Published in the February 2013 issue of Today’s Hospitalist
LAST YEAR produced a bumper crop of new studies in critical care research. According to Michael A. Gropper, MD, PhD, vice chair of the department of anesthesia and perioperative care and director of critical care medicine at University of California, San Francisco (UCSF), that barrage of new evidence has physicians like him rethinking several areas of their practice.
At a UCSF conference last fall on managing hospitalized patients, Dr. Gropper spelled out what those new findings may mean for treating patients in the ICU. Among the research presented was the latest thinking on what fluids to use in resuscitation, as well as evidence on a sedative you should consider using “and one you should probably avoid.
Tight glycemic control: end of story?
When it comes to glycemic control in the ICU, the now legendary Van den Berghe study published in the Nov. 8, 2001, New England Journal of Medicine (NEJM) has cast a very long shadow. Researchers in that trial found what Dr. Gropper called “unbelievable” benefits to using intensive insulin therapy to maintain blood glucose levels between 80 and 110 mg/dL in surgical ICU patients.
But “the walls have been crumbling” on those results ever since, he said, particularly after the NICE-SUGAR trial results in 2009 linked tight glycemic control to higher mortality in the ICU. As a result, Dr. Gropper noted, “we relaxed our glycemic protocols to where we wouldn’t really treat anybody who was much below 180.”
The same NICE-SUGAR team followed up on its findings with a new study in the Sept. 20, 2012, issue of NEJM “to try to understand a little better why these patients died.” Researchers found a dose-response relationship between intensive control and either moderate (41-70 mg/dL) or severe (40 mg/dL or below) hypoglycemia. Researchers also found that the risk of death was three times higher for patients with severe hypoglycemia.
“We still don’t have a definitive answer,” Dr. Gropper said, “and we don’t know what our glucose targets should be.” And physicians continue to struggle with the issue because “recommendations from the Joint Commission and others continue to want us to do some glycemic control. But we really need to back off on glycemic control as more and more evidence comes out that it may be dangerous.”
New evidence on sedatives
Another new study casts a pall on etomidate, a popular sedative used for intubations, particularly in code situations.
“People like etomidate because it causes very little hypotension,” said Dr. Gropper. “The problem is that studies now show it to be dangerous.”
A meta-analysis published in the November 2012 issue of Critical Care Medicine looked at both randomized controlled trials and observational studies on the use of the drug in sepsis patients for rapid intubation. Researchers found a 20% increased risk of death associated with the drug across all studies; among randomized controlled trials, that risk rose to 26%. The drug was also linked to higher rates of adrenal insufficiency, and among patients with adrenal insufficiency, the analysis found mortality risk jumped as much as 35%.
“We’ve known for a long time that etomidate causes adrenal suppression, but this is the first pretty good evidence that it may be associated with mortality,” Dr. Gropper pointed out.
“We need to think about using other drugs like propofol at very low doses. Etomidate probably shouldn’t even be available anymore in the hospital, especially with an alternative like ketamine being a good option.”
Both midazolam and propofol are effective sedatives, he added, but midazolam has been associated with delirium. And while propofol works well and is titratable, said Dr. Gropper, “many patients on it will develop hypertriglyceridemia.”
Two noninferiority trials published in the March 21, 2012, issue of the Journal of the American Medical Association (JAMA) compared a newer agent “dexmedetomidine (Precedex) “to both midazolam and propofol. Those trials found dexmedetomidine to be similar to the other agents for duration of mechanical ventilation and ICU length of stay.
But patients who received dexmedetomidine were much more interactive and arousable than those on either of the other sedatives, and had a significantly greater ability to communicate. Dr. Gropper noted that he and other researchers are now starting a head-to-head trial between propofol and Precedex, with results due perhaps in 2014.
“Those of you who interact with your pharmacists know they get angry when you use Precedex because it’s fairly expensive,” he said. “But if you can get these patients extubated and out of bed sooner, it may be highly cost effective.”
Back off sedation interruptions?
