Hybrid procedure included both endovascular revascularization and surgical revascularization during hospitalization. ICD-10-CM indicates International Statistical Classification of Diseases, Tenth Revision, Clinical Modification; NRD, Nationwide Readmissions Database.
A, Kaplan-Meier graphs showing major amputation in patients with critical limb ischemia. Compared with SR, ER had a slightly higher risk of major amputation at 6 months (997 of 10 090 [9.9%] vs 869 of 10 318 [8.4%]; hazard ratio [HR], 1.18; 95% CI, 1.08-1.29; P = .001). The HR is calculated from competing risk regression analysis. B, Kaplan-Meier graphs showing mortality among patients with critical limb ischemia. Compared with SR, ER had similar all-cause mortality at 6 months (517 [4.7%] vs 490 [4.4%]; HR, 1.06; 95% CI, 0.93-1.20; P = .39). C, Kaplan-Meier graphs showing unplanned all-cause readmission among patients with critical limb ischemia. Risk of unplanned readmission at 6 months was similar for ER and SR (4278 of 10 950 [39.1%] vs 4313 of 10 912 [39.5%]; HR, 0.98; 95% CI, 0.94-1.02; P = .39).
A, Forest plot comparing ER and SR for major amputation in different subgroups. Sex, hospital procedure volume, and diabetes showed interaction with ER. B, Spline graph for major amputation comparing ER and SR by annual procedure volume (P = .01 for interaction). The difference in risk of major amputation between the 2 procedures decreased as the procedure volume of a hospital increased. ER indicates endovascular revascularization; ESKD, end-stage kidney disease; HR, hazard ratio; and SR, surgical revascularization.
eTable 1. ICD-10 Codes of Variables Used for Cohort Selection
eTable 2. ICD-10 Procedure Codes of Devices and Procedures Used for Endovascular Revascularization
eTable 3. ICD-10 Codes of Variables Used for Clinical Presentation and Comorbidities
eTable 4. ICD-10 Codes for Outcomes
eTable 5. Details About Time-to-Event Analysis
eTable 6. Studies Comparing Endovascular vs Surgical Revascularization in Patients With Critical Limb Ischemia
eFigure. Balance of Covariates Between 2 Revascularization Strategies
Customize your JAMA Network experience by selecting one or more topics from the list below.
Majmundar M, Patel KN, Doshi R, et al. Comparison of 6-Month Outcomes of Endovascular vs Surgical Revascularization for Patients With Critical Limb Ischemia. JAMA Netw Open. 2022;5(8):e2227746. doi:10.1001/jamanetworkopen.2022.27746
What is the difference between endovascular and surgical revascularization regarding in-hospital safety, 6-month major amputation, and mortality for patients with critical limb ischemia?
In this cohort study of 11 106 matched pairs, endovascular revascularization was associated with an 18% higher risk of major amputation at 6 months and a 17% lower risk of in-hospital safety outcomes compared with surgical revascularization, with no difference in mortality between the 2 groups at 6 months. The difference in major amputation outcome was not observed in high-volume centers.
Findings suggest that endovascular revascularization has a better safety profile than surgical revascularization, without any difference in mortality; hence, procedure selection should be a shared and informed decision with the patient.
The Bypass Versus Angioplasty for Severe Ischemia of the Leg randomized controlled trial showed comparable outcomes between endovascular revascularization (ER) and surgical revascularization (SR) for patients with critical limb ischemia (CLI). However, several observational studies showed mixed results. Most of these studies were conducted before advanced endovascular technologies were available.
To compare ER and SR treatment strategies for 6-month outcomes among patients with CLI.
Design, Setting, and Participants
This retrospective, population-based cohort study used the Nationwide Readmissions Database to identify 66 277 patients with CLI who underwent ER or SR from January 1, 2016, to December 31, 2018. Data analyses were conducted from January 1, 2022, to February 8, 2022. A propensity score with 1:1 matching was applied. Patients with CLI who underwent ER or SR were identified, and those with missing information on the length of stay and/or younger than 18 years were excluded.
Endovascular or surgical revascularization.
