Effects of COPD on in‐hospital outcomes of transcatheter aortic valve implantation: Results from the National Inpatient Sample database

Abstract Background Comorbid chronic obstructive pulmonary disease (COPD) increases morbidity and mortality among aortic valve replacement patients undergoing conventional surgery. The impact of COPD in patients undergoing less invasive transcatheter aortic valve insertion (TAVI) is unclear. Hypothesis This study evaluates the in‐hospital outcomes of TAVI in patients with and without COPD. Methods This population‐based, retrospective study of 8466 TAVI patients (29.87% with COPD) evaluates the effects of COPD on short‐term clinical outcomes (in‐hospital mortality, length of hospital stay, and postoperative complications) using data from the National Inpatient Sample database from 2011 to 2014. Logistic regression analysis was used to determine factors associated with in‐hospital mortality and postoperative complications. Linear regression analysis was used to identify factors associated with length of hospital stay. Results COPD is significantly associated with increased risk of respiratory complications and pneumonia after TAVI (aOR = 1.43, 95% CI: 1.24‐1.64; P < .001) but not in‐hospital mortality, length of hospital stay, or non‐respiratory postoperative complications as compared to non‐COPD patients. Concomitant COPD is significantly associated with increased risk of respiratory complications or pneumonia after TAVI but may still be the best treatment option for some patients. Conclusions Patients with comorbid COPD who receive TAVI have greater risk of developing postoperative respiratory complications and pneumonia. Vigilance for specific respiratory complications is highly warranted when treating this subgroup. Treatment decisions must be individualized.


| INTRODUCTION
Chronic obstructive pulmonary disease (COPD) describes a constellation of conditions, including emphysema, chronic bronchitis, and refractory asthma, characterized by progressive airflow limitation and reduced gas exchange that is not fully reversible. In developed countries, the prevalence of COPD is high, estimated to be 8% to 10% among adults 40 years of age and older. 1 COPD patients have a high burden of co-morbid chronic illness, suggesting that overarching systemic processes underlie these conditions. 2 Aortic stenosis (AS) is a common cardiovascular comorbidity among COPD patients, and respiratory complications may compromise postoperative morbidity and mortality in those who must undergo surgical treatment for AS. 3 In AS, the increased LV pressure required to maintain cardiac output leads to ventricular hypertrophy, resulting in systolic and diastolic dysfunction, 4 making valve replacement the only effective treatment. 5 Comorbid COPD is associated with morbidity and mortality after open-chest cardiac surgery. 6,7 Transcatheter aortic valve implantation (TAVI) became available as an alternative to conventional surgical aortic valve replacement (SAVR) in 2012 with the FDA approval of a device indicated for transfemoral or transapical delivery in patients with AS in whom open aortic valve replacement is precluded. 8,9 In a randomized, controlled trial confirming that clinical outcomes of TAVI are similar to those of SAVR, 10 clinicians are advised to decide between these procedures, taking into account the effects of COPD on outcomes.
In pre-surgical assessments of patients with severe aortic stenosis, the presence of COPD can contribute to a high preoperative risk score, potentially influencing the treatment choice toward TAVI rather than conventional surgery. 11 However, the risk to COPD patients undergoing TAVI is still uncertain. Compared to conventional SAVR, TAVI results in significantly fewer respiratory complications in COPD patients because the duration of mechanical ventilation is shorter. 12 However, another study showed that among patients with COPD, TAVI did not reduce the occurrence of postoperative pulmonary complications as compared to SAVR. 3 Furthermore, several studies have shown that COPD is associated with higher mortality rates among TAVI patients. 11,13,14 Given these conflicting results and serious implications regarding mortality, further investigation in a large patient cohort is warranted. Therefore, we conducted this study to determine the impact of COPD on short-term outcomes (in-hospital mortality, length of hospital stay, and postoperative complications) of patients receiving TAVI, using data from the National Inpatient Sample, the largest all-payer US inpatient care database.   Table S1.

| METHODS
Patient characteristics for covariate analysis included age, gender, race, income, insurance status, and transapical access (ICD-9 procedure code 35.06). Comorbidities were identified from AHRQ comorbidity measures in the database determined by ICD-9 diagnostic codes using algorithms validated by Elixhauser comorbidity scores. 18 Hospital-related characteristics (bed size, location/teaching status, and TAVI caseload) also were extracted from the NIS database.

