HSR&D Home » Research » IIR 09-347 – HSR&D Study
Cardiac Risk and Stent Effect on Adverse Perioperative Events
Mary T Hawn, MD MPH BS
Birmingham VA Medical Center, Birmingham, AL
Funding Period: October 2010 - March 2014
Drug eluting stent (DES) implantation for the treatment of coronary artery disease has become the standard practice.(Malenka, Kaplan et al. 2008) While DES are associated with decreased rates of vessel restenosis due to intimal hyperplasia, they are associated with higher rates of stent thrombosis. Antiplatelet therapy is the mainstay for prevention of future thrombosis after coronary artery stenting. Initial recommendation for antiplatelet therapy duration was 3 to 6 months post stent implantation, depending on the brand of DES used. However, several publications have reported long term risk of subsequent myocardial infarction (MI) and death with cessation of antiplatelet therapy, even after 12 months of treatment post DES implantation, resulting in significant concern for the exact safe period to stop antiplatelet therapy.(Kaluza, Joseph et al. 2000, Eisenstein, Anstrom et al. 2007, Ho, Fihn et al. 2007)
The management of antiplatelet therapy in patients with implanted DES who need subsequent non-cardiac surgical procedures is complex. Most surgeons performing non-cardiac surgery stop antiplatelet medication 5 to 7 days prior to the procedure to minimize the bleeding complications. However, there are several case reports of acute stent thrombosis occurring in the perioperative period leading to significant morbidity and death.(Kaluza, Joseph et al. 2000, Wilson, Fasseas et al. 2003, Nasser, Kapeliovich et al. 2005, Brichon, Boitet et al. 2006, Conroy, Bolsin et al. 2007, Nuttall, Brown et al. 2008) These reports of stent complications in surgical patients, as well as those describing adverse events in patients with early cessation of antiplatelet therapy, led to a modification of the American College of Cardiology/American Heart Association (ACC/AHA) 2007 Guidelines on Perioperative Cardiovascular Evaluation and Care for Noncardiac Surgery to recommend that surgery in patients with DES be delayed 365 days if elective, and if urgent, that the procedure be performed without cessation of aspirin therapy.(Fleisher, Beckman et al. 2007) The level of evidence supporting these guidelines is graded Level C (very limited) and Class IIb recommendations (may be beneficial, effectiveness is not well established). In fact, there are only 3 case reports and 6 case series documenting a total of 28 reports of stent thrombosis, MI, death, or excessive bleeding following noncardiac surgery in patients with recent stent implantation.
While the risk of adverse events associated with DES and the benefit of delaying surgical intervention for non-cardiac procedures are not well known, the consequences of delaying necessary surgery, operating on antiplatelet therapy, or performing surgery within 365 days post stent implantation are essentially unknown. In order to provide necessary information to better guide the safety and timing of subsequent surgical intervention, we propose a nationwide study leveraging two robust databases in VA; 1) Cardiac Assessment Reporting and Tracking for Cath Labs (CART-CL) and 2) VA National Surgical Quality Improvement Program (VASQIP) to further study the incidence of surgery post stent implantation and the association between type of cardiac stent and adverse peri-operative events. CART-CL accrues cardiac specific patient risk variables as well as detailed information on catheterization findings and coronary intervention, including type and location of stent. VASQIP is a validated surgical outcomes database that has been used in numerous research studies and accrues detailed risk and outcome information for approximately 100,000 non-cardiac surgical procedures annually.
The ultimate goal of this project is to inform surgeons, cardiologists, anesthesiologists, and primary care physicians regarding the incidence of subsequent surgery in patients with implanted coronary stents, the incremental risk of adverse events for non-cardiac surgical procedures, and whether adherence to current guidelines attenuates that risk. As the management of these patients with regard to their cardiac and surgical risk of bleeding is often at conflict, we will characterize the extent of that conflict across provider groups and understand the implication for implementing guidelines.
