aCardiocine, MiamHospital, Sof MedicinSinai St Ludivision, SfCardiovasMaryland;Michigan;Tropical MIcahn Sch

0002-9149http://dx.do

Comparison of Outcomes of Balloon Aortic ValvuloplastyPlus Percutaneous Coronary Intervention Versus

Percutaneous Aortic Balloon Valvuloplasty Alone During the SameHospitalization in the United States

Vikas Singh, MDa, Nileshkumar J. Patel, MDb, Apurva O. Badheka, MDc,*, Shilpkumar Arora, MDd,Nilay Patel, MDe, Conrad Macon, MDa, Ghanshyambhai T. Savani, MDf, Sohilkumar Manvar, MDg,

Jay Patel, MDg, Badal Thakkar, MDh, Vinaykumar Panchal, MDh, Shantanu Solanki, MDi,Neil Patel, MDi, Ankit Chothani, MDj, Sidakpal S. Panaich, MDg, Vinny Ram, MDk,

Chad A. Kliger, MDl, Theodore Schreiber, MDg, William O’ Neill, MDm, Mauricio G. Cohen, MDa,Carlos E. Alfonso, MDa, Cindy L. Grines, MDg, Abeel Mangi, MDc, Steven Pfau, MDc,

John K. Forrest, MDc, Michael Cleman, MDc, and Raj Makkar, MDn

The use of percutaneous aortic balloon balvotomy (PABV) in high surgical risk patients has

vasculai, Flortaten Ise, Newke’s Roaint Pecular dgCardihCardioedicin

ool of

/14/$ -i.org/1

resurged because of development of less invasive endovascular therapies. We comparedoutcomes of concomitant PABV and percutaneous coronary intervention (PCI) with PABValone during same hospitalization using nation’s largest hospitalization database. Weidentified patients and determined time trends using the International Classification ofDiseases, Ninth Revision, Clinical Modification, procedure code for valvulotomy fromNationwide Inpatient Sample database 1998 to 2010. Only patients >60 years with aorticstenosis were included. Primary outcome included in-hospital mortality, and secondaryoutcomes included procedural complications, length of stay (LOS), and cost of hospitali-zation. Total 2,127 PABV procedures were identified, with 247 in PABV D PCI group and1,880 in the PABV group. Utilization rate of concomitant PABV D PCI during samehospitalization increased by 225% from 5.1% in 1998 to 1999 to 16.6% in 2009 to 2010(p <0.001). Overall in-hospital mortality rate and complication rates in PABVD PCI groupwere similar to that of PABV group (10.3% vs 10.5% and 23.4% vs 24.7%, respectively).PABV D PCI group had similar LOS but higher hospitalization cost (median [interquartilerange] $30,089 [$21,925 to $48,267] versus $18,421 [$11,482 to $32,215], p <0.001) incomparison with the PABV group. Unstable condition, occurrence of any complication, andweekend admission were the main predictors of increased LOS and cost of hospitaladmission. Concomitant PCI and PABV during the same hospitalization are not associatedwith change in in-hospital mortality, complications rate, or LOS compared with PABValone; however, it increases the cost of hospitalization. � 2014 Elsevier Inc. All rightsreserved. (Am J Cardiol 2014;-:-e-)

Coronary artery disease (CAD) and aortic stenosis (AS)share similar atherosclerosis risk factors and, therefore,coexist in up to 50% of the patients who present with symp-tomatic AS.1,2 The traditional management option of surgicalaortic valve replacement with coronary artery bypass graft isnot feasible in up to 60% of the patients with severe

r division, University of Miami Miller School of Medi-ida; bCardiovascular division, Staten Island Universityland, New York; cCardiovascular division, Yale SchoolHaven, Connecticut; dCardiovascular division, Mountosevelt Hospital, New York, New York; eCardiovascularter’s University Hospital, New Brunswick, New Jersey;ivision, Prince George’s Hospital Center, Cheverly,ovascular division, Detroit Medical Center, Detroit,vascular division, Tulane School of Public Health ande, New Orleans, Louisiana; iCardiovascular division,Medicine at Mount Sinai, New York, New York;

