Variability Analysis as a Predictor of Liberation From Mechanical Ventilation

Overview

The purpose of this study is to evaluate if the variability of biological signals, such as heart rate and temperature, can predict weaning from mechanical ventilation in patients with failure to wean.

Failure to wean is one of the most feared complications of mechanical ventilation. Prolonged mechanical ventilation occurs in 5-20% of all patients requiring MV in an intensive care unit. The published experience in the Cleveland Clinic in-hospital-weaning unit (ReSCU, Respiratory Special Care Unit) on long-term mechanical ventilation (\>14 days) is that on average, 60% of the patients achieve complete ventilator independence. The identification of factors that predict liberation from mechanical ventilation should improve outcomes and allocation of resources. Several attempts have been done to develop models to identify patients who will wean from prolonged MV; most rely on multiple measurements and their predictive ability is uncertain. Given the complexity of medical problems and the heterogeneity of the patients with prolonged MV, it is not unexpected that in order to achieve a good prediction, multiple variables are needed to encompass the whole population and relevant factors associated with the failure to wean. Heart rate variability (HRV) is obtained from the measurement of the interval between successive heart beats; its analysis has been used extensively in cardiovascular disease. HRV is interpreted as a manifestation of the neurohumoral and autonomic system influence over the heart. Some researchers think that HRV relates to overall variability and is a manifestation of health in a complex biological system. Interpretation apart, multiple studies have demonstrated that the loss of variability reflects a poor prognosis overall. This phenomenon is seen not only in the heart, but also in breathing patterns, blood pressure, leukocyte count, electroencephalogram, gait, and recently in temperature. Limitations in interpretation and difficulty in acquisition (ventilator influence, respiratory rate influence, arrhythmias, medication) make HRV less practical in the mechanically ventilated population. Surface temperature is not usually measured in clinical practice; it reflects skin thermoregulatory properties (where the autonomic system has a fundamental role). Recently, a novel marker to describe biological variability in body temperature has been studied. The temperature curve complexity (TCC) was used to predict mortality in critically ill patients. Its performance was comparable to that of scores like APACHE and SOFA, this has been described with HRV, hence suggesting a relation between TCC and HRV. Although no prior study has related TCC or HRV alone or to prolonged mechanical ventilation outcomes, there is reason to suspect that variability is decreased in ventilator-dependent patients. In summary, the analysis of HRV and TCC has provided prognostic information in critically ill patients. A relation between HRV and TCC is expected but has not been described. Given the multiple influences of MV on HRV, TCC may be a better indicator of of variability in patients on MV. Our primary hypothesis is that the analysis of HRV and TCC will predict failure to wean from prolonged MV. Our secondary aims are: to identify the relationship between HRV and TCC, to describe the TCC in patients with prolonged MV, and to describe the changes in HRV and TCC in patients with successful and failed weaning. This is an observational study in which measurements of HRV and TCC will be done in patients admitted to the ReSCU and compared to the ability to wean from mechanical ventilation.

Conditions

Eligibility

Locations (1)