In order to understand the incidence and risk of subsequent surgery in patients following cardiac stent implantation, we propose a retrospective cohort study leveraging the CART-CL and VASQIP databases, enhanced by administrative databases and medical chart abstraction to address the following Specific Aims:
1. To determine the incidence of surgically treated disease in patients with cardiac stents in the two years following implantation. In order to define the magnitude of the need for subsequent surgery in this population, we will determine the rate of referral to surgical clinics and the rate of surgical procedures in the two years following cardiac stent implantation.
2. To estimate the attributable risk of type of cardiac stent on adverse peri-operative events. We will compare the rate of adverse peri-operative events in a cohort of patients that have undergone percutaneous coronary intervention (PCI) in VA from 2000-2011 and had a subsequent non-cardiac surgical procedure with a stratified random sample generated cohort of similar age undergoing similar procedures without preceding PCI. We will assess whether the type of stent implanted and the time between PCI and surgical procedure modulates the association. Finally, we will compare the event rate in a stratified random sample cohort derived from CART-CL that does not undergo subsequent surgery to compare the underlying event rate of stent thrombosis and cardiac complications in this patient population.
3. To describe variation in knowledge, attitudes and adherence to current guidelines among surgeons, anesthesiologists, cardiologists and primary care physicians regarding the timing of elective noncardiac operations after stent placement and antiplatelet therapy cessation for invasive procedures. We will use quantitative techniques to assess variation in provider practices regarding the timing of subsequent surgical intervention and antiplatelet therapy cessation using survey methods, analyses of the dataset created in Specific Aims 1 & 2 supplemented by antiplatelet therapy information from Pharmacy Benefits Management data.
Using these robust databases, we propose a retrospective cohort study supplemented by provider survey to test the following hypotheses:
H1. DES are associated with increased risk of adverse peri-operative events including stent thrombosis, acute MI and death, compared to BMS or no stent, controlling for patient and procedure risks.
H2. The time between implantation of drug eluting stent and subsequent surgery is inversely related to the incidence of adverse peri-operative events.
H3. Surgeons, anesthesiologists, cardiologists and primary care physicians perceive the occurrence and magnitude of the risks of cardiac stent complications and bleeding complications differently.
H4. Patients with implanted coronary artery stents undergoing surgery based on the ACC/AHA recommended guidelines have fewer adverse peri-operative events than patients undergoing surgery outside the recommended guidelines.
H5. Patients with cardiac stents are at increased risk of late cardiac morbidity following surgical procedures, and this increased risk is associated with decreased resumption of antiplatelet therapy.
Design and Participants
This project utilizes a retrospective cohort study design in combination with a nested matched case-control study. In addition, a cross-sectional provider survey was used to describe variation in knowledge, attitudes, and adherence to current guidelines as described in objective 3. The study was solely observational, involving no interventions. For the retrospective cohort study, the primary cohort included all VA patients with a cardiac stent placed in the VA between fiscal year 2000 and fiscal year 2011. We identified all cardiac stents using codes from the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM, 36.06 for BMS or 36.07 for DES) and direct abstraction from the CART-CL data files. Percutaneous coronary intervention (PCI) care episodes were defined as a single visit to the catheterization laboratory for a PCI procedure, where 1 or more stents were implanted.
Patients were followed for 2 years following their cardiac stent placement to determine the incidence of non-cardiac surgery. Non-cardiac surgery occurring in the VA was identified in the VASQIP and noncardiac surgery occurring outside of the VA was identified using Centers for Medicare & Medicaid Services (CMS) data for the 73% of veterans in the cohort who had dual VA-Medicare eligibility. A secondary cohort of patients undergoing non-cardiac surgery without a history of cardiac stent was obtained to examine the attributable risk of cardiac stent on adverse perioperative events.
