Nearly a quarter of patients who experience in-hospital cardiac arrest (IHCA) survive to discharge if CPR quickly results in return to spontaneous circulation, Get With The Guidelines—Resuscitation registry data show. But with each passing minute of ongoing CPR, both mortality risk and the odds of poor functional outcome increase.

Currently, “guidelines do not specify how long to continue CPR, [so] the decision is based on clinical judgement,” lead investigator Masashi Okubo, MD (University of Pittsburgh School of Medicine, PA), told TCTMD in an email. Multiple factors, including patient age, comorbidities, prior advance directives, and shockable versus nonshockable rhythm, can influence whether efforts persist even when the pulse hasn’t returned.

Although the new results give resuscitation teams, patients, and their surrogates some idea of what to expect, CPR timing should not be set in stone, Okubo stressed. “If clinicians believe prolonged CPR is beneficial, it is reasonable to provide prolonged CPR.”

For their study, which was published online recently in the BMJ, the researchers identified 348,996 US adults who received CPR for IHCA between 2000 and 2021. Two-thirds achieved spontaneous circulation, at a median of 7 minutes after the start of chest compressions. The remainder failed to achieve spontaneous circulation despite receiving a median of 20 minutes’ worth of CPR before resuscitation was terminated.

In all, 22.6% of patients survived to hospital discharge and, among those with available data, 15.6% had a good functional outcome (defined as a Cerebral Performance Category score of 1 or 2, indicating good performance or moderate disability).

If clinicians believe prolonged CPR is beneficial, it is reasonable to provide prolonged CPR. Masashi Okubo

Fully 99% of those who survived saw spontaneous circulation return within 44 minutes. Similarly, nearly all of those with favorable functional outcome had return of spontaneous circulation within 43 minutes.

At just 1 minute of CPR, 22.0% of patients survived and 15.1% had a favorable functional outcome. Yet their prognosis worsened as time passed—in an analysis designed to assume clinicians’ decisions to terminate resuscitation were 100% accurate, survival was less than 1% at 39 minutes of CPR and the rate of favorable functional outcome was less than 1% at 32 minutes of CPR.

When the researchers excluded patients who had termination of resuscitation before or at each time point from the denominator, “the probabilities of survival and favorable functional outcome plateaued above 5% and above 3%, respectively, after 20 minutes’ cardiopulmonary resuscitation,” the paper reports. Two factors could be at play, either that CPR duration itself informed the choice to stop resuscitation or there was confounding by indication.

In other words, their findings suggest that “patients for whom clinicians decided to provide longer CPR had much higher chance of survival/good functional recovery even if CPR duration was long,” Okubo explained.

Even when considering CPR time, the odds of survival and favorable function tended to be higher for patients younger than 60 years, those with witnessed arrest, and those with initial shockable rhythm.

The investigators say that altogether the data “imply that the decision to terminate resuscitation should not be solely dependent on duration of cardiopulmonary resuscitation . . . but should be based on clinical judgment of treating providers.”

Okubo acknowledged that no single study will be able to provide the answer on what the optimal CPR duration is for individual patients but said he’s “motivated to take steps” in the right direction.

First, there needs to be research into the characteristics of patients who receive prolonged CPR, Okubo suggested. Additionally, “a clinical trial comparing a strategy with CPR duration at provider’s discretion versus a strategy with prespecified CPR duration could provide more evidence about how long CPR should be done before termination.” The challenge with the latter approach is that the best CPR duration—one that both minimizes the risks of stopping early and maximizes the chance of positive outcomes—isn’t universal. Such a trial could only estimate the average treatment effect in the overall population.