see front matter � 2014 Elsevier Inc. All rights reserved.0.1016/j.amjcard.2014.11.032

symptomatic AS because of high operative risk owing to theirmultiple co-morbidities.3 With the development of trans-catheter aortic valve replacement (TAVR), studies have notedpercutaneous aortic balloon balvotomy (PABV) as a bridge tovalve replacement in up to 37% of the patients who wereinitially considered inoperable.4,5 Percutaneous management

jCardiovascular division, MedStar Washington Hospital Center, Washing-ton, District of Columbia; kCardiovascular division, University of Arizona,Tucson, Arizona; lCardiovascular division, Lenox Hill Heart and VascularInstitute-North Shore LIJ Health System, New York, New York; mCar-diovascular division, Henry Ford Hospital, Detroit, Michigan; and nCar-diovascular division, Cedars Sinai Medical Center, Los Angeles, California.Manuscript received October 30, 2014; revised manuscript received andaccepted November 20, 2014.

See page 6 for disclosure information.*Corresponding author: Tel: (408) 324-4516; fax: (203) 737-2437.E-mail address: apurva.badheka@yale.edu (A.O. Badheka).

www.ajconline.org

Table 1Baseline characteristics of the study population, according to undergoing concomitant percutaneous aortic balloon valvotomy (PABV) and percutaneouscoronary intervention (PCI) during the same hospitalization versus PABV alone, in the United States over the study period of 13 years from 1998 e 2010

PABV Only PABV þ PCI P-Value

Demographic variableTotal no. of procedures (Unweighted no.) 1880 247Total no. of procedures (Weighted no) 9411 1230Patient level variablesAge (years) 0.9Median 83 (78-88) 84 (78-88)65-79 28.9% 26.4%�80 68.6% 69.4%Gender <0.001Male 45.5% 50.8%Female 54.5% 49.2%Race* 0.5White 66.9% 65.9%Non-white 8.1% 9.1%Missing 25.0% 25.0%Co-morbidities†

Charlson/Deyo co-morbidity indexz <0.0010 12.3% 7.4%1 28.5% 22.0%�2 59.2% 70.6%

Obesityx 4.0% 3.6% 0.1History of hypertension 44.2% 53.5% 0.02History of diabetes 20.8% 31.5% <0.001History of congestive heart failure 6.6% 2.8% <0.001History of chronic pulmonary disease 20.7% 23.1% 0.7Peripheral vascular disease 11.1% 20.1% <0.001Renal failure 29.3% 35.4% 0.1Neurological disorder or paralysis 3.7% 1.5% <0.001Anemia or coagulopathy 18.2% 25.8% <0.001Hematological or oncological malignancy 6.4% 3.9% <0.001Weight loss 3.0% 2.9% 0.4Rheumatoid arthritis or other collagen vascular Disease 2.0% 3.4% 0.02Admission types <0.001Emergent/Urgent 53.9% 55.9%Elective admission 35.0% 25.1%Unstable Patient{ 7% 12.4% 0.003Admission day 0.03Weekdays 88.7% 86.6%Weekend 11.3% 13.4%Length of stay [Median (interquartile range)] 5 (2-12) 7 (3-14) 0.06Mortality, cost and outcomeTotal charges ($)[Median (interquartile range)] 18,421 (11,482-32,215) 30,089 (21,925-48,267) <0.01Disposition 0.003Home 63.3% 59.7%Facility/others 26.1% 30.1%Death 10.5% 10.3% 0.8

* Race was missing in 25% of the study population and hence excluded in the multivariable analysis.† Variables are AHRQ co-morbidity measures, which were only available from 2002 through 2010.z Charlson/Deyo co-morbidity index was calculated as per Deyo classification.x Obesity was defined as (Body Mass Index >30).{ Unstable was defined as (DX1-DX25) having a listed code for shock (ICD-9-CM code 785.5) or ventilator dependence (ICD-9-CM code V461).

2 The American Journal of Cardiology (www.ajconline.org)

of CAD and AS with percutaneous coronary intervention(PCI) and PABV is now increasingly performed in these high-risk patients as a bridge to TAVR. Currently, there is paucityof data regarding feasibility and safety of combined PABVand PCI with previous studies being limited by small samplesize. The aim of this study was to determine in-hospital out-comes of concomitant PABV plus PCI and compare them

with PABV alone using the nation’s largest hospitalizationdatabase.

