Association of Anemia With Outcomes in Patients Undergoing Percutaneous Peripheral Vascular Intervention: Insights From the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2 VIC)

Abstract: Objectives. To evaluate the clinical features and outcomes of patients with anemia undergoing percutaneous peripheral vascular intervention (PVI) in a contemporary registry. Methods. We evaluated the differences in the clinical features and outcomes of patients with and without anemia undergoing PVI in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium (BMC2 VIC) registry. Anemia was defined using World Health Organization criteria. Results. Baseline anemia was present in 42.3% of 15,683 patients undergoing PVI. Compared to patients without anemia, those with anemia were older (mean age, 67 years vs 71 years), were more often black (16% vs 29%), and had higher comorbidities. Anemic patients were twice as likely to present with acute limb ischemia (5% vs 11%) and undergo urgent PVI (6% vs 15%) or below-the-knee PVI (18% vs 35%). Many in-hospital adverse events were higher in anemic patients. In a propensity-matched cohort, any adverse outcome (3.4% vs 8.4%; odds ratio [OR], 2.58; 95% confidence interval [CI], 1.94-3.42) or major cardiovascular event, defined as death, myocardial infarction, stroke, or amputation (1.1% vs 3.2%; OR, 2.96; 95% CI, 1.83-4.79) was more likely in anemic patients. Of all adverse events, the highest odds were observed for post-PVI transfusions and amputations in anemic patients. Multivariable logistic regression showed that baseline hemoglobin (1 g/dL below the normal value) was associated with greater risk of any adverse event (OR, 1.57; 95% CI, 1.47-1.68). Conclusion. The prevalence of anemia was high among PVI patients and was associated with significantly greater likelihood of amputation, any adverse event, and major cardiovascular events. Whether preprocedure correction of anemia has the potential to decrease post-PVI adverse events remains to be studied.

J INVASIVE CARDIOL 2018;30(1):35-42.

Key words: anemia, peripheral arterial disease, peripheral vascular intervention, outcomes

Baseline anemia has been shown to be associated with increased short-term and long-term morbidity and mortality in patients with heart failure or with coronary artery disease (CAD) managed either conservatively or using invasive strategies.^1-5 Although the majority of the outcomes associated with anemia are mediated by advanced age and underlying increased comorbidity burden among anemic patients, baseline anemia remains an independent correlate of adverse events in patents with CAD and peripheral vascular disease (PVD).^6-8 However, the prevalence, demographic, clinical, and angiographic features, treatment, and outcomes of patients with anemia undergoing percutaneous intervention for PVD have been less well elucidated. 

The goal of this study was to characterize these features and outcomes of patients with anemia and PVD using information from the multihospital quality-improvement statewide Blue Cross Blue Shield of Michigan Cardiovascular Consortium Vascular Intervention Collaborative (BMC2 VIC) registry.⁹ 

Methods

Study population and data collection. The study population consisted of consecutive patients who underwent peripheral vascular intervention (PVI) between January 1, 2012, and December 31, 2014, at one of 47 hospitals and outpatient centers in Michigan that participate in the BMC2 VIC quality improvement registry. Details of the BMC2 VIC registry have been described elsewhere.^9,10 All patients who underwent lower-extremity percutaneous PVI (defined as a percutaneous endovascular procedure performed on an artery in the aortoiliac, femoropopliteal, and below-the-knee arterial bed) and whose data were collected in the registry, were included in this analysis. Excluded from the study were patients undergoing “hybrid” procedures that involved an open surgical and PVI procedure during the same hospitalization. Data were collected on demographic and clinical characteristics, procedural details, and outcomes of patients undergoing PVI procedures and submitted electronically to the registry by onsite registered nurse coordinators. Data quality and the inclusion of consecutive procedures were ensured by ad hoc queries, random chart reviews, and a series of diagnostic routines included in the database.^9,10 The registry has been approved or the need for approval waived by the institutional review board of each participating hospital. 

