| Study | Reason for exclusion |
|---|---|
| Acosta 2010 | Possibles excluded |
| Acosta 2011 | Duplicate of Acosta 2012 |
| Asano‐Mori 2008 | Possibles excluded |
| Bellanger 2011 | Possibles excluded |
| Bergeron 2010 | No control group |
| Bergeron 2012 | No control group |
| Bochenek 2007 | Polish; no translation available |
| Bretagne 1997 | In‐house test |
| Castagnola 2010 | Possibles excluded |
| Cesaro 2008 | Possibles excluded |
| Challier 2004 | Case‐control design; included healthy controls |
| Cuenca‐Estrella 2009 | Possibles excluded |
| Desai 2009 | Insufficient information provided |
| Fortun 2001 | Possibles excluded |
| Fujiuchi 2009 | Japanese; no translation available |
| Guinea 2008 | Possibles excluded |
| Hachem 2009 | Possibles excluded |
| Hadrich 2011 | Possibles excluded |
| Hayden 2008 | Possibles excluded |
| He 2011 | Possibles excluded |
| Hu_H 2012 | Possibles excluded |
| Husain 2004 | Possibles excluded |
| Jarque 2003 | Possibles excluded |
| Jathavedam 2009 | Insufficient information to derive 2 x 2 tables |
| Ji 2007 | Full text unavailable |
| Ji 2011 | Not a diagnostic accuracy study |
| Jin_J 2010 | Case‐control study with healthy controls |
| Khanna 2013 | Case‐control design including healthy blood donors as controls |
| Kimura 2009 | Possibles excluded |
| Kitasato 2009 | Insufficient information to derive 2 x 2 tables |
| Leng_Y 2010 | No EORTC criteria (or something similar) used |
| Li_L 2011 | No EORTC criteria (or something similar) used |
| Lim 2004 | Insufficient information to derive 2 x 2 tables |
| Lopes 2010 | Sample‐based analysis |
| Maertens 2004 | Possibles excluded |
| Maertens 2007 | Possibles excluded |
| Maertens 2007a | Possibles excluded |
| Marr 2005 | Possibles excluded |
| Meersseman 2008 | Possibles excluded |
| Pazos 2005 | Possibles excluded |
| Penack 2008 | How specificity was estimated and whether possibles were included or excluded is not reported |
| Pereira 2005 | Cases were patients with invasive fungal infections, including candida infections and other fungal infections than aspergillosis; cases were from different department |
| Perkins 2007 | Letter |
| Racil 2008 | Czech; no translation available |
| Rogers 2013 | Possibles excluded |
| Sarrafzadeh 2010 | Control group was made up from probables and possibles; no patients without IA included |
| Steinbach 2007 | Possibles excluded |
| Sulahian 1996 | In‐house test |
| Sulahian 2001 | In‐house test |
| Tabone 1997 | In‐house test |
| Uryu 2006 | Japanese; full text not available |
| Verweij 1995 | In‐house test |
| Wang_Y 2013 | 2 x 2 tables could not be derived |
| Xiang_J 2010 | 2 x 2 tables could not be derived in a correct way (the numbers in the tables and the sensitivity/specificity estimates did not match) |
| Yao 2009 | 2 x 2 tables could not be derived |
| Yu 2010 | Possibles and no IA excluded |
| Zedek 2006 | Insufficient data |
| Zeng 2011 | Case‐control design with healthy volunteers |
EORTC: European Organization for Research and Treatment of Cancer; IA: invasive aspergillosis