Methods

Data were obtained from the Nationwide InpatientSample (NIS), which is the largest available database of

Figure 1. Trend of concomitant PCI and PABV during same hospitalizationin United States from 1998 to 2010.

Valvular Heart Disease/Comparison of Outcomes of PABV þ PCI vs PABV Alone 3

hospital inpatient stays in the United States. The 2010 NIScontains all discharge data from 1,051 hospitals located in45 states, approximating a 20% stratified sample of UScommunity hospitals. Data from the NIS have previouslybeen used to identify, track, and analyze national trends inhealth care usage, patterns of major procedures, access,disparity of care, trends in hospitalizations, charges, quality,and outcomes.6,7 Each individual hospitalization is de-identified and maintained in the NIS as a unique entrywith 1 primary discharge diagnosis and <24 secondary di-agnoses during that hospitalization. Each entry also carriesinformation on demographic details, insurance status, co-morbidities, primary and secondary procedures, hospitali-zation outcome, and length of stay (LOS) with safeguards toprotect the privacy of individual patients, physicians, andhospitals. Annual data quality assessments are performed toassure the internal validity of the database. To maintain theexternal validity, database is compared with the followingdata sources: the American Hospital Association AnnualSurvey Database, the National Hospital Discharge Surveyfrom the National Center for Health Statistics, and theMedPAR inpatient data from the Centers for Medicare andMedicaid Services.8

We queried the Health Care Cost and Utilization Pro-ject’s NIS between 1998 and 2010 using the InternationalClassification of Diseases, Ninth Revision, Clinical Modifi-cation (ICD-9-CM) procedure codes of 35.96 for percuta-neous valvuloplasty, 36.06 for nonedrug-eluting coronaryartery stents, 36.07 for drug-eluting coronary artery stents,and 00.4� in any of the procedural fields. Only patients>60 years with AS (424.1, 395.0, 395.2, 396.2, and 746.3)were included. Patients with concomitant mitral, tricuspid,or pulmonary stenosis were excluded (394.0, 394.2, 396.0,396.1, 396.8, 397.0, 397.1, 746.0, 746.1, 746.5, 424.2, and424.3).

We defined severity of co-morbid conditions using Deyomodification of Charlson comorbidity index (CCI).9 Thisindex contains 17 co-morbid conditions with differentialweights. The score ranges from 0 to 33, with higher scorescorresponding to greater burden of co-morbid diseases(Supplementary Table 1).

Procedural complications were identified by Patient SafetyIndicators (PSIs), version 4.4, March 2012, which have beenestablished by the Agency for Healthcare Research andQuality to monitor preventable adverse events during hospi-talization. These indicators are based on ICD-9-CM codes andMedicare severity Diagnosis-Related Groups, and each PSIhas specific inclusion and exclusion criteria.9,10 Proceduralcomplications not included in PSI were identified using ICD-9-CM codes (Supplementary Table 2). To prevent classifi-cation of a pre-existing condition (e.g., stroke or heart block)as a complication, cases with the ICD-9-CM code for acomplication listed as the principal diagnosis were excluded.Unstable patients were defined as those having a listed codefor shock (ICD-9-CM code 785.5) or ventilator dependence(ICD-9-CM code V461). This methodology of identifyingpatients who underwent procedures, co-morbid conditions,and associated complications has previously been used inseveral studies.11,12

Stata IC 11.0 (Stata-Corp, College Station, Texas) andSAS 9.2 (SAS Institute Inc, Cary, North Carolina) was used

for analyses, which accounted for the complex surveydesign and clustering. All analyses were performed usinghospital-level discharge weights provided by the NIS tominimize biases. Differences between categorical variableswere tested using the chi-square test, and differences be-tween continuous variables were tested using Student’st test, continuous variables with Gaussian distributions, andKruskal-Wallis rank-sum tests for continuous variables withnon-Gaussian distributions. p-Value less than 0.05 wasconsidered significant.

