Table 2.
Effects of interventions sorted by comparison.
| Comparison | Trial | Haemostatic effect | Healing | Adverse effects | Costs |
|---|---|---|---|---|---|
| 1. Epinephrine vs. thrombin | Brezel et al. 1987 | + | NS | NS | direct costs in favour of epinephrine |
| Carucci et al. 1984 | + | no data | NS | probably favouring epinephrine | |
| Netscher et al. 1996 | + | no data | NS | direct costs in favour of epinephrine* | |
| 2. Epinephrine vs. control | Carucci et al. 1984 | + | no data | NS | probably favouring control* |
| Gacto et al. 2008 | + | favouring epinephrine | NS | no data | |
| Netscher et al. 1996 | + | no data | NS | probably favouring control* | |
| Barret et al. 1999 | NS | NS | NS | no data | |
| 3. Fibrin sealant vs. control | Nervi et al. 2001 | + | no data | no data | no data |
| Greenhalgh et al. 1999 | +/NS/NS** | NS | NS | no data | |
| 4. Fibrin sealant vs. thrombin | Drake et al. 2003 | + | NS | NS | no data |
| 5. Thrombin vs. control | Carucci et al. 1984 | + | no data | NS | probably favouring control* |
| Netscher et al. 1996 | + | no data | NS | probably favouring control* | |
| 6. Alginate dressing vs. fine mesh gauze | Steenfos et al. 1998 | + | NS | NS | no data |
+ = significantly in favour of the former agent;
- = significantly in favour of the latter agent;
NS = not significant.
* Only direct costs per product or per patient were given, so no conclusion on overall costs per treatment could be conducted.
** Blood loss was estimated in three different ways.