| In-hospital mortality
Results of multivariate analysis showed that female sex, income level at the 51st to 75th percentile (vs 76th to 100th percentile), treatment in a hospital at the lowest percentile of annual TAVI caseload, coagulopathy, fluid/electrolyte disorders, paralysis, renal failure, and weight loss were associated with increased in-hospital mortality. Conversely, anemia, depression, and hypertension were associated with lower odds of patient in-hospital mortality (all P < .05) ( Table 3).

| Length of stay
Results of multivariate analysis showed that shorter hospital stays correlated with patient age ≥ 75 years, white race, income level at the 26th to 50th percentile (vs the highest percentile), treatment at hospitals with small bedsize, rural or urban non-teaching hospital, collagen-vascular disease, and hypertension. Conversely, longer hospital stays correlated with female sex, insurance status of self-pay/no-charge/other, congestive heart failure, coagulopathy, fluid/electrolyte disorders, paralysis, renal failure, weight loss, and atrial fibrillation (all P < .05) ( Table 3).

| Non-respiratory postoperative complications
Results of multivariate analysis showed increased odds of nonrespiratory postoperative complications associated with female sex, transapical access in TAVI, treatment in a hospital at the lowest percentile of annual TAVI caseload, and comorbidities, including coagulopathy, fluid/electrolyte disorders, paralysis, peripheral vascular disorders, renal failure, and weight loss. Hypertension and obesity were associated with lower odds of postoperative complications (all P < .05) ( Table 4).

| DISCUSSION
The present study evaluated the clinical characteristics and in-hospital outcomes of patients undergoing TAVI with and without COPD. Statistical analysis of data for this cohort of 8466 patients (29.87% with COPD) revealed that COPD is significantly associated with increased risk of respiratory complications and pneumonia after TAVI. However, no significant differences were observed regarding in-hospital mortality, length of hospital stay, or non-respiratory postoperative complications between COPD and non-COPD patients.
COPD is a common comorbidity among patients with AS and is a major predictor of adverse outcomes and greater mortality in patients undergoing open cardiac surgeries. 6 In one study, 28% to 43% of patients undergoing TAVI were reported to have COPD. 14  hospital stays and improved short-and long-term survival in selected COPD patients compared to the standard approach.
Studies that stratify patients by COPD severity, an option not possible using NIS data, report a higher risk of pulmonary complications and mortality after TAVI in patients with more severe COPD. 14 This finding may help to explain some of the differences in mortality and in-hospital outcomes reported between studies, as COPD cohorts may differ with respect to severity profile. For example, while one study reported no differences in 30-day mortality between TAVI patients with and without COPD, 19 other studies have reported higher post-TAVI mortality in patients with COPD than in those without COPD. 11,13 This may be associated with higher baseline risk profiles of TAVI patients that may mask the benefits portended by TAVI. 3 Thus, the complexity of COPD appears to confound direct comparison between study cohorts and underscores the importance of choosing treatments carefully on a case-by-case basis in this patient population.
An important strength of the present study is that the cohort includes the largest number of patients is such a study, selected from all geographical regions in the US and covering hospital admissions over a 4-year period. In addition, we analyzed non-respiratory and respiratory complications separately, an important distinction for COPD patients that has been determined less frequently in previous studies. This study also has several limitations. COPD and other comorbidities were defined based on the ICD-9 coding system, which does not include the severity of COPD and other comorbidities. Also, the NIS database 2011-2014 used for this study does not include current-generation TAVI valves, 22 which have improved over time. In addition, potential confounding variables such as types of anesthesia patients received and lifestyle factors were not collected by HCUP-NIS and therefore could not be included in the analyses of the present study. This study focused mainly on in-hospital outcomes. Late morbidity after discharge was not evaluated due to the nature of the database. Further well-designed studies that include analysis of COPD severity and late morbidities are highly warranted.
Concomitant COPD in TAVI patients is significantly associated with increased risk of respiratory complications and pneumonia but may still be the best treatment option for some patients. Nevertheless, findings of the present study suggest that vigilance for respiratory complications and pneumonia during postoperative care is still highly warranted while managing this subgroup of patients. Treatment decisions regarding aortic valve replacement for COPD patients are not straightforward and must be made with careful consideration of each patient.