Possible sources of bias
| Study . | Time from VTE diagnosis to enrollment . | Total follow-up . | Adjudicated outcomes?* . | Known ALPA excluded? . | Lost to follow-up . | Blind to APLA status . | When APLA sample obtained . | Overall appraisal of potential bias† . |
|---|---|---|---|---|---|---|---|---|
| Patients | ||||||||
| Health care team | ||||||||
| Assessors | ||||||||
| Adjudicators | ||||||||
| Bank, 2003 (Europe) | Standard immediate | Standard 4 mo | Yes | No | NS‡ | Uncertain | Unknown | Moderate (no adjustment for confounders, possible selection bias) |
| Uncertain | ||||||||
| Uncertain | ||||||||
| Uncertain | ||||||||
| Ginsberg, 1995 (N. America) | Standard immediate | Variable 1.5-6 mo on, and mean 8.6 mo after | No | No | NS | Y | At diagnosis (probably pretreat) | Moderate (no adjustment for confounders, outcomes not adjudicated) |
| Y | ||||||||
| Y | ||||||||
| NA | ||||||||
| Kearon, 1999 (N. America) | Standard 3 mo | Variable mean: 9 mo | Yes | Yes, prescreening not routine | NS | Y | At enrollment, on anticoagulation | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Kearon, 2004 (N. America) | Standard 1 mo | Standard 11 mo | Yes | Yes, prescreening not routine | 0 | Y | At enrollment, on anticoagulation | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Rodger, 2008 (N. America, Europe) | Standard 5-7 mo | Variable mean: 18 mo | Yes | Yes, prescreening not routine§ | 17 | Y | At enrollment, on anticoagulation | High (no adjustment for confounders, possible selection bias, positive APLA not defined a priori) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Wahlander, 2005 (N. and S. America, Europe, Africa) | Standard 6 mo | Standard 18 mo | Yes | Unknown, prescreening not routine | NS | Y | 0.5-3 mo after enrollment∥ | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Taliani, 2009 (Europe) | Standard 3 mo (150) or 12 mo (147) | Variable median: 50 mo | Yes | Yes, prescreening not routine | NS | Y | Blood drawn after stopping anticoagulants | High (no adjustment for confounders, criteria for positive APLA not described, incomplete participation among eligible patients) |
| Uncertain | ||||||||
| Uncertain | ||||||||
| Probable | ||||||||
| Schulman, 2006 (Europe) | Standard immediate | Standard ∼10 y | Yes | Unknown, prescreening not routine | 25 | Y | Blood drawn after stopping anticoagulants | Moderate (no adjustment for confounders, selective reporting possible) |
| Y | ||||||||
| Y | ||||||||
| Y |
| Study . | Time from VTE diagnosis to enrollment . | Total follow-up . | Adjudicated outcomes?* . | Known ALPA excluded? . | Lost to follow-up . | Blind to APLA status . | When APLA sample obtained . | Overall appraisal of potential bias† . |
|---|---|---|---|---|---|---|---|---|
| Patients | ||||||||
| Health care team | ||||||||
| Assessors | ||||||||
| Adjudicators | ||||||||
| Bank, 2003 (Europe) | Standard immediate | Standard 4 mo | Yes | No | NS‡ | Uncertain | Unknown | Moderate (no adjustment for confounders, possible selection bias) |
| Uncertain | ||||||||
| Uncertain | ||||||||
| Uncertain | ||||||||
| Ginsberg, 1995 (N. America) | Standard immediate | Variable 1.5-6 mo on, and mean 8.6 mo after | No | No | NS | Y | At diagnosis (probably pretreat) | Moderate (no adjustment for confounders, outcomes not adjudicated) |
| Y | ||||||||
| Y | ||||||||
| NA | ||||||||
| Kearon, 1999 (N. America) | Standard 3 mo | Variable mean: 9 mo | Yes | Yes, prescreening not routine | NS | Y | At enrollment, on anticoagulation | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Kearon, 2004 (N. America) | Standard 1 mo | Standard 11 mo | Yes | Yes, prescreening not routine | 0 | Y | At enrollment, on anticoagulation | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Rodger, 2008 (N. America, Europe) | Standard 5-7 mo | Variable mean: 18 mo | Yes | Yes, prescreening not routine§ | 17 | Y | At enrollment, on anticoagulation | High (no adjustment for confounders, possible selection bias, positive APLA not defined a priori) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Wahlander, 2005 (N. and S. America, Europe, Africa) | Standard 6 mo | Standard 18 mo | Yes | Unknown, prescreening not routine | NS | Y | 0.5-3 mo after enrollment∥ | Moderate (no adjustment for confounders, possible selection bias) |
| Y | ||||||||
| Y | ||||||||
| Y | ||||||||
| Taliani, 2009 (Europe) | Standard 3 mo (150) or 12 mo (147) | Variable median: 50 mo | Yes | Yes, prescreening not routine | NS | Y | Blood drawn after stopping anticoagulants | High (no adjustment for confounders, criteria for positive APLA not described, incomplete participation among eligible patients) |
| Uncertain | ||||||||
| Uncertain | ||||||||
| Probable | ||||||||
| Schulman, 2006 (Europe) | Standard immediate | Standard ∼10 y | Yes | Unknown, prescreening not routine | 25 | Y | Blood drawn after stopping anticoagulants | Moderate (no adjustment for confounders, selective reporting possible) |
| Y | ||||||||
| Y | ||||||||
| Y |
Each study used a standardized definition for VTE recurrence.
Two authors independently rated each study for possible bias using the Quality in Prognosis Studies tool.13
Only patients who completed the original randomized, controlled trial were included in this analysis.
To be excluded for known APLA, a patient had to have either a persistently positive ACLA >30 or a persistently positive LA.
124 in the original study did not provide blood; these patients had a much higher recurrence rate.
NA, not available; NS, not stated.