Main Outcomes and Measures
The primary outcome was a major amputation at 6 months. Significant secondary outcomes were in-hospital and 6-month mortality and an in-hospital safety composite of acute kidney injury, major bleeding, and vascular complication. Subgroup analysis was conducted for major amputation in high-volume centers.
A total of 66 277 patients were identified between 2016 and 2018 who underwent ER or SR for CLI. The Nationwide Readmissions Database does not provide racial and ethnic categories. The mean (SD) age of the cohort was 69.3 (12) years, and 62.5% of patients were male. A total of 54 546 patients (82.3%) underwent ER and 11 731 (17.7%) underwent SR. After propensity score matching, 11 106 matched pairs were found. Endovascular revascularization was associated with an 18% higher risk of major amputation compared with SR (997 of 10 090 [9.9%] vs 869 of 10 318 [8.4%]; hazard ratio, 1.18; 95% CI, 1.08-1.29; P = .001). However, no difference was observed in major amputation risk when both procedures were performed in high-volume centers. Endovascular revascularization and SR had similar mortality rates (517 of 11 106 [4.7%] vs 490 of 11 106 [4.4%]; hazard ratio, 1.06; 95% CI, 0.93-1.20; P = .39). However, the ER group had a 17% lower risk of in-hospital safety outcomes compared with the SR group (2584 of 11 106 [23.3%] vs 2979 of 11 106 [26.8%]; odds ratio, 0.83; 95% CI, 0.78-0.88; P < .001).
Conclusions and Relevance
The results of this study suggest that ER was safer, without any difference in mortality, but ER was associated with an increased risk of major amputation compared with SR. However, the risk of major amputation was similar when both procedures were performed at high-volume centers.
Peripheral artery disease affects more than 230 million people worldwide, and approximately 11% of patients with the disease present with critical limb ischemia (CLI).1,2 Hospitalizations for CLI increased significantly from 2011 to 2017 and are associated with higher rates of all-cause readmissions (27.1% at 30 days and 56.6% at 6 months).3,4 Such hospitalizations cost approximately $4.2 billion annually, with an additional 30-day readmission cost of $624 million.5 Surgical revascularization (SR) has been the mainstay therapy for CLI for decades; however, endovascular revascularization (ER) is emerging as an effective alternative therapy for CLI management. The introduction of newer drug-eluting stents, drug-coated balloons, pedal arch reconstruction, venous arterialization, and atherectomy devices has led to a steady increase in the success rate for endovascular interventions, given its advantages of better short-term outcomes and less resource use.3
According to 2016 American Heart Association/American College of Cardiology guidelines, SR is preferred (Class I) for patients with a common femoral artery lesion, a long segment lesion, and lesions at multiple anatomic levels in an artery, whereas ER is considered (Class IIa) for patients with femoropopliteal disease.6 The recommendations are mainly based on the Bypass Versus Angioplasty for Severe Ischemia of the Leg (BASIL) trial that showed similar amputation-free survival and mortality among patients between ER and SR.7 However, it is not easy to draw firm conclusions from the BASIL trial because it included balloon angioplasty without drug coating in the endovascular approach. Since the trial concluded, new endovascular therapies have emerged, and there is a paucity of data comparing new ER techniques with SR. The ongoing BASIL-3 and Best Endovascular vs Best Surgical Therapy in Patients With Critical Limb Ischemia trials will provide more clarity on the choice of revascularization. With this background, we aimed to compare ER and SR treatment strategies and analyze 6-month outcomes among patients with CLI from the most recently available Nationwide Readmissions Database (NRD).
