Abstract
Objective:
There are limited data evaluating the impact of postoperative hyperglycemia in patients undergoing vascular procedures. This study evaluated the relationship between suboptimal glucose control and adverse outcomes after carotid artery stenting (CAS) and carotid endarterectomy (CEA).
Methods:
Patients admitted for elective carotid procedures were selected from the Cerner Health Facts® (2008–2015) database using ICD-9-CM diagnosis and procedure codes. We examined the relationship between patient characteristics, postoperative hyperglycemia (any value >180 mg/dL), and complications with chi-square analysis. A multivariable model examined the association between patient characteristics, procedure type, and glucose control with infection, renal failure, stroke, respiratory and cardiac complications, and length of stay (LOS) over 10 days.
Results:
Of the 4,287 patients admitted for an asymptomatic carotid procedure, 788 (18%) underwent CAS and 3,499 (82%) underwent CEA. Most patients (87%) had optimal postoperative glucose control (80–180 mg/dL); 13% had suboptimal glucose control. On average, patients with suboptimal glucose control experienced: higher stroke rates (6.2% vs. 2.7%; p < 0.001); more cardiac complications (5.1% vs. 2.0%; p < 0.001); longer hospital stays (3.1vs.1.8 days; p< .001); higher rates of post-procedure infection (4.0% vs.1.8%; p=.001); and more complications than patients with optimal glucose control. Multivariable logistic regression demonstrated that patients with suboptimal glucose control had higher odds of having an infectious (pneumonia, cellulitis, surgical site etc.) complication (OR 1.91, 95% CI 1.10–3.34), renal failure (OR 3.36, 95% CI 1.95–5.78), respiratory complications (OR 1.81, 95% CI 1.21–2.71), stroke (OR 1.82, 95% CI 1.15–2.88), or LOS>10 days (OR 4.07, 95% CI 2.02–8.20).
Conclusions:
Suboptimal glucose control was associated with adverse events after CAS and CEA, independent of a diabetes diagnosis. Several adverse outcomes were associated with hyperglycemia, including stroke. Given the singular role of carotid procedures in preventing stroke, we suggest that incorporating rigorous post-operative glucose control into best medical treatment of carotid disease should be considered as standard practice.
INTRODUCTION
Hyperglycemia is a common occurrence following surgical procedures, with around 25% of non-cardiac surgical patients experiencing blood glucose elevations greater than 180 mg/dL in the postoperative period.1 There is a substantial body of evidence indicating that maintaining blood glucose levels between 80 and 180 mg/dL during the postoperative period is associated with improved surgical outcomes including reduced mortality, surgical site infection, readmission, length of stay and myocardial infarction (MI).2–7 Despite these data, it has been reported that glucose monitoring occurs in only 59% of hospitalized patients, with only 54% of these monitored patients receiving insulin therapy for elevated glucose levels.8 In patients with diabetes, the association of adverse outcomes with hyperglycemia is less clear. Studies have shown a reduction in adverse outcomes for diabetic patients to rates approximating those of non-diabetic patients following cardiac surgery when insulin therapy is used.9, 10 Other multicenter cohort studies, however, have reported that diabetic patients with a prior history of insulin use exhibit worse outcomes than non-insulin-treated diabetic patients,11 and furthermore, that non-diabetic patients have a higher risk of adverse events with hyperglycemia than diabetic patients following abdominal, vascular and spine surgery.12
Despite the high prevalence of diabetes diagnosis in patients with vascular disorders, few studies have focused on the association of postoperative hyperglycemia with complications following vascular surgery. Contemporary study of outcomes associated with postoperative hyperglycemia following elective abdominal aortic aneurysm repair (AAA) concluded that one in six patients experienced postoperative hyperglycemia and that both non-diabetic and diabetic patients had a higher chance of in-hospital mortality and infection when postoperative hyperglycemia was experienced.5 A prospective study of postoperative glycemic control using a standardized insulin infusion protocol after infrainguinal bypass and open AAA concluded that while postoperative hyperglycemia was common and could be effectively ameliorated with an insulin infusion protocol, complications including surgical site infections were not reduced in patients with optimal perioperative glycemic control.13 The current study evaluates the association of postoperative hyperglycemia with outcome following carotid procedures.
METHODS
Data source.
The International Classification of Disease, Ninth Edition, Clinical Modification (ICD-9-CM) diagnosis and procedure codes were used to identify asymptomatic patients who underwent carotid artery stenting (CAS) or endarterectomy (CEA) between 2008 and 2015 from the Cerner Health Facts® database. Health Facts is a proprietary database comprised of electronic medical records from hospitals and hospital systems that use Cerner Corporation’s electronic health record and that choose to contribute data. Participating hospitals include all encounters, but choose which data elements to contribute, potentially including billing data, diagnoses and procedures, encounter data, medication orders, and diagnostic test results. Cerner de-identifies and standardizes the data before including them in Health Facts using statistical methods that are compliant with the Health Insurance Portability and Accountability Act (HIPAA). Rigorous validity checks are applied to the data. Several outcomes have been successfully investigated using Health Facts.14–16 Informed patient consent was not needed as the Health Facts database is completely de-identified. The study was exempted by the Health Sciences Institutional Review Board at the University of Missouri.
Study population.
The study included patients undergoing elective asymptomatic CEA or CAS. Patients were excluded from the study if they were less than 21 years old at admission; had admissions during which both CEA and CAS procedures were performed; had admissions flagged as emergent or urgent, or were symptomatic; did not have postoperative medication or laboratory data in Health Facts; or had postoperative blood glucose levels below 80 mg/dL.
Covariates.
