Using Bispectral Index Monitoring to Detect Potential Breakthrough Awareness and Limit Duration of Neuromuscular Blockade

Determining any potentially treatable cause of agitation is paramount in optimal delivery of care to critically ill patients. The author presents a case report that illustrates the use of BIS monitoring in a patient initially treated with sedation/analgesia and neuromuscular blockade to manage dangerous agitation and dyssynchrony with controlled ventilation.Pain, anxiety, agitation, and ventilator dyssynchrony all may be consequences of critical illness, traumatic injury, or intubation and controlled ventilation.1,2 Controlled ventilation, airway manipulation, and suctioning can cause pain and an arousal response. Excessive intrathoracic pressure may also compromise hemodynamic stability and venous return and increase pulmonary vascular resistance3 and the risk for barotrauma and pneumothorax.3,4Pain, anxiety, and agitation also mobilize the stress response, a complex, neurohormonal response to physiological stress, including stress caused by critical illness. Consequences include elevation in circulating catecholamines, which may increase heart rate, blood pressure, and metabolic demands.1,5 In the short-term, the stress response helps maintain physiological stability. Over the long-term, it may increase levels of stress hormones such as cortisol and glucagon, cause a hypercoagulable state, and compromise immune function and gastrointestinal motility, all of which may increase morbidity.1,4,5Agitation is most commonly a consequence of a concurrent clinical state such as pain or severe anxiety. Determining any potentially treatable cause of agitation is paramount in optimal delivery of care to critically ill patients. Potentially treatable causes of agitation include pain, anxiety, and positive-pressure ventilation.1 Agitation is often treated with agents such as benzodiazepines and opioids.6 For patients with dangerous, life-threatening agitation and severe ventilator dyssynchrony after initial doses of sedatives and analgesics, neuromuscular blockade may be necessary to prevent further clinical compromise.7Risks associated with neuromuscular blockade include prolonged weakness and length of stay, potential for awake paralysis, and dramatically compromised neurological evaluation, which increases morbidity.8–11 Determining the effectiveness of sedation and analgesia in patients with chemically induced paralysis is difficult at best. Parameters such as heart rate, blood pressure, diaphoresis, and tearing, particularly in response to stimulation, may be affected by the underlying illness, circulating fluid volume, and concurrent drug therapy.1,7,8,12,13Risks associated with undersedation include prolonged mobilization of the stress response and risks of associated morbidity.1,5,12 With neuromuscular blockade, patients may be aware and experience pain while paralyzed.8 Even with concurrent sedation and analgesia, some patients can be awake, anxious, and in pain while chemically paralyzed. In one study,14 more than one third of patients following recovery reported having experienced these conditions during paralysis. Patients who survive critical or catastrophic illness that includes “awake paralysis” may experience posttraumatic stress after discharge from the intensive care unit (ICU).2Risks associated with oversedation include masking of neurological signs and symptoms and causing hemodynamic compromise, which may be more pronounced in volume-depleted patients or patients with higher illness severity.1,2,12 Oversedation also makes the differential diagnosis of agitation difficult and is associated with prolonged recovery of baseline level of consciousness and prolonged intubation or ventilation.1,12,15,16 Excessive sedation may also increase ICU length of stay,12 risk of airway injury, and overall cost of care.Each extreme (oversedation or undersedation) may increase morbidity and length of stay. This risk is more pronounced in patients receiving neuromuscular blockade in whom the clinical assessment of sedation is limited. In my experience, a patient given lorazepam for sedation at 10 mg/h for 3 days while chemically paralyzed reported, after recovery and extubation, accurate recollections of being awake while paralyzed. This experience was characterized by the patient as a “bad dream.” In a second example in my experience, a patient chemically paralyzed for management of status asthmaticus was given a lorazepam infusion of 6 mg/h for 3 days during