Use a light touch for ICU sedation

Take a stepwise approach to choose sedation and paralysis wisely.


Sedation in the ICU originated with operating room practices, but evidence from the past few decades supports a lighter approach.

At a session titled “Choosing Wisely in the ICU: Less Is More” at CHEST 2019, held in New Orleans in October, Vishesh Paul, MD, offered high-value care practices for sedation and paralysis in ICU patients.

Historically, use of sedation in ICU care followed the way anesthesia was used in the operating room, said Dr. Paul, who is a pulmonology and critical care subspecialist at Carle Foundation Hospital in Urbana, Ill. Older ventilators could not synchronize with patients' efforts to breathe, and patients stayed under deep sedation until they were ready for extubation. Today, ventilators are synchronized and newer, shorter-acting sedatives and analgesics are available. “Most importantly, there is recognition that oversedation really is associated with worse outcomes,” Dr. Paul said. “So things have changed.”

The tide began to turn in 1998, when a trial published in CHEST found that use of continuous IV sedation in the ICU was associated with prolonged mechanical ventilation and suggested that physicians look for strategies to end this practice. Subsequent trials over the next decade looked at various strategies to decrease continuous sedation and found improved outcomes with nursing-based light sedation protocols, daily sedation interruption, paired sedation interruption and spontaneous breathing trials, or sedatives on an as-needed basis, Dr. Paul noted.

“All of these sedation strategies do improve outcomes, and none of them were associated with any increase in adverse effects,” he said. Light sedation has been shown to lead to fewer days on mechanical ventilation, shorter ICU and hospital length of stay, and lower tracheostomy rates without increasing rates of delirium, post-traumatic stress disorder, or self-extubation. “It's a win-win, and any of the strategies can be used,” Dr. Paul said, depending on clinician preference and hospital policies.

While it might seem logical to combine sedation strategies, research to date has shown no benefit, he said. A randomized controlled trial published in JAMA in 2012 found no difference in outcomes between daily sedation interruption and protocolized sedation versus protocolized sedation alone in mechanically ventilated critically ill patients.

Even though evidence supports “light sedation,” the term was used loosely for a long time without a universally accepted definition, according to Dr. Paul. Today, he said, scales like RASS (Richmond Agitation and Sedation Scale) and SAS (Sedation-Agitation Scale) are universally accepted and used by guidelines and research studies to define depth of sedation.

On the RASS, which ranges from −5 (unarousable, no response to voice or physical stimulation) to 4 (combative, violent, immediate danger to staff), light sedation has been defined as a score of −2 to 1. The PADIS (Pain, Agitation/sedation, Delirium, Immobility and Sleep disruption) 2013 guidelines, published in Critical Care Medicine, recommended a RASS goal of −2 to 0 for sedation in the ICU.

However, the PADIS 2018 update, also published in Critical Care Medicine, suggested a goal RASS score of −1 to 0. “They described that −2 is probably too deep,” Dr. Paul said. “At a RASS of −2, patients cannot be awake enough to maintain eye contact for 10 seconds and cannot participate in therapy, and we know therapy does improve outcomes in the ICU.”

PADIS 2018 also emphasized the concept of analgosedation, or “analgesia first and sedation later” if needed, Dr. Paul said. He noted that analgosedation has been shown to improve outcomes as well as patient comfort. “Pain in ICU patients is often underrecognized and undertreated,” said Dr. Paul, “so that's what the emphasis is on: pain management first.”

Despite analgesia, many patients will still need sedation, and the choice of sedative matters, Dr. Paul said. Options include propofol, dexmedetomidine, and benzodiazepines (midazolam or lorazepam). However, studies over the past several years have found that benzodiazepines are associated with more days on mechanical ventilation and increased incidence of delirium versus the other commonly used drugs. “Dexmedetomidine as a sole agent is often insufficient, but again, that doesn't mean we need to use benzodiazepines,” Dr. Paul said. “There's data saying that benzodiazepines should be used last, unless there's a specific indication.”

It's best to manage ICU patients with a stepwise approach, Dr. Paul said. When a patient comes to the ICU in distress, the first thing to do is identify the cause, such as patient-ventilator dyssynchrony, hypoxia, or bladder distention, and treat that first. “Just don't sedate that patient, don't knock out that patient. If you can fix [the problem], it will lead to better outcomes,” he said. The next step is to address pain and add analgesics to the regimen. In addition, some evidence shows that using nonpharmacologic options such as verbal communication, family visits, or music can be helpful, Dr. Paul said. If sedation is still needed, proceed with sedative medicines (avoiding benzodiazepines, if possible), and pick any of the light sedation strategies.

Only a few patients require deep sedation, Dr. Paul stressed, such as those with severe acute respiratory distress syndrome (ARDS), refractory status epilepticus, or increased intracranial pressure, as well as those on neuromuscular blockers (NMBs). “For the rest, keep it light,” he said.

Paralysis using NMBs is not uncommon in the ICU, Dr. Paul noted. In addition to during short procedures like intubation, bronchoscopy, and endoscopy, paralysis is used in cases of severe hypoxemia (such as moderate to severe ARDS) and patient/ventilator dyssynchrony, for shivering due to therapeutic hypothermia, in patients with refractory status asthmaticus, and in patients with significantly increased intracranial pressure or intra-abdominal pressure.

Apart from immediate side effects, the most significant adverse effect of NMB use is ICU-associated weakness, which includes the spectrum of polyneuropathy, myopathy, and polyneuromyopathy, he said. While a direct causal effect has been questioned, recent evidence has suggested an association between use of NMBs and weakness in patients with sepsis or septic shock, renal or liver dysfunction, or concomitant steroid use, Dr. Paul said.

Paralysis in patients with ARDS has been an area of interest, Dr. Paul said. The ACURASYS trial, published in 2010 by the New England Journal of Medicine, found that patients with severe ARDS who received NMBs had decreased mortality rates. After that trial was published, more clinicians began using paralysis in these patients. “Use of paralysis went up quite a bit,” Dr. Paul said.

However, in a large multicenter trial published this year in the New England Journal of Medicine, NMB use was not associated with a significant difference in 90-day mortality versus usual care with light sedation in patients with moderate to severe ARDS. This called into question the practice of using paralysis for all cases of moderate to severe ARDS, Dr. Paul noted.

“So how can we choose paralysis wisely in the ICU?” he asked. Although some patients with severe ARDS will require use of paralytics to maintain patient-ventilator dyssynchrony and improve oxygenation, severe ARDS should not be an automatic indication for paralysis, Dr. Paul said. Patients with moderate to severe ARDS should not always be started on NMBs if oxygenation and ventilation can be managed without paralysis, he stressed. “One can still use it—we use it—but at least the use has gone down,” he said.

For patients with shivering due to therapeutic hypothermia, Dr. Paul recommended first trying other effective anti-shivering methods, such as the Columbia Anti-Shivering Protocol, before starting paralysis. If possible, try to avoid NMB use in patients getting corticosteroids, he said. And even when NMBs are used, Dr. Paul stressed, a majority of indications don't require complete paralysis. Most institutions use peripheral nerve stimulation testing with a train-of-four monitor, and scores of 0/4 are almost never needed, he said.

Finally, Dr. Paul reminded clinicians who prescribe paralysis that it is important to monitor and reassess patients daily. “Reassess,” he said, “and turn off the paralysis as soon as we can.”