Since key findings were published in a 2000 study, the gospel has been to interrupt patients’ sedation on a daily basis. Standard practice has been to stop patients’ sedatives, then start them up again (if patients need them) at half the rate, sparing patients excess sedation and unnecessary ventilator time.
“But a lot has changed since 2000,” said Dr. Gropper, “and we now do much more targeted sedation with sedation scores on every patient.” In an era of targeted sedation, a randomized controlled trial published in the Nov. 21, 2012, issue of JAMA debunked the benefits of daily sedation interruption.
Researchers compared protocolized sedation alone to protocolized sedation plus daily interruption. When they looked at a host of endpoints, including duration of ventilation, ICU length of stay and delirium, researchers “found no clear difference between these two groups,” Dr. Gropper said. But nurse workload was greater for patients having daily interruptions “and those patients actually had higher cumulative sedative dosages overall.
“It suggests that when you stop sedatives in patients receiving appropriate doses,” Dr. Gropper said, “you have to give them significantly more medication to get them back on target. It confirms that we need to use protocol-driven sedation with nurses doing regular sedation scoring, hourly ideally. But you don’t need to do daily sedation interruption.”
Fluids and resuscitation
As for what fluids physicians should use to resuscitate critically ill patients, Dr. Gropper highlighted several studies.
The SAFE trial, published in NEJM in 2004, compared albumin to saline resuscitation and found no difference in outcomes. “In our hospital, albumin is one of the most expensive products we use,” Dr. Gropper noted. “Why not use the salt water? It’s much cheaper.”
A 2008 NEJM study compared pentastarch (Pentaspan), a synthetic colloid, to modified Ringer’s lactate for resuscitation. “The patients who got Pentaspan had significantly lower survival than those who got Ringer’s lactate,” Dr. Gropper explained. “Those who got high-dose Pentaspan had a much higher mortality than those who got the lower dose.”
He admitted that many physicians may not be familiar with Pentaspan, which is used much more extensively in Europe than in the U.S. to resuscitate septic shock patients. But as an anesthesiologist, he said, “We use similar solutions quite a bit more in the operating room, and the options we have to rapidly expand blood volume are these synthetic colloids or albumin. For a long time, we stayed away from albumin because we were worried that it was a human blood product.”
But another new study is changing that practice. A randomized controlled trial published in the July 12, 2012, NEJM, compared another synthetic colloid “Tetraspan “to Ringer’s acetate for resuscitating patients with severe sepsis. It found “a marked difference in death and dialysis dependence between the two groups, “with mortality hitting 50% in the Tetraspan group vs. 43% in the Ringer’s acetate.
“While synthetic colloids are probably fine at low doses,” Dr. Gropper said, “we need to carefully consider whether the formulations used in the U.S. might carry some of the same risks.”
Another eye-opening study, this one published in the Oct. 17, 2012, JAMA, compared the standard use of resuscitation fluids containing liberal amounts of chloride “including saline and albumin “to a much more restrictive policy where physicians who wanted to use fluids with a lot of chloride for critically ill patients needed to specifically request them.
Researchers found a “statistically significant difference in the percentage of patients who developed end-stage renal disease or at least needed long-term dialysis “8% vs. 14% ” just from excluding chloride-containing solutions,” said Dr. Gropper. Those results have him thinking about making normal saline much less available.
“What we’ve already done in the operating room,” he said, “I think we need to do in the ICU: Use things like PlasmaLyte or Ringer’s lactate that have lower chloride concentrations.”
A big push for early mobility
Evidence continues to grow on the devastating and long-lasting effects that critical illness has on patients’ cognitive function. Adding to that literature, Dr. Gropper co-authored a review posted online in 2012 by Anesthesiology that looked at the weakness and neuropathy brought on by prolonged immobility in the ICU and the safety and benefits of early mobility.
At his hospital, he said, he and a group of physical therapists have championed early mobility for all patients in the ICU. “We’ve always done that for our cardiac surgery patients,” he explained, “but for multi-organ failure medical patients, we’ve basically said, ‘We can’t do that,’ and that’s unfortunate. If you put enough effort in, you’ll find something different.”
