Revascularization strategies for multivessel coronary artery disease based on sex and age

Abstract

Open in new tabDownload slide

OBJECTIVES

This study describes coronary revascularization strategies used by sex and age in the USA.

METHODS

A sex-stratified cohort study from the National Inpatient Sample from the Agency for Healthcare Research and Quality (USA) including patients admitted for coronary revascularization with primary or secondary diagnoses of chronic coronary syndrome or non-ST elevation myocardial infarction who underwent ≥3-vessel coronary artery bypass grafting or percutaneous coronary intervention from January 2019 to December 2020. The primary outcome was the use rate of coronary artery bypass grafting or multivessel percutaneous coronary intervention. Prespecified subgroups included age and non-ST elevation myocardial infarction.

RESULTS

Among 121 150 patients (21.7% women), there were no sex differences in age (women: 66.6 [66.5–66.7], men: 67.6 [67.5–67.7], standardized mean difference: 0.1) or non-ST elevation myocardial infarction incidence (women: 37.4%, men: 45.7%, standardized mean difference: 0.17). The majority of women (74.2%) and men (84.9%) underwent bypass grafting, which was unaffected by age, race or non-ST elevation myocardial infarction. Women were less likely to undergo bypass grafting than percutaneous intervention (adjusted odds ratio 0.49, 95% confidence interval 0.44–0.54; P < 0.001) and a disparity most pronounced in patients >80 years old (adjusted odds ratio 0.31, 95% confidence interval 0.22–0.45; P < 0.001).

CONCLUSIONS

Most patients with multivessel coronary artery disease needing revascularization undergo bypass grafting, irrespective of sex, age or clinical presentation. The sex disparity in the use of bypass grafting is mostly seen among patients >80 years old.

INTRODUCTION

Multivessel coronary artery disease (CAD) is highly prevalent among CAD patients [1] and is associated with worse outcomes when compared with less extensive CAD [2–4]. Revascularization strategies for the management of multivessel CAD include surgical [coronary artery bypass grafting (CABG)] and interventional [multivessel percutaneous coronary intervention (PCI)] revascularization. Current clinical practice guidelines [5, 6] recommend both CABG and PCI for multivessel CAD; as surgery has better outcomes in patients with severe disease [7–14], recommendations shift towards CABG with increasing complexity of coronary anatomy. While previous studies have described the relative use of CABG and PCI in the general population [15–18], there is limited evidence on physician revascularization choices by age and sex. This is particularly important as women and older patients have specific preoperative risk profiles that may drive referral to one or another coronary revascularization modality.

In the present analysis, we utilize a national database with the aim of evaluating modalities of coronary revascularization for multivessel CAD by sex and age in the current US clinical practice.

METHODS

Ethics statement

The study was reviewed by the Weill Cornell Medicine Institutional Review Board (# 23-09026517-01) and it was determined that the proposed study was not research involving human subjects per to Code of Federal Regulations on the Protection of Human Subjects and the need for further review was waived and thus individual consent was not needed.

Study design and data collection

This retrospective cohort study follows the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines [19], and the checklist is provided in Supplementary Material, Table S1.

The data were obtained via the National Inpatient Sample (NIS) from the Agency for Healthcare Research and Quality, which is the largest publicly available all-payer inpatient database in the USA [20]. It is comprised of ∼13 million discharges from the index years 2019–2020, representing 20% of all inpatient stays in the USA. The NIS includes information on demographics, inpatient diagnoses and procedures and hospital characteristics and has been used in the past to investigate sex differences in CAD [16, 21]. The Agency for Healthcare Research and Quality provides sampling weights, which are used to obtain national estimates of the sample size [20].

Study population

The NIS was queried for all adult (≥18 years) hospital stays with a primary or secondary diagnosis of non-ST elevation myocardial infarction (NSTEMI) or chronic coronary syndrome (CCS), and who underwent multivessel PCI or CABG (3 or more coronary arteries) from January 2019 to December 2020. The International Classification of Diseases, 10th Edition codes that were utilized to identify patients can be found in Supplementary Material, Table S2. While the standard definition of multivessel CCS encompasses the presence of obstructive disease in 2 or more epicardial coronary arteries, we specifically included PCI codes that indicate 3 or more arteries, as including two-vessel PCI in the study would have introduced additional complexity in interpreting the results. Patients were included if they underwent multivessel (3 or more vessels) PCI or CABG. Patients were excluded if they had concomitant CABG and PCI or if they had ST-elevation myocardial infarction. If a hospital stay had CCS and NSTEMI co-diagnoses, these patients were assigned to the NSTEMI group. Patients who did not have NSTEMI or CCS as their primary or secondary diagnosis code were excluded.