Data definitions and outcomes. Anemia was defined by the World Health Organization (WHO) definition of hemoglobin <12 g/dL in women and <13 g/dL in men.¹¹Vascular access complication was defined as a composite of retroperitoneal hematoma, pseudoaneurysm, hematoma requiring transfusion or associated with a decrease in hemoglobin >3 g/dL, arteriovenous fistula demonstrated by arteriography or ultrasound, acute thrombosis, or need for surgical repair of the access site. Procedural success was defined as vascular access, deployment of device(s), and ≤30% diameter residual stenosis after revascularization, as well as freedom from major periprocedural complications.¹² The major endpoint for this analysis was in-hospital composite of any peri-PVI adverse event (death, myocardial infarction [MI], stroke or transient ischemic attack [TIA], post-PVI amputations, repeat PVI, red blood-cell transfusions, vascular access-site complications, contrast-induced nephropathy, limb salvage surgery, or thrombosis and/or embolism). We also studied in-hospital major adverse cardiovascular outcome (composite of death, MI, and stroke/TIA) and major vascular event (composite of post-PVI amputations or death). 

Statistical analysis. First, patients were categorized into those with and without anemia; then, demographic variables, clinical history, clinical presentation, and procedural variables were compared across these two groups. Categorical variables were reported as frequencies and percentages and continuous variables as means ± standard deviations. Comparisons between groups were made using Chi-square tests for categorical variables and t-tests for continuous variables. No adjustments were made for multiple comparisons. Second, for each peri-PVI adverse event, major cardiovascular event, major vascular event, and composite of any peri-PVI event, frequencies and percentages were calculated across patients in both categories and compared using score decompositions (first derivatives of the log likelihood) between test statistics from multiple marginal models to account for multiple comparisons.¹³ Third, to determine if baseline hemoglobin values (as opposed to dichotomous classification as anemic/non-anemic) impacted the incidence of any peri-PVI adverse event, a multivariable logistic regression model was created to adjust for differences in potential confounders and estimate an adjusted odds ratio (OR) and 95% confidence interval (95% CI) of the effect of a per gram/dL reduction in baseline hemoglobin values from the normal values of 12 g/dL in women and 13 g/dL in men. Due to the clustering of patients within institutions, a “robust” covariance matrix was utilized for the CI calculations and statistical tests.¹⁴ Given the low incidence of adverse events, particularly in non-anemic patients, it was not possible to build reliable multivariable logistic regression models to calculate adjusted estimates of the effects of anemia for any individual adverse event. To be able to compare anemic and non-anemic patients on individual adverse events, patient matching with a propensity score was used to create anemic and non-anemic patient groups with balanced distributions of potential confounders, making the two groups directly comparable with unadjusted statistical tests.¹⁵ The propensity score was created using multivariable logistic regression with all demographic, clinical history, and presentation variables except procedure history variables (prior percutaneous coronary intervention [PCI], coronary artery bypass graft [CABG] surgery, PVI, or vascular surgery), as these variables had very little overlap in their distributions and resulted in very few matched patients. Patients with and without anemia were matched with a 1:1 ratio, without replacement, on the logit of the propensity score using calipers with the width equal to 0.25 of the standard deviation of the logit of the propensity score.¹⁶ Balance in the matched groups was assessed by comparing standardized differences of the prevalences and the means across the two groups.¹⁵

All statistical tests were two sided, with P-values <.05 considered statistically significant. The statistical analysis was conducted using the R statistical program (2015; The R Foundation) using the rms package for logistic regression, the MatchIt package for propensity score matching, and the multcomp package for multiple comparisons.

Results

Demographics and clinical history (Table 1). A total of 15,681 patients were included in the analysis, of which 42% were anemic. Patients with anemia were on average 4 years older and approximately twice as likely to be black and non-smokers, with lower body mass index compared to those without anemia. The prevalence of almost all comorbid conditions was higher among patients with anemia. Specifically, diabetes was 1.5-fold higher, atrial fibrillation was 1.8-fold higher, congestive heart failure was 2.2-fold higher, and renal failure requiring dialysis was 15-fold higher in anemic patients.