Hierarchical mixed-effects models were generated toidentify the independent multivariate predictors of in-hospital mortality, postprocedural complications, LOS, andcost of hospitalization. Two-level hierarchical models (withpatient-level factors nested within hospital-level factors)were created with the unique hospital identification numberincorporated as random effects within the model. Hierar-chical mixed-effects logistic regression models were usedfor categorical dependent variables like in-hospital mortalityand postprocedural complications, and hierarchical mixed-effects linear regression models were used for continuousdependent variables like LOS. LOS and cost of care werenot normally distributed in the population and were, there-fore, converted into logarithmic scale.

Variables with >10% missing data (race and type ofadmission) were not included in the multivariate model.Because majority of the procedures were performed in urbanhospitals, we did not include rural or urban location ofhospital in the model. In all multivariate models, weincluded hospital-level variables, such as hospital region,teaching versus nonteaching hospital, admission over theweekend, and patient-level variables like age, gender, CCI,unstable patient, and primary payer (with Medicare orMedicaid considered as referent). All interactions werethoroughly tested. Collinearity was assessed using varianceinflation factor.

Results

A total of 2,127 (weighted n ¼ 10,641) PABV pro-cedures were identified, of which 247 (weighted n ¼ 1,230)were PABV þ PCI (PABV þ PCI group) and 1,880(weighted n ¼ 9,411) were PABVs alone (PABV group).Table 1 demonstrates the baseline characteristics of the

Figure 2. Individual complication rates in PABV þ PCI versus PABV groups. CABG ¼ coronary artery bypass grafting; TIA ¼ transient ischemic attack.

Table 2Adverse clinical events* related to whether PCI performed in addition to PABV during the same hospitalization or not

Complications PABV þ PCI PABV only P Value

Mortality 10.3% 10.5% 0.9Any complications 23.4% 24.7% 0.8Mortality D any complications 28.7% 29.5% 0.9Death D Vascular D Stroke D Cardiac D Requirement of Open heart surgery 24.1% 24.3% 0.1Vascular complications 7.5% 6.7% 0.61) Postop-hemorrhage requiring transfusion† 2.5% 2.8% 0.82) Vascular injury 5.1% 4.3% 0.5Cardiac complications 6.7% 7.4% 0.81) Iatrogenic cardiac complications including periprocedural myocardial infarction 4.6% 4.5% 0.82) Pericardial complications 0.4% 0.8% 0.53) Complete Heart Block 1.6% 3.2% 0.2Requiring CABG 1.5% 2.5% 0.4Respiratory complications 8.3% 7.6% 0.6Postop-Stroke/TIA/Stroke effects 1.7% 3.2% 0.2Renal and metabolic complications 0.8% 1.9% 0.21) Acute renal failure requiring dialysis 0.8% 1.8% 0.22) Acute severe metabolic derangement 0% 0.1% 0.7

CABG ¼ coronary artery bypass grafting; PE ¼ pulmonary embolism; TIA ¼ transient ischemic attack.* Details in supplementary Table 2.† Hemorrhage requiring transfusion was identified as having any patient having post-operative hemorrhage and also got transfusion.

4 The American Journal of Cardiology (www.ajconline.org)

study population; 69% of the patients were >80 years inboth groups. At baseline, the patients in the PABV þ PCIgroup were noted to have significantly higher incidence ofhypertension, diabetes, peripheral vascular disease, anemia,or coagulopathy compared with the PABV group. ThePABV þ PCI group also had significantly more emergentadmissions and patients with CCI �2. Hemodynamic sup-port with devices, such as intra-aortic balloon pump, Impellaor TandemHeart, was used in 95 procedures: 32 (12.7%) inthe PABV þ PCI group and 63 (3.3%) in the PABV group.The utilization rate of concomitant PABV þ PCI during thesame hospitalization increased by 225% from 5.1% ofPABVs in 1998 to 1999 to 16.6% of PABVs in 2009 to2010 in the United States (p <0.001) (Figure 1).