For this cohort study, we obtained our study population from the NRD between January 1, 2016, and December 31, 2018. The NRD is a national, all-payer database sponsored by the Healthcare Cost and Utilization Project, a health care body established by the Agency for Healthcare Research and Quality. The NRD includes in-hospital and 1-year readmission data derived from US hospitals in 28 geographically dispersed states and is designed to represent approximately 58.7% of all US hospitalizations.8 The NRD uses a deidentified unique number for every patient, tracking patients and determining readmissions across hospitals within a calendar year. The present study was deemed exempt by the Indiana University institutional review board because the database contained deidentified data sets with prior ethical committee approval; thus, the need for informed consent was waived by the board. This study followed the reporting guidelines specified by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.9
We queried the NRD with the International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes to identify all adults (n = 825 896) who had a primary or secondary peripheral artery disease diagnosis. We extracted hospitalizations with CLI (n = 757 550), using appropriate ICD-10-CM codes as a primary or secondary diagnosis. The codes used to identify the cohort in this study have been summarized in eTable 1 in the Supplement and used in a previous study.3 We excluded patients younger than 18 years and missing length of stay (LOS) and derived the index hospitalizations of patients with CLI (n = 600 096). The flow diagram for patient selection has been summarized in Figure 1.
We used NRD variables for patients’ age, sex, type of admission (elective or nonelective), admission day (weekday or weekend), the primary payer, and median household income for patients’ zip code.10 We used the Elixhauser comorbidity variables to include hypertension, diabetes, hyperlipidemia, valvular heart disease, chronic heart failure, chronic pulmonary disease, chronic kidney disease, chronic liver disease, anemia, alcohol abuse, drug abuse, obesity, coagulopathy, cancer, fluid-electrolyte disorders, and depression.11 ICD-10-CM codes were used to define the history of nonadherence, smoking, stroke or transient ischemic attack, ischemic cardiomyopathy, carotid artery disease, percutaneous coronary intervention, coronary artery bypass graft, atrial fibrillation, and end-stage kidney disease. For clinical presentation, we applied ICD-10-CM codes to extract patients with rest pain, ulcer, osteomyelitis, gangrene, chronic total occlusion, diffuse atherosclerosis, nontraumatic ischemic muscle infarct, bacteremia or sepsis, history of lower extremity amputation, reduced mobility, and oxygen dependence to understand the patients’ level of severity. The ICD-10-CM codes used to identify these variables are summarized in eTable 3 in the Supplement. We used the NRD’s hospital variables to identify hospital size according to the number of beds, teaching status, hospital location, and procedure volume. The procedure volume at each hospital was calculated by adding all the admissions with ER or SR separately. Hospitals were divided into quartiles (first, second, third, and fourth quartile) to ensure an equal number of procedures. Hospitals in the fourth quartile were considered high-volume hospitals (median number of procedures, 52 [IQR, 43-73]), and those in the first to third quartiles were considered low-volume hospitals (median number of procedures, 16 [IQR, 9-26]).
We divided patients with CLIs into 2 cohorts, those who underwent ER and those who underwent SR. We excluded duplicates and patients who had both procedures (hybrid procedures; n = 9373) during the same hospitalization to arrive at the final cohort of index admissions of patients with CLI who underwent either ER or SR (N = 66 277). We applied ICD-10-CM procedure codes to derive ER and SR (eTable 1 in the Supplement) and types of devices and procedures used for ER (eTable 2 in the Supplement).
The study’s primary outcome was cumulative incidence of major amputation at 6 months after hospital discharge. In-hospital amputation was not included in the outcome because it could be a part of the initial revascularization strategy. We did not use amputation-free survival for the outcome because out-of-hospital mortality is not captured by the NRD. Secondary outcomes were cumulative incidence of mortality, in-hospital safety outcome (composite of acute kidney injury, major bleeding, and vascular complication), postdischarge major adverse cardiovascular events (composite of all-cause mortality, stroke, and myocardial infarction) at 6 months, and unplanned all-cause readmission at 6 months. Appropriate ICD-10-CM codes were used to identify all secondary outcomes (eTable 4 in the Supplement). For mortality, 6-month outcomes were cumulatively in-hospital and postdischarge events within 6 months. For unplanned all-cause readmissions, we excluded elective readmissions. Other outcomes were the postprocedure LOS and hospital cost of the index hospitalization.
We presented categorical variables as numbers and percentages and continuous variables as mean (SD) values or median (IQR) values based on distribution. Categorical variables were compared between the 2 groups with the χ2 test or the Fisher exact test, whereas continuous variables with normal distribution were compared with the t test, and those not normally distributed were compared with the Wilcoxon rank sum test.