Clinical presentations and procedural characteristics (Table 2). Patients with anemia were more likely to have presented with severe symptoms of PVI, ie, critical and/or acute limb ischemia, compared to patients without anemia. Location of the intervention was 1.9-times more likely to be below the knee and urgent or emergent procedures were 2.6-fold higher in anemic patients. Antegrade access was more likely to be used among these patients compared to those without anemia. Patients with anemia were more likely to be treated with angioplasty or atherectomy as opposed to stent implantation(s). The mean heparin dose was higher, whereas mean contrast agent dose was lower, in patients with anemia.

Procedural and in-hospital outcomes (Table 3). Both technical and procedural success rates were significantly lower, and the odds of any in-hospital adverse event were significantly higher in patients with anemia vs those without anemia. Among patients with anemia, the unadjusted odds for composite of death, MI, amputation, or stroke (OR, 5.95; 95% CI, 4.54-7.8), composite of death or amputation (OR, 6.98; 95% CI, 5.14-9.47), or any adverse event (OR, 3.68; 95% CI, 3.18-4.27) were all higher vs patients without anemia. Multivariable logistic regression demonstrated that baseline hemoglobin value of 1 g/dL below its normal value was associated with an increase in the odds of any adverse event by 1.57 (95% CI, 1.47-1.68). Specific adverse outcomes with statistically significant increased odds of occurring among anemic patients included: transfusions (OR, 7.11; 95% CI, 5.06-10.00); contrast-induced nephropathy (OR, 2.45; 95% CI, 1.96-3.08); major amputations (OR, 7.99; 95% CI, 2.80-22.78) and all amputations (OR, 7.77; 95% CI, 4.55-13.26); major cardiovascular events (OR, 3.77; 95% CI, 2.07-6.88); and death (OR, 4.92; 95% CI, 2.16-11.20).

Propensity-matched results of procedural and in-hospital outcomes (Table 4). Although propensity matching resulted in a loss of 2010 patients (30%) in the anemia group and 4527 patients (49%) in the no-anemia group, there were 4572 matched pairs of subjects across the two groups (total n = 9144). There were substantial reductions in the standardized differences between anemia and no-anemia groups for all matching variables (Figure 1), indicating balance across the two groups on potential confounders of in-hospital outcomes. In the matched sample, there was an attenuation of the impact of anemia on procedural and in-hospital adverse outcomes, but some effects remained large and statistically significant. Procedural success remained less likely among anemic patients (OR, 0.97; 95% CI, 0.83-1.14), but no difference was seen in terms of technical success (OR, 0.97; 95% CI, 0.83-1.14). Anemic patients matched with non-anemic patients had 2.58-times higher odds of having any in-hospital adverse outcome (95% CI, 1.95-3.42). The highest odds among any individual adverse outcomes were observed for peri-PVI transfusions (OR, 4.57; 95% CI, 2.95-7.09) and amputations (OR, 4.03; 95% CI, 2.15-7.56) among patients with anemia vs those without anemia, but not major amputations (OR, 3.44; 95% CI, 0.99-11.99). Also, contrast-induced nephropathy remained significantly higher among anemic patients (OR, 1.52; 95% CI, 1.12-2.05). 

Discussion

Study findings. Our study examined the differences in the prevalence, clinical features and presentation, angiographic findings, and many in-hospital outcomes of patients undergoing lower-extremity PVI in a community at large with anemia compared to those without anemia. Our data indicated that anemia was quite prevalent, and present in approximately 2 of every 5 patients undergoing PVI in the community at large. Most comorbid conditions were higher among patients with anemia, specifically diabetes, atrial fibrillation, congestive heart failure, and renal failure requiring dialysis. Anemic patients were more likely to present for below-the-knee PVI and with critical or acute-limb ischemia rather than claudication, and hence were more likely to need urgent or emergent PVI procedures. Procedural success rate was lower for PVI performed in anemic patients. Many adverse in-hospital events were higher among patients with anemia. Anemia remained associated with higher incidence of the composite of any in-hospital adverse event in the propensity-matched population, lending further credence to the prognostic significance of anemia among patients undergoing PVI. The odds for post-PVI transfusions and amputations were the highest of any adverse outcomes in anemic patients, and anemia was also significantly related to the risk of contrast-induced nephropathy. 

