Author + information
- Received May 10, 2018
- Revision received September 27, 2018
- Accepted October 9, 2018
- Published online February 18, 2019.
- Adam C. Salisbury, MD, MSca,b,∗ (, )
- Dimitri Karmpaliotis, MDc,
- J. Aaron Grantham, MDa,b,
- James Sapontis, MDd,
- Qingrui Meng, MSa,
- Elizabeth A. Magnuson, ScDa,b,
- Hemal Gada, MDe,
- William Lombardi, MDf,
- Jeffrey Moses, MDc,
- Haiyan Li, MSa,
- Suzanne V. Arnold, MD, MHAa,b,
- Suzanne J. Baron, MD, MSca,b,
- John A. Spertus, MD, MPHa,b and
- David J. Cohen, MD, MSca,b
- aSaint Luke’s Mid America Heart Institute, Kansas City, Missouri
- bUniversity of Missouri–Kansas City, Kansas City, Missouri
- cColumbia University Medical Center and New York Presbyterian Hospital, New York, New York
- dMonash Heart, Melbourne, Australia
- ePinnacleHealth Cardiovascular Institute, Wormleysburg, Pennsylvania
- fUniversity of Washington School of Medicine, Seattle, Washington
- ↵∗Address for correspondence:
Dr. Adam C. Salisbury, Saint Luke’s Mid America Heart Institute, 4401 Wornall Road, CV Research, 9th Floor, Kansas City, Missouri 64111.
Objectives The aim of this study was to describe the costs of chronic total occlusion (CTO) percutaneous coronary intervention (PCI) and the association of complications during CTO PCI with costs and length of stay (LOS).
Background CTO PCI generally requires more procedural resources and carries higher risk for complications than PCI of non-CTO vessels. The costs of CTO PCI using the hybrid approach have not been described, and no studies have examined the impact of complications on in-hospital costs and LOS in this population.
Methods Costs were calculated for 964 patients in the 12-center OPEN-CTO (Outcomes, Patient Health Status, and Efficiency in Chronic Total Occlusion Hybrid Procedures) registry using prospectively collected resource utilization and billing data. Multivariate models were developed to estimate the incremental costs and LOS associated with complications. Attributable costs and LOS were calculated by multiplying the independent cost of each event by its frequency in the population.
Results Mean costs for the index hospitalization were $17,048 ± 9,904; 14.5% of patients experienced at least 1 complication. Patients with complications had higher mean hospital costs (by $8,603) and LOS (by 1.5 days) than patients without complications. Seven complications were independently associated with increased costs and 6 with LOS; clinically significant perforation and myocardial infarction had the greatest attributable cost per patient. Overall, complications accounted for $911 per patient in hospital costs (5.3% of the total costs) and 0.2 days of additional LOS.
Conclusions Complications have a significant impact on both LOS and in-hospital costs for patients undergoing CTO PCI. Methods to identify high-risk patients and develop strategies to prevent complications may reduce CTO PCI costs.
Contemporary chronic total occlusion (CTO) percutaneous coronary intervention (PCI) using the hybrid approach has been shown to be safe and effective for relief of angina symptoms and is now performed with technical success rates of approximately 90% at leading centers (1,2). However, these procedures are more time consuming, and often require additional devices, compared with PCI of arteries that are not totally occluded. Only 1 prior study has examined the costs of CTO angioplasty, using a decision-analytic model that was constructed using transition probabilities for success rates and complications that were derived largely from studies of older CTO PCI techniques (3). Since that publication, however, the hybrid approach to CTO PCI has been adopted by the majority of high-volume CTO operators in the United States (3,4). To date, no studies have examined the costs of CTO PCI in contemporary practice, or the relationship between procedural complications and key outcomes of in-hospital costs and length of stay (LOS).
To address these gaps in knowledge, we conducted a prospective economic analysis alongside the OPEN-CTO (Outcomes, Patient Health Status, and Efficiency in Chronic Total Occlusion Hybrid Procedures) registry, which enrolled consecutive patients undergoing CTO PCI at 12 experienced U.S. centers (5). This study included careful assessment of procedural complications, including core laboratory adjudication for rates of clinically significant perforation and collection of detailed procedural resource use and hospital billing data to allow calculation of both procedural and nonprocedural costs. The insights from this study should provide a better understanding of the impact of complications, some of which are potentially avoidable, on the costs of CTO PCI, and should also help inform future studies of the cost-effectiveness of CTO PCI using the hybrid approach.