Outcomes

The primary outcome was the use rate of CABG or PCI.

Statistical analyses

Baseline demographics were stratified by sex. Categorical variables were reported as frequency counts and percentages, while continuous variables were reported as means with standard deviations. Age, PCI and CABG volumes were all analysed as continuous variables.

We calculated the standardized mean differences for continuous and categorical variables using the following formulas [22], respectively:

where μ stands for mean, σ stands for standard deviation, n1 and n2 stand for sample sizes and P stands for proportion [23]. These standardized mean differences are identical to Cohen's size effect, therefore standardized mean differences (SMDs) of <0.20, 0.20, 0.50 and 0.80 are considered small, medium and large [24].

CABG and PCI use rates were compared between groups using the chi-squared test and univariable logistic regression. A multivariable logistic regression model was fit for PCI and CABG use and included preoperative patient factors such as comorbidity, demographics and socioeconomic status. Covariate selection was based on clinical relevance and inclusion in commonly used risk assessment scores, such as the Society of Thoracic Surgeons score. Given the clinical relevance of these factors, we deemed it important to adjust for them, irrespective of whether collinearity was present or not. Results were reported as odds ratios and adjusted odds ratios with 95% confidence intervals (CIs). The factors included in risk adjustment are presented in Supplementary Material, Table S3. Significance was set at P-value 0.05 without multiplicity adjustment.

Given the described relationship between mortality and institutional volume of PCI and CABG [25–27], as a sensitivity analysis, another multivariable logistic regression model was fit for PCI and CABG use that included all the factors in the first model and added institutional volume of PCI and CABG. An additional sensitivity analysis was stratified by race (white versus non-white patients). Finally, we provided E-values for adjusted odds ratios to assess the potential effect of unmeasured confounders [28]. The E-value quantifies the odds ratio that an unmeasured confounder would need to have to nullify the observed odds ratio. A higher E-value indicates a lower likelihood of the existence of such a confounding factor.

Pre-specified subgroup analyses included analysis by NSTEMI status (non-NSTEMI versus NSTEMI) and by age (4 age subgroups: ≤60, 61–70, 71–80 and >80 years).

Missing data accounted for <1.0% of the overall dataset; therefore, no imputation was performed. Stata 16 was used for all statistical analyses [29].

RESULTS

Study cohort and baseline characteristics

Between January 2019 and December 2020, 121 150 patients (26 280 [21.7%] women) were admitted with primary or secondary diagnoses of NSTEMI or CCS and underwent multivessel PCI or CABG. The baseline patient and hospital characteristics are summarized in Table 1. All differences between sexes in race, age (women: 66.6 [66.5–66.7], men: 67.6 [67.5–67.7], SMD: 0.1), incidence of NSTEMI (women: 37.4%, men: 45.7%, SMD: 0.17), household income status, hospital teaching status, admission at large hospital, Medicare/Medicaid insurance status, elective hospital admission, incidence of prior sternotomy, prior cerebrovascular accident, congestive heart failure, valvular disease, hypertension, stage III or IV chronic kidney disease, peripheral vascular disease, coagulopathy, malignancy, liver disease, diabetes, chronic pulmonary disease, anaemia, obesity and pre-existing arrhythmia were very small.

Coronary artery bypass grafting versus percutaneous coronary intervention

The vast majority of both women (74.2%) and men (84.9%) underwent CABG, while a minority of women (25.8%) and men (15.1%) underwent multivessel PCI. After adjustment for both patient and hospital factors, women were less likely than men to undergo CABG [adjusted odds ratio (aOR) 0.49, 95% CI 0.44–0.54; P < 0.001]. Among CABG patients, the on-pump technique was used in 70.45% in males and 61.6% in females (P < 0.001). Use rates of CABG and PCI are summarized in Table 2. In-hospital mortality rates per age group, CABG versus PCI and NSTEMI/CCS status are available in Supplementary Material, Table S4.