Hospital characteristics such as bed size and designation as a teaching facility as well as patients’ demographics (age, sex, and race), and acute and chronic problems at the index admission (e.g., chronic heart disease, diabetes, hypertension) were examined. The Agency for Healthcare Research and Quality’s (AHRQ) Clinical Classifications Software was used to group diagnosis codes into clinically relevant groups. We used ICD-9 CM codes at admission to calculate the Charlson Comorbidity Index, a measure that has been associated with 1-year mortality.17 We used postoperative glucose values up to seven days after the procedure to define optimal or suboptimal control. Using the American Diabetes Association (ADA) and American Association of Clinical Endocrinologists (AACE) criteria, any postoperative glucose level >180 mg/dL was defined as postoperative hyperglycemia (suboptimal). Optimal glucose control was assigned to patients with all postoperative glucose levels between 80 and 180 mg/dL.
Statistical analysis.
All analyses were performed in SAS for Windows version 9.4 (SAS Institute, Cary, NC). We used the chi-square statistic to evaluate the relationship of postoperative hyperglycemia with patient and hospital characteristics, infections, acute and chronic problems, length of stay, and complications. We calculated unadjusted relative risks (RRs) and 95% confidence intervals (CIs). In addition, multivariable logistic regression was used to examine the association between postoperative hyperglycemia and any infection (includes cellulitis, pneumonia, sepsis, surgical site infection, urinary tract infection, and other infections), LOS>10 days, renal failure, and respiratory and cardiac complications after adjusting for patient and hospital characteristics. We calculated odds ratios (ORs) and 95% CIs, and assessed model discrimination with the c-statistic (or area under the curve), where 0.5 is no better than a coin toss and 1.0 indicates a perfect fit. Model calibration over the range of risk was assessed with the Hosmer-Lemeshow goodness-of-fit chi-square test, with p-values >.05 indicating adequate fit.
RESULTS
Patient characteristics.
There were 4,287 carotid procedures, of which 3,499 (82%) were CEA and 788 (18%) comprised CAS (Table I). Suboptimal glucose control was reported in one in seven patients after carotid procedures. Mean patient age was 73 years, and most were men (58%) or Caucasian (91%). One-third of patients had a diagnosis of diabetes during the index hospitalization (33%) and half carried a diagnosis of chronic heart disease (52%). Patients who underwent CEA had a higher percentage of suboptimal glucose control than CAS (13.7% vs. 8.8%, respectively; p < .001). Average length of stay was higher for patients with suboptimal glucose control compared to patients with optimal glucose control (3.1 vs. 1.8 days; p < .001). The Charlson Index was 2.2 for the overall cohort, but was higher for patients with suboptimal glucose control (3.0) than patients with optimal glucose control (2.1; p < .001).
Table I.
Characteristics of patients who underwent elective carotid stenting or endarterectomy, by post-operative blood glucose levels [frequency (column %) unless otherwise indicated]
| Total (N = 4287) | Optimal1 (n = 3736) | Suboptimal2 (n = 551) | p-value3 | ||||
|---|---|---|---|---|---|---|---|
| Patient characteristics: | |||||||
| Age (mean, SD) | 73.05 | (7.43) | 73.17 | (7.43) | 72.26 | (7.36) | .007 |
| 60–69 | 1524 | (35.55) | 1303 | (34.88) | 221 | (40.11) | |
| 70–79 | 1826 | (42.59) | 1594 | (42.67) | 232 | (42.11) | |
| 80 or older | 937 | (21.86) | 839 | (22.46) | 98 | (17.79) | |
| Gender | .49 | ||||||
| Male | 2494 | (58.18) | 2166 | (57.98) | 328 | (59.53) | |
| Female | 1793 | (41.82) | 1570 | (42.02) | 223 | (40.47) | |
| Race/ethnicity | .08 | ||||||
| African-American | 167 | (3.90) | 136 | (3.64) | 31 | (5.63) | |
| Caucasian | 3911 | (91.23) | 3416 | (91.43) | 495 | (89.84) | |
| Other/Unknown | 209 | (4.88) | 184 | (4.93) | 25 | (4.54) | |
| Hospital characteristics: | |||||||
| Bed size | .11 | ||||||
| <200 | 1346 | (31.40) | 1194 | (31.96) | 152 | (27.59) | |
| 200–299 | 422 | (9.84) | 373 | (9.98) | 49 | (8.89) | |
| 300–499 | 1216 | (28.36) | 1049 | (28.08) | 167 | (30.31) | |
| 500 or more | 1303 | (30.39) | 1120 | (29.98) | 183 | (33.21) | |
| Teaching facility | 2614 | (60.98) | 2251 | (60.25) | 363 | (65.88) | .007 |
| Procedural/stay characteristics: | |||||||
| Procedure type | <.001 | ||||||
| CEA | 3499 | (81.62) | 3018 | (80.78) | 481 | (87.30) | |
| CAS | 788 | (18.38) | 718 | (19.22) | 70 | (12.70) | |
| In-hospital mortality | 6 | (0.14) | 4 | (0.11) | 2 | (0.36) | .13 |
| Length of stay, mean (SD) | 1.96 | (2.43) | 1.79 | (1.93) | 3.12 | (4.37) | <.001 |
| > 10 days | 50 | (1.17) | 30 | (0.80) | 20 | (3.63) | |
| Charlson Index, mean (SD) | 2.23 | (1.36) | 2.11 | (1.28) | 2.99 | (1.62) | <.001 |
Optimal glycemic control (highest post-operative blood glucose fell between 80–180 mg/dl).
Suboptimal glycemic control (highest post-operative blood glucose was > 180 mg/dl).
Chi-square (t-test for continuous) comparison of having optimal vs. suboptimal post-operative glucose levels.