The OPEN-CTO (Outcomes, Patient Health Status, and Efficiency in Chronic Total Occlusion Hybrid Procedures) registry is a prospective, single-arm study that enrolled consecutive patients with CTO who underwent attempted CTO PCI at 12 U.S. centers between January 21, 2014, and July 22, 2015. A full description of the study methods has been published previously (5). Eligible patients were >18 years of age and had symptoms suggestive of ischemic heart disease. All patients had CTO, which was defined as a lesion with TIMI (Thrombolysis In Myocardial Infarction) flow grade 0 antegrade flow that was thought to have been present for at least 3 months. Patients were treated according to the previously published hybrid algorithm for CTO PCI (4). The CTO PCI operators were required to have a minimum of 2 years of experience performing CTO PCI and were required to have performed at least 100 CTO PCI procedures before participating in OPEN-CTO. Enrollment of all consenting patients undergoing CTO PCI at each center was confirmed via linkage of the study database to each hospital’s National Cardiovascular Data Registry CathPCI data to ensure that all patients undergoing CTO PCI during the study time frame were included in OPEN-CTO. The results of this audit have been published previously, indicating no evidence of selection bias (5). Each participating site obtained institutional research board approval, and all patients provided informed consent to participate in the study.
In-hospital complications were identified by the sites according to standardized definitions (Online Table 1) (5). Complications assessed included death, stroke (ischemic or hemorrhagic), myocardial infarction, vascular complication, access-site hematoma, gastrointestinal or other non-access-site bleeding, contrast-induced nephropathy, donor-vessel dissection or thrombosis, clinically significant perforation, urgent repeat PCI in-hospital, emergent cardiac surgery, septal hematoma, pericardial effusion, and acute radiation dermatitis.
Index hospitalization costs
The costs of the index hospitalization were assessed from the perspective of the U.S. health care system and are reported in 2016 U.S. dollars. Any costs incurred in years other than 2016 were converted to 2016 U.S. dollars using the medical care component of the Consumer Price Index. Costs were determined using a combination of hospital billing data and resource-based accounting methods (6). Prior studies have demonstrated that this approach correlates closely with costs derived from individual hospitals’ microcost accounting systems (7,8).
Each participating center recorded the duration of the procedure (from first injection of local anesthetic to the removal of the last guiding catheter) and collected counts for each device used, including guidewires, guiding catheters, balloons, stents, dissection and re-entry catheters (CrossBoss catheters and Stingray balloons), microcatheters, intravascular ultrasound or optical coherence tomographic catheters, atherectomy devices, pacing catheters, coils, hemodynamic support devices, snares, embolic protection devices, and closure devices. Sites also recorded total contrast volume and periprocedural anticoagulant use. Procedural costs were calculated by multiplying item counts by their respective acquisition costs, which were assessed on the basis of a survey of participating centers. Costs for additional disposable equipment, overhead, and depreciation for the cardiac catheterization laboratory were estimated on the basis of the average cost per procedure at Saint Luke’s Mid America Heart Institute in 2016, adjusted for actual procedure duration, which was defined as the time between first injection of lidocaine at the access site to removal of the last guiding catheter.
Costs for the remainder of each index hospitalization were calculated by multiplying all nonprocedural charges by the cost center–specific cost-to-charge ratio obtained from the hospital’s Medicare cost report (7). Physician fees during the hospital stay were derived from the Medicare fee schedule.
The analytic cohort for our study consisted of all patients with complete cost data for the index hospitalization. Continuous variables are described as mean ± SD along with median (interquartile range [IQR]), as appropriate. Categorical variables are described as proportions. Univariate comparisons between patients with and without procedural complications were performed using independent t tests for continuous variables and chi-square tests for categorical variables. The unadjusted incremental cost (or LOS) associated with each complication was defined as the difference in mean cost (or LOS) between patients who developed the complication of interest and those who did not develop that particular complication.
We used multivariate regression analysis to estimate the incremental cost associated with each complication, while adjusting for potential confounders. To identify potential confounders of the association between complications and costs, we fit a series of models to identify the factors that were either associated with in-hospital complications or were associated with costs among patients without complications (9). In the first step, we used logistic regression models with backward selection, with a threshold for variable retention of p < 0.10, with any complication as the dependent variable and examined the demographic and clinical variables associated with this outcome. For these models, we considered the following potential predictors: age, sex, prior PCI, current smoking, baseline hemoglobin value, lesion length, and J-CTO (Multicenter CTO Registry in Japan) score. Second, we used generalized linear models with a log link to examine the predictors of costs among the subset of patients without complications, using the same backward selection approach and potential covariates.
