Do heart and respiratory rate variability improve prediction of extubation outcomes in critically ill patients?

Abstract

Introduction: Prolonged ventilation and failed extubation are associated with increased harm and cost. The added value of heart and respiratory rate variability (HRV and RRV) during spontaneous breathing trials (SBTs) to predict extubation failure remains unknown.

Methods: We enrolled 721 patients in a multicenter (12 sites), prospective, observational study, evaluating clinical estimates of risk of extubation failure, physiologic measures recorded during SBTs, HRV and RRV recorded before and during the last SBT prior to extubation, and extubation outcomes. We excluded 287 patients because of protocol or technical violations, or poor data quality. Measures of variability (97 HRV, 82 RRV) were calculated from electrocardiogram and capnography waveforms followed by automated cleaning and variability analysis using Continuous Individualized Multiorgan Variability Analysis (CIMVA™) software. Repeated randomized subsampling with training, validation, and testing were used to derive and compare predictive models.

Results: Of 434 patients with high-quality data, 51 (12%) failed extubation. Two HRV and eight RRV measures showed statistically significant association with extubation failure (P <0.0041, 5% false discovery rate). An ensemble average of five univariate logistic regression models using RRV during SBT, yielding a probability of extubation failure (called WAVE score), demonstrated optimal predictive capacity. With repeated random subsampling and testing, the model showed mean receiver operating characteristic area under the curve (ROC AUC) of 0.69, higher than heart rate (0.51), rapid shallow breathing index (RBSI; 0.61) and respiratory rate (0.63). After deriving a WAVE model based on all data, training-set performance demonstrated that the model increased its predictive power when applied to patients conventionally considered high risk: a WAVE score >0.5 in patients with RSBI >105 and perceived high risk of failure yielded a fold increase in risk of extubation failure of 3.0 (95% confidence interval (CI) 1.2 to 5.2) and 3.5 (95% CI 1.9 to 5.4), respectively.

Conclusions: Altered HRV and RRV (during the SBT prior to extubation) are significantly associated with extubation failure. A predictive model using RRV during the last SBT provided optimal accuracy of prediction in all patients, with improved accuracy when combined with clinical impression or RSBI. This model requires a validation cohort to evaluate accuracy and generalizability.

Trial registration: ClinicalTrials.gov NCT01237886. Registered 13 October 2010.

Figures

Figure 1

Flow diagram of selection of patients. Beside standard exclusions due to protocol and technical violations, the diagram shows how the dataset was reduced to ensure proper variability computation. In particular, patients were excluded when (1) having less than two windows of both heart rate and respiratory rate variability to analyze prior and during the spontaneous breathing trial, and (2) variability was extracted from waveforms deemed to be poor quality.

Figure 2

Distributions of respiratory rate (RR),rapid shallow breathing index (RSBI) and variability. This figure shows the distribution of values for passed and failed of three different measures (from the left: respiratory rate, rapid shallow breathing index, and respiratory rate variability recurrence quantification analysis: maximal diagonal line). Each grey circle represents a subject. The black box with a white line in between represents the median with its 95% confidence interval.

Figure 3

Weaning and variability evaluation (WAVE) score quartile. This figure shows the risk/fold increase in risk of failing extubation associated with each quartile of the population. The risk is defined as the number of patients who failed divided by the total number of patients in a given quartile. The fold increase in risk is the risk divided by the average risk of failure of the dataset (approximately 12%). The total number of patients is 434, therefore each quartile is representative of 108 patients.

Figure 4

Weaning and variability evaluation (WAVE) score, rapid shallow breathing index (RBSI) and clinical impression. These figures show how the risk/fold increase in risk of failing extubation associated with positive WAVE score (that is above 0.5) increases with increasing RSBI during SBT (above), or the clinical impression of the physician at the end of the SBT (below). The risk is defined as the number of patients who failed divided by the total number of patients in a given group (for example, above 0.5). The fold increase in risk is the risk divided by the average risk of failure of the dataset (approximately 12%). There are 396 patients with low RSBI (45 failed, 351 passed), and 26 patients with high RSBI (6 failed, 20 passed), while 12 passed had no RSBI reported. There is no statistically significant difference between the number of failed and passed that had no RSBI reported (P value = 0.2, chi-squared test for proportions). There are 330 patients with low/average risk of failure (32 failed, 298 passed), and 45 with high risk of failure (12 failed, 33 passed), while 7 failed and 52 passed have no perceived risk of failure reported. There is no statistically significant difference between the number of failed and passed that had no perceived risk of failure reported (P value = 0.98, chi-squared test for proportions).