Overall in-hospital mortality rate in the PABV þ PCIgroup was not statistically different from that of the PABVgroup (10.3% vs 10.5%). The significant predictors of in-hospital mortality in the PABV þ PCI group wereincreasing co-morbidities (odds ratio [OR] 1.17; 95% confi-dence interval [CI] 1.01 to 1.35, p ¼ 0.03), unstable patient(OR6.19; 95%CI 3.45 to 11.09, p<0.001), occurrence of anycomplication (OR 2.81; 95% CI 1.85 to 4.28, p<0.001), andadmission over the weekend (OR 2.12; 95% CI 1.23 to 3.67,p ¼ 0.007; Supplementary Table 3).

The overall complications rates were also similar inPABV þ PCI group versus PABV group (23.4% vs 24.7%,p ¼ 0.80). Figure 2 demonstrates the comparison of variouscomplications in the 2 groups. Most common complications

Table 4Multivariate predictors of increased length of stay and cost of hospitalization for PABV þ PCI during same hospitalization

PredictorsLength of Stay Cost of Hospitalization

OR (95% CI) P value OR (95% CI) P value

Presence of any complications 1.82 (1.59-2.10) <0.001 1.74(1.54 e 1.99) <0.001Unstable patient* 1.57 (1.23-2.00) <0.001 1.85 (1.46 -2.34) <0.001Charlson score† 1.11(1.07 e 1.16) <0.001 1.07(1.03 e 1.11) 0.001Weekend vs Weekdays admission 1.62 (1.33 e 1.97) <0.001 1.40 (1.18 e 1.66) <0.001Hospital procedure volume (per 10

procedures increase per year)0.95 (0.90 e 0.99) 0.045 0.99 (0.94 e 1.05) 0.9

Two levels hierarchical mixed effects models were generated (patient level factors nested within institutional level factors) with the unique hospitalidentification number incorporated as random effects. Models were adjusted for age, sex, Deyo’s modification of Charlson Comorbidity Index, unstablepatients, occurrence of any post-procedural complication, primary payer, weekend admission, hospital region, hospital teaching status and hospital volume.CI ¼ confidence interval; OR ¼ odds ratio.* Unstable patients were defined as those having a listed code for shock (ICD-9-CM code 785.5) or ventilator dependence (ICD-9-CM code V461).† Charlson’s/Deyo comorbidity index was calculated as per Deyo classification.

Table 3Multivariate adjusted Odds ratios of end-points of PABV plus PCI during same hospitalization when compared with PABV alone

Outcomes OR (95% CI) P value

Mortality 0.82 (0.51-1.32) 0.4Any complications 0.83 (0.60-1.15) 0.3Mortality þ any complications 0.83 (0.60-1.14) 0.2Death þ Vascular þ stroke þ Cardiac þ requirement of Open heart surgery 0.87 (0.63-1.20) 0.4Length of Stay 1.23 (0.99-1.51) 0.05Cost of care 1.54 (1.28-1.85) <0.001

Two levels hierarchical mixed effects models were generated (patient level factors nested within institutional level factors) with the unique hospitalidentification number incorporated as random effects. Models were adjusted for age, sex, Deyo’s modification of Charlson Comorbidity Index, unstablepatients, occurrence of any post-procedural complication, primary payer, weekend admission, hospital region, hospital teaching status and hospital volume.

Valvular Heart Disease/Comparison of Outcomes of PABV þ PCI vs PABV Alone 5

in the PABV þ PCI group were postprocedure respiratoryfailure (8.3%), vascular (7.5%), and cardiac (6.7%).Unstable patients were a significant predictor of highercomplications (OR 5.09; 95% CI 2.97 to 8.70, p <0.001;Supplementary Table 3).

Overall, there were no significant differences in clinicallyrelevant end points when PABV þ PCI group wascompared with the PABV group (Tables 2 and 3). However,patients in the PABV þ PCI group had significantlyincreased cost of hospitalization (median [interquartilerange] $30,089 [$21,925 to $48,267] versus $18,421[$11,482 to $32,215], p <0.001) compared with those of thePABV group (Table 3). The significant predictors ofincreased LOS and cost of hospital admission were unstablecondition, occurrence of any complication, weekendadmission, and increasing co-morbidities (Table 4).

