Sex as a prognostic factor for mortality in adults with acute symptomatic pulmonary embolism

Pulmonary embolism (PE) is relatively common worldwide. It is a serious condition that can be life-threatening. Studies on the relationship between adverse outcomes of this condition and whether a patient is male or female have yielded inconsistent results. Determining whether there is an association between sex and short-term mortality in patients with acute PE is important as this information may help guide different approaches to PE monitoring and treatment.

To determine whether sex (i.e. being a male or a female patient) is an independent prognostic factor for predicting mortality in adults with acute symptomatic pulmonary embolism.

The Cochrane Vascular Information Specialist searched the Cochrane Vascular Specialised Register, CENTRAL, MEDLINE, Embase, and CINAHL databases, and the World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov trials register up to 17 February 2023. We scanned conference abstracts and reference lists of included studies and systematic reviews. We also contacted experts to identify additional studies. There were no restrictions with respect to language or date of publication.

We included phase 2-confirmatory prognostic studies, that is, any longitudinal study (prospective or retrospective) evaluating the independent association between sex (male or female) and mortality in adults with acute PE.

We followed the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of prognostic factor studies (CHARMS-PF) and the Cochrane Prognosis Methods Group template for prognosis reviews. Two review authors independently screened the studies, extracted data, assessed the risk of bias according to the Quality in Prognosis Studies (QUIPS) tool, and assessed the certainty of the evidence (GRADE). Meta-analyses were performed by pooling adjusted estimates. When meta-analysis was not possible, we reported the main results narratively.

We included seven studies (726,293 participants), all of which were retrospective cohort studies with participants recruited and managed in hospitals between 2000 and 2018. Studies took place in the USA, Spain, and Japan. Most studies were multicentre. None were conducted in low- or middle-income countries. The participants' mean age ranged from 62 to 69 years, and the proportion of females was higher in six of the seven studies, ranging from 46% to 60%. Sex and gender terms were used inconsistently. Participants received different PE treatments: reperfusion, inferior vena cava filter, anticoagulation, and haemodynamic/respiratory support. The prognostication time (the point from which the outcome was predicted) was frequently omitted. The included studies provided data for three of our outcomes of interest. We did not consider any of the studies to be at an overall low risk of bias for any of the outcomes analysed. We judged the certainty of the evidence as moderate to low due to imprecision and risk of bias. We found moderate-certainty evidence (due to imprecision) that for female patients there is likely a small but clinically important reduction in all-cause mortality at 30 days (odds ratio (OR) 0.81, 95% confidence interval (CI) 0.72 to 0.92; I² = 0%; absolute risk difference (ARD) 24 fewer deaths in women per 1000 participants, 95% CI 35 to 10 fewer; 2 studies, 17,627 participants). However, the remaining review outcomes do not indicate lower mortality in female patients. There is low-certainty evidence (due to serious risk of bias and imprecision) indicating that for females with PE, there may be a small but clinically important increase in all-cause hospital mortality (OR 1.11, 95% CI 1.00 to 1.22; I² = 21.7%; 95% prediction interval (PI) 0.76 to 1.61; ARD 13 more deaths in women per 1000 participants, 95% CI 0 to 26 more; 3 studies, 611,210 participants). There is also low-certainty evidence (due to very serious imprecision) indicating that there may be little to no difference between males and females in PE-related mortality at 30 days (OR 1.08, 95% CI 0.55 to 2.12; I² = 0%; ARD 4 more deaths in women per 1000 participants, 95% CI 22 fewer to 50 more; 2 studies, 3524 participants). No study data was found for the other outcomes, including sex-specific mortality data at one year. Moreover, due to insufficient studies, many of our planned methods were not implemented. In particular, we were unable to conduct assessments of heterogeneity or publication bias or subgroup and sensitivity analyses.

The evidence is uncertain about sex (being male or female) as an independent prognostic factor for predicting mortality in adults with PE. We found that, for female patients with PE, there is likely a small but clinically important reduction in all-cause mortality at 30 days relative to male patients. However, this result should be interpreted cautiously, as the remaining review outcomes do not point to an association between being female and having a lower risk of death. In fact, the evidence in the review also suggested that, in female patients, there may be a small but clinically important increase in all-cause hospital mortality. It also showed that there may be little to no difference in PE-related mortality at 30 days between male and female patients. There is currently no study evidence from longitudinal studies for our other review outcomes. Although the available evidence is conflicting and therefore cannot support a recommendation for or against routinely considering sex to quantify prognosis or to guide personalised therapeutic approaches for patients with PE, this Cochrane review offers information to guide future primary research and systematic reviews.