Long-Term Trends in Preoperative Cardiac Evaluation and Myocardial Infarction after Elective Vascular Procedures

Abstract

Introduction:

Vascular surgery has seen rapid increase in the use of less invasive endovascular therapies along with advancements in cardiac perioperative optimization in the past two decades. However, a recent NSQIP database study found no improvement in postoperative myocardial infarction (POMI) over a 10-year period in high-risk procedures. The national Society for Vascular Surgery Vascular Quality Initiative (VQI) registry provides a more in-depth characterization of vascular surgery procedures. Here we sought to evaluate long-term trends in POMI using VQI registry data for patients undergoing carotid endarterectomy (CEA), thoracic endovascular aortic repair (TEVAR), endovascular aortic repair (EVAR), open abdominal aortic aneurysm repair (oAAA), suprainguinal bypass (SIB), and infrainguinal bypass (IIB).

Methods:

A retrospective cohort study was performed using data on elective procedures from 2003-17. Procedures were subdivided by date of operation into 3-year era consecutive groups for subanalysis (2003-05, 2006-08, 2009-11, 2012-14, 2015-17). The incidence of POMI, preoperative risk factors (including individual patient VQI cardiac risk indices (CRI)), and demographics were determined over time.

Results:

A total of 227,837 elective procedures were identified: CEA (n=88,805, 39.0%), TEVAR (n=7,494, 3.3%), EVAR (n=34,376, 15.1%), oAAA (n= 7,568, 3.3%), SIB (n=11,354, 5.0%), and IIB (n=34,661, 15.2%). Across all procedures, the overall rate of POMI was 1.3%. POMI rates from 2003-05 to 2015- 17 for CEA decreased from 0.9% to 0.7% (p=0.21), EVAR from 2.0% to 0.7%, p= 0.003, oAAA from 6.8% to 5.1% (p=0.12), and IIB from 3.8% to 2.4% (p=0.003). SIB POMI decreased from 3.06% to 2.95%, p=0.85 from 2009-17. While POMI after TEVAR increased from 2.40% to 2.56% from 2009-17, p=0.91. Over these same time periods, only EVAR and IIB had a reduction in CRI (p=0.059 and p<0.001, respectively). CEA, EVAR, IIB, and oAAA all showed a significant (p<0.001) increase in preoperative statin use.

Conclusions:

Except for TEVAR, the incidence of POMI has remained unchanged or decreased over the past 15 years in VQI registries. Patients undergoing IIB and EVAR demonstrated decreases in POMI rates that correspond with a reduction in CRI and increased preoperative statin use. CEA and SIB had no significant change in POMI rates nor CRI. The etiology of decreased POMI rate is uncertain, but increasing statin use, patient-specific factors, and patient selection for procedures may be important drivers of this improvement.

Introduction:

Postoperative myocardial infarction (POMI) is one of the more common complications after vascular procedures. Risk for POMI is always considered when deciding to pursue elective interventions. Given the impact of POMI on long-term survival¹, determining each individual’s risk for this complication is imperative prior to proceeding with elective interventions. The American College of Cardiology/American Heart Association (ACA/AHA) guidelines report the risk of POMI and cardiac death is often greater than 5%² in patient’s undergoing aortic, peripheral artery or other major vascular surgery intervention, and that endovascular interventions significantly decrease the rate of major adverse cardiac events³.

There are currently several calculators used to stratify a patient’s risk of adverse cardiac event. One of the most widely used risk calculators is the Revised Cardiac Risk Index (RCRI). However, the predictive value of this index in patients undergoing vascular interventions has been questioned because of the underestimation of perioperative cardiac events in that patient population⁴. More recently, the Vascular Quality Initiative-Cardiac Risk Index (VQI-CRI) was developed to specifically risk stratify vascular surgery patients and has demonstrated improved predictive values⁵. The ACA/AHA has also published detailed guidelines on appropriate individualized preoperative cardiac workup⁶ which further assists the vascular surgeon in performing a more comprehensive risk stratification. Patients at high risk for suffering a cardiac event can be further risk stratified by cardiac stress testing^7,8 but studies have failed to show a benefit from preemptive cardiac intervention based on these findings⁹.