neuromuscular blockade. After discontinuation of neuromuscular blockade and recovery of neuromuscular function, the patient was aggressively weaned from the lorazepam infusion. This patient remained deeply sedated for 3 days, resulting in a significant delay in readiness to be weaned from controlled ventilation. These examples dramatically illustrate difficulties in monitoring level of sedation during neuromuscular blockade.Progress is being reported on use of electroencephalographic (EEG)-based parameters to monitor sedation and arousal states. Bispectral Index (BIS) monitoring has been used in the operating room to monitor the effects of anesthetic agents on the brain. Clinical experience and research in the ICU with this technology are being increasingly reported, indicating its effectiveness in monitoring the level of sedation and the response to stimulation. Because of the risks associated with both extremes in the administration of sedative and analgesic agents and with neuromuscular blockade therapy, reasonable goals of care can be determined. One goal is limiting duration and use of neuromuscular blockade therapy as clinically appropriate. A second goal is more precise, making real-time adjustments in the doses of sedative and analgesic agents, potentially reducing risks associated with undersedation and oversedation. The following case report illustrates use of BIS monitoring in a patient initially treated with sedation/analgesia and neuromuscular blockade to manage dangerous agitation and dyssynchrony with controlled ventilation.R.B., a 54-year-old man, came to the emergency department because he had had an increasingly severe sore throat for 2 days. He said that before admission and the onset of the illness, he worked outdoors and “felt fine.” He had a history of hypertension, which was treated with oral hydrochlorothiazide 50 mg/d, carvedilol (Coreg) 12.5 mg twice daily, a combination of amlodipine besylate 2.5 mg plus benazepril hydrochloride 10 mg (Lotrel), and controlled-onset verapamil (Covera) 180 mg at bedtime. He also had non–insulin-dependent diabetes mellitus, which was treated with oral metformin (Glucophage) 500 mg twice daily. He lived with his family, did not smoke, and drank alcohol only socially. His reported weight and height were 118 kg (260 lb) and 191 cm (6 ft 3 in), respectively.On initial assessment in the emergency department, mild dyspnea and neck tenderness were noted. R.B. had difficulty swallowing, and he and his family reported that his voice “sounded different.” Emergency department staff reported that his voice was soft and high-pitched and that he was not drooling. The admitting diagnosis was probable epiglottitis. After administration of oxygen by aerosol mask at a fraction of inspired oxygen of 0.4 and insertion of a peripheral intravenous access catheter, chest and neck radiographs were obtained. Immediately after these procedures, R.B. became increasingly short of breath and stridor developed. Because his airway was at risk, he was intubated with a No. 6 oral endotracheal tube. Physical examination by the anesthesiology provider during direct laryngoscopy for intubation revealed marked tissue inflammation in the posterior part of the pharynx, including marked edema of the vocal cords. These findings confirmed the admitting diagnosis of severe epiglottitis.After endotracheal intubation, R.B. was agitated and dyssynchronous with positive-pressure ventilation, causing marked elevations in peak inspiratory pressures and increasing the risk for barotrauma and pneumothorax. The agitation was potentially dangerous, with a potential for self-extubation and difficulty in reestablishing the airway. Because the clinical situation might deteriorate, and because of R.B.’s large body size, control of movement was essential. Therefore, he was given increased doses of lorazepam (Ativan) intravenously for sedation, up to 12 mg in divided doses within 20 minutes. He was also given morphine sulfate 8 mg intravenously over 20 minutes in divided doses. These interventions were ineffective in maintaining synchrony with controlled ventilation and in controlling agitation. He was then paralyzed with vecuronium bromide to ensure synchrony with controlled ventilation and prevent injury. After stabilization, his blood pressure was 120/76 mm Hg, his temperature was 38.4°C (101°F), and his heart rate was 70/min to 80/min. Respirations were 12/min, as set on the ventilator. A chest radiograph obtained after intubation showed an infiltrate in the left upper lung