We generated propensity score–matched cohorts for patients who underwent ER or SR. The propensity score was generated with all covariates described in Table 1, including patients’ demographic characteristics, comorbid conditions, hospital characteristics, clinical presentation, admission type, admission day, primary payer, and household income through multivariable logistic regression. Patients with similar propensity scores in the 2 groups were matched with a 1-to-1 scheme without replacement, using a Greedy method. Maximum propensity score differences (caliper width) of 0.2 were permitted between matched-pair observations to keep standardized differences less than 5%. Patients without matched observations were excluded. The appropriateness of all models was assessed by the receiver operating characteristic curve C statistic, which was greater than 0.75. The standardized difference was used to assess the balance of variables between 2 matched cohorts and depicted graphically (eFigure in the Supplement). For patients who died during the index hospitalization, time to death was calculated by postprocedure LOS. For patients who died after hospital discharge, time to mortality was calculated by adding postprocedure LOS of index hospitalization, time to readmission, and LOS of readmission. In calculating 6-month unplanned all-cause readmission, we excluded patients who died in the hospital. The total hospital costs were calculated by multiplying total hospital charges by the corresponding cost to charge ratio, and it represents the total cost of resources used for the inpatient hospital stay, including the cost of the procedure. All costs were adjusted for inflation or the consumer price index.
We conducted logistic regression for safety outcomes to compute the odds ratio with 95% CI in the matched cohort. Time-to-event analyses were conducted with Kaplan-Meier curves, and Kaplan-Meier estimates were generated for major amputation, all-cause mortality, and unplanned all-cause readmission (eTable 5 in the Supplement). We conducted Cox proportional hazards regression for major amputation, mortality, and unplanned all-cause readmission to compute the hazard ratio with 95% CI. For major amputation, we applied competing risk regression analysis to take concurrent mortality into account. The proportionality assumption was not violated for the Cox model for these outcomes, with the global test P > .05. For LOS and cost, we applied a generalized linear model with gamma family and log link to compute coefficients with 95% CI.
A subgroup analysis was performed for major amputation to look for sources of heterogeneity. Clinically important subgroups included age, sex, hypertension, diabetes, smoking, gangrene, osteomyelitis, chronic kidney disease, and low- and high-volume hospitals. We did not include rest pain as a subgroup because of the small sample in the included cohort; only 2094 of 54 456 patients (3.8%) in the ER group and 800 of 11 716 (6.8%) in the SR group had rest pain before intervention. We created a spline graph comparing ER with SR for a 6-month major amputation by hospital procedure volume.
We conducted previously validated falsification end point and E-value analysis to evaluate the robustness of our findings.12-16 In the falsification method, we selected an alternative outcome that may not be expected to be causally affected by the treatment being studied. Then, we assessed whether the study intervention (ER) was associated with alternative outcomes by using a similar method to assess other study outcomes. If no treatment effect was observed for the alternative outcome, it supported but did not prove that there may have been a causal treatment effect for the study outcomes. We chose a composite of pneumonia and gastrointestinal and urinary tract infection at readmission as an alternative outcome and studied the association with interventions (ER). The E-value identifies the minimum strength of association that unmeasured confounders need to have with both treatment and outcome, conditional on measured covariates, to fully explain the observed association. This strength of association estimates what the relative risk may have to be for any unmeasured confounder to overcome the observed association of study intervention (ER) with study outcomes.
All P values were 2 sided, with P < .05 considered statistically significant. All statistical analyses were conducted with unweighted samples with Stata, version 16.1 (StataCorp). Data analyses were conducted from January 1, 2022, to February 28, 2022.
A total of 66 277 patients were identified between 2016 and 2018 who underwent ER or SR for CLI. The mean (SD) age of the cohort was 69.3 (12) years, 34 036 patients were male (62.5%), and 20 420 were female (37.5%). The NRD does not provide racial and ethnic categories. Of the 66 277 patients identified, 54 546 (82.3%) underwent ER and 11 731 (17.7%) underwent SR. After propensity score matching, 11 106 matched pairs were found with a median follow-up of 180 days (IQR, 90-270 days).