Over the past decade, the field of vascular surgery has seen the rapid increase in use of endovascular therapies^10,11 along with advancements in cardiac perioperative optimization^2,6.A recent study looking at high-risk open vascular procedures in both elective and emergent cases found no difference in postoperative myocardial infarction (POMI) over the past decade using the NSQIP database¹². This finding is surprising given the improvements in perioperative risk stratification and advances in medical perioperative optimization. It is unclear from the NSQIP registry if the observed stasis in POMI rates is related to migration of high-risk surgical candidates to endovascular therapy. Our present study uses the VQI registry to evaluate long-term trends in rates of POMI and associated trends in pre-operative cardiac workup and anticipated cardiac risk of patients undergoing procedural intervention as reflected in the CRI.

Methods:

Database

The Vascular Quality Initiative (VQI) was established to evaluate outcomes after vascular procedures under the auspices of a patient safety organization (PSO) as designated by the Agency for Healthcare Quality and Research. The VQI was initially formed as a regional registry in 2002, but has since evolved to include over 700 participating centers¹³. VQI data entry occurs for all consecutive procedures for each registry in a prospective manner and data entered are cross-referenced with hospital claims to evaluate for missing procedures and ensure complete reporting. Data collection began in 2003 with carotid and abdominal aortic procedures and expanded in 2009 to include thoracic aortic and lower extremity bypass procedures. All data collected are matched to the Social Security Death Index to allow long-term assessment of mortality.

This project was screened and approved by the VQI Research Advisory Council and was reviewed and approved by the Institutional Review Board at the University of Alabama at Birmingham.

Cohort Selection

A retrospective cohort study was performed using data from elective procedures within the VQI database from 2003 to 2017. Patients within the following VQI registries were included: carotid endarterectomy (CEA), thoracic endovascular aortic repair and complex EVAR (TEVAR), endovascular aortic repair (EVAR), open abdominal aortic aneurysm repair (oAAA), supra inguinal bypass (SIB), and infra inguinal bypass (IIB) were included. Patients undergoing non-elective procedures or with missing post-operative MI data were excluded from analysis. Within each cohort, duplicate entries were condensed to ensure no procedural repetitions; however across cohorts duplicate patients were allowed given that the typical vascular patient has multilevel disease and may require several different interventions. On average there was a 1.7% overlap of patients across cohorts. In order to identify trends over time, patients within each procedural registry were subdivided by year of operation into the following time frames (2003-2005, 2006-2008, 2009-2011, 2012-2014, 2015-2017), depending on the age of the registry and availability of data. Not all registries have existed over the duration of the VQI, and therefore not all procedures have the same number of eras available for analysis. TEVAR/complex EVAR and suprainguinal bypass registries were not developed until 2009; therefore, the time frames for analysis in these two procedural groups begin at 2009 rather than 2003.

Covariates Examined

Patient demographics including the cardiac risk index derived from the VQI model coefficient⁵, peri-procedural characteristics, and post-operative cardiac outcomes were collected for each time of the five designated frames. Cardiac risk assessment was performed for each patient using a previously developed cardiac risk index (CRI)⁵. The CRI model was developed using procedure specific VQI datasets. CRI was only created and validated for CEA, EVAR, oAAA, SIB, and IIB procedural registries and was therefore not applied to the TEVAR registry. The average area under curve for each procedure specific cohort CRI was 0.705. Cardiac workup was evaluated by determination of patient undergoing either a preoperative stress testing or having documented ejection fraction assessment in the aortic modules where this variable is reported. Post-operative MI was defined per VQI standards as either: isolated troponin elevation, EKG changes along with abnormality in cardiac biomarkers (CKMB or troponin), or clinical evidence of an MI along with troponin abnormality in cardiac biomarkers (CKMB or troponin)^5,14. This definition of POMI has been used throughout VQI literature and has not changed throughout the years, giving consistency for outcome comparisons within vascular surgery literature. Of note the NISQIP definition of POMI does not include isolated troponin elevation, making direct comparisons with NIQIP outcomes difficult¹⁵.

