Rare and Highly Variable, Cath Lab Cardiac Arrests Need More Study, Guidance

Between hospitals, similar patients in the GWTG registry faced very different survival odds, but the reasons are unclear.

Rare and Highly Variable, Cath Lab Cardiac Arrests Need More Study, Guidance

The odds of surviving an in-hospital cardiac arrest that occurs in the cath lab vary significantly from one hospital to another, in some cases by more than 70% between patients with identical characteristics, registry data show.

“It’s eye-opening [and] it tells us we need to investigate, at the hospital level, why these variations are happening,” lead author Avnish Tripathi, MD, PHD, MPH (University of Kentucky College of Medicine, Bowling Green), told TCTMD.

While the reason behind the variation is unclear, it could hinge on the experience level of the cath lab staff in dealing with cardiac arrests. In the analysis of data from the Get With The Guidelines (GWTG)—Resuscitation registry, Tripathi and colleagues found that hospitals with higher volumes of in-lab cardiac arrests had better patient survival rates than those with low volumes.  

“That finding feeds into the logic or argument about whether we should try and centralize high-risk and specialty procedures [in] certain hospitals so that the volume of each hospital is high,” he added. “That way, the physician, and in a broader sense everyone who is part of the team, has a little bit more experience dealing with these situations.”

The variation will always be there . . . but I think our aim should be making that variation less. It should not be as drastic as it is. Avnish Tripathi

Compared with other sites for in-hospital arrests, only a minority (< 4%) occur in the cath lab. Just 1% of PCI and TAVI patients have a cardiac arrest, although short-term mortality for these patients is greater than 60%, notes Matthew I. Tomey, MD (Icahn School of Medicine at Mount Sinai, New York, NY), in an accompanying editorial.

Tomey, however, cautions that cath lab arrests are overlooked or systemically excluded from published evidence and guidelines and says that shortfall, alongside the findings of these registry data, represent a “call to action” to learn more.

Between-Hospital Variations

For the study, published online ahead of the December 26, 2022, issue of JACC: Cardiovascular Interventions, Tripathi and colleagues analyzed data on 4,787 patients (mean age 66 years; 37% female; 12% Black) who had an in-hospital cardiac arrest in one of 231 GWTG-Resuscitation registry hospital cath labs between 2003 and 2017 and received resuscitation attempts. Modeling was used to generate risk-adjusted survival rates for each hospital that reported at least five cardiac arrest cases, and each hospital was categorized into a low, middle, or high tertile for survival.

Yearly case volume—considered here as a proxy for experience—was calculated by dividing the total number of in-hospital cardiac arrests in the cath lab by years of GWTG-Resuscitation registry participation.

The median risk-adjusted survival to hospital discharge across all hospitals was 36%, but that figure fluctuated from a low of 20% at low-tertile hospitals to 36% at middle-tertile hospitals and 52% at high-tertile hospitals. In subanalyses, immediate post-cardiac arrest survival rates were 53%, 65%, and 76%, respectively, in the low, middle, and high tertiles (P < 0.001). The findings were similar for 24-hour survival at 37%, 53%, and 67%, respectively, across the tertiles (P < 0.001).

Compared with the middle- and high-tertile hospitals, those in the low tertile had greater representation of nonwhite patients, more patients with serious comorbidities, and a greater number of patients with an initial nonshockable rhythm. Patients with a nonshockable rhythm had the poorest risk-adjusted survival at a median of 28% across all hospitals. A modest correlation was seen between being in a high-tertile hospital and surviving to discharge in patients with either a shockable or no-shockable rhythm (P = 0.004).

The extent of hospital variation in the overall cohort was 1.71 (95% CI 1.52-1.87), implying that survival to discharge for patients with identical covariates varied by 71% between any two randomly-selected hospitals in the study. Demographic and clinical characteristics associated with lower rates of survival to discharge included older age, Black race, and high-risk comorbidities like acute stroke, depressed central nervous system, hypotension/hypoperfusion, MI, and respiratory insufficiency. Having an initial shockable rhythm was associated with a greater than threefold increased odds of survival to discharge.

Among hospital-level factors, no differences in survival were seen based on hospital size, rural or urban location, teaching status, or geographic location. Annual volume of in-hospital cardiac arrest was the only hospital-level factor associated with higher risk-adjusted survival (P < 0.001).

A Distinct Entity

To TCTMD, Tripathi agreed that cardiac arrests that occur in the cath lab often are excluded from studies of in-hospital cardiac arrest because they represent a different subset of patients, and their outcomes are harder to extrapolate to a generalized patient population. 

Tomey took this one step further by suggesting that cath lab cardiac arrest is “a distinct entity differing in substantive ways from out-of-hospital cardiac arrest and in-hospital cardiac arrest that warrants its own categorization, dedicated study, new evidence, and specific guidance.”

Tripathi added that at various levels, not only should there be a different category, but there should possibly be different guidelines as well.

“The variation will always be there . . . but I think our aim should be making that variation less. It should not be as drastic as it is,” he said.

One reason Tomey urges caution, however, is that the study did not convey information on type of procedure, who did the resuscitation, how it was done, and other factors. He also notes that if brief, aborted cardiac arrests were included, they could have skewed the survival patterns.

For those reasons, he urges caution in interpreting the magnitude of the variation between sites and how to tackle it, but says the study serves to “support a hypothesis that meaningful variation exists in survival from [cath lab cardiac arrests] across institutions beyond that which may be explained by clinical circumstances.”

In order to better study cath lab cardiac arrest, “there is a need for our reporting instruments and registries to capture variables particular to [this scenario] and the performance of resuscitation in the cardiac catheterization laboratory,” Tomey says.

“A necessary first step is the development of consensus data elements for supplemental reporting” in these cases, he continues. “Once defined, these elements will be vital to harmonize data collection efforts in the context of institutional quality improvement, observational studies, and pragmatic trials of resuscitation care in the cardiac catheterization laboratory.”

  • Tripathi and Tomey report no relevant conflicts of interest.