Table 3. Heterogeneity and the effect of moderators on heterogeneity for overall and clinical feature-specific case fatality rates in descriptive studies of severe Plasmodium falciparum malaria.
| Number of studies with available data (n) | Overall heterogeneity from random effects model (I 2) | Total estimated heterogeneity in random effects model (τ2) | Effect of moderators on heterogeneity (τ2) | Final mixed effect model | ||||||
| Outcome of interest | Def, τ2 (%) | Region, τ2 (%) | Year, τ2 (%) | Age, τ2 (%) | Drug, τ2 (%) | Moderators included | Residual heterogeneity, τ2 (%) | |||
| Case fatality rates: | ||||||||||
| Overall | 45a | 98.0% | 0.0019 | 0.0017 (11)** | 0.0016 (16)* | 0.0020 (NS) | 0.0019(NS) | 0.0019 (NS) | Def+Region | 0.0015 (21)* |
| Deep coma | 50b | 81.3% | 0.0048 | NA | 0.0035 (27)** | 0.0049 (NS) | 0.0051 (NS) | 0.0045 (NS) | Region | 0.0045 (27)* |
| Metabolic acidosis | 10 | 93.3% | 0.0052 | NA | 0.0045 (NS) | 0.0034 (35)* | 0.0056 (NS) | 0.0053 (NS) | Year | 0.0034 (35)* |
| Severe anemia | 23 | 95.8% | 0.0035 | NA | 0.0026 (26)* | 0.0034 (NS) | 0.0038 (NS) | 0.0034 (NS) | Region | 0.0026 (27)* |
| Hypoglycemia | 20 | 93.6% | 0.0321 | NA | 0.0320 (NS) | 0.0178 (45)*** | 0.0214 (33)** | 0.0222 (31)* | Year+Age | 0.0132 (59)* |
| Coma/anemia/acidosis overlap syndromec | 6 | 0.03% | 0.0000 | NA | NA | NA | NA | NA | NA | NA |
NS, No statistically significant contribution to overall heterogeneity; Def, Definition; *, ** and ***, P-values<0.05, <0.01 and <0.0001, respectively.
a
Includes 7 studies where overall case fatality rates are calculated according to treatment regimen.
b
Includes 9 studies where case fatality rates due to deep coma are calculated separately according to treatment regimen.
c
Children with overlapping syndrome comprising coma, severe anemia and any of: metabolic acidosis, respiratory distress or hyperlactatemia.