Six studies evaluated efficacy: 2 RCTs (n=249), 2 prospective non-randomised controlled studies (n=128) and 2 observational studies (182 patients in a retrospective chart review and 41 episodes of catheter malfunction in a case series).
The table of safety studies reported on 6 studies: 3 three RCTs (n=336) and 3 prospective non-randomised controlled studies (n=367). The review also reported safety data from 3 additional studies.
One double-blind RCT (105 patients receiving first cuffed TCC randomised, 85 completed and analysed) reported no significant difference between minidose warfarin (1 mg/day) and placebo in the number of days of thrombosis-free TCC survival until TCC failure and removal because of late malfunction.
One non-randomised controlled study (35 high-risk patients receiving warfarin titrated to INR of 1.5 to 2.0 versus 30 low-risk patients not receiving warfarin, all prevalent haemodialysis patients) reported no significant difference between groups in TCC thrombotic events.
One RCT (144 patients with warfarin titrated to INR of 1.8 to 2.5) reported that patients who received warfarin early (12 hours) after a TCC insertion had a significantly lower rate of malfunction events per year than patients who did not receive warfarin until a catheter thrombosis or malfunction (0.16 events per year versus 1.65, p<0.001).
One prospective observational study (63 patients) reported that primary patency was significantly longer in patients already taking warfarin for other indications (INR titrated to 2.0 to 3.0) and patients allocated to aspirin (325 mg/day) than patients taking neither drug. There was no significant difference between warfarin and aspirin groups.
One case series (41 episodes of catheter malfunction in patients who had undergone fibrinolysis with urokinase before starting warfarin) reported 13 episodes of catheter dysfunction during treatment with warfarin, with most (11 episodes) occurring with a mean INR of 1.46.
One retrospective chart review (182 patients of which 21 had taken warfarin in the previous 4 years; indications for warfarin were not reported) reported a statistically significant increase in TCC thrombosis in patients receiving warfarin compared with those who had not (0.13 versus 0.03 events per 100 catheter days).
One RCT (n=85) reported minimal risk of bleeding with minidose warfarin (1 patient with cerebral emboli and peripheral haemorrhage).
One non-randomised study (n=65) reported no major bleeds in high-risk patients taking warfarin (target INR 1.5 to 2.0).
One RCT (n=107) reported no bleeds in patients taking warfarin (target INR 1.8 to 2.5) plus ticlopine.
One RCT (144 patients with arterio-venous grafts) reported a statistically significant increase in major bleeds in patients taking warfarin (target INR 1.4 to 1.9) compared with no warfarin (8.9% versus 0%); all 5 major bleeds were in patients taking concurrent antiplatelet therapy.
One non-randomised study (n=42) reported an increase in major bleeds in patients taking warfarin compared with no warfarin (3 out of 11 patients versus 0 out of 31).
One observational study (n=240) reported an increase in major bleeds in patients taking warfarin for cardiovascular reasons compared with no warfarin (relative risk 2.36, 95% confidence interval, CI: 1.19, 4.27).
One observational study (number of participants not reported) reported gastrointestinal bleeding in 8.0% of haemodialysis patients on warfarin and 8.7% on aspirin per year.
It was not reported whether patients in the following 2 studies were receiving haemodialysis. One study (number of participants not reported) reported a significant increase in calcification of surgically removed heart valves in patients taking a coumarin derivative compared with control (37% versus 16%, p=0.02). The other study reported significantly increased coronary and valvular calcification in patients taking long-term warfarin.