Table 1 shows baseline characteristics of patients who underwent ER and patients who underwent SR before and after matching. Before matching, the ER population was older and had more female patients than the SR group. The ER cohort had a higher percentage of patients with a history of nonadherence, hypertension, diabetes, valvular heart disease, heart failure, atrial fibrillation, chronic kidney disease, end-stage kidney disease, chronic liver disease, anemia, obesity, cancer, and fluid electrolyte imbalance. The SR cohort had a higher percentage of patients with hyperlipidemia, carotid artery disease, prior coronary artery bypass graft, chronic pulmonary disease, smoking, and alcohol abuse.
In terms of clinical presentation, the ER cohort had a higher percentage of patients with osteomyelitis, chronic total occlusion, systemic infection (bacteremia or sepsis), history of lower extremity amputation, and reduced mobility. In contrast, the SR cohort had a higher percentage of patients with rest pain, ulcer, and diffuse atherosclerosis.
Among the 54 456 patients who underwent ER, 15 988 (29.4%) received a drug-coated balloon, 8662 (15.9%) received a drug-eluting stent, 16 968 (31.2%) received a bare-metal stent, and 29 305 (53.8%) had undergone atherectomy. Most SR and ER procedures were conducted in hospitals with large bed sizes, teaching hospitals, and urban hospital locations. For ER and SR, more admissions were nonelective and on weekdays. Both ER and SR cohorts had Medicare as the primary payer and a lower median household income per zip code.
Endovascular revascularization was associated with an 18% higher risk of amputation compared with SR at 6 months (997 of 10 090 [9.9%] vs 869 of 10 318 [8.4%]; hazard ratio, 1.18; 95% CI, 1.08-1.29; P = .001) (Table 2; Figure 2A).
Endovascular revascularization and SR had similar all-cause mortality at 6 months (517 of 11 106 [4.7%] vs 490 [4.4%]; hazard ratio, 1.06; 95% CI, 0.93-1.20; P = .39) (Table 2; Figure 2B).
The ER group had better in-hospital safety outcomes (17% less risk of composite of acute kidney injury, major bleeding, or vascular complication) compared with the SR group (2584 of 11 106 [23.3%] vs 2979 [26.8%]; odds ratio, 0.83; 95% CI, 0.78-0.88; P < .001) (Table 2).
Compared with the SR group at 6 months, the ER group had a similar risk of major adverse cardiovascular events (637 of 11 106 [5.7%] vs 635 of 11 106 [5.7%]; hazard ratio, 1.00; 95% CI, 0.90-1.12; P = .95) and unplanned all-cause readmission (4278 of 10 950 [39.1%] vs 4313 of 10 912 [39.5%]; hazard ratio, 0.98; 95% CI, 0.94-1.02; P = .39) (Table 2; Figure 2C).
Regarding resource use, ER was associated with 25% shorter LOS (odds ratio, 0.75; 95% CI, 0.73-0.78; P < .001) and 10% lower hospitalization cost (odds ratio, 0.90; 95% CI, 0.88-0.92; P < .001) compared with SR (Table 2).
The falsification end point was similar between the 2 groups at 6 months, supporting well-balanced groups for unmeasured confounders (odds ratio, 0.84; 95% CI, 0.65-1.10; P = .19). For major amputation, E-value analysis showed that unmeasured confounders required a risk ratio of 1.6 (lower limit of the 95% CI, 1.4) with treatment and amputation to explain the effect. For in-hospital safety outcomes, E-value analysis showed that unmeasured confounders require a risk ratio of 1.7 (lower limit of the 95% CI, 1.5) with treatment and amputation to explain the effect (Table 2). For the subgroup analysis, none of the proposed subgroups showed an interaction except sex (P = .03 for interaction), diabetes (P = .06 for interaction), and volume of the hospital (P = .007 for interaction) (Figure 3).