field. Arterial oxygenation was adequate, and the fraction of inspired oxygen was adjusted to 0.5; oxygen saturation was maintained at greater than 95%.R.B.’s initial drug therapy included dexamethasone (Decadron) 10 mg intravenously and ceftriaxone (Rocephin) 2 g intravenously. Findings on clinical assessment were consistent with dehydration, and he was given isotonic sodium chloride solution: 1 L over 1 hour and then at a rate of 40 mL/h. Pending transfer to the medical ICU, he received multiple bolus doses of morphine and lorazepam for analgesia and sedation and was given cisatracurium by infusion for maintaining control over gross motor movement.R.B. was admitted to the medical ICU for ventilator management and further treatment of epiglottitis and probable pneumonia. He was given ceftriaxone 2 g/d intravenously and azithromycin (Zithromax) 500 mg/d intravenously for the infection and pneumonia, ranitidine 50 mg every 8 hours intravenously for gastrointestinal prophylaxis, heparin sodium 5000 IU subcutaneously every 12 hours to prevent deep vein thrombosis, and dexamethasone 8 mg intravenously every 8 hours to treat laryngeal/vocal cord inflammation and edema. Neuromuscular blockade was maintained with an infusion of cisatracurium at 1.2 μg/kg per minute. Initial orders for sedation and analgesia were lorazepam 4 mg intravenously as bolus doses every 2 to 3 hours and morphine sulfate 2 mg/h by continuous infusion.R.B. was given fluids intravenously for hydration. His vital signs at the time of admission to the ICU were heart rate 76/min, blood pressure 128/73 mm Hg, and respirations 15/min (ventilator rate). Additional intravenous access was established for managing drug therapy, including infusions of sedative, analgesic, and neuromuscular blocking agents as well as antibiotics and other medications.With stabilization of pulmonary and ventilator management and hemodynamic considerations, multiple critical issues were identified at this point. The first critical issue was preventing dangerous agitation and maintaining synchrony with controlled ventilation. The second was minimizing length of time on controlled ventilation as clinically appropriate. The third was assessing level of sedation, a particular challenge in patients with neuromuscular blockade.Preventing agitation remained a significant concern. Because R.B. was 191 cm tall and weighed 118 kg, he could cause serious injury to himself and to care-givers if he became severely agitated or combative. A related concern, maintaining synchrony with controlled ventilation, remained vital. Agitation may also cause dyssynchrony with positive-pressure ventilation. This dyssynchrony dramatically increases the risk for barotrauma and pneumothorax. Minimizing the duration of controlled ventilation was a concern. Prolonged periods of controlled ventilation increase the risk of airway injury, increasing morbidity.3 Assessing level of sedation remained of utmost concern. A patient, such as R.B., who is receiving neuromuscular blockers may, with relaxation of skeletal muscles, appear to be calm and asleep but may be awake and in pain.8 Inadequate sedation/analgesia could lead to awareness and pain while paralyzed, prolonged stress, and psychiatric complications after recovery. Overly aggressive sedation/analgesia predisposes patients to hemodynamic compromise, prolonged recovery, and increased length of stay.For R.B., the issue associated with intubation and controlled ventilation was one of airway maintenance. Control of movement and maintenance of synchrony with ventilation was facilitated by using sedation, analgesia, and neuromuscular blockade. A critical issue at this point was assessing the level of sedation. Using clinical assessment tools such as sedation scales is ineffective because they rely on a patient’s ability to physically respond to stimulation, a response lacking in patients with neuromuscular blockade. Hemodynamic parameters such as changes in heart rate and blood pressure indicating increased sympathetic outflow in response to stress were clearly not reliable indicators. This finding was determined by a careful review of R.B.’s drug therapy. Use of a β-blocker (carvedilol) and calcium channel blocking agents (amlodipine and verapamil) attenuated signs of increased sympathetic outflow such as hypertension and tachycardia in response to stress. Thus, using changes in vital signs as a measure of sedative/analgesic state was completely unreliable.As part of ongoing clinical research, the medical ICU had the