Table 5.

Logistic regression analysis of risk factors for intra-hospital death from laboratory-confirmed bloodstream infection and clinically suspected systemic infection

Characteristic	Laboratory-confirmed bloodstream infection		Clinically suspected systemic infection
HIV-status analyzed	No	Yes	No	Yes
No. of observations	n = 216	n = 128	n = 1527	n = 790
	OR (95%CI)	OR (95%CI)	OR (95%CI)	OR (95%CI)
Male sex	*	*	0.8 (0.60–1.05)	*
Neonate (≤ 1 month)	*	*	*	0.7 (0.45–1.15)
Malnutrition	1.9 (0.95–3.88)	*	2.1 (1.47–2.90)‡	1.6 (0.96–2.55)
HIV infection	NA	3.4 (1.22–9.40)†	NA	2.1 (1.29–3.26)†
Other underlying infectious disease	*	*	1.8 (1.13–2.87)†	*
Underlying non-infectious disease	*	*	*	*
Malaria	*	*	*	*
Hospital-acquired infection	*	*	*	*
Polymicrobial infection	*	*	*	*
Growth in blood-culture of:
- Enterobacteriaceae	3.5 (1.71–7.03)†	5.6 (2.06–14.95)†	3.3 (2.09–5.34)‡	4.2 (2.27–7.65)‡
- Non -Enterobacteriaceae	2.4 (0.92–6.13)	4.0 (0.99–16.45)	2.4 (1.02–5.46)†	2.7 (0.80–8.81)
- Candida spp.	2.6 (0.91–7.29)	2.9 (0.70–12.09)	2.5 (1.02–6.03)†	*
Inappropriate antimicrobial treatment	2.1 (1.09–4.16)†	2.3 (0.95–5.58)	2.1 (1.14–3.93)†	1.7 (0.76–3.73)

OR, odds ratio; 95%CI, 95% confidence interval. *Variables with P > 0.2 were removed from the models. Statistically significant risk factors are marked † P < 0.05 and ‡ P < 0.001.