Peri-procedural complications in women: an alarming and consistent trend

This editorial refers to ‘Sex-based differences in outcomes, 30-day readmissions, and costs following catheter ablation of atrial fibrillation: the United States Nationwide Readmissions Database 2010–14’^†, by J.W. Cheung et al., on page 3035.

Atrial fibrillation (AF) is the most common, sustained arrhythmia worldwide, with similar lifetime risks for men and women.¹ Since women have longer life expectancies, they account for over half of AF patients, particularly in older populations.² Research suggests that women with AF tend to be more symptomatic, less responsive to pharmacotherapy, and report a lower quality of life.³ It is therefore essential for the healthcare community not only to study AF therapies such as catheter ablation, but also to assess whether these treatments are equally safe and effective for both men and women.

In this issue of the European Heart Journal, Cheung et al. examine sex-based outcomes after AF ablation in a real-world cohort, including >50 000 patients in the US National Readmissions Database.⁴ Women accounted for a third of patients who underwent AF ablation, and were older with a higher overall comorbidity burden compared with men. The in-hospital complication rate was higher among women (8.3% vs. 5.6%), driven by increased bleeding, vascular complications, and cardiac perforations. In 30-day follow-up, the readmission rate was 4% higher among women (13.4% vs. 9.4%), driven by AF recurrence and heart failure.

This study is the most recent and largest of multiple studies looking at sex-based differences in outcomes of AF ablation. A consistent message emerges: complication rates are higher in women who undergo AF ablation compared with men. Kaiser et al., using a health insurance claims database, found increased rates of peri-procedural complications driven by perforation, bleed, and vascular complications, with a 5% absolute increase in repeat hospitalizations at 1 year.⁵ Zylla et al. evaluated the German AF ablation registry, with similarly increased complications in women, particularly vascular access complications.⁶

As clinicians and scientists, we are obligated to explain this disparity in order to find a remedy; the heart of the issue is methodology. One consideration is differential selection bias between men and women. If women are more likely to decline invasive procedures early on, the resulting ablation population may be sicker compared with men, leading to more adverse events. This explanation is simplistic and prematurely assigns the blame to women. Although bias is worthy of investigation, evaluation of the biological factors that influence risk is warranted.

Explaining sex-based differences is difficult, if not impossible, using administrative data, which lack detailed clinical information. Confounding plagues all observational analyses. Propensity score weighting accounts only for observed patient characteristics; unobserved imbalances remain. For example, women may have smaller left atria and with thinner walls, as well as smaller groin vasculature, both of which are harder to navigate with ablation catheters.^7,8 These anatomical differences, features which are absent from most analyses, could contribute to increased complication rates. In addition, frailty, social support, and socioeconomic status—unobserved confounders—often play a role in healthcare outcomes. Advances in healthcare informatics increasingly support large, granular data set curation across institutions. These advances should enable better patient phenotyping that will allow for improved observational analyses.

Another fundamental methodological issue is representation of women in the research and development patient populations. Women represent over half of the AF population, yet accounted for only a third of patients in landmark ablation trials.⁹ If women are more symptomatic with AF than men and experience more side-effects from anti-arrhythmic drugs, it follows that rhythm control through AF ablation would be both a reasonable and favourable option for women. Still, as seen in this study, women are less likely to undergo AF ablation. The ship has sailed on AF ablation technology, with new technologies being tested each year. Although the clinical, regulatory, and research communities have recognized that clinical studies, including cardiac device trials, have not traditionally enrolled a proportion of women that reflect the underlying disease distribution for men and women, enrollment barriers present ongoing challenges. The Food and Drug Administration has more recently encouraged companies to consider sex when designing a clinical trial, and when analysing the safety and effectiveness results in different subgroups.¹⁰ These efforts, if upheld by investigators and demanded by clinicians and patients, could lead to greater inclusion and a better understanding of sex-based disparities.

This important, large study from Cheung et al. reaffirms that the risk–benefit ratio for AF ablation is different for men and women. The higher rate of complications among women undergoing AF ablation is consistent and alarming. Sadly, this pattern is not confined to ablation. Women experience higher periprocedural complication rates in the setting of percutaneous coronary interventions (PCI) as well as pacemaker and defibrillator implantation.^11–15 Advances in data science may help us understand some of the contributing factors behind these observed differences in complication rates. However, that is hopefully just a first step. We can ‘adjust’ away the association between female sex and complications by accounting for left atrial structure or other features, but the fact remains that women experience higher complication rates. If we can delineate the factors that contribute to increased risk among women, then the procedures, catheters, and drugs can be designed and tested with these factors in mind. Millions of women undergo cardiac procedures. When the complications rates are 30–35% higher in women compared with men, we are dealing with a large number of suboptimally treated patients that deserve an equal standard of care.

Conflicts of interest: K.A.S. is an employee of the Veterans Health Administration. R.U.S. is supported by a Career Development Award from the NHLBI (K08 HL136850). L.W. has no conflicts to declare.

The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.