Methods to discriminate primary from secondary dengue during acute symptomatic infection

Abstract

Continuous variables are presented as median (IQR) and categorical variables are presented as number and percentage Primary Secondary Unknown p value Odds Ratio (95% CI) N = 105 N = 144 N = 44 (36%) (49%) (15%) Age (years) 13 (10.0–16.0) 14 (11.0–19.0) 14 (12.0–19.8) <  0.001 1.11 (1.05–1.17) Sex (% male) 79 (75) 76 (53) 35 (79) <  0.001 2.79 (1.61–4.83) Day of fever at enrolment, Number (%) 1 6 (6) 5 (4) 4 (9) 0.1 2 51 (49) 54 (38) 19 (43) 3 45 (43) 82 (57) 21 (48) 1.40 (0.94, 2.10) 4 2 (2) 3 (2) 0 (0) 5 1 (1) 0 (0) 0 (0) Serotype Number (%) 1 75 (71) 55 (38) 8 (18) <  0.001 – 2 18 (17) 40 (28) 17 (40) 3.03 (1.57–5.84) 3 9 (9) 13 (9) 7 (15) 1.96 (0.79–4.91) 4 3 (3) 36 (25) 12 (27) 16.36 (4.79–55.88) P-values are for comparisons between primary and secondary infections, using univariate logistic regression The kinetics of the antibody responses by illness day by immune status are shown in Fig. 2. The 1.4 cut-off used by Kuno is the same as the value we estimate on Day4, and the 1.2 used by Shu falls between our estimates from Day5 and Day6. [...]application of Innis’ algorithm overestimates secondary infections from Day2 onwards, while both Kuno and Shu algorithms overestimate primary infections early in illness, and overestimate secondary infections later in illness. [...]we chose to use PRNTs 6 months after infection as the gold standard for developing the models, not HI during infection as has been used conventionally in previous studies. [...]we gratefully acknowledge all the patients and their families for participating in this study.