Effect of DOACs vs LMWHs on patient-important outcomes at the end of 30- to 42-day prophylaxis
| Outcomes . | No. of participants (studies) followed up . | GRADE certainty in the evidence . | Relative effect (95% CI) . | Anticipated absolute effects . | |
|---|---|---|---|---|---|
| Risk with LMWHs . | Risk difference with DOACs . | ||||
| Mortality | 20 225 (3 RCTs) | ⊕⊕⊕◯ Moderate*,† | RR 1.01 (0.89-1.14) | Study population: 49 per 1000 High-risk patients: 99 per 1000‡ | Study population: 0 fewer per 1000 (5 fewer to 7 more) High-risk patients: 1 more per 1000 (11 fewer to 14 more) |
| PE | 18 827 (3 RCTs) | ⊕⊕⊕◯ Moderate§ | RR 0.67 (0.41-1.09) | Study population: 4 per 1000 Moderate-risk patients: 4 per 1000|| | Study population: 1 fewer per 1000 (2 fewer to 0 fewer) Moderate-risk patients: 1 fewer per 1000 (2 fewer to 0 fewer) |
| Symptomatic DVT | 18 838 (3 RCTs) | ⊕⊕⊕◯ Moderate¶ | RR 0.62 (0.36-1.05) | Study population: 6 per 1000 Low-risk patients: 2 per 1000#,** | Study population: 2 fewer per 1000 (4 fewer to 0 fewer) Low-risk patients: 1 fewer per 1000 (1 fewer to 0 fewer) |
| Major bleeding | 21 831 (3 RCTs)1-3 | ⊕⊕⊕⊕ High | RR 1.99 (1.08-3.65) | Study population: 4 per 1000 High-risk patients: 12 per 1000†† | Study population: 4 more per 1000 (0 fewer to 10 more) High-risk patients: 12 more per 1000 (1 more to 32 more) |
| Outcomes . | No. of participants (studies) followed up . | GRADE certainty in the evidence . | Relative effect (95% CI) . | Anticipated absolute effects . | |
|---|---|---|---|---|---|
| Risk with LMWHs . | Risk difference with DOACs . | ||||
| Mortality | 20 225 (3 RCTs) | ⊕⊕⊕◯ Moderate*,† | RR 1.01 (0.89-1.14) | Study population: 49 per 1000 High-risk patients: 99 per 1000‡ | Study population: 0 fewer per 1000 (5 fewer to 7 more) High-risk patients: 1 more per 1000 (11 fewer to 14 more) |
| PE | 18 827 (3 RCTs) | ⊕⊕⊕◯ Moderate§ | RR 0.67 (0.41-1.09) | Study population: 4 per 1000 Moderate-risk patients: 4 per 1000|| | Study population: 1 fewer per 1000 (2 fewer to 0 fewer) Moderate-risk patients: 1 fewer per 1000 (2 fewer to 0 fewer) |
| Symptomatic DVT | 18 838 (3 RCTs) | ⊕⊕⊕◯ Moderate¶ | RR 0.62 (0.36-1.05) | Study population: 6 per 1000 Low-risk patients: 2 per 1000#,** | Study population: 2 fewer per 1000 (4 fewer to 0 fewer) Low-risk patients: 1 fewer per 1000 (1 fewer to 0 fewer) |
| Major bleeding | 21 831 (3 RCTs)1-3 | ⊕⊕⊕⊕ High | RR 1.99 (1.08-3.65) | Study population: 4 per 1000 High-risk patients: 12 per 1000†† | Study population: 4 more per 1000 (0 fewer to 10 more) High-risk patients: 12 more per 1000 (1 more to 32 more) |
Concern about applying the data to a “real-life” population both with regard to baseline risks in the RCTs but baseline risk estimates from observational studies are realistic.
Serious imprecision. The relative estimate of effect is compatible with important harm and important benefit for the intervention that probably crosses the relevant decision threshold.
Spencer et al11 reported on incidence rates of all-cause mortality in older adults based on a community-based study (n = 1223) prospective and retrospective).
Serious imprecision. Wide CI with only 66 events in total, and important harm or benefit is still likely or cannot be excluded.
Guijarro17 reports on the incidence of PE in acutely ill hospitalized medical patients (n = 1 148 301) based on findings from the Spanish National Discharge Database from October 2005 to September 2006 (retrospective database study).
Serious imprecision. Wide CI with only 85 events in total, and important harm or benefit is still likely or cannot be excluded.
Guijarro17 reports on the incidence of DVT in acutely ill hospitalized medical patients (n=1 148 301) based on findings from the Spanish National Discharge Database from October 2005 to September 2006 (retrospective database study).
We applied the assumption that ∼20% of symptomatic DVTs are proximal, 80% are distal, and 100% of each is of moderate severity.
Spencer et al11 reported on incidence rates of major bleeding in older adults based on a community-based study (n = 1223) (prospective and retrospective).