Skip to main content

Table 2 Estimates of important time-to-event percentiles for three parameter distributions based on patients with symptoms and total cases

From: Demand for longer quarantine period among common and uncommon COVID-19 infections: a scoping review

Analysis

Distribution

Percentiles (95% CI), d

50

95

97.5

99

AIC

Symptomatic and presymptomatica

n = 1765

Log-normal

4.2 (4.0‒4.4)

18.7 (17.7‒19.7)

24.9 (23.3‒26.5)

34.8 (32.1‒37.5)

8934.1

Gamma

4.6 (4.4‒4.9)

15.5 (14.8‒16.1)

18.5 (17.6‒19.3)

22.5 (21.4‒23.6)

8722.9

Weibull

4.8 (4.6‒5.0)

15.1 (14.4‒15.7)

17.7 (16.9‒18.6)

21.1 (20.0‒22.2)

8709.5

Total

(sensitivity analysis)b

n = 1920

Log-normal

4.4 (4.2‒4.6)

20.2 (19.1‒21.2)

27.0 (25.3‒28.6)

37.8 (35.1‒40.6)

9816.0

Gamma

4.9 (4.7‒5.1)

16.7 (15.9‒17.4)

20.0 (19.0‒20.9)

24.3 (23.1‒25.5)

9605.7

Weibull

5.0 (4.8‒5.3)

16.3 (15.6‒17.1)

19.3 (18.4‒20.3)

23.1 (21.9‒24.4)

9599.4

Total

n = 1920

Log-normal

4.4 (4.2‒4.6)

20.0 (18.9‒21.0)

26.7 (25.0‒28.3)

37.4 (34.7‒40.1)

9788.5

Gamma

4.9 (4.7‒5.1)

16.5 (15.8‒17.2)

19.8 (18.9‒20.7)

24.1 (22.9‒25.3)

9579.0

Weibull

5.0 (4.8‒5.2)

16.2 (15.5‒17.0)

19.2 (18.2‒20.2)

22.9 (21.7‒24.3)

9573.0

  1. CI confidence interval
  2. aBy using the information of patients with symptom onset time, we obtained the incubation period estimates as its original definition
  3. bSensitivity analysis: To examine the difference between using the date of getting negative testing results and the hospital discharge for recurrent positive cases, we separately estimated the time-to-event distribution by using different dates and applied the former as the sensitivity analysis