Comparisons with prior studies. The inverse association of baseline hemoglobin with worse outcomes in PAD patients has been demonstrated in previous smaller studies. Toor et al⁸ demonstrated in a single-center experience of patients with severe PAD (n = 101) undergoing PVI that preprocedural hemoglobin in the first tertile compared with the third tertile was associated with an increased risk of adverse peripheral vascular outcomes (composite of target-lesion revascularization, amputation, or death: OR, 4.17; 95% CI, 1.56-11.16). A study by Jaffery et al¹⁷ on 346 patients undergoing PVI also showed that a preprocedural hemoglobin drop of every 1 g/dL was associated with a higher risk of 9-month all-cause mortality (OR, 1.56; 95% CI, 1.19-2.04). These findings were confirmed by Oshin et al,¹⁸ who studied 360 patients undergoing PAD surgery and demonstrated that each drop of hemoglobin of 1 g/dL below the mean was associated with higher risks of major adverse cardiovascular events (defined as MI, coronary revascularization, sudden death left ventricular failure: OR, 1.4; 95% CI, 1.13-1.70) and death at 30 days (OR, 1.5; 95% CI, 1.14-1.86). 

In contrast to these studies that evaluated the relationship of baseline hemoglobin with outcomes of revascularization among patients with PAD, some other investigations have evaluated the relationship of baseline anemia among patients with PAD with adverse clinical events. The Cohorte des Patients Artéritiques (COPART) registry investigators¹⁹ evaluated 925 patients hospitalized between June 2004 and January 2010 for revascularization or medical management of PAD at three University Hospitals in southern France and showed that approximately one-half of these patients were anemic. At 1-year follow-up, the composites of death or major amputation (adjusted hazard ratio [HR], 1.44; 95% CI, 1.15-1.80) and major amputation (adjusted HR, 1.47; 95% CI, 1.10-1.96) were significantly higher for anemic patients with PAD. Dunkelgrun et al⁶ studied 1211 patients scheduled for PAD surgery with suspected or known CAD who were referred for preoperative testing between February 1990 and August 2006 at a single major university center in the Netherlands. They showed that anemia was present in one-third of these patients and its severity was directly related to 5-year adjusted cardiac death or MI in this cohort (mild: HR, 2.4; 95% CI, 1.5-4.2; moderate: HR, 3.6; 95% CI, 2.4-5.6; and severe: HR, 6.1; 95% CI, 4.1-9.1). Similarly, Gupta et al²⁰ evaluated 31,857 patients from over 200 participating sites enrolled in the National Surgical Quality Improvement Program between 2007-2009 undergoing either open vascular interventions or PVI. Approximately one-half of the patients in their study had anemia (47%) and a percentage point decrease in hematocrit below normal was associated with an adjusted 4.2% increase in the risk of 30-day death (95% CI, 1.9%-6.5%) . 

Finally, Diehm et al²¹ evaluated 711 patients undergoing endovascular repair of abdominal aortic aneurysm at a single center between March 1994 and November 2006 and demonstrated that baseline hemoglobin (adjusted HR for death, 0.87; 95% CI, 0.78-0.96) as well as anemia (log-rank test P<.001) were associated with lower long-term survival (follow-up 48 ± 32 months).

While our study findings are consistent with these studies in confirming that baseline anemia was common and was associated independently with adverse outcomes among patients undergoing PVI, some differences between our study and these studies need to be highlighted. Besides using more recent contemporary data, our study provided information on angiographic data and procedural and other outcomes missing in prior investigations. These additional data suggested that anemic patients were more likely to present for below-the-knee PVI and with critical or acute-limb ischemia rather than claudication, making them more likely to need urgent or emergent PVI procedures. These more complex procedures were associated with a lower procedural success rate. Not only was anemia associated with higher risks of peri-PVI death, MI, or major amputation rates, but also with other important peri-PVI events, such as stroke and contrast-induced nephropathy, and (not surprisingly) with a high prevalence of blood transfusion. 

