Background: Diagnostic delay in patients with pulmonary embolism (PE) is typical, yet the proportion of patients with PE that experienced delay and for how many days is less well described, nor are determinants for such delay.
Objectives: This study aimed to assess the prevalence and extent of delay in diagnosing PE.
Methods: A systematic literature search was performed to identify articles reporting delays in diagnosing PE. The primary outcome was mean delay (in days) or a percentage of patients with diagnostic delay (defined as PE diagnosis more than seven days after symptom onset). The secondary outcome was determinants of delay. Random-effect meta-analyses were applied to calculate a pooled estimate for mean delay and to explore heterogeneity in subgroups.
Results: The literature search yielded 10,933 studies, of which 24 were included in the final analysis. The pooled estimate of the mean diagnostic delay based on 12 studies was 6.3 days (95% prediction interval 2.5 to 15.8). The percentage of patients having more than seven days of delay varied between 18% and 38%. All studies assessing the determinants of coughing (n = 3), chronic lung disease (n = 6) and heart failure (n = 8) found a positive association with diagnostic delay. Similarly, all studies assessing recent surgery (n = 7) and hypotension (n = 6), as well as most studies assessing chest pain (n = 8), found a negative association with diagnostic delay of PE.
Conclusion: Patients may have symptoms for almost one week before PE is diagnosed and in about a quarter of patients, the diagnostic delay is even longer.
Keywords: Pulmonary embolism; delay; diagnosis; meta-analysis; systematic review; venous thromboembolism.