We then constructed multivariate linear regression models that included potential confounders identified during the first 2 steps, along with age, sex, and all periprocedural complications of interest. To avoid model overfitting, complications that occurred in fewer than 3 patients were excluded as potential predictors. Model reduction was performed using backward selection with a significance threshold of p < 0.10 for variable retention. This cost model was fit using both log-transformed and untransformed costs as the dependent variable, but because both models fit and the magnitude of the association between complications and cost were similar in the 2 analyses, we report only the untransformed costs for ease of interpretation. Attributable costs for each complication were calculated by multiplying the independent cost of each event (as derived from regression model coefficients) by the frequency of the event in the study population. Finally, all analyses were repeated with LOS as the dependent variable. All statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, North Carolina).
Contribution margin estimation
To compare the costs of CTO PCI with hospital reimbursement for these procedures, we conducted additional analyses to estimate the contribution margin of CTO PCI. Using Medicare Provider Analysis and Review data for 2014 (the year patients were enrolled in OPEN-CTO), we determined the median Medicare reimbursement paid to each study hospital for diagnosis-related group 246. We then estimated a contribution margin for each patient by subtracting the total hospital cost (excluding physician fees) for each OPEN-CTO patient from these site-specific reimbursement rates. In addition to the overall analysis, we performed a subgroup analysis restricted to those patients ≥65 years of age (who were eligible for Medicare). Finally, we repeated the analysis under the assumption that hospital payments were at the mean Medicare rate for all participants ≥65 years of age and at 1.5 times Medicare rates for all participants <65 years of age (because those hospitalizations would have been reimbursed by private payers ). Confidence intervals (CIs) for these estimates were obtained by bootstrap replication (1,000 samples).
Complete index hospitalization billing data were available for 964 of 1,000 patients (96.4%) enrolled in OPEN-CTO. There were no clinically important differences in clinical or angiographic characteristics between those patients with versus without hospital billing data (Online Table 2). The mean age of the analytic cohort was 65 years, and 80% were men. Overall, 36.2% of patients had undergone previous coronary artery bypass graft, 65.4% had undergone prior PCI, and 22.7% had histories of chronic heart failure. The baseline characteristics of patients with and without complications are compared in Table 1. Patients with complications were more likely to be women and to be current smokers. Mean J-CTO scores were higher among patients with complications compared with patients who did not experience complications (2.7 ± 1.1 vs. 2.3 ± 1.1; p < 0.001). Successful completion of the procedure was accomplished using an antegrade approach in 576 patients (65.0%) and a retrograde approach in 367 patients (35.0%). The complication rate was higher in procedures requiring retrograde access compared with antegrade procedures (57 [22.2%] vs. 59 [10.2%]; p < 0.001).
The frequency of each periprocedural complication is listed in Table 2. At least 1 complication occurred in 140 patients (14.5%); 2.7% experienced 2 complications, and 1.2% experienced 3 or more complications. The most frequent complications were clinically significant perforation (4.8%), access-site hematoma (4.3%), and myocardial infarction (2.7%).
Initial treatment costs and LOS
Procedural resource utilization and costs for the initial revascularization procedure and the index hospitalization are summarized in Table 3, and specific device costs and components of nonprocedural costs are presented in Online Table 3. On average, procedures lasted 120 ± 64 min and required 3.4 ± 1.9 guiding catheters, 10.1 ± 6.4 wires, 5.1 ± 3.1 angioplasty balloons, and 2.6 ± 1.5 stents. Index procedure costs averaged $12,280 ± $5,972 per patient (median $11,420; IQR: $8,428 to $14,514), while nonprocedural hospital costs were $3,424 ± $6,188 (median $1,802; IQR: $934 to $2,627) and physician fees were $1,344 ± 427 (median $1,279; IQR: $1,179 to $1,283). Overall, the mean cost of the initial PCI hospitalization was $17,048 ± 9,904 per patient (median $14,729; IQR: $11,377 to $18,799), and mean length of hospital stay was 1.7 ± 2.4 days (median 1 day; IQR: 1 to 1 day). The occurrence of 1 or more complications was associated with significantly higher in-hospital costs, with an unadjusted incremental cost of $8,603, driven by both increases in procedural resource utilization and LOS (Table 4). The unadjusted incremental cost of complications ranged from −$428 for septal hematoma to $28,247 for death. The unadjusted incremental LOS ranged from −0.2 days for septal hematoma to 10.6 days for emergent cardiac surgery (Table 4).
