Table 2.

Logistic models with logit links of epidemiological, virological and immunological variables influence on dengue mortality

Variable (Factor vs.)	Factor	n (fatal/non-fatal)	OR (95% CI)	p value
Gender	Male	2165 (97/2068)	1.12 (0.88–1.44)	0.42
Female vs
n = 2279 (91/2188)
Serotypes	DENV-2	1047 (83/964)	5.67 (3.823–8.68)	< 0.0001
DENV-1 vs.	DENV-3	1279 (62/1217)	3.36 (2.23–5.50)	< 0.0001
n = 1405 (21/1384)
	DENV-4	802 (23/779)	1.95 (1.17–3.23)	0.029
DENV-2 vs.	DENV-3	1279 (62/1217)	0.59 (0.42–0.83)	0.0024
n = 1047 (83/964)
	DENV-4	802 (23/779)	0.34 (0.21–0.54)	< 0.0001
DENV-3 vs.	DENV-4	802 (23/779)	0.58 (0.35–0.93)	0.0281
n = 1279 (62/1217)
Age (years old)	0–15	710 (38/672)	1.74 (1.17–2.60)	0.021
16–30 vs.	31–50	1205 (44/1161)	1.16 (0.79–1.71)	0.522
n = 1075 (34/1041)
	51–96	604 (53/551)	2.94 (2.04–4.30)	< 0.0001
0–15 vs.	31–50	1205 (44/1161)	0.66 (0.43–1.04)	0.07
n = 710 (38/672)
	51–96	604 (53/551)	1.68 (1.11–2.61)	0.017
31–50 vs.	51–96	604 (53/551)	2.53 (1.68–3.86)	< 0.0001
n = 1205 (44/1161)
Immune Responses	Secondary	293 (74/219)	0.97 (0.71–1.34)	0.89
Primary
n = 265 (67/198)

To calculate each logistic GLM, Death/Alive outcome was coded as binary variable. Odd ratios (COR), 95% confidence intervals (95%CI) and P-values were calculated using one GLM for each studied variable separately. Values highlighted in bold presented: COR > 1, values of OR contained into the 95%CI range and p-values < 0.05