Table 2.
Effect of diagnostic testing on paediatric fever treatment in 12 studied countries in 2010-2012
| Country | n Febrile under-fives taken to any care | n any anti-malarial drug use | Any anti-malarial use | n ACT use | ACT use | n any antibiotic drug use | Any antibiotic use | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| COR (95% CI) | AOR (95% CI) | p value | COR (95% CI) | AOR (95% CI) | p value | COR (95% CI) | AOR (95% CI) | p value | |||||
| Benin | 620 | 298 | 2.61 (1.51-4.51) | 1.65 (0.92-2.98) | 0.096 | 109 | 2.37 (1.20-4.70) | 1.96 (0.91-4.19) | 0.084 | 177 | 1.83 (1.06-3.17) | 1.15 (0.64-2.08) | 0.636 |
| Burkina Faso | 1,823 | 875 | 2.08 (1.39-3.11) | 1.32 (0.84-2.05) | 0.225 | 228 | 1.64 (0.99-2.71) | 1.45 (0.84-2.52) | 0.180 | 803 | 1.22 (0.82-1.81) | 0.89 (0.57-1.40) | 0.616 |
| Burundi | 1,432 | 371 | 3.62 (2.64-4.96) | 3.71 (2.63-5.25) | <0.001 | 258 | 2.62 (1.79-3.83) | 2.78 (1.81-4.27) | <0.001 | 782 | 0.62 (0.47-0.81) | 0.53 (0.40-0.72) | <0.001 |
| Cote d’Ivoire* | 965 | 220 | 3.29 (2.05-5.25) | 1.89 (1.14-3.13) | 0.013 | 32 | 7.09 (2.45-20.54) | 16.83 (1.03-276.13) | 0.048 | 334 | 1.99 (1.31-3.01) | 1.08 (0.68-1.74) | 0.737 |
| Gabon | 738 | 194 | 2.25 (1.40-3.61) | 2.00 (1.16-3.44) | 0.013 | 78 | 2.74 (1.45-5.16) | 2.45 (1.13-5.33) | 0.024 | 436 | 0.88 (0.58-1.35) | 0.84 (0.52-1.35) | 0.467 |
| Guinea* | 931 | 412 | 1.70 (1.08-2.67) | 1.28 (0.78-2.11) | 0.330 | 20 | 4.29 (1.25-14.68) | 2.42 (0.43-13.68) | 0.319 | 378 | 1.76 (1.11-2.78) | 1.05 (0.63-1.75) | 0.862 |
| Malawi | 4,337 | 2,384 | 1.65 (1.40-1.94) | 1.34 (1.11-1.61) | 0.002 | 2,019 | 1.26 (1.07-1.48) | 1.12 (0.94-1.34) | 0.206 | 1,285 | 1.12 (0.94-1.33) | 1.00 (0.82-1.22) | 1.000 |
| Mozambique | 888 | 371 | 2.85 (2.02-4.02) | 2.79 (1.92-4.05) | <0.001 | 225 | 3.65 (2.45-5.42) | 3.54 (2.33-5.39) | <0.001 | 107 | 1.04 (0.67-1.61) | 1.01 (0.64-1.59) | 0.966 |
| Rwanda* | 657 | 134 | 0.93 (0.57-1.52) | 0.83 (0.48-1.44) | 0.506 | 129 | 0.96 (0.59-1.56) | 0.88 (0.51-1.51) | 0.633 | 322 | 3.70 (2.38-5.74) | 2.95 (1.82-4.79) | <0.001 |
| Senegal | 1,275 | 180 | 1.75 (1.11-2.75) | 1.69 (1.04-2.76) | 0.036 | 70 | 2.54 (1.24-5.19) | 2.99 (1.32-6.79) | 0.009 | 547 | 1.90 (1.27-2.85) | 1.50 (0.97-2.31) | 0.070 |
| Uganda | 2,440 | 1,704 | 1.50 (1.19-1.89) | 1.24 (0.96-1.61) | 0.097 | 1,158 | 1.13 (0.92-1.39) | 0.84 (0.66-1.06) | 0.133 | 860 | 1.45 (1.18-1.78) | 1.37 (1.09-1.72) | 0.007 |
| Zimbabwe* | 217 | 11 | 13.23 (1.56-112.52) | 170.9 (0.30-98480.04) | 0.113 | 6 | 12.18 (1.94-76.45) | 25.55 (1.69-385.68) | 0.019 | 84 | 0.62 (0.24-1.60) | 0.55 (0.20-1.51) | 0.244 |
CI = confidence interval. AOR = adjusted odds ratio. COR = crude odds ratio. AORs based on mixed-effects logistic regression models in individual country datasets adjusted for data clustering and confounding covariates (malaria endemicity; transmission season; public/private source; level of care; child’s age and sex; maternal age and education; residence; household wealth and size; health care access (money); health care access (distance); symptoms; health card).
*Rwanda’s model does not include the ‘level of care’ covariate due to multi-collinearity with the public/private source covariate. Guinea’s model does not include ‘money or distance as problems accessing care’ covariates. Some results should be interpreted with caution due to few observations and few positive outcomes (e.g., Cote d’Ivoire, Guinea, Zimbabwe).