ORBIT classification | ORBIT assessment of risk of bias | Number (%) of instances of unreported or incomplete reporting of outcomes | Example |
A | High | 5 (26) | Posner et al13 only report results for the intervention versus control group comparison for water safety as p<0.33 (it is likely they mean p=0.33). No further information is provided to allow calculation of the effect estimate beyond the p value. |
B | None | 4 (21) | Posner et al13 only report results for the intervention versus control group comparison for poison safety score as p<0.02. No further information is provided to allow calculation of the effect estimate beyond the p value. |
C | Low | 0 (0) | |
D | High | 0 (0) | |
E | High | 7 (37) | Paul et al14 collected data on the installation of lockable cabinets for poisons, but the results are not reported separately for the intervention and control group. Judgement suggests that this is due to statistically non-significant results since all reported results are statistically significant. |
F | Low | 3 (16) | In Mallonee et al,15 comparison of smoke alarm installation is only made pre- and postintervention in the intervention group, even though smoke alarm installation data for a control group were reported to have been collected. However, the focus of the paper was on pre- versus postintervention comparison in the intervention group and so lack of comparison with the control group is not believed to be due to results of such a comparison. |
G | High | 0 (0) | |
H | Low | 0 (0) | |
I | None | 0 (0) | |
Total | 19 |
ORBIT, Outcome Reporting Bias in Trials.