Statistical Analysis

Patients characteristics and outcomes were stratified by procedure type and then divided into groups based on operative year eras. Patient preoperative characteristics and outcome variables were compared using t-tests for continuous variables, χ2 for categorical variables, and logistic regression models were created to adjust for confounding variables and allow for subgroup analysis. Statistical significance was determined using a two-tailed test with an alpha of <0.05. Statistical analysis was performed using SPSS (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.).

Results:

A total of 227,837 procedures were identified from 2003-2017: CEA (n=88,805, 39.0%), EVAR (n=34,376, 15.1%), IIB (n=34,661, 15.2%), oAAA (n= 7,568, 3.3%), SIB (n=11,354, 5.0%), and TEVAR (n=7,494, 3.3%). Overall, the rate of POMI was 1.3% over the entire data set. Missing data for POMI was <1% for all procedure groups over time. Figure 1 shows the trends in POMI across all procedural subgroups. Univariate analysis between demographic factors, over all time periods, between individuals with and without POMI are described in Table 1.

Figure 1:

Postoperative myocardial infarction rates over time.

Table 1:

Patient Preoperative Demographics

Variable	CEA (N=88,805)		TEVAR (N=7,494)		EVAR (N=34,376)		oAAA (N=7,568)		SIB (N=11,354)		IIB (N=34,661)
	POMI n=7,104	No POMI n=81,709	POMI n=188	No POMI n=7,305	POMI n=447	No POMI n=33,929	POMI n=431	No POMI n=7,137	POMI n=340	No POMI n=11,014	POMI n=1,039	No POMI n=33,622
Age (year)	70.4±9.3	72.6±8.7	72.9±8.8	69.8±11.4	77.2±8	73.4±8.6	72.1±7.6	69.3±8.4	67.7±9.7	63.4±10.5	71±9.6	66.7±10.9
Male (%)	57%	60%	66.30%	65.10%	74%	81.50%	73.90%	73.90%	57.80%	59%	65.50%	69%
Baseline Disease
HTN (%)	92%	89.2%%	91.60%	89.10%	89.4%%	83.3%%	95.40%	88.00%	91.30%	82%	89.50%	84%
COPD (%)	24%	23%	36.80%	31.60%	41%	32.90%	25.80%	26.70%	40.20%	35%	41.40%	33%
CAD (%)	46.30%	28.10%	42.60%	23.60%	49.80%	29.60%	52.70%	49.80%	43.40%	25.90%	37.40%	26.30%
CHF (%)	19.80%	10.40%	21.10%	12.60%	24.70%	11.90%	32.10%	15.80%	20.20%	10.20%	15.70%	7.10%
PreOp Medications
Statin (%)	80.40%	81.20%	74.70%	65.00%	75.30%	70.60%	74.30%	70.70%	80.10%	72.20%	76.90%	71.80%
APA (%)	90.90%	88.50%	68.40%	64%	69%	77.70%	74.90%	68%	83.10%	77%	82.70%	79%
BB (%)	71.3	55.9	71.60%	67.90%	68.20%	55.90%	77%	58.10%	70.20%	51.50%	70.40%	62.50%
PreOp Imaging
Abnormal EF (%)			15%	22.90%	25%	36.30%	14.40%	21.40%
Abnormal Stress Test (%)	16.90%	8.10%	20%	6.60%	19.50%	9.40%	21.80%	11.40%	21.80%	9.90%	19.50%	9.40%

