Background: Acute pulmonary embolism (PE) severity index (PESI) well predicts 30-day mortality in PE patients. However, improvements have been advocated because it sometimes overestimates early mortality risk in stable PE patients, especially in shorter simplified forms (sPESI). (1, 2) We aimed to evaluate predictability of PESI in our patients and to explore interaction with other possible prediction tools as a potential improvement in stratifying low risk patients eligible for ambulatory treatment.

Patients and Methods: Retrospective analysis of demographic, clinical and laboratory variables in consecutive 299 adults with MSCT confirmed acute PE admitted to a single institution over a 3-year period with 30-day mortality as a main outcome.

Results: There were 19 (6.4%) severely unstable patients who died within 48 hours and were excluded from further analyses. Among remaining stable patients, 30-day mortality was 12.1% (34/280). There were 131 patients with PESI <105 and 185 patients with PESI <125 with 30-day mortality rates of 3.1% and 4.9%, respectively. Among all variables analyzed, only estimated glomerular filtration rate, D-dimer value, platelet-to-lymphocyte ratio, and red blood cell distribution width (RDW) value showed significant effect on PESI predicted mortality in multivariate regression analysis. RDW moderation of PESI effect on mortality was most notable: there were 140/280 (50%) patients identified with both PESI < 125 and RDW < 15% and a 30-day mortality rate of only 0.7% (1/140). In addition, there were no deaths in the first 30 days among 97/280 (35%) patients with both PESI < 105 and RDW < 14.5%.

Conclusions: In stable acute PE patients RDW strongly moderates 30-day mortality risk associated with PESI. It could be used to improve PESI accuracy and identify a larger proportion of stable patients eligible for safe ambulatory treatment and quick and safe discharge from emergency room with appropriate anticoagulation therapy.