Skip to main content
. 2024 Jan 30;14:1330788. doi: 10.3389/fcimb.2024.1330788

Table 3.

Statistical performance of different types of samples (n = 297, composite standard).

Sample type TP FP TN FN PPA NPA PPV NPV
BALF 30 5 5 0 100.0
(85.9–100)
50.0
(20.1–79.9)
85.7
(69.0–94.6)
100
(46.3–100)
Sputum 31 3 2 0 100.0
(86.3–100)
40.0
(7.3–83.0)
91.2
(75.2–97.7)
100
(19.8–100)
Urine 17 6 1 0 100.0
(77.1–100)
14.3
(0.8–58.0)
73.9
(51.3–88.9)
100
(5.5–100)
Abscess 22 0 1 1 95.7
(76.0–99.8)
100.0
(5.5–100)
100
(81.5–100)
50.0
(2.7–97.3)
Plasma 45 17 34 13 77.6
(64.4–87.1)
66.7
(52.0–78.9)
72.6
(59.6–82.8)
72.3
(57.1–83.9)
CSF 15 1 22 2 88.2
(62.3–97.9)
95.7
(76.0–99.8)
93.8
(67.7–99.7)
91.7
(71.5–98.5)
SCE 5 3 13 3 62.5
(25.9–89.8)
81.3
(53.7–95.0)
62.5
(25.9–89.8)
81.3
(53.7–95.0)
Total 165 35 78 19 89.7
(84.1–93.5)
69.0
(59.5–77.2)
82.5
(76.4–87.4)
80.4
(70.9–87.5)

TP, FP, FN, and TN were shown as numbers of tests. PPA, NPA, PPV, and NPV were shown as % (95% confidence intervals) and were calculated with http://vassarstats.net/clin1.html#return.