. 2023 Oct 3;23:653. doi: 10.1186/s12879-023-08466-8

Table 3.

PPVs (95% CI) of claims-based algorithms for infectious diseases from two hospitals

Claims-based algorithms	Prevalent cases^a	Incident cases^b
Gold standard 1 (physician diagnosis)	67.4 (57.9–76.8)	90.0 (84.1–95.9)	63.4 (53.0–73.9)	45.0 (35.3–54.8)	79.6 (68.3–90.9)	65.2 (45.8–84.7)	49.5 (39.2–59.7)	69.1 (56.9–81.3)	87.9 (79.6–96.3)	56.0 (42.2–69.8)	48.8 (38.1–59.5)	80.0 (65.7–94.3)	50.0 (21.7–78.3)	51.3 (40.3–62.2)
Gold standard 2 (overall decision; confirmed or probable cases)	70.5 (61.4–79.7)	90.0 (84.1–95.9)	63.4 (53.0–73.8)	45.0 (35.3–54.8)	83.7 (73.3–94.0)	69.6 (50.8–88.4)	48.4 (38.1–58.6)	70.9 (58.9–82.9)	87.9 (79.6–96.3)	54.0 (40.2–67.8)	51.2 (40.5–61.9)	83.3 (70.0–96.7)	50.0 (21.7–78.3)	52.5 (41.6–63.4)
Gold standard 3 (overall decision; confirmed cases)	N/A^e	67.0 (57.8–76.2)	18.3 (9.9–26.7)	20.0 (12.2–27.8)	N/A^e	34.8 (15.3–54.3)	22.0 (13.5–30.5)	N/A^e	58.6 (46.0–71.3)	10.0 (1.7–18.3)	22.6 (13.7–31.6)	N/A^e	8.3 (0.0–24.0)	23.8 (14.4–33.1)

Claims-based algorithms

Prevalent cases^a

Incident cases^b

Main analysis

Sensitivity analysis^c

Ad hoc analysis^d

Main analysis

Sensitivity analysis^c

Ad hoc analysis^d

HZ
(n=95)

MTB
(n=100)

NTM
(n=82)

PJP
(n=100)

HZ
(n=49)

NTM
(n=23)

PJP
(n=91)

HZ
(n=55)

MTB
(n=58)

NTM
(n=50)

PJP
(n=84)

HZ
(n=30)

NTM
(n=12)

PJP
(n=80)

Gold standard 1 (physician diagnosis)

67.4

(57.9–76.8)

90.0

(84.1–95.9)

63.4

(53.0–73.9)

45.0

(35.3–54.8)

79.6

(68.3–90.9)

65.2

(45.8–84.7)

49.5

(39.2–59.7)

69.1

(56.9–81.3)

87.9

(79.6–96.3)

56.0

(42.2–69.8)

48.8

(38.1–59.5)

80.0

(65.7–94.3)

50.0

(21.7–78.3)

51.3

(40.3–62.2)

Gold standard 2 (overall decision; confirmed or probable cases)

70.5

(61.4–79.7)

90.0

(84.1–95.9)

63.4

(53.0–73.8)

45.0

(35.3–54.8)

83.7

(73.3–94.0)

69.6

(50.8–88.4)

48.4

(38.1–58.6)

70.9

(58.9–82.9)

87.9

(79.6–96.3)

54.0

(40.2–67.8)

51.2

(40.5–61.9)

83.3

(70.0–96.7)

50.0

(21.7–78.3)

52.5

(41.6–63.4)

Gold standard 3 (overall decision; confirmed cases)

N/A^e

67.0

(57.8–76.2)

18.3

(9.9–26.7)

20.0

(12.2–27.8)

N/A^e

34.8

(15.3–54.3)

22.0

(13.5–30.5)

N/A^e

58.6

(46.0–71.3)

10.0

(1.7–18.3)

22.6

(13.7–31.6)

N/A^e

8.3

(0.0–24.0)

23.8

(14.4–33.1)

PPVs of claims-based algorithms for each infectious disease were calculated versus the gold standard diagnosis definitions

^aNumber of cases regardless of baseline HZ-, MTB-, NTM-, or PJP-free period

^bNumber of cases preceded by a 12-month HZ-, MTB-, NTM-, or PJP-free period

^cSensitivity analysis reporting PPVs for claims-based algorithms of HZ and NTM, including treatment

^dExcludes period prior to August 2012

^eConfirmed cases only, no criteria for probable HZ cases

CI, confidence interval; HZ, herpes zoster; MTB, Mycobacterium tuberculosis infection; NTM, nontuberculous mycobacteria infection; PJP, Pneumocystis jirovecii pneumonia; PPV, positive predictive value