HTN=Hypertension, COPD=Chronic Obstructive Pulmonary Disease, CAD=Coronary Artery Disease, CHF=Congestive Heart Failure, APA=Anti-Platelet Agent, BB=Beta-Blocker, EF=Ejection Fraction

Carotid Endarterectomy (CEA)

A total of 88,805 CEA procedures were recorded within the VQI registry between 2003 and 2017. Demographic changes across time are presented in Table 2. Patients undergoing a CEA in the most recent era (2015-2017) compared to the earliest era (2003-2005) were significantly older (71 years vs. 69 years, p=0.004) , with higher rates of baseline diabetes (36% vs. 29.8%, p<0.001) and CHF (11.1% vs 7.4%, p<0.001), and were more likely to be receiving appropriate preoperative statin therapy (82.6% vs 65.1%; p<0.001). In the most recent era (2015-2017), patients were less likely to be on antiplatelets (89.4% vs. 91.1%, p=0.007), carry a diagnosis of CAD preoperatively (27.6% vs. 38.3%, p<0.001), and be a current smoker (33% vs 25%, p<0.001, Figure 4). Relevant changes to key preoperative medication usage changes within the CEA cohort across time are shown in Figure 4.

Table 2:

Preoperative Patient Demographics Long-Term Trends

Variable	Timeframe	CEA (n=88,805)	TEVAR (n=7,494)	EVAR (n=34,376)	oAAA (n=7,568)	SIB (n=11,354)	IIB (n=34,661)
AGE mean(SD)	Early Era	69.1 (9.2)	68.8 (12.3)	73.5 (7.7)	70.8 (7.9)	63.0 (10.6)	68.2 (11.4)
	Late Era	70.5 (9.4)	70.0 (11.4)	73.2 (8.7)	69.4 (8.3)	63.6 (10.5)	66.7 (11.0)
	p-value	0.04	0.32	0.03	0.35	0.24	<0.001
Male Gender (%)	Early Era	58.6	55.2	80.7	73.1	60.7	69.2
	Late Era	60.6	65.6	81.7	74.6	59.5	69.5
	p-value	0.07	0.16	0.65	0.47	0.47	0.81
Diabetes (%)	Early Era	29.8	14.4	17.8	15.2	22.9	51.2
	Late Era	36	16.7	20.6	17.6	27.6	49.2
	p-value	<0.001	0.5	0.2	0.19	<0.001	0.22
Preop CABG +/− PCI (%)	Early Era	34.4	21.6	29.5	30.8	23.7	34.1
	Late Era	35.4	27.1	34.7	31	27.7	35.7
	p-value	0.36	0.17	0.04	0.92	0.01	0.3
Preop APA (%)	Early Era	91.1	59.2	77.3	78.4	77.8	78.6
	Late Era	89.4	64.5	68.5	64.8	77.9	80.5
	p-value	0.01	0.22	<0.001	<0.001	0.94	0.11
Preop Statin (%)	Early Era	65.1	63.2	56.8	69.8	73.9	52.6
	Late Era	82.6	64.5	70.9	71.9	74.1	73.4
	p-value	<0.001	0.77	<0.001	<0.001	0.9	<0.001
Preop CAD (%)	Early Era	38.3	22.4	42.6	35.9	26	42.2
	Late Era	27.6	24.2	29.5	26.9	27	31.7
	p-value	<0.001	0.6	<0.001	<0.001	0.51	<0.001
Preop CHF (%)	Early Era	7.4	8.8	13.3	7.4	9.4	17.6
	Late Era	11.1	12.7	12.2	8.1	10.4	16.9
	p-value	<0.001	0.19	0.54	0.58	0.34	0.61
Preop COPD (%)	Early Era	23.4	27.2	41.6	39.1	33.9	29.6
	Late Era	23.1	33.2	33.6	33	36.1	27.8
	p-value	0.77	0.16	0	0.01	0.16	0.22
Preop Dialysis (%)	Early Era	0.4	2.4	1.1	0.6	2	8.6
	Late Era	1.1	2.1	1	0.5	1	5.1
	p-value	0.17	0.82	0.11	0.69	0	0.58

Early Era = 2003-2005, Late Era = 2015-2017, CABG=Coronary Artery Bypass Graft, PCI=Percutaneous Coronary Intervention, APA=Antiplatelet Agent, CAD=Coronary Artery Disease, CHF= Congestive Heart Failure, COPD=Chronic Obstructive Pulmonary Disease.