The nested case-control subset was restricted to (1) VA operations, (2) MI or revascularization end points, and (3) surgeries occurring more than 6 weeks after stent placement. Operations that occurred in the first 6 weeks after stent placement and operations followed by death alone were excluded. After exclusions, we matched patients experiencing a cardiac MACE by fiscal year of operation, CPT category, work relative value unit, stent type, revised Cardiac Risk Index, and time from stent to operation using 24,691 potential controls from VA surgeries that were not followed by a MACE.
To conduct the cross-sectional provider survey, we designed a web-based survey and administered it nationwide to Veterans Health Administration (VHA) surgeons, anesthesiologists, and cardiologists with a valid e-mail address.
The main outcome variable for this study was the occurrence of a major adverse cardiac event (MACE) within 30 days of non-cardiac surgery. MACE was a composite variable including death from any cause, MI (ICD-9-CM codes 410.xx or VASQIP nurse-abstracted MI), or coronary revascularization (ICD-9-CM 00.66, 36.01-36.09; CPT: 33510-33519, 33520-33523, 33530-33536, 92973-92984, 92995-92998).
The 6-point revised cardiac risk index (rCRI) was used to assess the patient's cardiac risk at the time of noncardiac surgery. The rCRI was calculated from ICD-9-CM diagnostic codes for congestive heart failure (CHF), stroke, MI, and diabetes; CPT codes associated with high-risk surgery; and laboratory data identifying 1 or more serum creatinine values greater than 2 mg/dL in the year prior to surgery.(Lee, Marcantonio et al. 1999) An insulin prescription in the Decision Support System pharmacy data within 12 months of surgery was used to identify insulin-dependent diabetes in patients with an ICD-9 code for diabetes. The rCRI was analyzed as both an ordinal and categorical variable: low risk (1 point), moderate risk (2 points), or high risk ( 3 points).
Demographics and comorbidities were obtained from the VA National Patient Care Datasets or CMS inpatient, outpatient, and carrier base files. Death was obtained from the VA Vital Status File. Additional laboratory results and medication prescriptions were obtained from the VA Decision Support System. Demographics, co-morbidities, labs, and postoperative outcomes assessed by VASQIP were used to describe the secondary cohort of patients undergoing non-cardiac surgery without a history of cardiac stent who were matched to patients with a cardiac stent undergoing surgery in the VASQIP dataset.
For the nested case-control study, medical chart abstract was used to assess the exposure of preoperative antiplatelet medication management as well as to confirm the postoperative outcome for the patient. Separate abstraction forms were assigned for exposure and outcome so that an abstractor did not assess both for the same patient. Uncertainty of an exposure or outcome variable was adjudicated by 2 of the senior investigators.
For the provider survey, the survey design was based on the conceptual framework put forth by Kitson et al on successful implementation of evidence into practice.(Kitson, Harvey et al. 1998, Kitson, Rycroft-Malone et al. 2008) Information collected on the survey included provider demographics, awareness of and agreement with the 2007 ACC/AHA guidelines for dual anti-platelet management during the peri-operative period, characteristics of guideline dissemination at the facility level, risk perceptions of bleeding and stent thrombosis for both major and minor procedures, personal experience with peri-operative stent thrombosis, and anti-platelet therapy practice patterns. In order to assess practice patterns, the survey examined five different scenarios of a stented patient in need of subsequent surgery - 1) a patient without a stent, 2) BMS placed < 6 weeks prior to surgery, 3) BMS placed > 6 weeks prior to surgery, 4) DES placed < 12 months prior to surgery, and 5) DES placed > 12 months prior to surgery. The time periods chosen for the patient scenarios correspond to the time periods identified in the ACC/AHA guidelines for delaying elective surgery for each stent type (DES: 365 days and BMS: 30-45 days).(Fleisher, Beckman et al. 2007)
To address objective 1, determine incidence, univariate frequencies and bivariate statistics were used to describe the patient population. For all surgeries, differences by stent type, operation type, and operation complexity (work RVU) were examined using univariate frequencies and median months to surgery as well as with bivariate frequencies by time to surgery category. Chi-square tests or Wilcoxon's signed ranks tests were used to test for differences between categorical and continuous variables, respectively. Log-rank tests and Kaplan-Meier estimates were used to examine differences in time to first non-cardiac surgery by stent type.