In this multicenter, retrospective cohort study of patients with CLI that compared ER with SR, we derived 4 important findings; compared with SR, ER was associated with (1) a higher risk of major amputations at 6 months, but this risk was similar when revascularization was performed in high-volume centers; (2) similar mortality, major adverse cardiovascular events, and unplanned all-cause readmission at 6 months; (3) better safety outcomes, including a lower risk of acute kidney injury, major bleeding, and vascular complications; and (4) lower resource use (shorter LOS and lower cost).
Although it is well known that CLI causes significant health care burden and mortality, to our knowledge, recent data on CLI peripheral interventions and outcomes are lacking. However, for the past 2 decades, the trend of SR has decreased with an increase in ER procedures.3,17,18 The Inter-Society Consensus for the Management of Peripheral Arterial Disease, along with the most recent guidelines from the American College of Cardiology and the American Heart Association, recommend revascularization as the optimal treatment.6,19 However, there is still debate between surgical and endovascular approaches as the first choice, given the lack of contemporary comparative data and the mixed results of observational studies.
We found an association between increased risk of major amputation and the ER group compared with the SR group at 6 months, which differs from the results of the BASIL trial, the Registry of First-Line Treatments in Patients With Critical Limb Ischemia (CRITISCH), the California state database, and the Surgical Reconstruction Versus Peripheral Intervention in Patients With Critical Limb Ischemia (SPINACH) study.7,20-22 The BASIL trial, CRITISCH, and the SPINACH study showed similar amputation-free survival rates between the 2 groups, whereas the California database showed better amputation-free survival in the ER group. The BASIL trial included few patients with acute limb ischemia, and the SPINACH study had a higher percentage of patients (55%) undergoing dialysis. Moreover, we studied major amputation rather than amputation-free survival because the NRD does not capture out-of-hospital mortality, as already mentioned. The sample sizes of the BASIL trial, CRITISCH, and the SPINACH study were smaller than ours and applied several exclusion criteria to select the final cohort, whereas this study included a patient population without stringent selection criteria. Moreover, Menard et al23 reported in 2016 that endovascular treatment in the BASIL trial was limited to percutaneous transluminal angioplasty alone, as was standard practice in the United Kingdom where the trial was undertaken, and the applicability of the trial result has been questioned regarding practice in the United States, which typically includes the use of stents, atherectomy, and drug-coated balloons. In accordance with the types of ER devices and procedures used in our study (Table 1) and the years of our database, our study included advanced ER. Thus, differences in patient population, duration of studies, ER techniques, and nature of outcome from our study might be the reason for the observed difference. The California state database included nonfederal California hospitals, limiting its generalizability. In addition, the California state database incorporated a few variables, such as age, sex, race and ethnicity, insurance, coronary artery disease, and kidney failure in propensity score matching. In contrast, we included more variables in propensity score matching that were associataed with CLI outcomes directly, giving more robust results for all comers than the prespecified cohort with profuse exclusion criteria (eTable 6 in the Supplement). The reason for the higher level of major amputation associated with ER is not clear; however, previous studies have shown that major tissue loss and severe infection are well-known risk factors for delayed wound healing and major amputation. It often requires abundant blood flow for limb salvage, and SR is better at restoring sufficient blood flow supply to tissue.24-26 Moreover, the subgroup analysis showed a significant interaction for sex and the volume of hospitals. In the high-volume centers, the risk of major amputation was similar between ER and SR, which could be explained by improved operator experience. Studies showed that in men, multiple sites of the arterial tree are affected compared with those in women, which could make ER less effective in men.27,28 Moreover, women have a higher incidence of limb-related complications after SR, which could be the reason for comparable major amputation rates between the 2 procedures among women.29
In this study, there was no difference in mortality between ER and SR groups at 6 months, which aligns with the results of the BASIL trial,7 showing similar 12-month all-cause mortality between the 2 groups. However, it remains unclear whether the evidence presented by the randomized clinical trial would remain valid in current clinical practice with the availability of newer endovascular devices and revascularization techniques. For 6-month outcomes, our result is also consistent with recently published CRITISCH20 results and the California state database,21 showing no difference in mortality.