opportunity to use a BIS monitor (model A-2000, Aspect Medical Systems, Newton, Mass) as an objective means to assess R.B.’s level of sedation while he was paralyzed. Before the BIS sensor was placed on R.B.’s forehead, a thorough skin preparation was done to remove skin oils and ensure optimal electrical contact between the skin and the monitoring system. The BIS sensor was applied to the forehead, with electrode 1 approximately 3.8 cm (1½ in) above the nose, electrode 2 above the eye, and electrode 3 on the temple area between the hairline and the corner of the eye. The BIS sensor was connected to the A-2000 monitoring system. Information derived from BIS monitoring consists of the BIS reading, on a scale from 0 to 100. Values along this scale correspond with clinical states of various levels of sedation/hypnosis and of characteristic EEG states in response to drug administration17–20 (see TableT1). Secondary parameters that also may be detected with the BIS monitor include electromyographic (EMG) activity in the facial muscles and EEG suppression ratio (SR). SR is the percent of suppressed EEG within the previous 63 seconds of EEG data. SR may be elevated during high-dose barbiturate therapy or severe cerebral injury.20 EMG activity may be of concern because the BIS may be affected by high-frequency activity in the facial muscles, which can result in an increase in the BIS independent of the hypnotic state of the brain.21 Because R.B. was being given a neuromuscular blocker, the BIS provided effective feedback on the hypnotic state of the brain. EMG activity may be more pronounced during BIS monitoring in lighter hypnotic states and in the absence of neuromuscular blockade.One goal in R.B.’s case was to discontinue neuromuscular blockade as soon as possible, when clinically appropriate. Abrupt removal of neuromuscular blockade is associated with a risk for severe agitation and injury. With paralyzed patients, another option is to administer very large doses of sedatives and opioids to ensure, to the extent possible, that the patients are not awake and in pain while paralyzed. Although this approach may reduce the risks of awareness and pain, it may also predispose patients to markedly prolonged recovery times from overly aggressive sedation/analgesia and other adverse effects, including hemodynamic compromise.Shortly after R.B.’s arrival in the ICU, BIS monitoring was started; the baseline BIS was 79 to 80. This reading indicated that R.B. might be aware and at high risk of recall and pain while paralyzed. His caregivers concluded that his agitation noted before neuromuscular blockade most likely was due to pain and anxiety. Anxiety could also be attributed to awake paralysis and sudden hospital/ICU admission. Pain could be attributed to severe pharyngeal and vocal cord inflammation, tracheal intubation, airway suctioning, and controlled ventilation.From this point, opioid analgesia was increased for more aggressive treatment of pain. During the next 45 minutes, R.B. was given a total of 16 mg of morphine sulfate along with a continuous infusion of morphine at a rate of 3 mg/h. More aggressive treatment of anxiety was provided with a 6-mg bolus of lorazepam given over 3 minutes and then a lorazepam infusion at a rate of 3 mg/h. With these increased doses, the BIS decreased significantly. From the initial reading of 79 to 80 (potential awareness and arousal state related to pain), it decreased dramatically to 50 to 60, indicating a much deeper level of sedation and less risk of recall and awareness (Figure 1). By the end of hour 2 after ICU admission, R.B. was weaned off neuromuscular blockade and remained off. This step preserved the reliability of clinical examination and, in addition to BIS monitoring, facilitated identifying clinical states of undersedation or oversedation, making possible real-time adjustments in drug therapy. After removal of neuromuscular blockade, BIS monitoring was maintained to continue tracking the depth of sedation and to indicate if R.B. was progressing to a deeper level of sedation. With a deeper level of sedation, appropriate reductions could be made in the doses of sedative and analgesic agents. The healthcare providers also wanted to assess the degree of correlation between the BIS and clinical assessment of sedation (Sedation-Agitation Scale). In R.B., the correlation between the 2 measures was strongly positive (Figure 2). During the next 18 hours, doses of sedative and analgesic agents were adjusted to maintain a calm, comfortable patient