SD=standard deviation, CEA = carotid endarterectomy, CAS = carotid artery stenting
We found significant bivariable associations of acute and chronic problems and infections during the index hospital encounter with suboptimal postoperative glucose control (Table II). Suboptimal glucose control was associated with acute renal failure (RR 4.67, 95% CI 2.98–7.32), respiratory complications (RR 2.37, 95% CI 1.70–3.30), infections (RR 2.16, 95% CI 1.35–3.46), stroke (RR 2.28, 95% CI 1.56–3.33), and cardiac complications (RR 2.53, 95% CI 1.66–3.87).
Table II.
Unadjusted association of selected diagnoses during the index hospital encounter with post-operative blood glucose levels for patients who underwent elective carotid stenting or endarterectomy [frequency (column %)]
| Postoperative blood glucose level | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total (N = 4287) | Optimal1 (N = 3736) | Suboptimal2 (N = 551) | RR (95% CI) | p-value3 | |||||
| Acute problems | |||||||||
| Fluid and electrolyte disorders | 208 | (4.85) | 162 | (4.34) | 46 | (8.35) | 1.93 | (1.41 – 2.64) | <.001 |
| Acute renal failure | 76 | (1.77) | 45 | (1.20) | 31 | (5.63) | 4.67 | (2.98 – 7.32) | <.001 |
| Respiratory complications | 170 | (3.97) | 126 | (3.37) | 44 | (7.99) | 2.37 | (1.70 – 3.30) | <.001 |
| Stroke | 135 | (3.15) | 101 | (2.70) | 34 | (6.17) | 2.28 | (1.56 – 3.33) | <.001 |
| Cardiac complications (inc. MI) | 103 | (2.40) | 75 | (2.01) | 28 | (5.08) | 2.53 | (1.66 – 3.87) | <.001 |
| Chronic problems | |||||||||
| Anemia | 277 | (6.46) | 204 | (5.46) | 73 | (13.25) | 2.43 | (1.89 – 3.12) | <.001 |
| Chronic heart disease | 2212 | (51.60) | 1879 | (50.29) | 333 | (60.44) | 1.20 | (1.12 – 1.29) | <.001 |
| Congestive heart failure | 289 | (6.74) | 227 | (6.08) | 62 | (11.25) | 1.85 | (1.42 – 2.42) | <.001 |
| Chronic kidney disease | 452 | (10.54) | 353 | (9.45) | 99 | (17.97) | 1.90 | (1.55 – 2.33) | <.001 |
| Diabetes | 1414 | (32.98) | 1007 | (26.95) | 407 | (73.87) | 2.74 | (2.55 – 2.95) | <.001 |
| Infections | |||||||||
| Any infection4 | 91 | (2.12) | 69 | (1.85) | 22 | (3.99) | 2.16 | (1.35 – 3.46) | .001 |
| Cellulitis | 5 | (0.12) | 4 | (0.11) | 1 | (0.18) | 1.70 | (0.19 – 15.1) | .63 |
| Pneumonia | 24 | (0.56) | 13 | (0.35) | 11 | (2.00) | 5.74 | (2.58 – 12.7) | <.001 |
| Sepsis | 7 | (0.16) | 4 | (0.11) | 3 | (0.54) | 5.09 | (1.14 – 22.6) | .02 |
| Surgical site infection | 2 | (0.05) | 1 | (0.03) | 1 | (0.18) | 6.78 | (0.42 – 108) | .11 |
| Urinary tract infection | 49 | (1.14) | 39 | (1.04) | 10 | (1.81) | 1.74 | (0.87 – 3.46) | .11 |
| Other Infection | 30 | (0.70) | 25 | (0.67) | 5 | (0.91) | 1.36 | (0.52 – 3.53) | .53 |
| Other complications | |||||||||
| Posthemorrhagic anemia | 133 | (3.10) | 109 | (2.92) | 24 | (4.36) | 1.49 | (0.97 – 2.30) | .07 |
| Post-operative medications | |||||||||
| Type I Diabetes (insulin) | 1693 | (39.49) | 1277 | (34.18) | 416 | (75.50) | 2.21 | (2.07 – 2.36) | <.001 |
| Type II Diabetes (other meds) | 568 | (13.25) | 386 | (10.33) | 182 | (33.03) | 3.20 | (2.75 – 3.72) | <.001 |
| Steroids | 209 | (4.88) | 171 | (4.58) | 38 | (6.90) | 1.51 | (1.07 – 2.12) | .02 |
Optimal glycemic control (post-operative blood glucose levels between 80–180 mg/dl).
Suboptimal glycemic control (post-operative blood glucose > 180 mg/dl).
Chi-square comparison of having optimal vs. suboptimal post-operative glucose levels.
Any infection includes (cellulitis, pneumonia, sepsis, surgical site infection, urinary tract infection, and other infection).
RR = relative risk; CI = confidence interval.
Multivariable analysis.
Analyses were performed on six post-operative outcomes – infection, LOS>10 days, renal failure, respiratory complications, stroke and cardiac complications (Table III). The Hosmer-Lemeshow statistic demonstrated adequate fit for all models. Suboptimal glucose control was associated with an increased odds of having a postoperative infection (OR 1.91, 95% CI 1.10–3.34), LOS>10 days (OR 4.07, 95% CI 2.02–8.20), renal failure (OR 3.36, 95% CI 1.95–5.78), respiratory complications (OR 1.81, 95% CI 1.21–2.71), and stroke (OR 1.82, 95% CI 1.15–2.88). Compared to patients who underwent CEA, patients who underwent a CAS procedure had lower odds of respiratory complications (OR 0.51, 95% CI 0.30–0.87), but increased odds of stroke (OR 1.55, 95% CI 1.01–2.38). Patients with a higher Charlson Index (3 or higher) had increased odds of infection (OR 3.08, 95% CI 1.66–5.73), LOS>10 days (OR 2.28, 95% CI 1.09–5.16), renal failure (OR 13.98, 95% CI 4.83–40.4), respiratory complications (OR 4.40, 95% CI 2.71–7.17), stroke (OR 2.65, 95% CI 1.62–4.34), and cardiac complications (OR 3.33, 95% CI 1.81–6.14) compared to those with scores of 0 or 1. Hospitals with more than 500 beds demonstrated higher rates of prolonged length of stay (OR 5.0, 95% CI 1.65–15.2) as well as higher odds of stroke (OR 2.5; 95% CI 1.49–4.19) and cardiac complications (OR 2.73; 95% CI 1.46–5.10) than hospitals with fewer than 200 beds.