Discussion

In summary, our data suggest a significant increase in thenumber of concomitant PABV þ PCI procedures beingperformed in the United States over the last decade. Wedemonstrate that performing concomitant PCI (if deemednecessary) in addition to PABV during the same hospitali-zation does not affect the rate of in-hospital mortality,complications, or length of hospital stay; however, it is

associated with significantly increased cost of hospitaliza-tion compared with performing PABV alone.

With the development of novel endovascular therapies,such as TAVR, the last decade has witnessed a significantincrease in the volume of PABVs being performed in theUnited States.12 This is also associated with a paralleldecrease in hospital mortality in PABV.3,12,13 We recentlyreported a remarkable expansion in the utilization of PABVthrough 2007 to 2010, a time period that coincides with thepublication of the early TAVR experience and initiation ofclinical trials in the United States.12,14,15 Initial TAVRprotocols obligated treatment of any significant CAD beforethe procedure that may have required interventional cardi-ologists to strive for complete revascularization and to un-dertake concomitant PABV and PCI as a bridge to definitivetreatment. In the present study, we noted a significant in-crease in performance of concomitant PCI þ PABVsespecially through 2007 to 2010 (9.4 to 16 procedures/million elderly) during the same hospitalization.

There is limited data looking at the outcomes ofconcomitant PABV and PCI. Previous studies are eitheroutdated and/or limited by small sample size.16,17 McKayet al16 reported their experience in the pre-TAVR era with 9combined procedures and demonstrated the safety of thisapproach for patients with AS and single-vessel CAD. Inanother small retrospective study of 17 patients (>70 years),Pedersen et al reported no incidence of periprocedural

6 The American Journal of Cardiology (www.ajconline.org)

mortality, myocardial infarction, or stroke. They also foundthat performing PCI in addition to PABV increased thelength of hospital stay (4.1 � 2.8 days in the PABV þ PCIgroup compared with 3.3 � 2.1 days for the PABV group).Ben-Dor et al18 recently published their single-centerexperience comparing PABV with concomitant PABV þPCI. In this study, 80 concomitant PABV þ PCI procedureswere compared with 329 PABVs alone.18 They found thatserious adverse events (defined as intraprocedural death,stroke, moderate-to-severe aortic regurgitation, profoundhypotension requiring resuscitation and intubation or car-dioversion, tamponade, permanent pacemaker requirement,and vascular complication requiring intervention) occurredin 13.7% in the PABV þ PCI group compared with 17.3%in the PABV group (p ¼ 0.44).18 Similar to our study, theyfound no difference in length of hospital stay between the 2groups. They also found a high incidence of vascularcomplications in the PABV þ PCI group (7.5%) consistentwith the present study.18 However, it must be noted that thiswas a single-center study, where all procedures were per-formed by highly experienced operators, and thus, the re-sults may have limited external validity. Our present study isthe largest cross-sectional “real-world” US experience ofconcomitant PABV þ PCI involving both high- and low-volume centers and operators. In addition, the predictorsof mortality, complications, LOS, and cost in this patientpopulation have not been reported earlier.

The population in this study is very high risk, elderly, andwithmultiple co-morbidities. This could be an explanation forthe high associatedmortality rate (10.3%). Despite significantdifferences in the baseline characteristics of the 2 groups(sicker population in the PABV þ PCI group), the rates ofmortality, complications, and length of hospital stay weresimilar in the 2 groups. The most important predictors ofadverse events (mortality or any complication) wereincreasing Charlson’s co-morbidity score, hemodynamicinstability, and weekend admissions. These results are similarto a recently published large study on PABV alone.12 Anotherinteresting finding of this study is increased cost of hospital-izationwhen PCI is performed in addition to PABVduring thesame hospitalization. Previous studies have not evaluated thisincremental effect of PCI on the cost compared with PABValone. This finding may be relevant in the current scenariowhere complete revascularization before definitive treatmentfor AS has been challenged. Treatment of significant CADwith PCI before PABV seems enticing and self-evident as therapid right ventricular pacing technique used during PABVproduces a period of hypotension that may worsen coronaryflow in patients with CAD. In addition, left ventricular systoleagainst the occlusive balloon increases left ventricular wallstress and, hence, myocardial oxygen demand. This isparticularly seen in patients with depressed left ventricularsystolic function.19 In contrast, in a study of 136 patients withAS including 104 patients with CAD,Masson et al20 noted nodifference in mortality or adverse events among 5 groups(created on the basis of extent of CAD) of patients who un-derwent TAVR. Their findings suggest that complete revas-cularizationmay not be a prerequisite for TAVR. Thefindingsof our study regarding the effect of PCI on cost burden mayfurther add to the current debate and decision making.