Subgroup analyses

Stratified by all age groups, the majority of the patients of both sexes underwent CABG rather than PCI (Fig. 1). In patients younger than 61 years old, 75.3% of women and 84.1% of men underwent CABG, in patients aged 61–70 years old, 79.0% of women and 88.2% of men underwent CABG, in patients aged 71–80 years, 77.1% of women and 85.9% of men underwent CABG, and in the oldest patients (age >80), 44.8% of women and 67.1% of men underwent CABG. In all age groups, women were less likely than men to undergo CABG, but this disparity was most pronounced among the oldest patients (>80 years, aOR 0.31, 95% CI 0.22–0.45; P < 0.001). These results were consistent in the CCS and NSTEMI cohorts (Figs 2 and 3). Results of coronary revascularization strategies by age and clinical presentation are summarized in Table 2.

Within each sex, patients aged 61–70 and 71–80 years were equally, if not more, likely than the youngest patients (age <60 years) to undergo CABG. In both sexes, however, the oldest patient cohort (age >80 years) was less likely to undergo CABG when compared with patients aged <60 years (men: aOR 0.35, 95% CI 0.30–0.41; P < 0.001, women: aOR 0.22, 95% CI 0.17–0.29, P < 0.001). This finding was consistent in the CCS and NSTEMI cohorts. Results of within-sex comparisons of CABG use rates are presented in Table 3.

Sensitivity analysis

The sensitivity analysis adjusted for institutional PCI and CABG volume was consistent with the main analysis (aOR for women to undergo CABG 0.49, 95% CI 0.44–0.54; P < 0.001). This finding was similar for patients with CCS (aOR 0.0.45, 95% CI 0.39–0.53; P < 0.001) and for those with NSTEMI (aOR 0.5, 95% CI 0.44–0.57; P < 0.001). The sensitivity analysis for race was consistent with the main analysis (aOR for white women to undergo CABG 0.47, 95% CI 0.42–0.53; P < 0.001, aOR for non-white women to undergo CABG 0.51, 95% CI 0.42–0.62; P < 0.001).

DISCUSSION

In this analysis of 121 150 patients (26 280 [21.7%] women) from a US national database, the majority of men and women with multivessel CAD who underwent revascularization received CABG rather than PCI. This finding was consistent at all ages including octogenarians, and among patients presenting with NSTEMI, and was consistent among white and non-white patients.

Overall, women received CABG less frequently than men, but the absolute difference was most evident in the subgroup of patients aged over 80 years. This sex disparity was not associated with large differences in demographics, socioeconomic factors, comorbidities, or hospital procedural volumes between the sexes, suggesting a possible systematic bias in the referral of women for surgical revascularization. CABG has been associated with worse outcomes in women compared with men, including shorter-term mortality and major adverse cardiac events [30–33], which may discourage healthcare providers from offering CABG to women. However, overall, the vast majority of women received CABG, demonstrating that CABG remains recognized as the standard of care for multivessel CAD in both sexes. It is also reassuring to note that in 2014, the sex disparity in CABG use among Medicare beneficiaries in the USA was more pronounced than in our analysis (127 per 100 00 person-years for women and 392 per 100 person-years for men) [33], and this may be the signal of a progressive reduction of the sex-related gap in referral to surgery.

The difference between sexes in CABG use rate was greatest in the oldest patients (80 years and older). It is important to note that there is evidence that the sex difference in adverse CABG outcomes disappears in older patients [30]. An individual patient data meta-analysis including 4 randomized trials and 13 193 patients (2714 women) found that overall, women had a significantly higher risk of major adverse cardiac and cerebrovascular events (adjusted hazard ratio 1.12, 95% CI 1.04–1.21; P = 0.004) [30], but age was a significant sex effect modifier (P for interaction <0.001), and the excess risk was not seen in women aged 75 and above (adjusted hazard ratio 1.29, 95% CI 1.17–1.43 for women <75 years and 0.94, 95% CI 0.83–1.05 for women 75 years and above) [30]. This finding may be related to the lower prevalence of non-atherosclerotic CCS in older women [34, 35] and highlights the need to investigate the factors that deny older women access to the potentially life-prolonging surgery.

