Table 2.
Multivariable Analysis for Factors Associated With Bacteremia and Mortalitya
| Variable | Odds Ratio (95% CI) | P Value |
|---|---|---|
| Analysis for Bacteremia | ||
| Male gender | 0.37 (0.18–0.77) | .01 |
| Admission to MRCG at LSHTM hospital | 2.73 (1.28–5.82) | .01 |
| Normal hydration status | 0.31 (0.12–0.83) | .02 |
| Normal musculoskeletal exam | 0.13 (0.05–0.32) | <.01 |
| Sepsis | 12.75 (4.96–32.77) | <.01 |
| Analysis for Mortality | ||
| History of a comorbidity | 3.66 (1.26–10.61) | .02 |
| Clinical pallor | 4.56 (1.45–14.40) | .01 |
| Tachycardia for age | 0.39 (0.18–0.84) | .02 |
| Tachypnea for age | 2.93 (1.22–7.01) | .02 |
| Altered consciousness | 7.15 (3.21–15.93) | <.01 |
Abbreviations: CI, confidence interval; MRCG at LSHTM, Medical Research Council The Gambia at London School of Hygiene & Tropical Medicine; OR, odds ratio.
aClinical parameters were assessed on admission at the time of diagnosis. Top performing correlates of bacteremia and mortality were derived using a multivariate logistic regression approach, followed by a variable selection approach in which variables with the largest P values were iteratively excluded one at a time. The table includes those with those with P < .05.