Algorithm for Quantifying Frontal EMG Responsiveness for Sedation Monitoring

Novel insights on association and reactivity of Bispectral Index, frontal electromyogram, and autonomic responses in nociception-sedation monitoring of critical care patients

Background

Assessing nociception and sedation in mechanically ventilated patients in the ICU is challenging, with few reliable methods available for continuous monitoring. Measurable cardiovascular and neurophysiological signals, such as frontal EEG, frontal EMG, heart rate, and blood pressure, have potential in sedation and nociception monitoring. The hypothesis of this explorative study is that derived variables from the aforementioned signals predict the level of sedation, as described by the Richmond Agitation-Sedation score (RASS), and respond to painful stimuli during critical care.

Methods

Thirty adult postoperative ICU patients on mechanical ventilation and receiving intravenous sedation, excluding patients with primary neurological disorders, head injury, or need for continuous neuromuscular blockage. Bispectral Index (BIS), EMG power (EMG), EMG-derived Responsiveness Index (RI), and averaged blood pressure variability (ARV) were tested against RASS measurements. The aforementioned variables together with blood pressure and Surgical Pleth Index (SPI) were explored before and after painful stimuli (for example bronchoscopy, or pleural puncture) at varying RASS levels, to test variable responsiveness.

Results

BIS, EMG, and RI predicted RASS levels with a prediction probability (PK) of 0.776 for BIS, 0.761 for EMG, and 0.763 for RI. In addition, BIS, EMG, and ARV demonstrated responsiveness to painful stimuli during deep sedation (RASS score ≤ -3).

Conclusion

Variables derived from EEG and EMG are associated with sedation levels, as described by the RASS score. Furthermore, these variables, along with ARV, react with consistency to painful stimuli during deep sedation (RASS -5 to -3), offering novel tools for nociception-sedation monitoring of mechanically ventilated ICU patients requiring deep sedation.

Staff education, regular sedation and analgesia quality feedback, and a sedation monitoring technology for improving sedation and analgesia quality for critically ill, mechanically ventilated patients: a cluster randomised trial

BACKGROUND

Optimal sedation of patients in intensive care units (ICUs) requires the avoidance of pain, agitation, and unnecessary deep sedation, but these outcomes are challenging to achieve. Excessive sedation can prolong ICU stay, whereas light sedation can increase pain and frightening memories, which are commonly recalled by ICU survivors. We aimed to assess the effectiveness of three interventions to improve sedation and analgesia quality: an online education programme; regular feedback of sedation-analgesia quality data; and use of a novel sedation-monitoring technology (the Responsiveness Index [RI]).

METHODS

We did a cluster randomised trial in eight ICUs, which were randomly allocated to receive education alone (two ICUs), education plus sedation-analgesia quality feedback (two ICUs), education plus RI monitoring technology (two ICUs), or all three interventions (two ICUs). Randomisation was done with computer-generated random permuted blocks, stratified according to recruitment start date. A 45 week baseline period was followed by a 45 week intervention period, separated by an 8 week implementation period in which the interventions were introduced. ICU and research staff were not masked to study group assignment during the intervention period. All mechanically ventilated patients were potentially eligible. We assessed patients' sedation-analgesia quality for each 12 h period of nursing care, and sedation-related adverse events daily. Our primary outcome was the proportion of care periods with optimal sedation-analgesia, defined as being free from excessive sedation, agitation, poor limb relaxation, and poor ventilator synchronisation. Analysis used multilevel generalised linear mixed modelling to explore intervention effects in a single model taking clustering and patient-level factors into account. A concurrent mixed-methods process evaluation was undertaken to help understand the trial findings. The trial is registered with ClinicalTrials.gov, number NCT01634451.

FINDINGS

Between June 1, 2012, and Dec 31, 2014, we included 881 patients (9187 care periods) during the baseline period and 591 patients (6947 care periods) during the intervention period. During the baseline period, optimal sedation-analgesia was present for 5150 (56%) care periods. We found a significant improvement in optimal sedation-analgesia with RI monitoring (odds ratio [OR] 1·44 [95% CI 1·07-1·95]; p=0·017), which was mainly due to increased periods free from excessive sedation (OR 1·59 [1·09-2·31]) and poor ventilator synchronisation (OR 1·55 [1·05-2·30]). However, more patients experienced sedation-related adverse events (OR 1·91 [1·02-3·58]). We found no improvement in overall optimal sedation-analgesia with education (OR 1·13 [95% CI 0·86-1·48]), but fewer patients experienced sedation-related adverse events (OR 0·56 [0·32-0·99]). The sedation-analgesia quality data feedback did not improve quality (OR 0·74 [95% CI 0·54-1·00]) or sedation-related adverse events (OR 1·15 [0·61-2·15]). The process evaluation suggested many clinicians found the RI monitoring useful, but it was often not used for decision making as intended. Education was valued and considered useful by staff. By contrast, sedation-analgesia quality feedback was poorly understood and thought to lack relevance to bedside nursing practice.