who was appropriately responsive and accepting of care. R.B. was easily weaned from mechanical ventilation and was extubated early the next day, within 24 hours of his admission to the ICU. He was discharged to a general medical unit and quickly went home, with appointments for follow-up.For R.B., without objective feedback on his level of sedation and arousal while he was paralyzed, use of large doses of sedative and analgesic agents to reduce risk of awareness and pain while paralyzed would have been justified. Administration of large doses might have resulted in a far deeper level of sedation than was necessary and could have predisposed R.B. to markedly prolonged recovery and length of stay. By using objective data on R.B.’s level of sedation, the extremes of oversedation and undersedation were avoided. A prolonged length of stay may have been avoided in R.B.’s case because of the availability of objective monitoring of his level of sedation and arousal while he was chemically paralyzed.Complications of sedative and neuromuscular blocking agents are receiving increasing and well-deserved attention.1,2,4,10–12,14–16 Multiple important goals of care can be identified. One goal is to limit dosages and duration of neuromuscular blockade to the minimum necessary to meet clinical goals. A second is to ensure adequate sedation and analgesia during neuromuscular blockade. A third goal for sedation and analgesia is to make optimal adjustments in therapy to meet but not exceed clinical goals. A fourth goal is to determine the cause of physiological stress and optimal tailoring of drug therapy to the disturbance.Patients who require neuromuscular blockade or deeper levels of sedation cannot be optimally assessed by using physical examination to determine sedative or analgesic needs. For this reason, EEG-based monitoring of sedation may provide real-time feedback on the hypnotic or arousal state of the brain. EEG waveforms closely reflect cerebral metabolism and can provide feedback on the state of the brain when the clinical examination is compromised,22 as in patients with deep sedation/analgesia or neuromuscular blockade.23–25In addition to reflecting cerebral metabolic state,26 EEG waveforms also change in response to treatment with sedative and analgesic agents. For example, benzodiazepines decrease cerebral metabolism. In addition, an agent such as midazolam, after initial high-frequency augmentation, augments slowing of EEG waveforms to lower frequencies in a dose-related fashion.27 A second example is the effect of opioids, which, in high doses, may produce sedation as a side effect and slow EEG waveform frequencies.27 Using multiple channels of a diagnostic EEG study and examining relative waveform frequencies distributed within a large volume of EEG data may not be optimal as a measure of sedation in critically ill adults. EEG-based technology such as BIS monitoring may be an option for evaluating deeper levels of sedation or levels of sedation and arousal in patients receiving neuromuscular blockers. The BIS has been evaluated in limited studies involving critically ill adults as effective for monitoring sedation.28Beginning stages of research that culminated in the BIS included gathering a significant database of EEG recordings and related clinical state. The respective hypnotic states were determined by using clinical assessment, responsiveness to stimuli, plasma concentrations of sedative/hypnotic agents, and memory testing.18–20 Research indicated that specific EEG features correlated with clinical end points after administration of sedative or anesthetic agents.18–20 Multivariate statistical analysis was performed and relevant EEG features were combined to produce the BIS, a linear scale (0–100) corresponding with hypnotic state18–20 (see TableT1).BIS determination begins with the EEG signal obtained from a set of electrodes placed on the forehead and one temple area as described in the case report. In the management of R.B., the BIS A-2000 monitoring system was used. In the most current version of BIS technology, the BIS A-2000 XP platform, the BIS sensor has 4 electrodes. The additional electrode facilitates management of EMG artifact. Determination of the BIS requires multiple steps, and proprietary algorithms are used for EEG signal processing. EMG analysis of muscle activity within the facial/forehead musculature is incorporated.First, the EEG signal is filtered and digitized within the amplifier head box near the patient’s head. The EEG signal is evaluated and