, , , , , , , , , . Reasons for and consequences of vitamin K antagonist discontinuation in very elderly patients with non‐valvular atrial fibrillation. J Thromb Haemost 2016; 14: 2124– 31.
PubMed link: https://www.ncbi.nlm.nih.gov/pubmed/27471198
Full text article: https://onlinelibrary.wiley.com/doi/full/10.1111/jth.13427
| Methods | Study design: Retrospective case control
Number of groups: 2 |
| Participants | Number of participants: 798
Active group: 148 Control group: 650 Age: 84.4 ± 3.25 Female: 485 (61%) Participants with dementia: Unclear Inclusion criteria: Non valvular atrial fibrillation ≥80 years Referred to the hospital anticoagulation clinic Maintaining an INR between 2 – 3 Exclusion criteria: Previously been treated with warfarin Discontinued warfarin within 1 month of initiation Country: Italy Setting: Hospital outpatient clinic |
| Interventions | Medicine: Warfarin
Withdrawal schedule: Not described Comparator: Persisted on warfarin |
| Outcomes | Thrombotic events
Major bleeding Death |
| Dates | Dates: January 2007 – January 2012
Follow-up duration: Median 25 months |
| Funding sources | Nil reported |
| Notes | Deaths only reported in the warfarin cessation group. |
Risk of bias table
| Bias | Authors’ judgment | Support for judgment |
| Random sequence generation (selection bias) | Low | N/A-retrospective study design |
| Allocation concealment (selection bias) | Low | N/A-retrospective study design |
| Blinding of participants and personnel (performance bias) | Low | N/A-retrospective case-control study |
| Blinding of outcome assessment (detection bias) | Low | No blinding, however data was cross-checked by multiple investigators |
| Incomplete outcome data (attrition bias) | High | Participants followed for varying lengths of time, loss to follow up not described in detail |
| Selective reporting (reporting bias) | High | Deaths not reported completely, particularly in continuing use group |
| Confounding (non-randomized) | High | Some confounders described, however factors precipitating treatment withdrawal for end of life care are not considered as potentially influencing events such as mortality. |
| Other bias | High | Unclear where data was retrieved from, particularly regarding patient follow-up |
| Selection bias | Representativeness of the exposed cohort | Consecutive or representative series of cases |
| Selection of the non-exposed cohort | Drawn from the same community as the cases | |
| Ascertainment of exposure | No description | |
| Demonstration that outcome of interest was not present at start of study | Yes | |
| Comparability bias | Comparability of cohorts on the basis of the design or analysis | Study controls for frailty, comorbidity and life expectancy |
| Outcome bias | Assessment of outcome | No description |
| Was follow-up long enough for outcomes to occur | Yes | |
| Adequacy of follow-up of cohorts | Same for both groups | |
Leave a Reply