Mechanisms for worse outcomes and clinical implications. We are unable to provide mechanistic insights into the reasons for the adverse prognostic significance of anemia among patients undergoing PVI based on our data. Differences in measured baseline confounders may partially explain, but failed to account fully for the increased risk associated with baseline anemia in these patients. Other speculated mechanisms include impaired oxygen delivery leading to tissue hypoxia, most importantly resulting in reduced coronary flow reserve and myocardial ischemia,²² and increase preload and cardiac output leading to left ventricular hypertrophy, known to increase cardiovascular mortality.^23,24 Subclinical or documented CAD is more common in patients with PAD, making them even more susceptible to these adaptive changes. Anemia is associated with reduced renal oxygen delivery, thus hampering the internal mechanisms to maintain the medullary oxygen balance and increasing the propensity of renal tubules for radiographic contrast toxicity.²⁵ Similarly, systemic inflammation is one of the leading features of PAD that may contribute to anemia. This inflammatory response is secondary to a variety of factors, including elevated levels of inflammatory cytokines, volume overload, and oxidative stress. Increased cytokine levels are associated with impaired bone-marrow function and significantly alter iron metabolism.²⁶ 

Possible clinical implications. Currently, no data exist to suggest that improving anemia at baseline with various therapies is associated with improved post-PVI outcomes. Many approaches to improve pre-PVI anemia may have potential for improving outcomes, although none has been tested for this purpose. Iron and vitamin replacements have been recommended by the United States Food and Drug Administration for iron-deficiency anemia instead of blood transfusions before elective procedures.²⁷ The role of preoperative blood transfusions has not been established, and intraoperative or postoperative blood transfusions have been associated with increased risk of adverse events.^28-30 Presurgical autologous blood donation may be better than other volunteer donor transfusions for elective procedures, but is less practical because of its high cost and wastage, and increased safety of allogeneic (volunteer) blood. Erythropoiesis-stimulating agents have been shown to be safe, improve hemoglobin, reduce postoperative need for transfusions, and improve quality of life, but their relationship with hard clinical events has not been established.^31-33 Additionally, concerns regarding elevation of blood pressure, increased cardiovascular ischemic events, and increase in thrombotic risks (particularly among patients with renal failure) remain problematic with these agents.³⁴ Alternatively, hypoxia-inducible factor prolyl-hydroxylase inhibitors (Roxadustat; Fibrogen), which increase hemoglobin by a mechanism that mimics high altitude, may not have the side effects of erythropoiesis-stimulating agents and may hold promise for future applications.³⁵

Study limitations. Some limitations should be considered when interpreting the current study findings. Given the observational, retrospective nature of our study, causality should be inferred with caution. As in all observational analyses, association of missing data and unmeasured confounders on outcomes cannot be ascertained. The relationship of the duration of anemia (acute vs chronic) and its treatment before PVI with outcomes cannot be ascertained. Given low individual event rates despite the relatively large number of patients in the study, we were unable to perform multivariable analysis for individual components of some composite outcome variables. We were also unable to distinguish between amputations that were planned (where PVI was performed to improve healing) and amputations that were a result of ineffective, failed, or complicated procedures. Finally, the directional similarity of all outcomes suggesting higher events among anemic patients lends credence to our study results of adverse prognostic significance of anemia in patients undergoing PVI.

Conclusion

In a large, contemporary, consecutively collected cohort of patients with severe lower-extremity PAD who presented for percutaneous lower-extremity revascularization, the prevalence of anemia was high and was associated with significantly greater risk of post-PVI amputation, nephropathy, transfusion, and major cardiovascular events. Whether preprocedure correction of anemia has the potential to decrease post-PVI adverse events remains to be studied.

