Systematic Review and Meta-analysis of Outcomes of Patients With Subsegmental Pulmonary Embolism With and Without Anticoagulation Treatment

Background: This systematic review addresses the controversy over the decision to anticoagulate patients with subsegmental pulmonary embolism (SSPE). Methods: We searched Ovid MEDLINE, PubMed, Embase, the Cochrane Library, Scopus, Web of Science, ClinicalTrials.gov, Google Scholar, and bibliographies in March 2017. Two authors reviewed and retained papers with symptomatic patients who underwent computerized tomographic pulmonary angiography and had sufficient information to determine SSPE; decision to treat (or not) with systemic anticoagulation; and outcomes of bleeding, venous thromboembolism (VTE) recurrence, and death. Papers were assessed for selection and publication bias and heterogeneity, with Eggers and the inconsistency indexes (I-2). Results: From 1,512 papers screened, we included 14 studies comprising 15,563 patients for full-length review and analysis. Pooled data demonstrated I-2=99% with an Eggers p<0.001, suggesting significant publication bias. The pooled prevalence of SSPE was 4.6% (95% confidence interval [CI]=1.8%-8.5%). The frequency of bleeding in SSPE patients treated with anticoagulation (n=589) was 8.1% (95% CI=2.8%-15.8%), with no available bleeding data in untreated patients (n=126). The frequency of VTE recurrence within 90 days was 5.3% (95% CI=1.6%-10.9%) for treated versus 3.9% (95% CI=4.8%-13.4%) for untreated, while the frequency of death was 2.1% (95% CI=3.4%-5.2%) for treated versus 3.0% (95% CI=2.8%-8.6%) for untreated. Conclusion: This systematic review highlights the lack of any clinical trial to make a clear inference about harm or benefit of anticoagulation for SSPE. Comparison of pooled data from uncontrolled outcome studies shows no increase in VTE recurrence or death rates for patients who were not anticoagulated. These data suggest clinical equipoise for decision to anticoagulate or not anticoagulate patients with SSPE. However, this inference is limited by small numbers, imprecision, and the lack of a controlled clinical trial.