Quality features of 17 reports of RCT assessed in the meta-analysis
| Study | Year | Design | Reporting of randomisation | Generation of random sequence | Completeness of follow-up | Description of withdrawals | Quality score |
| Gunes et al21 | 2009 | SB-P | Y | NA | Y | N | 2 |
| Marazzi et al22 | 2009 | DB-P | Y | NA | Y | Y | 4 |
| Belardinelli et al23 | 2008 | SB-P | Y | NA | Y | Y | 3 |
| Tuunanen et al24 | 2008 | SB-P | Y | NA | Y | Y | 3 |
| Belardinelli et al25 | 2008 | DB-P | Y | Computer generated | Y | Y | 5 |
| Sisakian et al37 | 2007 | Open | Y | NA | Y | N | 2 |
| Di Napoli et al26 | 2007 | DB-P | Y | Sealed envelope | Y | Y | 5 |
| Fragasso et al27 | 2006 | Open | Y | Computer generated | Y | Y | 4 |
| Fragasso et al28 | 2006 | DB-C | Y | Computer generated | Y | Y | 5 |
| Di Napoli et al29 | 2005 | DB-P | Y | NA | Y | Y | 3 |
| El-Kady et al35 | 2005 | SB-P | Y | NA | Y | N | 2 |
| Vitale et al30 | 2004 | DB-P | Y | NA | Y | Y | 4 |
| Thrainsdottir et al31 | 2004 | DB-C | Y | NA | Y | Y | 4 |
| Rosano et al32 | 2003 | DB-P | Y | NA | Y | Y | 4 |
| Fragasso et al33 | 2003 | DB-C | Y | NA | Y | Y | 4 |
| Belardinelli et al34 | 2001 | DB-P | Y | NA | Y | Y | 4 |
| Brottier et al36 | 1990 | DB-P | Y | Double-blind | Y | Y | 5 |
DB-C, double-blind crossover; DB-P, double-blind parallel; NA,unable to assess; RCT, randomised controlled trial; SB-P, single-blind parallel.