filtered to eliminate artifacts such as ECG and pacemaker spikes in low-and high-frequency ranges. The filtered EEG signal is then divided into 2-second epochs or segments. Each epoch is grouped as suppressed or undulating (nonzero). Fast Fourier transformation is applied to the 2-second EEG epochs, and then power spectral and bispectral analysis processing techniques provide specific frequency domain parameters. Power spectral analysis determines the beta ratio component and is associated with clinical findings of light sedation. Bispectral analysis is used to determine the Synch-FastSlow component, which correlates with moderate sedation/light anesthesia.18–20 The near-suppression component (low-amplitude/low-frequency activity) correlates with very deep anesthesia.18–20 Last, the level of EEG suppression, associated with a very deep anesthetic state, is evaluated. A proprietary method is used to combine these EEG features; the result is a composite parameter, a single number termed the BIS. The BIS changes with increases in hypnotic state and decreases with decreases in the level of consciousness. Because the BIS incorporates information on multiple EEG features, it reflects changes in the level of consciousness.18–20EMG activity due to muscle activity within the facial muscles is a potential source of interference with BIS monitoring.29 This activity can easily be monitored as a secondary parameter on the BIS monitoring system. Because of its high frequency, EMG activity can cause an increase in the BIS reading.21,29,30 For example, in a patient receiving neuromuscular blockade, increased EMG activity may indicate a lightening of blockade independent of a change in hypnotic state. Or, breakthrough pain or emergence from deep sedation/analgesia may lead to movement of facial muscles and increased EMG activity. Shivering is another example of movement that can increase EMG activity. For these reasons, BIS interpretation must be done within the context of clinical examination and overall goals of therapy.31Because the BIS is based on EEG findings, it may also be sensitive to clinical states that affect level of consciousness independent of administration of sedative/hypnotic agents. Decreases in the BIS can be caused by sleep,28 neurological injury,32 and, potentially, hypothermia.33 In my clinical experience, in a patient receiving neuromuscular blockade, a significantly decreased BIS value was an initial indication of catastrophic cerebral injury. Neurological examination was compromised because the patient was chemically paralyzed. Cerebral injury was confirmed by the results of diagnostic EEG performed at the bedside.20Availability of EEGs and EEG-based parameters such as the BIS may facilitate adjustments in drug therapy and provide an indication of cerebral injury. For example, adjusting the dosage of sedation/analgesia on the basis of the BIS in patients with neuromuscular blockade may enable clinicians to decrease the duration of neuromuscular blockade. BIS monitoring of patients with neuromuscular blockade may lead to shorter lengths of stay and facilitate earlier recovery and extubation, as illustrated in the case study, because excessively deep sedation can be avoided during chemical paralysis.R.B. received aggressive and appropriate therapy to manage upper airway compromise due to severe epiglottitis and the resulting dyssynchrony with controlled ventilation. Appropriate to the clinical findings, his treatment included neuromuscular blockade. Break-through awareness and pain during neuromuscular blockade can and does occur. Adequately monitoring sedation and arousal states in patients with neuromuscular blockade is challenging. Using an EEG-based parameter to more directly assess the levels of sedation and arousal presented an opportunity to detect possible breakthrough awareness and pain. On the basis of the objective data, drug therapy for R.B. was adjusted in real time and withdrawal of neuromuscular blockade was facilitated; this led to earlier weaning and extubation. In the process, risks of both oversedation and undersedation were minimized. Further study is needed on ICU applications of BIS monitoring to determine additional impact on patients’ outcomes.Aspect, Bispectral Index, and BIS are trademarks of Aspect Medical Systems and are registered in the United States, the European Union, and other countries. The author is on the speaker’s bureau for Aspect Medical Systems. Aspect Medical Systems provided the BIS monitor used in this clinical application.