beer 81.3 12.5
4 0  ≤ 3 PST-B (none) 75 25

Smell subsets are ordered according to their respective sensitivities, then by specificity. The most discriminating smells are highlighted in bold