Figure 4:

Preoperative Patient Risk Factors and Medication Use Over Time

For patients undergoing CEA, the overall rate of POMI was 0.8%. The rate of POMI in CEA patients did not significantly decrease across the studied timeframe (0.9% to 0.7%, p=0.21) (Figure 1). The average cardiac risk index for CEA patients was 0.9%, which did not significantly decrease overtime (Figure 2). The use of preoperative stress testing to screen patients undergoing CEA significantly decreased (42.7% to 34.4 %, p<0.001) (Figure 3).

Figure 2.

Cardiac Risk Index Over Time.

Figure 3:

PreOperative Cardiac Imaging Usage in Early and Late Procedural Groups

Thoracic Endovascular Aneurysm Repair (TEVAR)

A total of 7,494 TEVAR procedures were recorded within the VQI registry between 2009 and 2017. Demographic changes across time are presented in Table 2. There were no statistically significant changes in pre-operative statin use, rates of CAD, or smoking within the TEVAR cohort (Figure 4). Patient preoperative variables across time were not significantly different in any time period.

For patients undergoing TEVAR, the overall rate of POMI was 2.5%. The rate of POMI in TEVAR patients did not significantly change across the studied timeframe (2.40% to 2.56%, p=0.91) (Figure 1). Similarly, no significant change in preoperative cardiac imaging (stress test, or ejection fraction estimate known from echo, nuclear scan, or cardiac catheterization) utilization was noted (p=0.478, Figure 3).

Endovascular Aneurysm Repair (EVAR)

A total of 34,376 EVAR procedures were documented within the VQI registry between 2003 and 2017. Demographic changes within the EVAR cohort across time are shown in Table 2. Patients undergoing EVAR within the most recent era (2015-2017) are younger (73.3 years vs. 73.5 years, p=0.032), with lower rates of preoperative CAD (29.5% vs. 42.6%), and were more likely to be on appropriate preoperative statin therapy (70.9% vs. 56.8%, p<0.001). These patients were overall less likely to be on preoperative antiplatelet agent (68.5% vs. 77.3%, p<0.001) and to have preoperative COPD (33.6 % vs 41.6%, p=0.002). There was no significant change in rate of smoking within the EVAR cohort. Relevant changes to key preoperative medication usage changes within the EVAR cohort across time are shown in Figure 4.

For patients undergoing EVAR, the overall rate of POMI across all timeframes was 1.3%. The rate of POMI decreased significantly across the studied timeframe (2% to 0.7%, p=0.003) (Figure 1). Correspondingly, the cardiac risk index showed a similarly significant decrease across the studied timeframe (0.50% to 0.43%, p= 0.059, Figure 2). These changes were accompanied by increasing use of preoperative cardiac imaging (stress test, or ejection fraction estimate known from echo, nuclear scan, or cardiac catheterization) (52.0% to 61.3%, p<0.001, Figure 3).

Open Abdominal Aortic Aneurysm (oAAA)

A total of 7,568 oAAA procedures were recorded within the VQI registry between 2003 and 2017. Demographic changes within the oAAA cohort across time are shown in Table 2. Patients undergoing oAAA most recently (2015-2017) were more likely to be younger (69.4 years vs. 70.8 years, p=0.038), with lower rates of preoperative COPD (33% vs. 39.1%, p=0.008), and less likely to be on preoperative APA (64.8% vs 78.4%, P<0.001). There was no significant change in rate of pre-operative smoking within the oAAA cohort (Figure 4). Relevant changes to key preoperative medication usage changes within the oAAA cohort across time are shown in Figure 4.

