Table 4

Point estimates and 90%credible intervals (CrI) fromthe log-linear hierarchical modelsof PRNT titers
Model A Model B
Intercept Absolute Titer (90% CrI)
Primary exposure Testing serotype
Uninfected DENV1 0.42 (0.16, 1.17) 0.43 (0.15, 1.29)
DENV2 0.44 (0.15, 1.18) 0.43 (0.13, 1.35)
DENV3 0.44 (0.16, 1.24) 0.44 (0.14, 1.26)
DENV4 0.41 (0.14, 1.17) 0.42 (0.14, 1.19)
DENV1 DENV1 5.44 (2.91, 10.7) 15.1 (9.52, 23.8)
DENV2 0.59 (0.31, 1.14) 1.19 (0.75, 1.89)
DENV3 0.55 (0.28, 1.08) 1.17 (0.73, 1.85)
DENV4 0.13 (0.06, 0.27) 0.29 (0.17, 0.48)
DENV2 DENV1 0.13 (0.06, 0.25) 0.46 (0.26, 0.81)
DENV2 9.17 (4.84, 16.9) 26.1 (17.0, 40.9)
DENV3 0.22 (0.11, 0.44) 0.65 (0.38, 1.10)
DENV4 0.13 (0.06, 0.24) 0.37 (0.22, 0.63)
DENV3 DENV1 0.11 (0.05, 0.23) 0.36 (0.17, 0.77)
DENV2 0.13 (0.06, 0.27) 0.32 (0.16, 0.62)
DENV3 7.98 (3.96, 15.4) 32.6 (19.0, 55.2)
DENV4 0.11 (0.05, 0.26) 0.30 (0.14, 0.65)
DENV4 DENV1 0.22 (0.11, 0.48) 0.52 (0.24, 1.07)
DENV2 0.46 (0.21, 1.00) 0.98 (0.48, 1.99)
DENV3 0.30 (0.13, 0.69) 0.72 (0.37, 1.43)
DENV4 12.1 (5.93, 24.0) 29.8 (16.7, 52.4)
Testing serotype among secondaryexposures
DENV1 26.5 (13.8, 53.8) 76.1 (43.0, 131)
DENV2 6.50 (3.26, 12.7) 17.6 (10.2, 30.3)
DENV3 5.46 (2.90, 10.4) 25.8 (15.3, 42.8)
DENV4 1.69 (0.79, 3.69) 2.63 (1.49, 4.57)
Fixed Effect Relative Titer (90% CrI)
10% increase in neutralization percentage 1.23 (1.14, 1.33)
Vaccination 0.91 (0.70, 1.16) 0.89 (0.85, 0.94)
Time category
Unexposed Reference Reference
1-11 days post-exposure (d.p.e.) 4.16 (2.57, 6.71) 2.86 (1.59, 5.01)
12-30 d.p.e. 8.51 (5.47, 13.2) 7.11 (4.46, 11.5)
31 -365 d.p.e. 5.57 (3.79, 8.24) 5.44 (3.68, 8.02)
>365 d.p.e. 3.13 (2.05, 4.81) 1.80 (0.98, 3.24)
Random Effect Relative Titer (90% CrI)
Testing serotype Testing strain
DENV1 16007 Reference
Hawaii 1.18 (0.79, 1.75)
Other DENV1 strains 0.97 (0.63, 1.50)
DENV2 16681 Reference
New Guinea B 0.90 (0.53, 1.53)
New Guinea C 0.64 (0.40, 1.00)
PR-159 0.48 (0.25, 0.85)
Other DENV2 strains 0.90 (0.60, 1.31)
DENV3 16562 Reference
H87 0.76 (0.49, 1.15)
116/00 1.33 (0.91, 1.96)
Other DENV3 strains 0.62 (0.43, 0.90)
DENV4 1036 Reference
H241 0.73 (0.45, 1.17)
Dominica/814669 0.32 (0.16, 0.62)
4328-S 1.60 (1.02, 2.55)
Other DENV4 strains 1.13 (0.66, 1.88)
Variance Components Proportion of the totalvariance (90% CrI)
Strain 8.04% (3.05%, 15.7%)
Article 50.7% (30.8%, 71.6%)
Deviance Information Criteria 14910 14830

The two models account for variance in PRNT titers due to differences between strains (Model A) or articles (Model B). The average PRNT titer of sera collected from individuals with specific characteristics can be calculated as the product of the exponential of the coefficients listed. For example, using estimates from model A, the average titer for convalescent individuals with primary infections from DENV2, tested against DENV2 strain New Guinea B using a PRNT60 would be: 9.17 [primary DENV2 exposure, tested against DENV2] * 0.90 [tested against New Guinea B] * 1.236 [PRNT60] * 8.51 [12–30 days post-exposure] = 243, or 1:243. Similarly, under the same conditions, secondary DENV2 exposure via vaccination would result in an average titer of 1:157 (6.50 [secondary DENV2 exposure, tested against DENV2] * 0.90 * 0.91 [vaccination] * 1.236 * 8.51 = 157).

Rainwater-Lovett et al.

Rainwater-Lovett et al. BMC Infectious Diseases 2012 12:233   doi:10.1186/1471-2334-12-233

Open Data