Table III.
Multivariable logistic regression models for risk factors for several outcomes following carotid stenting or endarterectomy.
| Infection (N = 91) | Length of Stay > 10 days (N = 50) | Renal Failure (N = 76) | Respiratory Complications (n = 170) | Stroke (n = 135) | Cardiac Complication (includes MI) (n = 103) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | (95% CI) | p | OR | (95% CI) | p | OR | (95% CI) | p | OR | (95% CI) | p | OR | (95% CI) | p | OR | (95% CI) | p | |
| Procedure type (CAS) | 0.94 | (0.52 – 1.71) | .84 | 0.38 | (0.13 – 1.11) | .07 | 0.92 | (0.48 – 1.75) | .79 | 0.51 | (0.30 – 0.87) | .01 | 1.55 | (1.01 – 2.38) | .04 | 0.76 | (0.41 – 1.40) | .37 |
| Post-operative hyperglycemia1 | 1.91 | (1.10 – 3.34) | .02 | 4.07 | (2.02 – 8.20) | <.0001 | 3.36 | (1.95 – 5.78) | <.0001 | 1.81 | (1.21 – 2.71) | .003 | 1.82 | (1.15 – 2.88) | .01 | 1.65 | (1.00 – 2.73) | .05 |
| Diabetes | 0.71 | (0.40 – 1.28) | .25 | 0.35 | (0.15 – 0.85) | .02 | 0.86 | (0.46 – 1.60) | .62 | 0.77 | (0.50 – 1.18) | .22 | 0.88 | (0.54 – 1.46) | .63 | 0.96 | (0.55 – 1.68) | .88 |
| Age (years) | 1.03 | (1.00 – 1.06) | .03 | 1.09 | (1.04 – 1.13) | <.0001 | 1.05 | (1.01 – 1.08) | .004 | 1.02 | (1.00 – 1.04) | .05 | 1.01 | (0.98 – 1.03) | .63 | 1.02 | (1.00 – 1.05) | .08 |
| Gender (Female vs. Male) | 2.10 | (1.37 – 3.22) | .0007 | 0.71 | (0.39 – 1.30) | .26 | 0.84 | (0.52 – 1.35) | .46 | 0.96 | (0.70 – 1.32) | .81 | 0.89 | (0.63 – 1.27) | .52 | 1.03 | (0.69 – 1.54) | .89 |
| Race (reference = Caucasian) | ||||||||||||||||||
| African-American | 0.25 | (0.03 – 1.85) | .17 | 2.62 | (0.86 – 7.98) | .08 | 2.99 | (1.29 – 6.90) | .01 | 0.98 | (0.41 – 2.32) | .95 | 1.13 | (0.53 – 2.44) | .75 | 0.78 | (0.28 – 2.23) | .64 |
| Other/Unknown | 2.02 | (0.95 – 4.30) | .06 | 2.14 | (0.73 – 6.24) | .16 | 2.17 | (0.95 – 4.96) | .06 | 1.89 | (1.06 – 3.38) | .03 | 0.74 | (0.30 – 1.84) | .51 | 0.75 | (0.27 – 2.09) | .58 |
| Charlson Index (reference = 1) | ||||||||||||||||||
| 2 | 1.93 | (1.03 – 3.64) | .04 | 1.03 | (0.43 – 2.47) | .94 | 3.84 | (1.23 – 11.9) | .02 | 2.03 | (1.21 – 3.39) | .006 | 1.19 | (0.70 – 2.02) | .51 | 1.67 | (0.88 – 3.16) | .11 |
| 3+ | 3.08 | (1.66 – 5.73) | .0004 | 2.38 | (1.09 – 5.16) | .02 | 14.0 | (4.83 – 40.4) | <.0001 | 4.40 | (2.71 – 7.17) | <.0001 | 2.65 | (1.62 – 4.34) | <.0001 | 3.33 | (1.81 – 6.14) | .0001 |
| Hospital bed size (reference = <200) | ||||||||||||||||||
| 200–299 | 1.18 | (0.51 – 2.70) | .70 | 9.07 | (2.67 – 30.8) | .0004 | 1.59 | (0.73 – 3.48) | .24 | 1.02 | (0.56 – 1.85) | .95 | 2.33 | (1.20 – 4.56) | .01 | 2.92 | (1.32 – 6.45) | .008 |
| 300–499 | 1.34 | (0.76 – 2.37) | .31 | 5.26 | (1.71 – 16.2) | .003 | 1.37 | (0.74 – 2.55) | .31 | 0.95 | (0.62 – 1.44) | .79 | 2.54 | (1.49 – 4.32) | .0006 | 3.37 | (1.82 – 6.24) | .0001 |
| 500 or more | 1.40 | (0.81 – 2.43) | .23 | 5.00 | (1.65 – 15.2) | .004 | 0.82 | (0.41 – 1.61) | .55 | 1.03 | (0.68 – 1.54) | .89 | 2.50 | (1.49 – 4.19) | .0005 | 2.73 | (1.46 – 5.10) | .001 |
| Post-operative medications | ||||||||||||||||||
| Steroids | 2.47 | (1.30 – 4.69) | .005 | 4.51 | (2.14 – 9.47) | <.0001 | 1.95 | (0.93 – 4.10) | .07 | 3.11 | (1.94 – 4.96) | <.0001 | 2.05 | (1.16 – 3.63) | .01 | 2.27 | (1.22 – 4.22) | .009 |
| Type I Diabetes (insulin) | 1.69 | (1.02 – 2.80) | .04 | 2.05 | (1.01 – 4.16) | .04 | 1.25 | (0.71 – 2.21) | .43 | 1.58 | (1.09 – 2.29) | .01 | 1.19 | (0.78 – 1.83) | .42 | 1.55 | (0.96 – 2.52) | .07 |
| Type II Diabetes (other meds) | 0.57 | (0.27 – 1.21) | .14 | 0.70 | (0.25 – 2.01) | .51 | 0.60 | (0.29 – 1.21) | .14 | 0.70 | (0.42 – 1.16) | .16 | 0.87 | (0.50 – 1.52) | .62 | 0.84 | (0.46 – 1.52) | .55 |
Any post-operative blood glucose above 180 mg/dl
OR = odds ratio; 95% CI = 95% confidence interval.
