Background: Seasonality in pulmonary embolism (PE) has been described and debated for a long time as we try to understand the risk factors leading towards its presentation.

Aims: We aim to replicate the findings of other works regarding PE seasonality in hospital admissions of the United States of America during the years of 2006 to 2014.

Methods: The NIS database was queried from 2006 to 2014 for adult patients with a discharge diagnosis of PE by ICD-9 CM codes. A time series analysis was performed with the monthly admission frequency for every year. All data was weighted and analyzed with a Winter’s Additive model as well as Auto-correlation Function (ACF) and Partial Auto-correlation Function (PACF) to identify a seasonality pattern. The population was then stratified per age group between patients ≤64 years of age and ≥ 65 years old. Finally a parallel analysis was performed in the described population involving a discharge diagnosis of both PE and deep vein thrombosis interpreted as a total of VTE patients.

Results: A total of 4’239.406 patients were analyzed for VTE and 2’545.320 (60% of total VTE) for PE. Of them 52.4% were females and 51.3% were ≥65 years old. The analysis demonstrated a seasonal variability with a statistically significant Ljung-Box Q test of < 0.001. 7% more admissions were observed during the month of January and 5% less admissions in June compared to the central tendency. Results obtained from the VTE analysis showed a similar trend. During the subgroup analysis a seasonality could not be demonstrated on the ≥ 65 years old population. There was an exponentially rising trend from 2006 to 20014 demonstrating a positive correlation (R²=0.974) over the years.

Conclusions: Our data contributes to the building literature in favor of seasonality as an underlying factor driving PE and VTE incidence over the years.

« Back to ISTH 2020 Congress

ISTH Congress Abstracts - https://abstracts.isth.org/abstract/seasonality-in-hospitalizations-for-pulmonary-embolism-between-2006-2016/