“…There was in general a low risk of bias in the cohort studies, except that 21 27 -29,31-33,35,37-41,45,49,50,54,61-63,65,67 of 41 had less than 90 per cent follow-up or did not account adequately for the lost subjects. Besides that, there were three 36,40,41 studies with a selected group that was not rep- Table 4 Risk ratios for recurrence in studies included in the network meta-analysis Christoffersen et al 51 Absorbable tacks Permanent tacks 1⋅37 (1⋅04, 1⋅81) Beldi et al 19 Sutures Permanent tacks 0⋅89 (0⋅06, 13⋅08) Eriksen et al 21 Fibrin Permanent tacks 5⋅00 (0⋅64, 38⋅87) Grubnik et al 22 Sutures Permanent tacks 0⋅22 (0⋅03, 1⋅89) Bansal et al 17 Sutures Permanent tacks and sutures 0⋅20 (0⋅01, 4⋅15)* Muysoms et al 24 Permanent tacks and sutures Permanent tacks 3⋅00 (0⋅36, 25⋅34)…”