To assess the influence of diabetes and hyperglycemia on outcome, an interaction term was tested in each of the models. Because diabetes and hyperglycemia were both coded as 0=no and 1=yes, the interaction terms resulting from multiplying the two variables represented participants who had both diabetes and hyperglycemia. None of the interaction terms was statistically significant, and were therefore not retained. As shown in Table III, postoperative receipt of insulin was associated with prolonged LOS (OR 2.05, 95% CI 1.01–4.16) as well as increased odds of infection (OR 1.69, 95% CI 1.02–2.80) and respiratory complications (OR 1.58, 95% CI 1.09–2.29). These associations with adverse outcomes were not seen in patients who were taking non-insulin oral medications in the postoperative period.
Table IV details the characteristics of patients undergoing elective CEA or CAS with respect to diabetes diagnosis. Non-diabetic patients tended to be older (>70; p<0.001) and were more likely to exhibit optimal glucose control (95.0% vs. 71.2%; p<0.001) and were more likely to undergo CAS rather than CEA (19.5 vs. 16.1; p=0.007).
Table IV.
Characteristics of patients who underwent elective carotid stenting or endarterectomy, by diabetes diagnosis [frequency (column %)]
| Total (N = 4287) | Non-Diabetic (N = 2873) | Diabetic (N = 1414) | p-value3 | ||||
|---|---|---|---|---|---|---|---|
| Age | <.001 | ||||||
| 60–69 | 1524 | (35.55) | 963 | (33.52) | 561 | (39.67) | |
| 70–79 | 1826 | (42.59) | 1239 | (43.13) | 587 | (41.51) | |
| 80 or older | 937 | (21.86) | 671 | (23.36) | 266 | (18.81) | |
| Gender | .12 | ||||||
| Male | 2494 | (58.18) | 1648 | (57.36) | 846 | (59.83) | |
| Female | 1793 | (41.82) | 1225 | (42.64) | 568 | (40.17) | |
| Post-operative glucose | <.001 | ||||||
| Optimal (80–180 mg/dL)1 | 3736 | (87.15) | 2729 | (94.99) | 1007 | (71.22) | |
| Suboptimal (> 180 mg/dL)2 | 551 | (12.85) | 144 | (5.01) | 407 | (28.78) | |
| Procedure type | .007 | ||||||
| CEA | 3499 | (81.62) | 2313 | (80.51) | 1186 | (83.88) | |
| CAS | 788 | (18.38) | 560 | (19.49) | 228 | (16.12) | |
| Outcomes: | |||||||
| Infection | 91 | (2.12) | 56 | (1.95) | 35 | (2.48) | .26 |
| LOS > 10 days | 50 | (1.17) | 33 | (1.15) | 17 | (1.20) | .98 |
| Renal failure | 76 | (1.77) | 33 | (1.15) | 43 | (3.04) | <.001 |
| Respiratory problems | 170 | (3.97) | 95 | (3.31) | 75 | (5.30) | .002 |
| Stroke | 135 | (3.15) | 77 | (2.68) | 58 | (4.10) | .01 |
| Cardiac complication | 60 | (1.40) | 28 | (0.97) | 32 | (2.26) | <.001 |
Optimal glycemic control (highest post-operative blood glucose fell between 80–180 mg/dl).
Suboptimal glycemic control (highest post-operative blood glucose was > 180 mg/dl).
Chi-square (t-test for continuous) comparison of having a diabetes diagnosis or not.
CEA = carotid endarterectomy, CAS = carotid artery stenting
DISCUSSION
This retrospective cohort study found that suboptimal glucose control was associated with several adverse events after CAS and CEA, independent of a diabetes diagnosis. We found that one in seven patients undergoing carotid procedures exhibited suboptimal glucose control in the postoperative period, which was independently associated with infection, prolonged LOS, renal failure, respiratory complications and stroke. We do stress that due to the nature of the data reporting, causality cannot be reasonably asserted between the post-operative glucose levels and the complications observed, however, the odds ratios and confidence intervals are indicators of the strengths of the association, or correlation, between postoperative glucose levels and the presence of complications. Consequently, we believe that these correlations are not due to chance and overall suboptimal glucose control is linked to poorer outcomes, despite the inability to definitively ascribe a direct causation to the relationship. Whether hyperglycemia in the postoperative period is a marker for a predisposition to adverse events or a true driver of adverse events is a challenging distinction to draw. We believe that it may be both, in that elevated glucose levels may result from underlying pathologic mechanisms that predispose to adverse events, but also that the direct metabolic consequences of hyperglycemia may also lead to further precipitation of the decline, leading to the occurrence of a complication.