The biggest barrier he’s found to early mobility is the nursing staff, which views getting patients up and about earlier as more work. But Dr. Gropper said that given the benefits “less time in the ICU and on a ventilator, fewer pressure ulcers “”it’s ultimately less work, and the results are extraordinary.”
He has convinced his administration that “a physical therapist is as much a part of an ICU as the intensivist and the nurse. As a result, two of the four ICUs at UCSF now have dedicated physical therapists. ICU staff now also maintain a daily mobility score for each patient, just like a sedation score.
“If you have a dedicated therapist, the mobility score goes up over time,” Dr. Gropper noted. “If you don’t have one, it’s flat or it goes down. You need to have these people there to help the nurses.”
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
Who should be staffing your ICU?
FOR MICHAEL A GROPPER, MD, PHD, director of critical care medicine at University of California, San Francisco (UCSF), one study published last year offered some welcome guidance on how to staff ICUs.
“I have a group of almost 40 intensivists, and the one thing we live in fear of is a really well-done study that shows that we should all be working nights in the ICU,” he noted at a UCSF meeting last fall on managing hospitalized patients. Instead, a study published in the May 31, 2012, issue of the New England Journal of Medicine put his mind at ease.
To gauge the impact of intensivist staffing on patient outcomes, researchers conducted a retrospective analysis of APACHE data from 200910 on nearly 66,000 patients. (They also did some follow-up surveys.) They found that in ICUs without intensivists, adding intensivist staffing at night “lowered mortality by 17%,”Dr. Gropper said.
However, in ICUs run by intensivists during the day or units in which intensivist consults are mandatory during the day, adding intensivist coverage at night didn’t bring any significant benefits, at least in terms of patient mortality. The researchers referred to such ICUs as “high-intensity.”
Because Dr. Gropper and his intensivist colleagues fall into a high-intensity group, the results came as good news. “I’ll sleep better for right now,” he said.
New thinking on mechanical ventilation
WHAT MICHAEL A GROPPER, MD, PHD, director of critical care medicine at University of California, San Francisco (UCSF), called “the single most powerful critical care trial” was published in the May 4, 2000, New England Journal of Medicine. That randomized controlled trial, which was stopped early, found lower mortality and fewer ventilator days in patients with acute lung injury or acute respiratory distress syndrome (ARDS) being ventilated with a lower tidal volume (6 ml/kg of predicted body weight) than with a traditional total volume of 12 ml/kg.
A study in the May 2012 issue of Anesthesiology returned to the issue of lung injury with higher tidal volumes. Looking at tidal volumes used for cardiac surgery patients, researchers found that “when people’s tidal volume was set by predicted body weight, they actually were getting relatively high tidal volumes,” Dr. Gropper explained. “For patients with a BMI of more than 30, the odds ratio of having high tidal volume was six, and it was four for females.”
The higher the tidal volume, he added, the more likely patients were to have multisystem organ failure, not just respiratory failure. Those results “point out the importance of the rigor with which we determine the tidal volume, which should be done purely by height to accurately determine predicted body weight.”
But Dr. Gropper added that he is constantly being asked these questions at medical meetings: What about low tidal volume in patients who don’t have acute lung injury or ARDS? Shouldn’t physicians just opt for low tidal volume all the time?
While he admitted that he often takes just such an approach, he said he’s always been quick to add the following caveat: “We don’t really know what’s right because it hasn’t been well-studied.”
But a meta-analysis of 20 articles covering close to 3,000 patients that was published in the Oct. 23/31, 2012, JAMA, seems to validate that choice. Looking at studies that enrolled patients without ARDS, researchers compared what they called “protective ventilation” (a mean tidal volume of 6.45 ml/kg of ideal body weight) to conventional ventilation, with a mean tidal volume of 10.60 ml/kg. They found that protective ventilation delivered statistically significant reductions for just about every endpoint: the rate of lung injury, mortality and infection, as well as atalectesis, which Dr. Gropper said was “probably the most surprising.”
“This suggests that our standard protocol for tidal volume should be 6 ml/kg for ideal body weight in all patients,” he added, “and maybe that makes our life easier.”