番茄社区

For in-hospital cardiac arrest, the use of therapeutic hypothermia was associated with poorer survival rates and worse neurologic outcomes, Get With The Guidelines registry data showed.

In a propensity-score matched analysis, therapeutic hypothermia was associated with a relative 12% lower likelihood of in-hospital survival compared with usual care (27.4% versus 29.2%, P=0.01), , Saint Luke's Mid America Heart Institute in Kansas City, Mo., and colleagues found.

The likelihood of 1-year survival without severe neurological disability (Cerebral Performance Category score of 1 or 2) likewise favored usual care over the cooling intervention (20.5% versus 17.0%, P<0.001), the group reported in the Oct. 4 issue of the Journal of the American Medical Association.

The disadvantage to cooling was similar for patients with non-shockable (asystole and pulseless electrical activity) and shockable cardiac arrest rhythms (ventricular fibrillation and pulseless ventricular tachycardia) both for survival and neurologic outcomes (P=0.74 and P=0.88, respectively, for interaction).

"Current use of therapeutic hypothermia for in-hospital cardiac arrest may warrant reconsideration," Chan's group concluded.

"It's not so surprising. We do suspect that hypothermia works, but it's a different type of cardiac arrest," commented, co-director, University of Arizona Sarver Heart Center in Tucson.

Therapeutic hypothermia has generally shown advantages in out-of-hospital cardiac arrest, in which the cause of cardiac arrest is more typically cardiac and response times slower than in in-hospital cases.

"Maybe we should study this with a randomized trial, rather than just expect that outpatient findings will fit the inpatient setting, which has sort of happened," Kern concluded. No randomized trials of therapeutic cooling have been done for in-hospital arrest.

"In the end, probably what we're going to find is there'll be a role for it but it will be much more narrow patient selection," commented , chair of the American College of Emergency Physicians' public relations committee and medical director of the Lexington Fire/EMS in Lexington, Ky.

"We probably don't know that exact temperature yet, and our technology is limited in being able to control it," he noted in an interview with 番茄社区. "We've been so aggressive that we've gotten way ahead of ourselves."

The study used Get With the Guidelines-Resuscitation registry data linked to Medicare files for 26,183 patients with in-hospital cardiac arrest, among whom 1,568 (6.0%) were treated with therapeutic hypothermia.

The propensity-matched analysis included 1,524 of the hypothermia-treated patients and 3,714 of those treated without hypothermia (mean age 61.6 versus 62.2; 58.5% versus 57.1% male).

The researchers noted the possibility of residual confounding despite "a particular focus" on assessing indication bias.

A sensitivity analysis excluding patients who died within the first 24 hours after return of spontaneous circulation suggested "that either therapeutic hypothermia was protective during the first 24 hours after return of spontaneous circulation but was not effective overall, or that any indication bias was in favor of patients treated with hypothermia," they noted.

The database didn't have data on comatose status (although patients were required to be on mechanical ventilation at the time of, or after, cardiac arrest as a surrogate for comatose status) or detailed data on therapeutic hypothermia protocols and treatments for each patient, such that temperature data were not available for most.

"The possibility remains that the null findings for therapeutic hypothermia seen in this study may reflect poor implementation (e.g., insufficient duration of hypothermia), even though the median lowest achieved temperature was 33.1°C among patients treated with hypothermia for whom temperature data were available," Chan's group noted.

"These observational findings warrant a randomized clinical trial to assess efficacy of therapeutic hypothermia for in-hospital cardiac arrest," they concluded.

Comment