Postoperative glucose levels have been studied in cardiac surgery, most notably,4, 18–20 and more recently in non-cardiac surgery.21–23 Patients with vascular disease often carry a diagnosis of diabetes and the impact of glucose control upon outcomes following vascular surgery has not been well studied. We found that several adverse outcomes were associated with suboptimal glucose control irrespective of a diagnosis of diabetes. Thus, hyperglycemia itself, rather than a diagnosis of diabetes, appeared to be the most important driver of poor outcome following elective carotid procedures.
This finding of independent association with adverse postoperative outcome has been previously described in patients undergoing cardiopulmonary bypass.24 Doenst et al. describe their institutional experience with 6,280 patients undergoing cardiac surgery that involved cardiopulmonary bypass and divided them into two groups: diabetic patients and non-diabetic patients. In-hospital mortality tripled when glucose levels were greater than 360 mg/dL compared with lower levels, and surprisingly, the relationship between mortality and hyperglycemia was identical for patients with and without diabetes. Hyperglycemia itself occurs with relative frequency in the postoperative period and has been purported to occur as a result of the post-surgical stress response.25 This is a complex mechanism of counter-regulatory hormones such as glucagon, growth hormone and catecholamines that drive blood glucose levels upwards by propagating insulin resistance as well as increasing free glucose through endogenous hepatic production.26, 27 The effect of hyperglycemia is not as well elucidated, in that it is unclear whether this is a normal stress response that should be monitored, or whether it is pathologic and attempts to attain euglycemia should be a post-operative goal.28
The majority of carotid procedures performed in the United States are performed for asymptomatic carotid stenosis.29 As the foundation of carotid procedures is to reduce and prevent future strokes, every effort should be extended to reduce the primary adverse outcome of an elective procedure. Whilst we cannot ascribe causality between suboptimal glucose control and stroke, 13% of patients undergoing elective carotid procedures in our study exhibited suboptimal glucose control, which was associated with a significantly increased odds of postoperative stroke. To achieve maximum benefit from elective surgery, the potentially modifiable risk factors for poor outcome should be aggressively managed in conjunction with adhering to best procedural techniques. We suggest that optimizing glucose control should be viewed as important as antiplatelet agents, cholesterol modification, and hypertension management in best medical therapy for carotid disease.
The role for optimal glucose control is starkly demonstrated with particular respect to adverse outcome following stroke. Parson et al. describe the association of hyperglycemia with worse stroke outcome.30 The authors examined the relationship between hyperglycemia, lactic acidosis and stroke outcome using perfusion-weighted magnetic resonance imaging and blood glucose measurements. They concluded that acute hyperglycemia increases brain lactate production and facilitates conversion of hypoperfused, at-risk tissue into infarction, which may adversely affect stroke outcome. In terms of achieving stroke prevention following elective carotid procedures, it would stand to reason that controlling hyperglycemia should be fundamental to medical management following surgery.
The association between suboptimal glucose control and infection has been demonstrated for vascular procedures, particularly lower extremity procedures and abdominal aortic aneurysm repair.5, 6 This increased risk for infectious complications with hyperglycemia has also been reported following general and non-cardiac surgery, and was similarly associated with increased hospital stay, nosocomial infections and increased in-hospital mortality.1, 31 The predisposition to infection related to hyperglycemia is clearly multifactorial; one of the experimentally-studied mechanisms involves compromise to the immune system. A study of staphylococcus aureus isolates from diabetic and non-diabetic foot ulcers demonstrated that the bacteria associated with ulceration in diabetic patients exhibited a lower virulence level than S. aureus isolated from non-diabetic infections.32 The authors point towards defective killing activity by peripheral blood neutrophils, reduced respiratory burst, as well as impaired clearance by macrophages as likely causes for the promotion of chronic S. aureus infection in diabetic patients. While surgical site infections are rare following carotid procedures, the presence of suboptimal glucose control was associated with a doubled relative risk. Given the overall predisposition of vascular patients to infection and the devastating complications associated with vascular prosthetic infections, optimizing glycemic control may have a significant impact on reducing this infectious burden.
Similarly, suboptimal glucose control was significantly associated with renal failure in multivariable analysis. Experimental evidence of the link between renal failure and hyperglycemia has been postulated in the setting of uncontrolled hyperglycemic crisis with an increase in expression of glucose transporters (GLUT-1, GLUT-2 and GLUT-3) leading to a net influx of glucose into renal tubules.33 This overload causes upregulation of cytokines, growth factors and inflammatory mediators that contribute to renal failure.34 In the present study, the odds of an adverse outcome associated with suboptimal glucose control was highest for renal failure, with a three and a half times the odds compared with optimal glucose control.
Interestingly, a higher percentage of CEA patients exhibited suboptimal glucose control than CAS patients. This difference may be attributable to a higher burden of comorbidities in the CAS group, with increased vigilance and more attention to medical management including antiplatelet agents, cholesterol-lowering agents and glucose control. Conversely, the CEA group may be an overall less comorbid group that is perceived as having good surgical risk and therefore already ‘optimized’ for surgery. Similarly, patients who had a Charlson Index of 3 or greater, indicating a high level of comorbidity, fared worse in multivariable analysis for all complications we studied. This is in stark contrast to patients with an index of 2, indicating a lesser burden of comorbidities, who experienced fewer post-operative complications. This provides further evidence that strict selection criteria should be applied before recommending prophylactic carotid operations, and these should be clearly defined with respect to sustained quality of life and minimization of perioperative risks.
