Are Meteorological Parameters a Risk Factor for Pulmonary Embolism? A Retrospective Analysis of 530 Patients

Abstract

Background: The influence of meteorological condi- tions on cardiovascular morbidity and mortality hasbeen known for a long time. However, few reportshave been published on the influence of meteorologi- cal parameters on the occurrence of acute pulmonaryembolism (PE). Aims: In this retrospective study, we compared themeteorological parameters between PE patients withrisk factors and idiopathic PE patients.Study Design: Cross-sectional study. Methods: Medical documentation of 1180 patientswith suspected acute pulmonary embolism diagnosedbetween January 2010 and December 2012 was ret- rospectively analyzed. A total of 530 patients with PEconfirmed by computed tomography pulmonary an- giography and/or ventilation/perfusion scan were in- cluded for further analysis. We divided the patients intotwo groups: PE with risk factors (provoked) and PEwithout risk factors (unprovoked). The meteorologicaldata were collected from the relevant time period: tem- perature, humidity, pressure, and wind velocity. As theexact time of PE onset was unknown, the meteorologi- cal values attributed to each patient were the means ofthe values in the months or weeks at the time of diag- nosis of PE. Results: The highest numbers of cases were seen inautumn (29.8%), followed by summer (28.9%), spring(22.1%), and winter (19.2%). In terms of months, thegreatest number of cases occurred in June (57), fol- lowed by November (56) and October (54). Case dis- tribution according to the months and seasons were sta- tistically significant. The wind direction also affectedthe incidence of PE. There was a statistically signifi- cant positive correlation between case frequency andair temperature (r=0.300; p=0.031). No correlation wasfound between the unprovoked PE cases monthly dis- tribution and pressure, humidity, or temperature. How- ever, there was a statistically significant positive cor- relation between the monthly distribution of the groupwith provoked PE cases and air temperature (r=0.586;p=0.045).Conclusion: A statistically significant inverse correla- tion between atmospheric pressure and temperature andthe number of all PE cases was observed in our study,which is in accordance with other reports. However, inunprovoked PE cases, there was no correlation betweenmeteorological parameters and case incidence.