| Study | Reason for exclusion |
|---|---|
| Belhabri 1999 | Subtype data for newly diagnosed AML patients were not available and most patients were not with newly diagnosed AML (only 5 out of 53 had newly diagnosed AML) |
| Buchner 2012 | Not a RCT |
| Candoni 2009 | Not a RCT |
| Castaigne 2004 | RCT：comparison arms not treated with idarubicin |
| Chan‐Lam 1992 | Not a RCT (comment) |
| Creutzig 2000 | Not a RCT |
| Creutzig 2005 | Not a RCT |
| Dluzniewska 2005 | Not a RCT |
| Gardin 2013 | Not a RCT |
| Keldsen 1990 | Not a RCT |
| Lambertenghi‐Deliliers 1989 | Not a RCT |
| Lange 2008 | Induction was not randomised |
| Leone 1999 | Not a RCT (review) |
| Li 2013a | Not a RCT |
| Liu 2005 | RCT: both arms treated with idarubicin |
| Morita 2010 | Subtype data for newly diagnosed AML patients were not available and most patients were not with newly diagnosed AML (only 54 out of 120 had newly diagnosed AML) |
| O'Brien 2002 | Not a RCT |
| Oriol 2003 | Not a RCT |
| Pashko 1991 | Cost‐effectiveness study of a RCT comparing IDA versus DNR |
| Reinhardt 2005 | Not a RCT (correspondence) |
| Shi 2013 | Not a RCT |
| Volkova 1993 | Not a RCT |
| Wheatley 2001 | Not a RCT (comment) |
| Witz 1995 | Induction was not randomised |
| Xia 2013 | Not a RCT |
AML: acute myeloid leukaemia; DNR: daunorubicin; IDA: idarubicin; RCT: randomised controlled trial.