Adjusted costs and LOS
After adjusting for age, sex, prior PCI, smoking, baseline hemoglobin, and J-CTO score, 7 complications were independently associated with in-hospital costs (Table 5): death, myocardial infarction, clinically significant perforation, vascular complications, emergent cardiac surgery, contrast nephropathy, and donor-vessel dissection or thrombosis. Of these, clinically significant perforation and myocardial infarction were the most important drivers of in-hospital costs, with attributable costs of $250/patient and $212/patient, respectively. Overall, $911 of the total cost of CTO PCI was attributable to periprocedural complications, which represents 5.3% of total hospitalization costs for the procedure.
After risk adjustment, 6 complications were independently associated with LOS (Table 5). Death was associated with shorter LOS, while myocardial infarction, vascular complications, contrast nephropathy, cardiac surgery, and donor-vessel thrombosis or dissection were each associated with longer lengths of stay. Cardiac surgery (11.4 days) and contrast nephropathy (4.2 days) were associated with the greatest incremental LOS, and cardiac surgery was associated with the greatest attributable LOS increase (0.06 days per patient).
Contribution margin of CTO PCI
Although individual patient-level contribution margin data were not available, we estimated the contribution margin for each OPEN-CTO patient by comparing his or her hospital cost with site-specific reimbursement levels for diagnosis-related group 246. One hospital was missing Medicare Provider Analysis and Review reimbursement data in 2014, and skewed values for reimbursement data for 3 centers led to the exclusion of 168 patients from this analysis. For the overall study population (n = 818), the contribution margin was $7,289 ± $10,151 (95% CI: $6,497 to $8,004). When the analysis was restricted to Medicare-eligible patients (n = 430), the contribution margin was $6,164 ± $10,922 (95% CI: $5,018 to $7,147) (Online Table 4). Finally, the estimated contribution margin increased to $12,842 ± $12,754 (95% CI: $11,892 to $13,729) under the assumption that reimbursement rates would be 1.5 times those of Medicare rates for patients <65 years of age.
This is the first study to formally assess the cost of CTO PCI using the hybrid algorithm and to examine the impact of procedure-related complications on hospital costs for this challenging patient population. Using detailed resource utilization and cost data collected alongside the OPEN-CTO registry, we found that across 12 high-volume CTO centers in the United States, mean costs for the index procedure and associated hospitalization were $17,048, of which $12,280 were related to the PCI procedure itself. In addition, we found that procedure-related complications were associated with substantial cost and increased LOS. On a per event basis, we found that the most costly complications were cardiac surgery and death (with adjusted incremental costs of $15,592 and $13,279 per event, respectively). On a per patient basis, however, given their greater frequency, the most costly complications were clinically significant vessel perforation and periprocedural myocardial infarction (with adjusted attributable costs of the complications of $212 and $250 per patient, respectively). Overall, these complications added more than $900 to the cost of each CTO PCI hospitalization.
Comparison with previous studies
The complication rate observed in OPEN-CTO was higher than in prior studies of CTO PCI such as the PROGRESS CTO (Prospective Global Registry for the Study of Chronic Total Occlusion Intervention) registry (11) and in a meta-analysis of prior studies of CTO PCI (12). These apparent differences in complication rates mainly reflect differences in the level of detail of procedural complications captured in OPEN-CTO compared with most previous studies (which have focused exclusively on major adverse cardiovascular events). Moreover, the careful inclusion of all patients undergoing CTO PCI at OPEN-CTO centers is likely to have resulted in a more complete ascertainment of complications compared with retrospective studies in which selection bias (i.e., “cherry picking”) may have resulted in a biased assessment of procedural outcomes. Finally, our use of an angiographic core laboratory resulted in the detection of a number of angiographic complications that were not reported by site investigators and likely led to more complete ascertainment of complications compared with other registry studies that did not include a core laboratory assessment.
Although there has been increasing interest in CTO PCI in recent years, little attention has been paid to the cost of these procedures. A prior economic analysis that was conducted using a decision-analytic model found that CTO PCI was likely a cost-effective intervention in patients with chronic stable angina and severe symptoms (3). However, inputs for that model were obtained from a host of prior PCI studies, all predating the hybrid approach to CTO PCI, which has been adopted by the majority of CTO operators in the United States. Moreover, that analysis was conducted by modeling of costs from prior studies rather than the collection of empirical resource utilization and cost data, the approach used in our study. Thus, we believe that the present study contributes substantial new high-quality data to our understanding of the costs of CTO PCI in current U.S. practice.
