As assessed in review | |||||
---|---|---|---|---|---|
High risk | Low risk | Unclear risk/not assessed | Total | ||
As assessed in this study on the primary outcomes of the review only | High risk excluding G | 10 | 18 | 50 | 78 (28%) |
High risk (based on G classifications only) | 3 | 17 | 64 | 84 (30%) | |
Low risk | 14 | 24 | 80 | 118 (42%) | |
Total | 27 (10%) | 59 (21%) | 194 (69%) | 280 |
Note that ‘As assessed in this study on the primary outcomes of the review only’ is split into three categories: high risk excluding G; high risk (based on G classifications only) and low risk. This is because G classifications, although having a high risk of bias, are subjective as they are given based on clinical judgement only when there are no details mentioned in the trial report. However, as shown in the original ORBIT study (Kirkham et al 7), the sensitivity and specificity of assigning G and H classifications was high.