Pulmonary Embolism Rule-out Criteria: Diagnostic Accuracy and Impact of COVID-19

Background

The Pulmonary Embolism Rule-Out Criteria (PERC), developed to minimize unnecessary testing in low-risk pulmonary embolism (PE) cases, lacks clear validation in the context of COVID-19.

Objectives

To assess the validity of the PERC in emergency department patients having undergone computed tomography pulmonary angiography (CTPA) during the COVID-19 pandemic.

Methods

We conducted a retrospective analysis of emergency department patients who underwent CTPA for suspected PE. COVID-19 status was based on the results of a reverse transcription-polymerase chain reaction (RT-PCR) test performed in the emergency department, or within 30 days prior to visiting the emergency department. We collected data on demographics, symptoms, d-dimer levels, and medical history relevant to thrombosis and conducted the PERC test using the criteria including age, oxygen saturation, heart rate, and the absence of hemoptysis or recent trauma. We categorized outcomes based on the concordance between the PERC results and CTPA findings, with specific definitions for true positive and negative, as well as false positive and negative results. We also evaluated the impact of COVID-19 status on the diagnostic performance of the PERC by analyzing the prevalence of PE in patients testing positive and negative for COVID-19.

Results

Among the 2.430 participants, 45.1% tested negative for COVID-19, 43.4% tested positive, and 11.5% were untested. The PERC identified 91.2% of the cases as positive, 6.9% of which were confirmed to have PE. Overall, 84.9% of cases (n = 2.062) showed a discordant result between the PERC and CTPA findings. The lack of significant correspondence between the PERC positivity and actual PE presence (p = 0.001; p < 0.01) indicated low diagnostic concordance. In patients with a positive COVID-19 test result, the PERC demonstrated a sensitivity of 95.3% (95% CI: 86.91–99.02), a specificity of 9.1% (95% CI: 7.46–11.15), a positive predictive value of 6.3% (95% CI: 6.01–6.70), a negative predictive value of 96.8% (95% CI: 90.81–98.94), and an accuracy of 14.4% (95% CI: 12.34–16.67). In patients who tested negative for COVID-19, the sensitivity was 95.4% (95% CI: 88.64–98.73), the specificity was 7.8% (95% CI: 6.25–9.66), the positive predictive value was 8.1% (95% CI: 7.83–8.57), the negative predictive value was 95.1% (95% CI: 88.11–98.14), and the accuracy was 14.7% (95% CI: 12.73–17.02).

Conclusion

The study demonstrates that the sensitivity and negative predictive value of the PERC are comparable in COVID-19 positive and negative patients. Furthermore, the incidence of PE among patients presenting to the emergency department did not significantly differ based on COVID-19 status. While this study highlights the relevance of the PERC in clinical decision-making, caution is advised as the PERC may not always provide reliable results when used as the sole diagnostic test.