It is critical to understand the costs of CTO PCI in the context of prior studies of complex PCI. The procedural costs of CTO PCI in the present study ($12,280 ± $5,972) are similar to the costs of multivessel or left main PCI in the SYNTAX (Synergy Between PCI With Taxus and Cardiac Surgery) ($11,919 ± $6,162) and FREEDOM (Future Revascularization Evaluation in Patients With Diabetes Mellitus: Optimal Management of Multivessel Disease) ($13,014 ± $5,173) trials (13,14). These costs reflect the fact that more procedural devices are typically required to treat CTOs, including multiple drug-eluting stents, guidewires, guiding catheters, and, frequently, specialized devices to facilitate dissection and re-entry. However, the higher reimbursements provided for these complex procedures appear to account for these additional costs, yielding a highly favorable net “contribution margin” for these procedures, as described in our analysis of the costs of CTO PCI in context of reimbursement for CTO PCI. These results are consistent with a favorable contribution margin for CTO PCI reported in a prior single-center analysis of the costs and contribution margins of CTO PCI (15).
In OPEN-CTO, we found that the incremental cost of a procedure-related complication was about $8,600, similar to the cost of a complication among an unselected population of Medicare patients undergoing PCI (16). There were important differences in the complications that drove costs in these 2 studies, however. Aside from death and cardiac surgery, which were expensive complications in both studies, medical complications such as renal failure, septicemia, and adult respiratory distress syndrome were major drivers of the cost of complications in the Medicare cohort, whereas we found that procedural complications such as donor-vessel dissection and thrombosis, perforation, and periprocedural myocardial infarction were major drivers of cost in OPEN-CTO.
There are several likely explanations for these differences. Because OPEN-CTO included stable outpatients admitted for elective PCI whereas the Medicare analysis included patients undergoing elective procedures as well as inpatients undergoing PCI for acute coronary syndromes, it is not surprising that costly complications in OPEN-CTO predominantly reflected procedural complications, rather than acute noncardiac medical conditions. In addition, it is likely that the use of endpoint adjudication and oversight by an angiographic core laboratory led to greater detection of cardiac-specific complications in OPEN-CTO. Finally, the differences in complications between the 2 studies likely reflects the differences in the technical challenges of conventional PCI versus CTO PCI procedures.
Clinical and economic implications
Our findings have several important implications for clinical care and quality improvement. Some of the complications that were associated with high costs in OPEN-CTO are potentially avoidable, such that careful attention to their prevention could represent an opportunity to improve patient outcomes while also reducing costs. For example, approaches such as the routine use of risk prediction tools to identify patients at high risk for contrast nephropathy (17) or the use of ultrasound-guided access to minimize the risk for access-site vascular complications (18) could reduce the risk for these complications and their associated costs.
These data may also be useful to inform future studies of the cost-effectiveness of CTO PCI. As noted previously, the only available study of the cost-effectiveness of CTO PCI relied upon cost inputs derived mainly from non-CTO patients, thus raising questions about the validity of the study projections (3). In contrast, the results of our study are directly applicable to patients undergoing CTO PCI using contemporary techniques and include a broad range of complications, some of which are unique or enriched in the CTO population (such as coronary perforation and donor-vessel dissection or thrombosis). In light of growing constraints on hospital resources, these insights are also key to understand the likely impact of these procedures on hospitals pursuing CTO PCI programs.
First, this study focused exclusively on in-hospital costs and complications of CTO PCI. Longer term studies will be necessary to develop a more complete understanding the impact of complications on the long-term costs of CTO PCI from a societal perspective.
Second, OPEN-CTO did not mandate routine post-PCI collection of cardiac biomarkers or serum creatinine values. It is therefore possible that we have underestimated the true incidence of these complications in the study population. However, given consistent definitions of clinical myocardial infarction and acute renal failure across OPEN-CTO sites, and careful collection of clinically apparent events, we believe these results are reasonable reflections of current clinical practice.
Third, although OPEN-CTO is among the largest prospective registries of CTO PCI, its sample size was relatively modest. Consequently, some of our incremental cost estimates have wide confidence intervals, and it is possible that some rare, but costly, complications were not identified in our study.
