ICEMAN online tool

Study type

Preliminary considerations

Effect modifier

Was the effect modifier measured before or at randomization?

Select… Yes, continue No, but an exception applies (see user manual) No, stop here and refer to user manual

Exception: ICEMAN still applies to an effect modifier measured after randomization if it is a non-modifiable characteristic such as sex or age. See user manual →

Note: ICEMAN applies to effect modifiers measured before or at randomization. If the effect modifier was measured after randomization (e.g. an intermediate outcome), the analysis of effect modification is complicated and potentially misleading, and requires different methods not covered by ICEMAN. See user manual for further instructions →

Outcome and time-point

Effect measure

Does the analysis suggest a possible effect modification (interaction p ≤ 0.1)?

Select… Yes No, stop here

Note: ICEMAN was developed to assess the credibility of a suggested effect modification (interaction p ≤ 0.1). It does not apply when there is no suggestion of effect modification. Assessing the credibility of an absence of effect modification requires a different approach not covered by ICEMAN. See user manual for scope of ICEMAN →

Summary of responses

— Fill in the items above to see a summary —

Overall rating

How would you rate the overall credibility of the proposed effect modification?

Click scale to rate manually

Very lowLowModerateHigh

Preliminary considerations

Effect modifier

Was the effect modifier measured before or at randomization?

Select… Yes, continue No, but an exception applies (see user manual) No, stop here and refer to user manual

Exception: ICEMAN still applies to an effect modifier measured after randomization if (1) it is a non-modifiable characteristic such as sex or age, or (2) it is a study characteristic such as risk of bias, length of follow-up, or mean received dose. See user manual →

Note: ICEMAN applies to effect modifiers measured before or at randomization. If the effect modifier was measured after randomization (e.g. an intermediate outcome), the analysis of effect modification is complicated and potentially misleading, and requires different methods not covered by ICEMAN. See user manual for further instructions →

Outcome and time-point

Effect measure

Does the analysis suggest a possible effect modification (interaction or meta-regression p ≤ 0.1)?

Select… Yes No, stop here

Note: ICEMAN was developed to assess the credibility of a suggested effect modification (interaction or meta-regression p ≤ 0.1). It does not apply when there is no suggestion of effect modification. Assessing the credibility of an absence of effect modification requires a different approach not covered by ICEMAN. See user manual for scope of ICEMAN →

Summary of responses

— Fill in the items above to see a summary —

Overall rating

How would you rate the overall credibility?

Click scale to rate manually

Very lowLowModerateHigh

Required fields missing

Please complete all required fields before continuing.