INTERPRETATION

Combination of RI monitoring and online education has the potential to improve sedation-analgesia quality and patient safety in mechanically ventilated ICU patients. The RI monitoring seemed to improve sedation-analgesia quality, but inconsistent adoption by bedside nurses limited its impact. The online education programme resulted in a clinically relevant improvement in patient safety and was valued by nurses, but any changes to behaviours did not seem to alter other measures of sedation-analgesia quality. Providing sedation-analgesia quality feedback to ICUs did not appear to improve any quality metrics, probably because staff did not think it relevant to bedside practice.

FUNDING

Chief Scientist Office, Scotland; GE Healthcare.

Rationale, design and methodology of a trial evaluating three strategies designed to improve sedation quality in intensive care units (DESIST study)

OBJECTIVES

To describe the rationale, design and methodology for a trial of three novel interventions developed to improve sedation-analgesia quality in adult intensive care units (ICUs).

PARTICIPANTS AND SETTING

8 clusters, each a Scottish ICU. All mechanically ventilated sedated patients were potentially eligible for inclusion in data analysis.

DESIGN

Cluster randomised design in 8 ICUs, with ICUs randomised after 45 weeks baseline data collection to implement one of four intervention combinations: a web-based educational programme (2 ICUs); education plus regular sedation quality feedback using process control charts (2 ICUs); education plus a novel sedation monitoring technology (2 ICUs); or all three interventions. ICUs measured sedation-analgesia quality, relevant drug use and clinical outcomes, during a 45-week preintervention and 45-week postintervention period separated by an 8-week implementation period. The intended sample size was >100 patients per site per study period.

MAIN OUTCOME MEASURES

The primary outcome was the proportion of 12 h care periods with optimum sedation-analgesia, defined as the absence of agitation, unnecessary deep sedation, poor relaxation and poor ventilator synchronisation. Secondary outcomes were proportions of care periods with each of these four components of optimum sedation and rates of sedation-related adverse events. Sedative and analgesic drug use, and ICU and hospital outcomes were also measured.

ANALYTIC APPROACH

Multilevel generalised linear regression mixed models will explore the effects of each intervention taking clustering into account, and adjusting for age, gender and APACHE II score. Sedation-analgesia quality outcomes will be explored at ICU level and individual patient level. A process evaluation using mixed methods including quantitative description of intervention implementation, focus groups and direct observation will provide explanatory information regarding any effects observed.

CONCLUSIONS

The DESIST study uses a novel design to provide system-level evaluation of three contrasting complex interventions on sedation-analgesia quality. Recruitment is complete and analysis ongoing.

TRIAL REGISTRATION NUMBER

NCT01634451.

A randomized controlled proof-of-concept trial of early sedation management using Responsiveness Index monitoring in mechanically ventilated critically ill patients

INTRODUCTION

Deep sedation is associated with adverse patient outcomes. We recently described a novel sedation-monitoring technology, the Responsiveness Index (RI), which quantifies patient arousal using processed frontal facial EMG data. We explored the potential effectiveness and safety of continuous RI monitoring during early intensive care unit (ICU) care as a nurse decision-support tool.

METHODS

In a parallel-group controlled single centre proof of concept trial, patients requiring mechanical ventilation and sedation were randomized via sequential sealed envelopes following ICU admission. Control group patients received hourly clinical sedation assessment and daily sedation holds; the RI monitor was connected but data were concealed from clinical staff. The intervention group received control group care, but RI monitoring was visible and nurses were asked to adjust sedation to maintain patients with an RI>20 whenever possible. Traffic-light colour coding (RI<20, Red; 20-40, Amber; >40, Green) simplified decision-making. The intervention lasted up to 48 hours. Sixteen nurses were interviewed to explore their views of the novel technology.

RESULTS

We analysed 74 patients treated per protocol (36 intervention; 38 control). The proportion of patients with RI<20 was identical at the start of monitoring (54% both groups). Overall, the proportion of time with RI<20 trended to lower values for the intervention group (median 16% (1-3rd quartile 8-30%) versus 33% (10-54%); P = 0.08); sedation and analgesic use was similar. A post hoc analysis restricted to patients with RI<20 when monitoring started, found intervention patients spent less time with low RI value (16% (11-45%) versus 51% (33-72%); P = 0.02), cumulative propofol use trended to lower values (median 1090 mg versus 2390 mg; P = 0.14), and cumulative alfentanil use was lower (21.2 mg versus 32.3 mg; P = 0.01). RASS scores were similar for both groups. Sedation related adverse event rates were similar (7/36 versus 5/38). Similar proportions of patients had sedation holds (83% versus 87 %) and were extubated (47% versus 44%) during the intervention period. Nurses valued the objective visible data trends and simple colour prompts, and found RI monitoring a useful adjunct to existing practice.

CONCLUSIONS

RI monitoring was safe and acceptable. Data suggested potential to modify sedation decision-making. Larger trials are justified to explore effects on patient-centred outcomes.

TRIAL REGISTRATION

NCT01361230 (registered April 19, 2010).