Limitations

The substantially lower number of women versus men included in this analysis suggests referral bias that preceded the diagnosis of multivessel CAD. In addition, the NIS does not provide detailed coronary artery anatomical or echocardiographic information, and therefore data that may predispose referral to 1 revascularization approach (for instance, target vessel and conduit quality, ejection fraction) were not captured. The NIS also fails to capture data on patient frailty, which may have predisposed physicians to refer patients to a less invasive revascularization strategy. The NIS also does not capture patient preference, although, in the Synergy between PCI with Taxus and Cardiac Surgery (SYNTAX) trial [7], anatomical inappropriateness and patient refusals accounted for only 16% of patients unsuitable for CABG. National estimates reported in this study are based on the stratified sampling design of the NIS and are subject to the inherent uncertainty associated with sampling. The NIS data are collected from billing claims, another possible source of error or bias. Lastly, the NIS captures in-hospital admissions only and PCIs that were performed in the outpatient setting are not captured in this database.

CONCLUSIONS

In this nationwide cohort study of 121 150 patients and 26 280 women with multivessel CAD, both men and women were more likely to undergo CABG rather than PCI for revascularization. This finding did not vary by age, race and clinical presentation. A sex disparity in the utilization of CABG was mostly evident among patients 80 years and older.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Funding

No funding is reported for this article.

Conflict of interest: Mario Gaudino receives research grants from the National Institutes of Health, the Canadian Health and Research Institutes and the Starr Foundation. C. Noel Bairey Merz is supported by the National Institutes of Health R01HL124649, U54 AG065141, the Edythe L. Broad and the Constance Austin Women's Heart Research Fellowships, Cedars-Sinai Medical Center, Los Angeles, California, the Barbra Streisand Women's Cardiovascular Research and Education Program, Cedars-Sinai Medical Center, Los Angeles, The Society for Women's Health Research (SWHR), Washington, DC, the Linda Joy Pollin Women's Heart Health Program, the Erika Glazer Women's Heart Health Project and the Adelson Family Foundation, Cedars-Sinai Medical Center, Los Angeles, California. Lisa Q. Rong is supported by the National Institutes of Health K23 HL153836. Lamia Harik is supported by the National Heart, Lung, and Blood Institute T32HL160520-01A1. Stephen E. Fremes receives research grants from the National Institutes of Health, the Canadian Institutes of Health Research, Medtronic, Boston Scientific and Amgen. Ruth Masterson Creber receives research funding from the National Institutes of Health (R01HL161458, R01NS123639, R01HL152021) and PCORI. The other authors report no conflict of interest.

DATA AVAILABILITY

The data were obtained via the National Inpatient Sample (NIS) from the Agency for Healthcare Research and Quality (AHRQ), which is the largest publicly available all-payer inpatient database in the USA (https://hcup-us.ahrq.gov/db/nation/nis/nisdbdocumentation.jsp) [20]. The data underlying this article will shared on reasonable request to the corresponding author.

Author contributions

Samian Sulaiman: Data curation; Formal analysis; Methodology; Software; Writing—original draft; Writing—review & editing. Lamia Harik: Conceptualization; Data curation; Methodology; Visualization; Writing—original draft; Writing—review & editing. C. Noel Bairey Merz: Methodology; Supervision; Validation; Visualization; Writing—original draft; Writing—review & editing. Stephen E. Fremes: Methodology; Supervision; Validation; Writing—original draft; Writing—review & editing. Ruth Masterson Creber: Methodology; Validation; Visualization; Writing—original draft; Writing—review & editing. Lisa Q. Rong: Methodology; Supervision; Validation; Writing—review & editing. Mohamad Alkhouli: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Visualization; Writing—original draft; Writing—review & editing. Mario Gaudino: Conceptualization; Data curation; Formal analysis; Methodology; Supervision; Validation; Visualization; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Ari Mennander and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• aOR

  Adjusted odds ratio

 
• CABG

  Coronary artery bypass grafting

 
• CAD

  Coronary artery disease

 
• CCS

  Chronic coronary syndrome

 
• CI

  Confidence interval

 
• NIS

  National Inpatient Sample

 
• PCI

  Percutaneous coronary intervention

 
• SMDs

  Standardized mean differences

Author notes