References

1.    Ezekowitz JA, McAlister FA, Armstrong PW. Anemia is common in heart failure and is associated with poor outcomes: insights from a cohort of 12 065 patients with new-onset heart failure. Circulation. 2003;107:223-225.

2.    McKechnie RS, Smith D, Montoye C, et al. Prognostic implication of anemia on in-hospital outcomes after percutaneous coronary intervention. Circulation 2004;110:271-277.

3.    Nikolsky E, Aymong ED, Halkin A, et al. Impact of anemia in patients with acute myocardial infarction undergoing primary percutaneous coronary intervention: analysis from the Controlled Abciximab and Device Investigation to Lower Late Angioplasty Complications (CADILLAC) trial. J Am Coll Cardiol. 2004;44:547-553.

4.    Reinecke H, Trey T, Wellmann J, et al. Haemoglobin-related mortality in patients undergoing percutaneous coronary interventions. Eur Heart J. 2003;24:2142-2150.

5.    Carson JL, Duff A, Poses RM, et al. Effect of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet. 1996;348:1055-1060.

6.    Dunkelgrun M, Hoeks SE, Welten GM, et al. Anemia as an independent predictor of perioperative and long-term cardiovascular outcome in patients scheduled for elective vascular surgery. Am J Cardiol. 2008;101:1196-1200.

7.    Kulier A, Levin J, Moser R, et al. Impact of preoperative anemia on outcome in patients undergoing coronary artery bypass graft surgery. Circulation. 2007;116:471-479.

8.    Toor IS, Jaumdally RJ, Moss MS, Babu SB. Preprocedural hemoglobin predicts outcome in peripheral vascular disease patients undergoing percutaneous transluminal angioplasty. J Vasc Surg. 2009;50:317-321.

9.    Mukherjee D, Munir K, Hirsch AT, et al. Development of a multicenter peripheral arterial interventional database: the PVD-QI2. Am Heart J. 2005;149:1003-1008.

10.    Diehm N, Baumgartner I, Jaff M, et al. A call for uniform reporting standards in studies assessing endovascular treatment for chronic ischaemia of lower limb arteries. Eur Heart J. 2007;28:798-805.

11.    Blanc B, Finch CA, Hallberg L, et al. Nutritional anaemias. Report of a WHO scientific group. World Health Organization technical report series. 1968;405:5-37.

12.    Jackson EA, Munir K, Schreiber T, et al. Impact of sex on morbidity and mortality rates after lower extremity interventions for peripheral arterial disease: observations from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium. J Am Coll Cardiol. 2014;63:2525-2530.

13.    Pipper CB, Ritz C, Bisgaard H. A versatile method for confirmatory evaluation of the effects of a covariate in multiple models. J Royal Stat Society: Series C (Applied Stat). 2012;61:315-326.

14.    Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics. 2000;56:645-646.

15.    Rosenbaum PR, Rubin DB. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician. 2012;39:33-38.

16.    Austin PC. Some methods of propensity-score matching had superior performance to others: results of an empirical investigation and Monte Carlo simulations. Biom J. 2009;51:171-184.

17.    Jaffery Z, Shamai L, White CJ. Preprocedural hemoglobin predicts mortality following peripheral vascular interventions. Catheter Cardiovasc Interv. 2011;78:599-603.

18.    Oshin OA, Torella F. Low hemoglobin concentration is associated with poor outcome after peripheral arterial surgery. Vasc Endovascular Surg. 2013;47:449-453.

19.    Desormais I, Aboyans V, Bura A, et al. Anemia, an independent predictive factor for amputation and mortality in patients hospitalized for peripheral artery disease. Eur J Vasc Endovasc Surg. 2014;48:202-207.

20.    Gupta PK, Sundaram A, Mactaggart JN, et al. Preoperative anemia is an independent predictor of postoperative mortality and adverse cardiac events in elderly patients undergoing elective vascular operations. Ann Surg. 2013;258:1096-1102.

