Table 17.

—[Section 3.1.1-3.1.4] Summary of Findings: Six or Twelve Months vs Three Months as Minimum Duration of Anticoagulation for VTE^{a,b,167,203,204}

Outcomes	No. of Participants (Studies), Follow-up	Quality of the Evidence (GRADE)	Relative Effect (95% CI)	Anticipated Absolute Effects
				Risk With 3 mo	Risk Difference With 6 or 12 mo (95% CI)
Recurrent VTE	2,061 (6 studies), 1-3 y	Moderatec-e due to imprecision	RR 0.89 (0.69-1.14)	115 per 1,000	13 fewer per 1,000 (from 36 fewer to 16 more)
Major bleeding	2,061 (6 studies), 1-3 y	Highf	RR 2.49 (1.2-5.16)	9 per 1,000	13 more per 1,000 (from 2 more to 37 more)
Mortalityg	1,331 (5 studies), 1-3 y	Moderated due to imprecision	RR 1.3 (0.81-2.08)	44 per 1,000	13 more per 1,000 (from 8 fewer to 47 more)

The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). Working group grades of evidence are as follow: High quality, further research is very unlikely to change our confidence in the estimate of effect; moderate quality, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; low quality, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; very-low quality, we are very uncertain about the estimate. See Table 1 and 3 legends for expansion of other abbreviations.

^aStudy populations varied across studies: Pinede et al¹⁶⁷ enrolled provoked and unprovoked proximal DVT and PE; Campbell et al²⁰³ enrolled provoked and unprovoked isolated distal DVT, proximal DVT, and PE; Agnelli et al¹⁹⁴ had separate randomizations for provoked PE (3 vs 6 mo) and unprovoked (3 vs 12 mo); and Agnelli et al²⁰⁴ enrolled unprovoked proximal DVT.

^bTiming of randomization relative to the start of treatment and length of treatment in the non-3 mo group varied across studies: Pinede et al¹⁶⁷ and Campbell et al²⁰³ randomized at diagnosis; and Agnelli et al^194,204 randomized after the initial 3 mo of treatment to stop or continued treatment. The longer duration of treatment was 6 mo in Agnelli et al¹⁹⁴ (provoked PE) and 12 mo in Agnelli et al²⁰⁴ and Agnelli et al¹⁹⁴ (unprovoked PE).

^cGenerally, study design was strong. No study stopped early for benefit; two stopped early because of slow recruitment (Campbell et al,²⁰³ Pinede et al¹⁶⁷) and one because of lack of benefit (Agnelli et al²⁰⁴). In one study (Campbell et al), 20% of VTE outcomes were not objectively confirmed. Patients and caregivers were not blinded in any study. Adjudicators of outcomes were blinded in all but one study (Campbell et al). All studies used effective randomization concealment, intention-to-treat analysis, and a low unexplained drop-out frequency.

^dCIs include both values suggesting no effect and values suggesting either benefit or harm.

^eLow number of events and a total number of participants < 2,000.

^fOne study may have confined the assessment of bleeding to when subjects were receiving anticoagulant therapy, which could have inflated the increase in bleeding associated with the longer duration of therapy (Campbell et al²⁰³).

^gDifferences in mortality are expected to be mediated by differences in recurrent VTE and bleeding.