To assess objective 2, determine factors associated with MACE, generalized additive models were used to determine the relationship between time from stent to surgery and MACE with adjustment for stent type, surgical characteristics, cardiac risk factors, and comorbid conditions. Generalized additive models were selected for this analysis to allow time between stent and surgery to be fit as a nonlinear term in assessing the relationship between surgical timing and MACE.(Wood 2006) The approximate p-values for spline terms are derived using a score test and algorithmically estimated degrees of freedom. To examine the relative contribution of variables in the adjusted models, we calculated the ANOVA chi-square for each variable minus its degrees of freedom ( 2-d.f.).(Harrell 2001) The statistical threshold for significance was set at p=0.05 for a two tailed test. In order to account for confounding by indication in choice of stent type, we conducted analyses using propensity score quintiles and inverse propensity weighting. We restricted this analysis to patients with stents placed after FY2003, when DES were widely available for implantation. Inverse probability weights were divided into quintiles and incorporated into our models.
All univariate and bivariate statistics were calculated using SAS v9.2 and generalized additive models used R package MGCV for spline models. Plots of unadjusted data were created with R package GGPLOT2(Wickham 2009) and smoothed trends were fitted using the loess algorithm. For the nested case/control study, univariate and bivariate statistics were calculated to examine differences in medication management by MACE. Odds ratios were calculated with conditional logistic regression to account for matched pairs.(Stokes, Davis et al. 1991)
To assess objective 3, univariate statistics in the form of frequencies and percentages were calculated to examine the study population in terms of demographics, guideline awareness and agreement, perceptions of risk, experience with perioperative stent thrombosis, and practice patterns. Bivariate statistics were calculated to examine differences by provider type, agreement with the guidelines, and experience with peri-operative stent thrombosis. Differences between categorical variables were compared using Chi-square and Fisher's Exact tests and between continuous variables using Wilcoxon Rank Sums tests. In order to determine the impact of patient scenario, provider specialty, guideline agreement, and experience with peri-operative stent thrombosis, generalized estimating equations were used to examine associations between these factors and reported practices for antiplatelet therapy management.
Objective 1: To determine the incidence of surgically treated disease in patients with cardiac stents in the two years following implantation.
Between fiscal year 2000 through fiscal year 2010, 102,678 distinct patients had at least one cardiac stent implanted at a VA facility. Of the 102,678 patients in our cohort, 19.0% underwent more than one stent placement; resulting in a total of 126,773 stent placements with follow-up available. The number of stents increased steadily from 6,665 in FY2000 to 15,971 in FY2010 and there was rapid adoption in the use of DES. Overall, 25,367 (24.7%) patients underwent at least one noncardiac surgical procedure in the ensuing 24 months after stent placement. Nearly two-thirds of the patients had stents implanted prior to the revision of the ACC/AHA guidelines and the rate of noncardiac surgery was higher before the guideline revision.(Hawn, Graham et al. 2012)
In FY2004, the first year after the DES was FDA approved, the incidence of surgery within 12 months was similar to BMS (17.2% vs. 18.6%, respectively). Although the incidence of surgery within 12 months remained stable after BMS, the incidence of surgery after DES steadily declined over the ensuing years to 6.6% in FY2010. Case reports of stent thrombosis with early cessation of antiplatelet therapy occurred in 2006 and the ACC/AHA guidelines were modified in late 2007. Accordingly, we observed that the overall incidence of noncardiac surgery within one year in the VA decreased from 16.7% for the FY2000-FY2007 period to 10.0% for the FY2008-FY2011 (p<0.0001). The decline in surgery during the first 12 months after publication of the guidelines is exclusively due to the decline that was observed for the DES cohort (14.9% pre vs 8.2% post) compared to the BMS cohort (18.0% pre vs 15.1% post).(Hawn, Graham et al. 2012, Graham, Singletary et al. 2014)
Objective 2: To estimate the attributable risk of type of cardiac stent on adverse peri-operative events.