The use of ICD-9 codes to identify procedures and diagnoses is a limitation, as coding can vary between institutions. Another limitation of the ICD-9 diagnosis codes found in these data is that they do not have a date and time associated with them and therefore cannot be distinguished as pre- or post-surgery conditions. Temporal relationships between glucose levels and adverse events cannot be discerned, as diagnosis coding is not time sensitive. Therefore, it is possible that suboptimal glucose values may occurs as a result of stress hyperglycemia due to adverse outcomes and not the other way around. Because Health Facts is a proprietary database comprised of data from hospitals and hospital systems that use Cerner’s electronic health record, the ethnicity and patient mix may not be representative of the wider US population. Proportions by ethnicity, however, are comparable to previously published evaluations of LE procedures utilizing Medicare data, which is considered a representative sample of the elderly US population.35 Certain types of data (e.g., laboratory or medication data) might differ between contributing hospitals and those that did not contribute these data.
Conclusions
Suboptimal glucose control occurs frequently in the postoperative period following carotid procedures. This study revealed an association between suboptimal glucose control and adverse outcomes following both CEA and CAS that was independent of a diagnosis of diabetes and appears to be a marker for an elevated odds of complications. Most concerning is the association between suboptimal glucose control and post-intervention stroke, and furthermore, the adverse effect of hyperglycemia on stroke-specific outcome. Given that carotid procedures are performed to reduce the risks of stroke, we believe that aggressive management of post-procedural hyperglycemia should form part of the best medical treatment for carotid disease. The authors feel that prospective investigation of the effect of optimal glucose control on outcome, with a special emphasis on stroke following carotid procedures, is warranted.
Acknowledgments
Support from the Agency for Healthcare Research and Quality was used to fund the research reported in this publication (R24HS022140). The authors take sole responsibility in the content of this report, which does not necessarily represent the official views of the Agency for Healthcare Research and Quality.
Footnotes
The authors declare no conflicts of interest
Presented at the Midwestern Vascular Surgery Society 42th Annual Meeting, September 13th-15th 2018, St. Louis, Missouri
References
- 1.Frisch A, Chandra P, Smiley D, Peng L, Rizzo M, Gatcliffe C, et al. Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac surgery. Diabetes Care. 2010;33(8):1783–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.King JT Jr., Goulet JL, Perkal MF, Rosenthal RA. Glycemic control and infections in patients with diabetes undergoing noncardiac surgery. Ann Surg. 2011;253(1):158–65. [DOI] [PubMed] [Google Scholar]
- 3.Dronge AS, Perkal MF, Kancir S, Concato J, Aslan M, Rosenthal RA. Long-term glycemic control and postoperative infectious complications. Arch Surg. 2006;141(4):375–80; discussion 80. [DOI] [PubMed] [Google Scholar]
- 4.Lazar HL, Chipkin SR, Fitzgerald CA, Bao Y, Cabral H, Apstein CS. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation. 2004;109(12):1497–502. [DOI] [PubMed] [Google Scholar]
- 5.Tarbunou YA, Smith JB, Kruse RL, et al. Outcomes associated with hyperglycemia after abdominal aortic aneurysm repair. J Vasc Surg 2019; 69(3): 763–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Vogel TR, Smith JB and Kruse RL. The association of postoperative glycemic control and lower extremity procedure outcomes. J Vasc Surg 2017; 66(4): 1123–1132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wiener RS, Wiener DC, Larson RJ. Benefits and risks of tight glucose control in critically ill adults: a meta-analysis. JAMA. 2008;300(8):933–44. [DOI] [PubMed] [Google Scholar]
- 8.Levetan CS, Passaro M, Jablonski K, Kass M, Ratner RE. Unrecognized diabetes among hospitalized patients. Diabetes Care. 1998;21(2):246–9. [DOI] [PubMed] [Google Scholar]
- 9.Schmeltz LR, DeSantis AJ, Thiyagarajan V, Schmidt K, O’Shea-Mahler E, Johnson D, et al. Reduction of surgical mortality and morbidity in diabetic patients undergoing cardiac surgery with a combined intravenous and subcutaneous insulin glucose management strategy. Diabetes Care. 2007;30(4):823–8. [DOI] [PubMed] [Google Scholar]
- 10.Furnary AP, Gao G, Grunkemeier GL, Wu Y, Zerr KJ, Bookin SO, et al. Continuous insulin infusion reduces mortality in patients with diabetes undergoing coronary artery bypass grafting. J Thorac Cardiovasc Surg. 2003;125(5):1007–21. [DOI] [PubMed] [Google Scholar]
- 11.Greco G, Ferket BS, D’Alessandro DA, Shi W, Horvath KA, Rosen A, et al. Diabetes and the Association of Postoperative Hyperglycemia With Clinical and Economic Outcomes in Cardiac Surgery. Diabetes Care. 2016;39(3):408–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kotagal M, Symons RG, Hirsch IB, Umpierrez GE, Dellinger EP, Farrokhi ET, et al. Perioperative hyperglycemia and risk of adverse events among patients with and without diabetes. Ann Surg. 2015;261(1):97–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Steely AM, Smith L, Callas PW, Nathan MH, Lahiri JE, Stanley AC, et al. Prospective Study of Postoperative Glycemic Control with a Standardized Insulin Infusion Protocol after Infrainguinal Bypass and Open Abdominal Aortic Aneurysm Repair. Ann Vasc Surg. 2017;44:211–20. [DOI] [PubMed] [Google Scholar]
