Documented surgical ineligibility associated with increased long-term mortality in PCI

Documented surgical ineligibility on surgical records is common and was associated with a sixfold increase in risk for in-hospital mortality in a cohort of patients with unprotected left main or multivessel coronary disease undergoing PCI.

“Decisions to proceed with surgical vs. percutaneous revascularization for multivessel CAD are often based on subtle clinical information that may not be captured in contemporary registries,” study researcher Stephen W. Waldo, MD, from Massachusetts General Hospital, and colleagues wrote.

To evaluate the association between surgical ineligibility and long-term mortality, the researchers identified 1,013 patients undergoing non-emergent PCI for left main or multivessel CAD at two centers from 2009 to 2012. Electronic medical records were used to determine surgical ineligibility.

Twenty-two percent of patients were deemed ineligible for CABG. A poor surgical target, which was reported on medical records of 24% of patients, was the most commonly reported reason for surgical ineligibility, followed by advanced age (16%) and renal insufficiency (16%).

Adjusted analysis indicated that documentation of surgical ineligibility remained independently associated with elevated risk for in-hospital mortality (OR=6.26; 95% CI, 2.16-18.15) and long-term mortality (HR=2.98; 95% CI, 1.88-4.72) in the PCI setting, according to the results.

Kirk N. Garratt

In an accompanying editorial, Kirk N. Garratt, MD, of Northshore LIJ and the Lenox Hill Heart and Vascular Institute of New York, wrote that registry analyses have limitations, but the current findings may affect clinical practice.

“The authors concluded that CABG ineligibility occurs frequently enough to be important to practice statistics, and appears to impact outcomes powerfully, but they note that it has not been incorporated into current risk adjustment models,” Garratt wrote.

Garratt drew comparisons between the current study and the ASCERT registry.

“Have [the researchers] really found a poison dart that threatens the validity of the many registry analyses that came before them, including ASCERT? It’s hard to be sure,” he wrote.

The two registries could lead to the conclusion that one unmeasured confounding factor may play a critical role in assessing risk in this patient population, according to Garratt. He raised the question of whether other single confounders may have a similarly important effect and suggested that future registries should attempt to manage such confounding factors as surgical turn-down.

“Better integration between clinical databases (with deep clinical information, such as laboratory values and imaging study findings) and administrative databases (with broad patient information, such as comorbidities and outpatient medication use) would allow better characterization of individual patients without overwhelming any single registry,” Garratt wrote. “Voluntary unblinding of patient identifiers in research databases, now forbidden, should be considered as a means of improving the accuracy of pooled database studies that depend currently on probabilistic matching techniques.”

Garratt noted that the current study is timely because [the National Cardiovascular Data Registry] is assembling a team to update the CathPCI Registry. “This assembly will provide an opportunity to discuss changes in the structure and operation of the NCDR CathPCI Registry, as much as how to improve its data element list,” he wrote. “Hopefully, [the Society of Thoracic Surgeons] will follow suit and look to make important changes to its Adult Cardiac Surgery Database as well.”

For more information:

Garratt KN. Circulation. 2014;doi: 10.1161/CIRCULATIONAHA.114.013678.

Waldo SW. Circulation. 2014;doi: 10.1161/CIRCULATIONAHA.114.011541.

Disclosure: The researchers report no relevant financial disclosures. Garratt reports associations with Abbott Vascular, Boston Scientific, CeloNova, Daiichi Sankyo/Lilly, Infarct Reduction Technologies and The Medicines Company.