Table 1.
Primary and secondary outcomes, continuous variables
| Participants with data available (n) | Mean estimate (SE or GSD) | Effect (95% CI) | |
|---|---|---|---|
| Haemoglobin at day 84 (g/L)* | |||
| Reference group | 139 | 110·1 (0·8) | ·· |
| 60 mg screen-and-treat approach | 131 | 107·9 (0·8) | –2·2 (–4·6 to 0·1) |
| 30 mg screen-and-treat approach | 145 | 107·4 (0·8) | –2·7 (–5·0 to −0·5) |
| Hepcidin at day 84 (μg/L)†‡ | |||
| Reference group | 140 | 6·3 (3·4) | ·· |
| 60 mg screen-and-treat approach | 132 | 3·3 (4·2) | 0·52 (0·37 to 0·75) |
| 30 mg screen-and-treat approach | 147 | 2·3 (4·7) | 0·37 (0·26 to 0·52) |
| Ferritin at day 84 (μg/L)† | |||
| Reference group | 139 | 34·6 (1·9) | ·· |
| 60 mg screen-and-treat approach | 130 | 23·1 (1·8) | 0·67 (0·58 to 0·77) |
| 30 mg screen-and-treat approach | 145 | 21·4 (1·7) | 0·62 (0·54 to 0·71) |
| Ferritin, inflammation-adjusted, at day 84 (μg/L)† | |||
| Reference group | 139 | 31·6 (0·1) | ·· |
| 60 mg screen-and-treat approach | 130 | 21·2 (0·1) | 0·67 (0·58 to 0·77) |
| 30 mg screen-and-treat approach | 145 | 19·3 (0·1) | 0·61 (0·53 to 0·70) |
| Ferritin index at day 84†§ | |||
| Reference group | 139 | 2·2 (1·5) | ·· |
| 60 mg screen-and-treat approach | 129 | 2·9 (1·5) | 1·35 (1·23 to 1·49) |
| 30 mg screen-and-treat approach | 145 | 3·1 (1·5) | 1·43 (1·31 to 1·58) |
Logarithmic transformation (ln) was done for hepcidin and ferritin variables. GSD=geometric SD.
*
Estimates are mean (SE); SE obtained by the Delta method. Effect is absolute difference in mean estimate versus the reference group.
†
Estimates are geometric mean (GSD). Exponentiation of ln-transformed variables yielded effects expressed as ratios of geometric mean estimates versus the reference group.
‡
Estimates obtained using Tobit regression on ln-transformed hepcidin concentrations were left-censored at 0·049 μg/L (limit of detection) and right-censored at 25 μg/L.
§
Ferritin index is the ratio soluble transferrin receptor: log10-ferritin.