2. ROBINS‐I assessment of risk of bias in included studies.
| Study: Bhandari 2015 | ||
| ROBINS‐I domain | Risk of Bias | Description: paired, contralateral‐eye study |
| Bias due to confounding | Serious risk | All participants received DSAEK before DMEK. Baseline BCVA, CCT, donor ECD and lens status were similar. No previous glaucoma surgery or treatment |
| Bias in selection of participants | Low risk | Intervention and follow‐up start were simultaneous as a rule for surgery. No evidence of selection into the study due to variables measured after the intervention since participants were included in the studies only if DSAEK was used in one eye and DMEK in the fellow eye. |
| Bias in classification of interventions | Low risk | Well‐defined surgical interventions |
| Bias due to deviations from intended interventions | Low risk | There is no mention of a difference in surgeon training in both techniques. There is no evidence of differences in co‐interventions such as concurrent cataract surgery. |
| Bias due to missing data | Low risk | No differential follow‐up or missing data reported; no participant selection due to missing data reported. |
| Bias in measurement of outcomes | Low risk | Retrospective study. BCVA measurement done routinely and not related to study objectives. Other outcomes and adverse events: objectively measured |
| Bias in selection of the reported result | Low risk | No selective reporting possible for our pre‐specified outcomes. |
| Overall bias | All outcomes: serious risk | |
| Study: Goldich 2015 | ||
| ROBINS‐I domain | Risk of Bias | Description: paired, contralateral‐eye study |
| Bias due to confounding | Serious risk | All participants underwent DSAEK before DMEK. Baseline BCVA and donor ECD or lens status were similar. CCT not reported. No previous glaucoma surgery or treatment |
| Bias in selection of participants | Low risk | Intervention and follow‐up start were simultaneous as a rule for surgery. No evidence of selection into the study due to variables measured after the intervention since participants were included in the studies only if DSAEK was used in one eye and DMEK in the fellow eye. |
| Bias in classification of interventions | Low risk | Well‐defined surgical interventions |
| Bias due to deviations from intended interventions | Low risk | There is no mention of a difference in surgeon training in both techniques. There is no evidence of differences in co‐interventions such as concurrent cataract surgery. |
| Bias due to missing data | Low risk | No differential follow‐up or missing data reported; no patient selection due to missing data reported |
| Bias in measurement of outcomes | Low risk | Retrospective study. BCVA measurement done routinely and not related to study objectives. Other outcomes and adverse events: objectively measured |
| Bias in selection of the reported result | Low risk | No selective reporting possible for our pre‐specified outcomes |
| Overall bias | All outcomes: serious risk | |
| Study: Guerra 2011 | ||
| ROBINS‐I domain | Risk of BIas | Description: paired, contralateral‐eye study |
| Bias due to confounding | Serious risk | All participants underwent DSAEK before DMEK. Baseline confounding variables as BCVA and donor ECD are similar. CCT not reported |
| Bias in selection of participants | Low risk | Intervention and follow‐up start were simultaneous as a rule for surgery. No evidence of selection into the study due to variables measured after the intervention since participants were included in the studies only if DSAEK was used in one eye and DMEK in the fellow eye. |
| Bias in classification of interventions | Low risk | Well‐defined surgical interventions |
| Bias due to deviations from intended interventions | Low risk | There is no mention of a difference in surgeon training in both techniques. There is no evidence of differences in co‐interventions such as concurrent cataract surgery. |
| Bias due to missing data | Low risk | No differential follow‐up or missing data reported; no patient selection due to missing data reported |
| Bias in measurement of outcomes | Low risk | Retrospective study. BCVA measurement done routinely and not related to study objectives. Other outcomes and adverse events: objectively measured |
| Bias in selection of the reported result | Low risk | No selective reporting possible for our pre‐specified continuous outcomes |
| Overall bias | All outcomes: serious risk | |
| Study: Maier 2015b | ||
| ROBINS‐I domain | Risk of BIas | Description: paired, contralateral eye study |
| Bias due to confounding | Serious risk | All participants underwent DSAEK before DMEK. Baseline confounding variables as BCVA and donor ECD are similar. CCT not reported |
| Bias in selection of participants | Low risk | Intervention and follow‐up start were simultaneous as a rule for surgery. No evidence of selection into the study due to variables measured after the intervention since participants were included in the studies only if DSAEK was used in one eye and DMEK in the fellow eye. |
| Bias in classification of interventions | Low risk | Well defined surgical interventions |
| Bias due to deviations from intended interventions | Low risk | There is no mention of a difference in surgeon training in both techniques. There is no evidence of differences in co‐interventions such as concurrent cataract surgery. |
| Bias due to missing data | Serious risk | Differential follow‐up reported: 21 months for DSAEK and 7 month for DMEK; this could have favoured DSAEK since visual acuity continues to improve during follow‐up with this technique. |
| Bias in measurement of outcomes | Low risk | Retrospective study. BCVA measurement done routinely and not related to study objectives. Other outcomes and adverse events: objectively measured |
| Bias in selection of the reported result | Low risk | No selective reporting possible for our pre‐specified continuous outcomes |
| Overall bias | All outcomes: serious risk | |
BCVA: best corrected visual acuity CCT: central corneal thickness DMEK: Descemet’s membrane endothelial keratoplasty DSAEK: Descemet’s stripping automated endothelial keratoplasty ECD: endothelial cell density