Of the 124,844 PCI episodes of care occurring between fiscal year 2000 through fiscal year 2010, a total of 28,029 (22.5%) patients met study inclusion criteria and underwent 41,989 surgical procedures within 24 months. A total of 1,980 (4.7%) MACE occurred within 30 days of surgery; 1,170 MI or repeat revascularization without death, 141 MI or repeated revascularization with death and 669 death alone. In unadjusted analyses, MACE rates differed significantly by stent type; BMS (5.1%) vs. DES (4.3%, p<0.001). In the overall model of MACE, acuity of presentation was the most explanatory determinant, followed by markers of cardiac disease including recent MI or CHF and higher rCRI score, whereas stent type was not significantly associated with MACE and ranked 12th of the 12 variables. There was no significant interaction between stent type and time to surgery (p=0.56 for BMS and p=0.20 for DES).(Hawn, Graham et al. 2013)
Time from stent to surgery was correlated with MACE with higher rates observed for surgery closer to stent implantation, non-elective admission source, rCRI category and recent MI. After adjustment, the odds of a MACE for surgery between 6 weeks and 6 months following DES was lower than for BMS (ORadj=0.8, 95% CI 0.6-0.9) and not significantly different for surgery less than 6 weeks (ORadj=1.1, 95% CI 0.8-1.5) or more than 6 months following stent (ORadj=0.9, 95% CI 0.8-1.1).(Hawn, Graham et al. 2013)
To investigate the association between APT management around the time of surgery and MACE, we performed a case-control study on the subset of VA surgical procedures.(Hawn, Graham et al. 2013) Of the 369 abstracted VA cases, a MACE was confirmed in 284 (77.0%). There was no significant difference in the likelihood of being on dual APT prior to surgery (59.9% cases vs. 55.6% controls, p=0.43) or to have complete cessation of APT for at least 5 days (22.9% cases vs. 25.4% controls. p=0.49). In matched analyses, there was no association between complete APT cessation and adverse cardiac events (OR=0.86, 95% CI 0.57-1.29). Post-hoc power analyses indicated that our cohort had 80% power to detect an odds ratio of 1.68 with alpha of 0.05.
Additional Secondary Analyses
Attributable MACE Risk of Non-Cardiac Surgery in Patients with Cardiac Stents. This paper is under review and found that patients with coronary stents are at higher risk of adverse cardiac events out to one year following PCI compared with a matched surgical cohort of similar cardiac risk without stent history.
Attributable MACE Risk of Cardiac Stent in Patients with Non-Cardiac Surgery. This paper is in press with the Journal of the American College of Cardiology. When compared to a matched cohort of patients with coronary stents that did not undergo surgery, the incremental risk of surgery was approximately half of the total observed risk in the 30-day postoperative period.
Outcomes of Transfusion. This study is examining transfusion thresholds in post-operative patients with significant cardiac risk. The analyses are ongoing.
MI Timing and Postoperative MACE Risk. This study examines the risk of recent MI and whether the patient received pre-operative revascularization on the risk of post-operative MI. The paper is in the final phases of manuscript preparation.
Triple Antithrombotic Therapy and Outcomes in Patients Undergoing Noncardiac Surgery. This paper examines a unique subset of patients undergoing noncardiac surgery and describes the increased risk in this cohort. The manuscript is in the final phases of preparation.
Echocardiography in Patients with Cardiac Stents. This study examines facility level variation in the use of pre-operative echocardiography in risk assessment of patients undergoing non-cardiac surgery. It is in the final phases of data analyses.