The risk of major adverse cardiovascular events was similar among both groups in our study, which was also observed in the prospective multicenter observational SPINACH study conducted in Japan.22 However, the SPINACH study included only 548 Japanese patients, and there was a higher prevalence of patients undergoing dialysis (55%), who typically have extensive calcification and severe peripheral artery disease compared with those without advanced kidney disease. Moreover, novel endovascular devices, including drug-coated balloons and atherectomy devices, were not used. The 30-day unplanned readmission rate between the 2 groups was similar in our analysis, whereas the overall rate of major bleeding, vascular complications, or acute kidney injury was lower in the ER group, which is similar to the results of the 2013-2014 NRD data analysis using ICD-9-CM codes.5 Moreover, the higher cost of SR hospitalization compared with ER hospitalization can be explained by longer LOS and a higher rate of in-hospital complications.
Our study needs to be interpreted in light of the following limitations. First, the NRD is an administrative database that carries the potential for miscoding. However, the Agency for Healthcare Research and Quality has quality control measures to secure best coding practices, to ensure that linkage to state-level data is verified and reliable, and to establish internal validation of diagnosis codes through multiple audits. We assumed that the initial peripheral artery disease diagnosis may have been miscoded or that the severity was not enough for patients to be considered for ER or SR because only 10% of the cohort underwent such procedures and were included in the analysis. However, ER, SR, and amputation procedures are less likely to be miscoded. Second, we did not have anatomic data, including lesion length, lesion complexity, types of bypasses used (such as the level of bypass and conduit used), and severity of the disease at presentation (ie, Rutherford or Fontaine classification), so we could not stratify results based on disease severity. However, we did show clinical presentation based on coding, and subgroup analyses were conducted for gangrene and osteomyelitis. We also showed types of ER devices and procedures (Table 1) based on coding; however, we did not have specific codes for laser atherectomy. Third, we could not calculate the Project of Ex Vivo Vein Graft Engineering via Transfection risk score, which can help to estimate the risk of amputation-free survival after CLI. Fourth, we could not assess outcomes such as a major adverse limb event, which includes major amputation and major target limb reintervention, because reintervention could occur in a different limb, which was not possible to discern from the target limb. Fifth, the NRD does not capture out-of-hospital death; hence, postdischarge mortality includes death during readmission. This limitation applies to both cohorts; thus, the calculation of the ratio would not be affected. Sixth, patients treated in an office-based outpatient center were not captured. Despite these limitations, this analysis provides valuable data on 6-month outcomes for patients with CLI from the most current, large, all-payer, nationally available database, and the multi-institutional sample makes our results generalizable. Application of falsification outcome and E-value analysis significantly decreased the possibility of selection bias and change of result due to unmeasured confounders.
In a propensity score–matched cohort of patients with CLI from the NRD, ER was associated with better safety without any difference in mortality but was also associated with an increased risk of major amputation compared with SR. However, the risk of major amputation was similar when both procedures were performed at high-volume centers. Hence, the procedure selection for CLI management should be a shared and informed decision, with the patient understanding all the risks and benefits as discussed in this study.
Accepted for Publication: July 1, 2022.
Published: August 19, 2022. doi:10.1001/jamanetworkopen.2022.27746
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Majmundar M et al. JAMA Network Open.
Corresponding Author: Ankur Kalra, MD, Cardiovascular Institute, Kalra Hospitals, Tulsi Dass Kalra Marg, New Delhi 110015 India (email@example.com).
Author Contributions: Drs Majmundar and Kalra had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Majmundar, Patel, Doshi, Anantha-Narayanan, Kumar, Reed, Bhatt, Kalra.
Acquisition, analysis, or interpretation of data: Majmundar, Patel, Doshi, Anantha-Narayanan, Reed, Puri, Kapadia, Jaradat, Bhatt, Kalra.
Drafting of the manuscript: Majmundar, Patel, Doshi, Anantha-Narayanan, Kumar, Kalra.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Majmundar, Patel, Doshi, Anantha-Narayanan, Kumar, Reed, Kalra.
Obtained funding: Patel, Kalra.
Administrative, technical, or material support: Majmundar, Patel, Doshi, Reed, Kalra.
Supervision: Majmundar, Patel, Anantha-Narayanan, Puri, Kapadia, Jaradat, Kalra.
