Table 2 Assessment of credibility of subgroup difference based on ICEMAN
Variable | Q1 | Q2 | Q3 | Q4 | Q5 | Q6 | Q7 | Q8 | Overall interpretation | |
|---|---|---|---|---|---|---|---|---|---|---|
LVEF | Completely between | NA | Large | Definitely yes | Chance a very likely explanation (0.1) | Definitely yes | Definitely yes | NA | Maximum usually moderate | Likely effect modification. Use separate effects for each subgroup but note remaining uncertainty |
LVEF (sensitivity analysis) | Completely between | NA | Rather small or unclear | Definitely yes | Chance a very likely explanation (0.06) | Definitely yes | Definitely yes | NA | Maximum usually low | Likely no effect modification. Use overall effect for each subgroup but note remaining uncertainty |
LVESV | Completely between | NA | Rather large | Definitely no | Chance a very likely explanation (0.92) | Definitely yes | Definitely yes | NA | Maximum usually low | Likely no effect modification. Use overall effect for each subgroup but note remaining uncertainty |
LVESV (sensitivity analysis) | Completely between | NA | Very small | Definitely yes | Chance a very likely explanation (0.15) | Definitely yes | Definitely yes | NA | Maximum usually low | Likely no effect modification. Use overall effect for each subgroup but note remaining uncertainty |
LVEDV | Completely between | NA | Rather large | Definitely no | Chance a very likely explanation (0.84) | Definitely yes | Definitely yes | NA | Maximum usually low | Likely no effect modification. Use overall effect for each subgroup but note remaining uncertainty |
LVEDV (sensitivity analysis) | Completely between | NA | Very small | Definitely yes | Chance may not explain (0.007) | Definitely yes | Definitely yes | NA | Maximum usually moderate | Likely effect modification. Use separate effects for each subgroup but note remaining uncertainty |
Hospitalization | Completely between | NA | Rather large | Definitely no | Chance a very likely explanation (0.26) | Definitely yes | Definitely yes | NA | Maximum usually low | Likely no effect modification. Use overall effect for each subgroup but note remaining uncertainty |