Table 4.
Population attributable risk for a fractional fall in hematocrit of > 25% before and after malaria treatment
| Risk factor | Frequency (%) | AOR (95% confidence interval)* | PAR† (%) | P value |
|---|---|---|---|---|
| Before treatment‡ | ||||
| Aged ≤ 5 years | 17 | 2.6 (2.2–3.1) | 21.4 | <0.001 |
| Recrudescent infections | 27 | 1.9 (1.5–2.4) | 19.5 | 0.005 |
| Palpable spleen | 23 | 2.7 (2.3–3.2) | 28.1 | <0.001 |
| Palpable liver | 16 | 1.8 (1.4–2.3) | 11.3 | 0.01 |
| After treatment§ | ||||
| Aged ≤ 5 years | 17 | 2.2 (1.7–2.7) | 16.9 | <0.001 |
| High parasitemia (μ 100,000 parasites/μL) | 10 | 2.4 (1.9–2.9) | 12.3 | <0.001 |
| Prolonged preadmission history | 65 | 2.9 (2.4–3.5) | 55.3 | <0.001 |
| Parasitemia not cleared in 24 hr | 64 | 2.7 (2.1–3.2) | 52.1 | <0.001 |
AOR = adjusted odds ratio. AORs and 95% confidence intervals of the risk associated with the presence of the factor when compared with its absence after adjustment for the other factors in the table. AORs were calculated by logistic regression analysis.
PAR = population attributable risk. PAR is for a fractional drop in hematocrit from normal values of > 25%. The overall PAR was calculated as 1 minus the product of 1 minus each individual PAR.
Overall PAR before treatment is 59.6.
Overall PAR after treatment is 84.4.