For patients undergoing an oAAA procedure, the overall rate of POMI across all time periods was 5.7%. Across the studied timeframe the rate of POMI decreased from 6.8% to 5.1% but this change did not reach statistical significance (p=0.12). The calculated cardiac risk index from the earliest time period (2012-2014) showed no significant change (6.51% to 6.67%, p=0.39) (Figure 2). From the 2003-2005 to 2015-2017 era, use of preoperative cardiac imaging (stress test, or ejection fraction estimate known from echo, nuclear scan, or cardiac catheterization) significantly decreased (86.2% to 79.1%, p<0.001) (Figure 3).

Supra-Inguinal Bypass (SIB)

A total of 11,354 SIB procedures were recorded within the VQI registry between 2009 and 2017. Demographic changes within the SIB cohort across time are shown in Table 2. Patients undergoing SIB recently (2015-2017) had higher incidence of diabetes (27.6% vs 22.9%, p<0.001) and were more likely to have undergone CABG and/or PCI (27.7% vs 23.7%, p=0.008). Patients within the SIB cohort were significantly less likely to be current smokers (65% vs 55%, p<0.001)(Figure 4).

For patients undergoing a SIB procedure, the overall rate of POMI across all time periods was 3%. From the 2009-2011 to 2015-2017 eras, the rate of POMI in the SIB cohort did not significantly change (3.06% to 2.95%, p=0.85) (Figure 1). Similarly, the calculated CRI did not significantly change between these eras (3.16% to 3.09%, P=0.326, Figure 2). The use of preoperative cardiac stress testing significantly decreased from 54.0% to 48.0% (p<0.001, Figure 3) over the same eras.

Infra-inguinal bypass (IIB)

A total of 34,661 IIB were recorded within the VQI registry between 2003 and 2017. Demographic changes within the SIB cohort across time are shown in Table 2. Patients undergoing IIB in the most recent era (2015-2017) were younger (66.7 ± 11.4 years vs. 68.2 ± 11.4 years, p<0.001), with lower rates of preoperative CAD (31.7% vs. 42.3%, p<0.001), and were more likely to be on appropriate preoperative statin therapy (73.4% vs 52.6%, p<0.001). There was no significant change in the rate of smoking within the IIB cohort. Relevant changes to key preoperative medication usage changes within the IIB cohort across time are shown in Figure 4.

For patients undergoing an IIB procedure, the overall rate of POMI across all time periods was 3%. Across the studied timeframe, the rate of POMI decreased significantly (3.8% to 2.4%, p=0.003) (Figure 1). The decrease in POMI rate was associated with a decrease in the calculated cardiac risk index (2.9% to 2.3%, p<0.001) (Figure 2). During this same period, there was no significant change in the use of preoperative cardiac stress testing (36.1% to 35.8%, p=0.82) (Figure 3).

Discussion:

Over the past 15 years the field of vascular surgery has seen changes in perioperative cardiac optimization^6,16 along with countless advancements in endovascular therapies^17–19. The impact of these advancements on altering the risk benefit ratio for vascular surgery patients has not been well elucidated. Surprisingly, a recent publication evaluating the NSQIP database showed no improvement in rates of POMI from 2003-2015 among patients undergoing open aortic or infra-inguinal bypass surgery⁵, which may be related to a migration of patients from high risk open operations to lower risk minimally invasive procedures. In our current study, we evaluated the temporal changes of POMI over the past 15 years, the associated trends in patient risk factors, preoperative medications, cardiac workup and concomitant expected cardiac risk associated.

Importantly, our study has shown that the rates of POMI within patients undergoing EVAR and IIB in the VQI registry has significantly decreased over the past 15 years. Though not statistically significant, the CEA, oAAA, and SIB procedure groups showed a trend towards decreased POMI rates. The TEVAR cohort was the only cohort to show an increase in POMI rates, however this change did not reach statistical significance. The POMI rates for each cohort across each time period are shown in Figure 1. The decrease in POMI rates were associated with reductions in preoperative cardiac risk factors, calculated cardiac risk index and increased use of preoperative medications for optimization.