Conflict of Interest Disclosures: Dr Bhatt reported receiving personal fees from AngioWave, Assistance Publique-Hôpitaux de Paris, DRS.LINQ, Endotronix, High Enroll, Oakstone CME, Rutgers University, and Wiley; serving on the advisory board of Boehringer Ingelheim, Cardax, CellProthera, Cereno Scientific, Elsevier Practice Update Cardiology, Janssen, Level Ex, Medscape Cardiology, MyoKardia, NirvaMed, Novo Nordisk, PhaseBio, PLx Pharma, Regado Biosciences, and Stasys; serving on the board of directors for the Boston VA Research Institute, Society of Cardiovascular Patient Care, and TOBESOFT; chairing the American Heart Association Quality Oversight Committee; serving on the data monitoring committees for Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St Jude Medical, now Abbott), Boston Scientific (chair, PEITHO trial), Cleveland Clinic (including for the ExCEED trial, funded by Edwards), Contego Medical (chair, PERFORMANCE 2 trial), Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the ENVISAGE trial, funded by Daiichi Sankyo), Novartis, and Population Health Research Institute; receiving honoraria from the American College of Cardiology (senior associate editor, Clinical Trials and News; chair, ACC Accreditation Oversight Committee), Arnold and Porter law firm (work related to Sanofi/Bristol Myers Squibb clopidogrel litigation), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute; RE-DUAL PCI clinical trial steering committee funded by Boehringer Ingelheim; AEGIS-II executive committee, funded by CSL Behring), Belvoir Publications (editor in chief, Harvard Heart Letter), Canadian Medical and Surgical Knowledge Translation Research Group (clinical trial steering committees), Cowen and Company, Duke Clinical Research Institute (clinical trial steering committees, including for the PRONOUNCE trial, funded by Ferring Pharmaceuticals), HMP Global (editor in chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (guest editor; associate editor), K2P (cochair, interdisciplinary curriculum), Level Ex, Medtelligence/ReachMD (CME steering committees), MJH Life Sciences, Piper Sandler, Population Health Research Institute (for the COMPASS operations committee, publications committee, steering committee, and US national coleader, funded by Bayer), Slack Publications (chief medical editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (secretary/treasurer), WebMD (CME steering committees); serving as the Clinical Cardiology deputy editor; chairing the NCDR-ACTION Registry Steering Committee and VA CART Research and Publications Committee; receiving research funding from Abbott, Acesion Pharma, Afimmune, Aker BioMarine, Amarin, Amgen, AstraZeneca, Bayer, Beren, Boehringer Ingelheim, Bristol Myers Squibb, Cardax, CellProthera, Cereno Scientific, Chiesi, CSL Behring, Eisai, Ethicon, Faraday Pharmaceuticals, Ferring Pharmaceuticals, Forest Laboratories, Fractyl, Garmin, HLS Therapeutics, Idorsia, Ironwood, Ischemix, Janssen, Javelin, Lexicon, Lilly, Medtronic, Moderna, MyoKardia, NirvaMed, Novartis, Novo Nordisk, Owkin, Pfizer, PhaseBio, PLx Pharma, Recardio, Regeneron, Reid Hoffman Foundation, Roche, Sanofi, Stasys, Synaptic, The Medicines Company, and 89Bio; receiving royalties from Elsevier (editor, Cardiovascular Intervention: A Companion to Braunwald’s Heart Disease); serving as site coinvestigator for Abbott, Biotronik, Boston Scientific, CSI, St Jude Medical (now Abbott), Philips, and Svelte; serving as a trustee for the American College of Cardiology; and conducting unfunded research for FlowCo, Merck, and Takeda. Dr Kalra reported serving as the chief executive officer and creative director of makeadent.org and receiving grants from the makeadent.org Ram and Sanjita Kalra Aavishqaar Fund during the conduct of the study. No other disclosures were reported.
Funding/Support: This study was supported by the makeadent.org Ram and Sanjita Kalra Aavishqaar Fund.
Role of the Funder/Sponsor: The makeadent.org Ram and Sanjita Kalra Aavishqaar Fund had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.