21.    Diehm N, Benenati JF, Becker GJ, et al. Anemia is associated with abdominal aortic aneurysm (AAA) size and decreased long-term survival after endovascular AAA repair. J Vasc Surg. 2007;46:676-681.

22.    Most AS, Ruocco NA Jr, Gewirtz H. Effect of a reduction in blood viscosity on maximal myocardial oxygen delivery distal to a moderate coronary stenosis. Circulation. 1986;74:1085-1092.

23.    Lipsic E, van der Horst IC, Voors AA, et al. Hemoglobin levels and 30-day mortality in patients after myocardial infarction. Int J Cardiol. 2005;100:289-292.

24.    Levy PS, Kim SJ, Eckel PK, et al. Limit to cardiac compensation during acute isovolemic hemodilution: influence of coronary stenosis. Am J Physiol. 1993;265:H340-H349.

25.    Heyman SN, Rosen S, Khamaisi M, Idee JM, Rosenberger C. Reactive oxygen species and the pathogenesis of radiocontrast-induced nephropathy. Invest Radiol. 2010;45:188-195.

26.    Viana MB. Anemia and infection: a complex relationship. Rev Bras Hematol Hemoter. 2011;33:90-92.

27.    Goodnough LT, Shander A, Spivak JL, et al. Detection, evaluation, and management of anemia in the elective surgical patient. Anesth Analg. 2005;101:1858-1861.

28.    Bursi F, Barbieri A, Politi L, et al. Perioperative red blood cell transfusion and outcome in stable patients after elective major vascular surgery. Eur J Vasc Endovasc Surg. 2009;37:311-318.

29.    O’Keeffe SD, Davenport DL, Minion DJ, Sorial EE, Endean ED, Xenos ES. Blood transfusion is associated with increased morbidity and mortality after lower extremity revascularization. J Vasc Surg. 2010;51:616-621, 621.e1-621e3.

30.    Society of Thoracic Surgeons Blood Conservation Guideline Task Force, Ferraris VA, Brown JR, et al. 2011 update to the Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists blood conservation clinical practice guidelines. Ann Thorac Surg. 2011;91:944-982.

31.    Corwin HL, Gettinger A, Fabian TC, et al. Efficacy and safety of epoetin alfa in critically ill patients. N Engl J Med. 2007;357:965-976.

32.    van Veldhuisen DJ, Dickstein K, Cohen-Solal A, et al. Randomized, double-blind, placebo-controlled study to evaluate the effect of two dosing regimens of darbepoetin alfa in patients with heart failure and anaemia. Eur Heart J. 2007;28:2208-2216.

33.    Drueke TB, Locatelli F, Clyne N, et al. Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med. 2006;355:2071-2084.

34.    McMurray JJV, Parfrey PS, KDIGO board members. KDIGO clinical practice guideline for anemia in chronic kidney disease. Kidney Int Suppl. 2012;2:279-330.

35.    Besarab A, Belo DS, Diamond S, et al. Evaluation of hypoxia-inducible factor prolyl hydroxylase inhibitor FG-4592 for haemoglobin correction and maintenance in non-dialysis chronic kidney disease patients for 16 and 24 weeks. Nephrol Dial Trans. 2012;27:ii144.

36.    Abidia A, Laden G, Kuhan G, et al. The role of hyperbaric oxygen therapy in ischaemic diabetic lower extremity ulcers: a double-blind randomised-controlled trial. Eur J Vasc Endovasc Surg. 2003;25:513-518.

From the ¹St. John Hospital and Medical Center, Detroit, Michigan; ²University of Michigan, Ann Arbor, Michigan; ³Henry Ford Medical Center, Detroit, Michigan; ⁴Detroit Medical Center, Detroit, Michigan; and ⁵Duke University Medical Center and Duke Clinical Research Institute, Durham, North Carolina.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted October 20, 2017, final version accepted November 28, 2017.

Address for correspondence: Hussein Othman, MD, 22101 Moross Road, Van Elslander Pavilion, 2nd Floor, Department of Cardiology, Detroit, MI 48236. Email: Hussein.othman@ascension.org