In this multicenter registry of CTO PCI using the hybrid approach, we found that complications occurred in about 14% of patients and were associated with higher hospital costs and greater LOS, accounting for about 5% of the overall costs of CTO PCI. On a per event basis, the most costly events were death and cardiac surgery, while on a patient level, clinically significant perforation and periprocedural myocardial infarction had the highest attributable costs. Future efforts to reduce potentially preventable complications should further reduce the costs of contemporary CTO PCI and enhance the cost-effectiveness of these revascularization procedures.
WHAT IS KNOWN? CTO angioplasty using the hybrid approach has been shown to be safe and effective and can be performed with high success rates. However, PCI for CTOs typically requires more procedural resources than vessels that are not totally occluded, and no contemporary studies have reported the costs of CTO PCI using the hybrid approach or the costs associated with procedural complications.
WHAT IS NEW? Mean costs for the index hospitalization were $17,048 ± 9,904; 14.5% of patients experienced at least 1 complication. Patients with complications had higher mean hospital costs (by $8,603) and LOS (by 1.5 days) than patients without complications. Overall, complications accounted for $911 per patient in hospital costs (5.3% of the total costs) and 0.2 days of additional LOS.
WHAT IS NEXT? The costs of CTO PCI using the hybrid approach were similar to costs that have been reported in prior studies of complex PCI patients. Complications have a significant impact on both LOS and in-hospital costs for patients undergoing CTO PCI. Methods to identify high-risk patients and develop strategies to prevent complications may reduce the costs of CTO PCI.
The OPEN-CTO study was funded by an unrestricted grant from Boston Scientific. Dr. Salisbury has received institutional research grants from Boston Scientific and Gilead. Dr. Grantham has received speaking fees and honoraria from Boston Scientific, Abbott Vascular, and Asahi Intecc; institutional research grant support from Boston Scientific; and institutional educational grant support from Abbott Vascular, Vascular Solutions, Boston Scientific, and Asahi Intecc. Dr. Grantham is a part-time employee of Corindus Vascular Robotics. Dr. Sapontis has received speaking fees and honoraria from Boston Scientific. Dr. Lombardi has received speaking fees and honoraria from Boston Scientific, Abbott Vascular, and Abiomed; and consultancy fees from Vascular Solutions, Abbott Vascular, Boston Scientific, Abiomed, and Roxwood Medical. Dr. Lombardi has equity in Roxwood Medical and Bridgepoint Medical; and his wife is an employee of Spectranetics. Dr. Karmpaliotis has received speaking fees, honoraria, and consulting fees from Abbott Vascular, Boston Scientific, and Medtronic. Dr. Baron has received research grant support from Boston Scientific. Dr. Spertus has received research grants from Lilly, Novartis, and Abbott Vascular; has served as a consultant for Novartis, Cytokinetics, Bayer, and United Healthcare; owns the copyright to the SAQ; and has an equity interest in Health Outcomes Sciences. Dr. Cohen has received institutional research grant support from Boston Scientific, Abbott Vascular, and Medtronic; and consulting fees from Medtronic and Abbott Vascular. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery disease
- confidence interval
- chronic total occlusion
- interquartile range
- length of stay
- percutaneous coronary intervention
- Received May 10, 2018.
- Revision received September 27, 2018.
- Accepted October 9, 2018.
- 2019 American College of Cardiology Foundation
- Kandzari D.E.,
- Kini A.S.,
- Karmpaliotis D.,
- et al.
- Gada H.,
- Whitlow P.L.,
- Marwick T.H.
- Brilakis E.S.,
- Grantham J.A.,
- Rinfret S.,
- et al.
- Sapontis J.,
- Marso S.P.,
- Cohen D.J.,
- et al.
- Cohen D.J.,
- Krumholz H.M.,
- Sukin C.A.,
- et al.,
- for the Stent Restenosis Study Investigators
- Arnold S.V.,
- Lei Y.,
- Reynolds M.R.,
- et al.
- ↵America’s Health Insurance Plans. National comparisons of commercial and Medicare fee-for-service payments to hospitals. Available at: https://www.ahip.org/national-comparisons-of-commercial-and-medicare-fee-for-service-payments-to-hospitals/. Accessed October 25, 2018.
- Danek B.A.,
- Karatasakis A.,
- Karmpaliotis D.,
- et al.
- Patel V.G.,
- Brayton K.M.,
- Tamayo A.,
- et al.
- Tsai T.T.,
- Patel U.D.,
- Chang T.I.,
- et al.
- Seto A.H.,
- Abu-Fadel M.S.,
- Sparling J.M.,
- et al.