This study has limitations that are inherent to any largepost hoc analysis. We used an administrative database, anddata were not independently verified. Variables that werenot collected as part of the NIS could not be included in theanalysis. Compared with previous studies, we lack infor-mation such as baseline functional status of patients, cardiacoutput, aortic valve area or hemodynamic data, and so on. Inaddition, we did not have postdischarge data on mortality orfunctional status as the NIS does not track individual pa-tients over time. We could not determine if a PABV wasperformed as “destination” palliative therapy or as a“bridge” to definitive therapy (surgical aortic valvereplacement or TAVR). The ICD-9-CM procedure codes forTAVR were not available during the study period to confirmits effect on procedures. We also could not determine thetiming of PCI in relation to PABV. In addition, we could notdetermine the timing during the hospitalization that couldhave affected the total LOS. Acknowledging these limita-tions, the present study has important strengths including thelargest sample size and the use of standardized definitions ofpreventable adverse events that are established by theAgency for Healthcare Research and Quality. This is alsothe first study to investigate the predictors of mortality,complications, LOS, and cost of hospital stay associatedwith combining these 2 procedures.

Disclosures

None of the other authors have any disclosures relevantto the content of the manuscript.

Supplementary Data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.amjcard.2014.11.032.

1. Ortlepp JR, Schmitz F, Bozoglu T, Hanrath P, Hoffmann R. Cardio-vascular risk factors in patients with aortic stenosis predict prevalenceof coronary artery disease but not of aortic stenosis: an angiographicpair matched case-control study. Heart 2003;89:1019e1022.

2. Rapp AH, Hillis LD, Lange RA, Cigarroa JE. Prevalence of coronaryartery disease in patients with aortic stenosis with and without anginapectoris. Am J Cardiol 2001;87:1216e1217.

3. Iung B, Cachier A, Baron G, Messika-Zeitoun D, Delahaye F, TornosP, Gohlke-Barwolf C, Boersma E, Ravaud P, Vahanian A. Decision-making in elderly patients with severe aortic stenosis: why are so manydenied surgery? Eur Heart J 2005;26:2714e2720.

4. Saia F, Marrozzini C, Moretti C, Ciuca C, Taglieri N, Bordoni B,Dall’ara G, Alessi L, Lanzillotti V, Bacchi-Reggiani ML, Branzi A,Marzocchi A. The role of percutaneous balloon aortic valvuloplasty asa bridge for transcatheter aortic valve implantation. EuroIntervention2011;7:723e729.

5. Ben-Dor I, Maluenda G, Dvir D, Barbash IM, Okubagzi P, TorgusonR, Lindsay J, Satler LF, Pichard AD, Waksman R. Balloon aorticvalvuloplasty for severe aortic stenosis as a bridge to transcatheter/surgical aortic valve replacement. Catheter Cardiovasc Interv 2013;82:632e637.

6. Kumar G, Kumar N, Taneja A, Kaleekal T, Tarima S, McGinley E,Jimenez E, Mohan A, Khan RA, Whittle J, Jacobs E, Nanchal R.Nationwide trends of severe sepsis in the 21st century (2000-2007).Chest 2011;140:1223e1231.

7. Deshmukh A, Kumar G, Kumar N, Nanchal R, Gobal F, Sakhuja A,Mehta JL. Effect of Joint National Committee VII report on hospital-izations for hypertensive emergencies in the United States. Am JCardiol 2011;108:1277e1282.

Valvular Heart Disease/Comparison of Outcomes of PABV þ PCI vs PABV Alone 7

8. Nationwide Inpatient Sample (NIS) related reports. Available at: http://www.hcupus.ahrq.gov/db/nation/nis/nisrelatedreports.jsp. Accessed onSeptember 1, 2012.