The significant decrease in rates of POMI rates in EVAR and IIB correlate with similar significant reductions in cardiac preoperative risk factors, calculated cardiac risk index, and improved trends in perioperative medical optimization including statin therapy. Within all cohorts, except for TEVAR, the rate of preoperative CAD was significantly lower within the newer era procedures (Figure 4). This reduction in baseline coronary disease was appropriately associated with significant decreases in the calculated cardiac risk index in both the EVAR and IIB cohorts and non-significant but still down-trending rates in all other cohorts (Figure 3). Within the infrainguinal bypass cohort (IIB), we surmise that the observed reduction in cardiac risk may be due to the shift of high-risk patients away from open bypass surgeries to catheter-based interventions. We suspect that a similar mechanism of change has occurred in the EVAR population. Historically EVAR was reserved for high risk patients that were poor surgical candidates and therefore likely to have significant baseline coronary disease, but the treatment paradigm has shifted and EVAR is now widely used for treatment in high and low risk patient populations^18,19. Improved perioperative medical optimization may also account for stasis or decline of POMI in selected modules. The use of preoperative statin therapy was higher among all the studied cohorts, but was significantly higher within the CEA, EVAR, oAAA, and IIB (Figure 4). While details of anesthetic regimens are limited in VQI, it is possible that the observed trends could also be attributed to improved intra-operative anesthesia techniques and resuscitative measures and this could be a further area of focus. Although we are unable to prove causation, we postulate that the improved rates of POMI are caused in part by improved patient selection and medical optimization.

The associated changes in preoperative imaging between procedure groups are possibly explained, at least in part, by changing treatment paradigms. As there has been migration of patients from oAAA towards EVAR there have been opposing effects on preoperative utilization of advanced cardiac evaluation such as stress tests and echocardiography. oAAA has shown significant decreases in advanced cardiac imaging whereas EVAR has seen a significant increase. It is unclear if these changes in preoperative advanced cardiac imaging are due to secular shifts in typical preoperative evaluation for the procedure or if cardiac imaging is a major driver behind what procedure is offered (i.e. oAAA versus EVAR). We theorize changes in preoperative cardiac imaging is impacting the migration of patients away from higher risk oAAA to lower risk EVAR. IIB showed a non-significant decrease in preoperative cardiac imaging and CEA had a significant decrease in preoperative cardiac imaging utilization. Overall, the reasoning behind these secular shifts in cardiac imaging is unclear and will require further inquiry.

Limitations:

A major limitation to this study is the lack of a true denominator for the patients in any specific procedural registry. It is possible that any trends that we see in higher or lower cardiac risk patients in open/endovascular procedures may be unrelated to a shift from one procedural paradigm to another and could be from changes in the pool of individuals being considered for any procedural intervention. Additionally, the effect of region/center/surgeon level variation may be important but is not feasibly addressed in this manuscript. We also anticipate that some patients will undergo multiple interventions over many of the studied years and among the different procedure groups. The total number of patients within these groups we suspect is small but cannot be evaluated with the deidentified VQI dataset.

Conclusion:

The incidence of POMI has remained unchanged or decreased over the past 15 years in all procedural VQI registries except for the TEVAR registry. The trends seen in POMI were accurately reflected in changes to the calculated CRI. Specifically, patients undergoing IIB and EVAR have seen decreased POMI rates that correspond with reduced CRI. The decrease in POMI rates were associated with decreased rates of preoperative CAD in all groups except for TEVAR, and with increased use of preoperative statins. While this study does not demonstrate a causative etiology of decreased POMI, it has demonstrated that increasing statin use, patient-specific factors, and patient selection serve as important drivers of this improvement.