Beta-Blocker and Statins. This study assesses peri-operative medication management and risk mitigation in our study cohort. The manuscript is in the final phases of preparation.
Failure to Rescue in Patients with an In-Hospital Myocardial Infarction. This study looks at risk factors for mortality following post-operative MI and specifically addresses whether PCI mitigates that risk. The manuscript is in the final phases of preparation.
Characteristics of PCI. This study examines post-operative adverse cardiac events by index PCI characteristics including vessel treated, number of stents deployed and other cath lab findings. The manuscript is in the final phases of preparation.
Objective 3: To describe variation in knowledge, attitudes and adherence to current guidelines among surgeons, anesthesiologists, cardiologists and primary care physicians regarding the timing of elective noncardiac operations after stent placement and antiplatelet therapy cessation for invasive procedures.
A total of 315 VA providers at 80 VA hospitals completed the online survey.(Graham, Maddox et al. 2013) Respondents were evenly distributed between the provider categories with 110 (37%) cardiologists (58 cardiologists and 52 interventional cardiologists), 94 (32%) anesthesiologists and 91 (31%) surgeons. Guideline awareness and agreement was high among all providers (93% and 82%, respectively), but varied significantly by provider type. Surgeons were significantly less likely to be aware of and to agree with the guidelines (78% and 64%, respectively) as compared to cardiologists (100% and 90%, respectively), interventional cardiologists (98% and 94%, respectively) and anesthesiologists (100% and 87%, respectively).
Reported provider guideline agreement was significantly associated with perception of risk and reported routine practice patterns. Providers that agreed with the guidelines, compared to those who were neutral or disagreed, were more likely to report that peri-operative stent thrombosis was difficult to manage (p<0.01) and bleeding complications were challenging to manage (p<0.01). They were also significantly less likely to stop all anti-platelet therapy when the surgery was within 6 weeks for BMS (p<0.01) or within 12 months for DES (p<0.01).(Graham, Maddox et al. 2013)
In adjusted models for practice patterns, patient scenario of stent type and timing of surgery (p<0.01) and provider type (p=0.01) were significantly associated with a provider reporting the continuation of dual antiplatelet therapy. Providers were far more likely to report the continuation of dual therapy for early surgeries regardless of stent type (BMS p<0.01; DES, p=0.03). For early surgeries, interventional cardiologists and anesthesiologists were significantly more likely to report the continuation of dual therapy as compared to surgeons. While guideline agreement was not a significant factor of continuing all antiplatelet therapy in any patient scenario after adjustment, personal experience with stent thrombosis was marginally associated with continuing dual therapy in patients with a DES in need of surgery beyond 12 months after stent placement (OR=2.3, 95% CI=1.0-4.9).(Graham, Maddox et al. 2013)
The major impact of this study has been the change in the ACC/AHA guidelines on management of patients with drug eluting stents undergoing non cardiac surgery. The 2014 guidelines now recommend delaying surgery for 6 months (formerly one year) and cite our publication as the reference for this change in recommendation.
Secondary knowledge gained from this study is that cessation of antiplatelet therapy for surgery was not significantly associated with post-operative adverse cardiac events. Additional important observations include that approximately 50% of the post-operative adverse event risk can be attributed to the underlying cardiac disease and not related to the surgical procedure (Holcomb JACC). That there was early dissemination and broad awareness of the ACC/AHA guidelines and the implementation of the guidelines led to fewer adverse cardiac events.(Graham, Singletary et al. 2014) A complete list of publications is provided.
External Links for this Project
NIH ReporterGrant Number: I01HX000371-01A1
Dimensions for VADimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.
If you have VA-Intranet access, click here for more information vaww.hsrd.research.va.gov/dimensions/
VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address. Search Dimensions for this project
DRA: Health Systems, Cardiovascular Disease
DRE: Treatment - Comparative Effectiveness
MeSH Terms: none