9. McDonald KM, Romano PS, Geppert J, Davies SM, Duncan BW,Shojania KG, Hansen A. Measures of Patient Safety Based on HospitalAdministrative Data—the Patient Safety Indicators, StructuredAbstract. Vol 02-0038. Rockville, MD: Agency for Healthcare Reseachand Quality, 2002.

10. Romano PS, Geppert JJ, Davies S, Miller MR, Elixhauser A, McDonaldKM. A national profile of patient safety in U.S. hospitals. Health Aff(Millwood) 2003;22:154e166.

11. Deshmukh A, Patel NJ, Pant S, Shah N, Chothani A, Mehta K, GroverP, Singh V, Vallurupalli S, Savani GT, Badheka A, Tuliani T,Dabhadkar K, Dibu G, Reddy YM, Sewani A, Kowalski M, Mitrani R,Paydak H, Viles-Gonzalez JF. In-hospital complications associatedwith catheter ablation of atrial fibrillation in the United States between2000 and 2010: analysis of 93801 procedures. Circulation 2013;128:2104e2112.

12. Badheka AO, Patel NJ, Singh V, Shah N, Chothani A, Mehta K,Deshmukh A, Ghatak A, Rathod A, Desai H, Savani GT, Grover P,Patel N, Arora S, Grines CL, Schreiber T, Makkar R, Rihal CS, CohenMG, De Marchena E, O’Neill WW. Percutaneous aortic balloon val-votomy in the United States: a 13-year perspective. Am J Med2014;127:744e753.e3.

13. Charlson E, Legedza AT, Hamel MB. Decision-making and outcomesin severe symptomatic aortic stenosis. J Heart Valve Dis 2006;15:312e321.

14. Cribier A, Eltchaninoff H, Tron C, Bauer F, Agatiello C, Nercolini D,Tapiero S, Litzler PY, Bessou JP, Babaliaros V. Treatment of calcificaortic stenosis with the percutaneous heart valve: mid-term follow-up

from the initial feasibility studies: the French experience. J Am CollCardiol 2006;47:1214e1223.

15. Webb JG, Pasupati S, Humphries K, Thompson C, Altwegg L, MossR, Sinhal A, Carere RG, Munt B, Ricci D, Ye J, Cheung A,Lichtenstein SV. Percutaneous transarterial aortic valve replacementin selected high-risk patients with aortic stenosis. Circulation2007;116:755e763.

16. McKay RG, Safian RD, Berman AD, Diver DJ, Weinstein JS, WymanRM, Cunningham MJ, McKay LL, Baim DS, Grossman W. Combinedpercutaneous aortic valvuloplasty and transluminal coronary angio-plasty in adult patients with calcific aortic stenosis and coronary arterydisease. Circulation 1987;76:1298e1306.

17. Pedersen WR, Klaassen PJ, Pedersen CW, Wilson JA, Harris KM,Goldenberg IF, Poulose AK, Mooney MR, Henry TD, Schwartz RS.Comparison of outcomes in high-risk patients >70 years of age withaortic valvuloplasty and percutaneous coronary intervention versusaortic valvuloplasty alone. Am J Cardiol 2008;101:1309e1314.

18. Ben-Dor I, Maluenda G, Looser PM, Okubagzi P, Torguson R, Xue Z,Lindsay J, Satler LF, Pichard AD, Waksman R. Outcomes ofconcomitant percutaneous coronary intervention and balloon aorticvalvuloplasty. Catheter Cardiovasc Interv 2013;82:E835e841.

19. Martinez CA, Singh V, Londono JC, Cohen MG, Alfonso CE, O’NeillWW, Heldman AW. Percutaneous retrograde left ventricular assistsupport for interventions in patients with aortic stenosis and left ven-tricular dysfunction. Catheter Cardiovasc Interv 2012;80:1201e1209.

20. Masson JB, Lee M, Boone RH, Al Ali A, Al Bugami S, Hamburger J,John Mancini GB, Ye J, Cheung A, Humphries KH, Wood D,Nietlispach F, Webb JG. Impact of coronary artery disease on out-comes after transcatheter aortic valve implantation. Catheter Car-diovasc Interv 2010;76:165e173.