- 14.Amin AP, Salisbury AC, McCullough PA, Gosch K, Spertus JA, Venkitachalam L, et al. Trends in the incidence of acute kidney injury in patients hospitalized with acute myocardial infarction. Arch Intern Med. 2012;172(3):246–53. [DOI] [PubMed] [Google Scholar]
- 15.Goyal A, Spertus JA, Gosch K, Venkitachalam L, Jones PG, Van den Berghe G, et al. Serum potassium levels and mortality in acute myocardial infarction. JAMA. 2012;307(2):157–64. [DOI] [PubMed] [Google Scholar]
- 16.Kosiborod M, Inzucchi SE, Goyal A, Krumholz HM, Masoudi FA, Xiao L, et al. Relationship between spontaneous and iatrogenic hypoglycemia and mortality in patients hospitalized with acute myocardial infarction. JAMA. 2009;301(15):1556–64. [DOI] [PubMed] [Google Scholar]
- 17.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. [DOI] [PubMed] [Google Scholar]
- 18.Blaha J, Mraz M, Kopecky P, Stritesky M, Lips M, Matias M, et al. Perioperative Tight Glucose Control Reduces Postoperative Adverse Events in Nondiabetic Cardiac Surgery Patients. J Clin Endocrinol Metab. 2015;100(8):3081–9. [DOI] [PubMed] [Google Scholar]
- 19.Verma S, Mazer CD, Fitchett D, Inzucchi SE, Pfarr E, George JT, et al. Empagliflozin reduces cardiovascular events, mortality and renal events in participants with type 2 diabetes after coronary artery bypass graft surgery: subanalysis of the EMPA-REG OUTCOME(R) randomised trial. Diabetologia. 2018;61(8):1712–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Duncan AE, Sessler DI, Sato H, Sato T, Nakazawa K, Carvalho G, et al. Hyperinsulinemic Normoglycemia during Cardiac Surgery Reduces a Composite of 30-day Mortality and Serious In-hospital Complications: A Randomized Clinical Trial. Anesthesiology. 2018;128(6):1125–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Shanks AM, Woodrum DT, Kumar SS, Campbell DA Jr., Kheterpal S. Intraoperative hyperglycemia is independently associated with infectious complications after non-cardiac surgery. BMC Anesthesiol. 2018;18(1):90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Meister KM, Hufford T, Tu C, et al. Clinical significance of perioperative hyperglycemia in bariatric surgery: evidence for better perioperative glucose management. Surg Obes Relat Dis 2018; 14(11): 1725–1731. [DOI] [PubMed] [Google Scholar]
- 23.Oh TK, Han S, Oh AY, et al. Chronic hyperglycemia with elevated glycated hemoglobin level and its association with postoperative acute kidney injury after a major laparoscopic abdominal surgery in diabetes patients. J Anesth 2018; 32(5): 740–747. [DOI] [PubMed] [Google Scholar]
- 24.Doenst T, Wijeysundera D, Karkouti K, Zechner C, Maganti M, Rao V, et al. Hyperglycemia during cardiopulmonary bypass is an independent risk factor for mortality in patients undergoing cardiac surgery. J Thorac Cardiovasc Surg. 2005;130(4):1144. [DOI] [PubMed] [Google Scholar]
- 25.McCowen KC, Malhotra A, Bistrian BR. Stress-induced hyperglycemia. Crit Care Clin. 2001;17(1):107–24. [DOI] [PubMed] [Google Scholar]
- 26.Fayfman M, Davis G, Duggan EW, et al. Sitagliptin for prevention of stress hyperglycemia in patients without diabetes undergoing general surgery: a pilot randomized study. J Diabetes Complications 2018; 32(12): 1091–1096. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD, et al. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med. 2009;360(2):129–39. [DOI] [PubMed] [Google Scholar]
- 28.Booth G, Stalker TJ, Lefer AM, Scalia R. Elevated ambient glucose induces acute inflammatory events in the microvasculature: effects of insulin. Am J Physiol Endocrinol Metab. 2001;280(6):E848–56. [DOI] [PubMed] [Google Scholar]
- 29.Abbott AL. Medical (nonsurgical) intervention alone is now best for prevention of stroke associated with asymptomatic severe carotid stenosis: results of a systematic review and analysis. Stroke. 2009;40(10):e573–83. [DOI] [PubMed] [Google Scholar]
- 30.Parsons MW, Barber PA, Desmond PM, Baird TA, Darby DG, Byrnes G, et al. Acute hyperglycemia adversely affects stroke outcome: a magnetic resonance imaging and spectroscopy study. Ann Neurol. 2002;52(1):20–8. [DOI] [PubMed] [Google Scholar]
- 31.Duggan EW, Klopman MA, Berry AJ, Umpierrez G. The Emory University Perioperative Algorithm for the Management of Hyperglycemia and Diabetes in Non-cardiac Surgery Patients. Curr Diab Rep. 2016;16(3):34. [DOI] [PubMed] [Google Scholar]
- 32.Tuchscherr L, Korpos E, van de Vyver H, Findeisen C, Kherkheulidze S, Siegmund A, et al. Staphylococcus aureus requires less virulence to establish an infection in diabetic hosts. Int J Med Microbiol. 2018;308(7):761–9. [DOI] [PubMed] [Google Scholar]
- 33.Kao Y, Feng IJ, Hsu CC, Wang JJ, Lin HJ, Huang CC. Association of hyperglycemic crisis with an increased risk of end-stage renal disease: A nationwide population-based cohort study. Diabetes Res Clin Pract. 2018;138:106–12. [DOI] [PubMed] [Google Scholar]
- 34.Van den Berghe G, Wilmer A, Hermans G, Meersseman W, Wouters PJ, Milants I, et al. Intensive insulin therapy in the medical ICU. N Engl J Med. 2006;354(5):449–61. [DOI] [PubMed] [Google Scholar]
- 35.Vogel TR, Smith JB and Kruse RL. Hospital readmissions after elective lower extremity vascular procedures. Vascular 2018; 26(3): 250–261. [DOI] [PMC free article] [PubMed] [Google Scholar]