2. Primary outcome 1.2 Pregnancy related deaths.
| Study | Measure of effect | Statistical approach1 | Intervention | Control |
| Lewycka 20132 women's groups only | Adjusted | 14/4773 | 15/2530 | |
| Lewycka 2013a peer counselling only | Adjusted | 18/4690 | 14/2529 | |
| Lund 2014 | Adjusted | 4/1351 | 1/1286 | |
| Majoko 2007 | OR 1.94 (0.31‐15.2) | Adjusted raw data because the reported OR is asymmetrical. No reply from trial authors regarding our query. | 4/6696 | 2/6483 |
| Manandhar 2004 | Adjusted | 2/2899 | 11/3226 | |
| More 2012 | Adjusted | 20/7656 | 24/7536 | |
| Mori 2015 | Continuity correction for no events in either treatment arm. Unpublished data provided by trial author | 0/252 | 0/248 | |
| Persson 2013 | Adjusted | 1/11,906 | 4/10655 | |
| Fottrell 2013 | Adjusted | 14/8819 | 23/8602 | |
| Villar 2001 | Adjusted | 7/11672 | 6/11121 |
1All data adjusted using an ICC of 0.00247 as suggested by Professor J P Souza by email.
2Mortality data from Trial Profile (Lewycka 2013, p. 1724). Control group (n = 5059) split for this analysis.