Table 2.
Results
| Number of patients in intervention group | Physicians increase survival | Not significant | Physicians decrease survival |
| <100 | Dickinson et al., 1997 (n = 9) | Suominen et al., 1998 (n = 49)*2 | |
| (Suominen et al., 1998 (n = 25))*2 | Di Bartolomeo et al., 2005 (n = 56) | ||
| Soo et al., 1999 (n = 38 et 37) | Iirola et al., 2006 (n = 81) | ||
| Nardi et al., 1994 (n = 42) | Di Bartolomeo et al., 2001 (n = 92) | ||
| Garner et al., 1999 (n = 67) | (Mitchell et al., 1997 (n < 100)) *1 | ||
| Sipria et al., 2000 (n = 70) | |||
| Frandsen et al., 1991 (n = 85) | |||
| Schwartz et al., 1990 (n = 93) | |||
| 100–1.000 | (Frankema et al., 2004 (n = 103))*3 | Frankema et al., 2004 (n = 107)*3 | Graf et al., 1993 (n = 107) |
| Osterwalder, 2003 (n = 196) | Hamman et al., 1991 (n = 145) | (Lee et al., 2003 (n = ?))*6 | |
| Oppe et al., 2001 (n = 210) | Cameron et al., 2005 (n = 211) | Liberman et al., 2003 (n = 801) | |
| Mitchell et al., 1997 (n = 306)*1 | Schmidt et al., 1992 (n = 221) | ||
| (Nicholl et al. 1995/Younge et al., 1997 (n = 337))*4 | Lee et al., 2003 (n = 224)*6 | ||
| (Christensen et al., 2003 (n = 177 et 388))*5 | Ringburg et al., 2007 (n = 260) | ||
| Nicholl et al., 1995 (n = 337)*4 | |||
| >1.000 | Roudsari et al., 2007 (n = 14.702) | Lechleutner et al., 1994 (n = 2.013) | |
| Christensen et al., 2003 (n = 2.869)*5 |
Studies are sorted according to overall result and ranked according to number of patients in the intervention group. Subgroup results are placed in parentheses and ranked according to number of patients in the sub-intervention group
*1 Due to a significant difference in witnessed events and patients receiving bystander CPR, the authors made a sub analysis on patients with witnessed collapse, bystander CPR and presenting rhythm of VF/VT. In this group, only a trend towards increased survival was found – how many patients this group comprised was not given, but it was less than 100 in the intervention group.
*2 Significantly higher survival in a group of patients with ISS from 25 to 49 – these comprised 51% (25/49) in the intervention group and 31% (22/72) in the control group
*3 Significantly higher survival only in a subgroup of patients with blunt trauma – these comprised 82% (195/239) in the control group and 96% (103/107) in the intervention group
*4 Not significant when calculated by Nicholl et al., later analysis using Ws by Younge et al. suggested higher survival.
*5 Significant results only in subgroups of patients with AMI and respiratory diseases – these groups comprised 3% (177/5819) and 7% (388/5819) of the included patients
*6 Significantly lower survival in a subgroup of patients not admitted to intensive care unit – this group comprised 50% of the included patients.
Abbreviations: See additional file 2