# Ultra Short Heart Rate Variability as a Prognostic Marker in Pulmonary Embolism: A Retrospective Cohort Study

## TL;DR

This study explores whether ultra-short heart rate variability measurements can predict outcomes in patients with pulmonary embolism, finding some value in non-oncological patients.

## Contribution

The study introduces the use of ultra-short HRV indices for risk stratification in acute pulmonary embolism patients.

## Key findings

- Higher SDNN in non-oncological PE patients was independently linked to increased 90-day mortality.
- HRV indices showed no predictive value for mortality in oncological PE patients.
- Lower HRV was associated with worse survival in both the overall cohort and non-oncological subgroup.

## Abstract

Background/Objectives: Pulmonary embolism (PE) remains a significant cause of cardiovascular mortality, with risk stratification being critical for optimizing treatment decisions. Heart rate variability (HRV), a measure of autonomic nervous system function, had been explored as a prognostic index in various cardiovascular conditions, yet has received limited investigation regarding PE prognosis. Our objective was to evaluate the prognostic value of ultra-short HRV indices, obtained at the emergency department (ED), in patients presenting with PE. Methods: A retrospective cohort study, conducted at Rambam Health Care Campus, Haifa, Israel. All eligible patients diagnosed with acute PE at the ED, between the years 2010 and 2012 were included. Further, a subgroup analysis was performed to differentiate between oncological (n = 118) and non-oncological (n = 115) patient populations. Ten-seconds electrocardiogram was used to compute ultra-short HRV indices, specifically SDNN (standard deviation of normal-to-normal RR intervals) and RMSSD (root mean square of successive differences). Multivariate logistic regression models were created to assess HRV’s independent predictive value for 30-day and 90-day mortality. In addition, a survival analysis was carried out utilizing Cox regression and Kaplan-Meier curves. Results: 233 patients (42% male; age 65 ± 17) were included in the analysis. Ultra-short HRV indices did not significantly correlate with short-term mortality. However, in non-oncological patients (n = 115), multivariate analysis demonstrated that higher SDNN (as a continuous variable), was independently associated with increased 90-day mortality (AOR 1.018, 95% CI 1.000–1.037; p = 0.044). In contrast, HRV showed no predictive value for mortality in oncological patients. In both the entire cohort and the non-oncological sub-group, Kaplan-Meier plots established statistically significant differences, with lower HRV indices correlating with worse survival. This finding is paradoxical. The issue of context-dependent HRV (i.e., based on ECG obtained during rapid shallow breathing, which reduces HRV on the one hand, but is possibly adaptive during an acute PE, to increase oxygen supply and prevent shock in the short run, on the other hand), may explain these findings. Conclusions: Ultra-short HRV shows some promise in short term risk stratification of non-oncological PE patients. As for oncological patients, HRV was not found to have short term prognostic relevance.