Abstract
Objective –
We evaluated the association between postoperative hyperglycemia and outcomes following abdominal aortic aneurysm (AAA) repair.
Methods –
We used diagnosis and procedure codes (ICD-9-CM) to identify patients who underwent open or endovascular repair (EVAR) of a non-ruptured AAA from September 2008 to March 2014 from the Cerner Health Facts® database. We evaluated the association between post-operative hyperglycemia (> 180 mg/dL) and infections, in-hospital mortality, readmission, patient characteristics, length of hospital stay, and medications. Multivariable logistic models examined the association of postoperative hyperglycemia with in-hospital infection and mortality.
Results –
Of 2,478 patients, 2,071 (83.5%) had good post-operative glucose control (80–180 mg/dL), and 407 (16.5%) had suboptimal control (hyperglycemia). Patients who had postoperative hyperglycemia experienced longer hospital stays (9.5 vs. 4.7 days, p <. 0001), higher infection rates (18% vs. 8%, p < .0001), higher in-hospital mortality (8.4 vs. 1.2%, p <.0001), and more acute complications (i.e., acute renal failure, fluid and electrolyte disorders, respiratory complications). After adjusting for patient characteristics and medications, multivariable logistic regression models demonstrated that patients receiving postoperative insulin had nearly 1.6 times the odds of having an infectious complication (OR 1.6, 95% CI 1.12–2.2, p = .007) than those who did not. Hyperglycemic patients had 3.5 times the odds of in-hospital mortality (OR 3.48, 95% CI 1.78–6.80, p = .0003; 2.3% vs. 1.2%, p <0.001). When stratified by procedure type, patients with hyperglycemia who underwent EVAR had nearly 2 times the odds of an infectious complication (OR 1.85, 95% CI 0.98–3.51, p = .05), and 7.5 times the odds of in-hospital mortality (OR 7.54, 95% CI 1.95–29.1, p = .003). Patients who underwent an open AAA repair and who had hyperglycemia had 3 times the odds of dying in the hospital (OR 3.05, 95% CI 1.29–7.21, p = .01).
Conclusions –
Among patients undergoing elective AAA repair, approximately one in six had post-operative hyperglycemia. Following AAA repair in patients with and without diabetes, postoperative hyperglycemia was associated with adverse events, including in-hospital mortality and infections. Compared to those who had open surgery, patients undergoing EVAR who had post-operative hyperglycemia had greater risk of infection and death. After controlling for insulin administration and post-operative hyperglycemia, a diabetes diagnosis was associated with lower odds of both infection and in-hospital mortality. Our study suggests that hyperglycemia may be used as a clinical marker as it was found to be significantly associated with inferior outcomes after elective AAA repair. This retrospective study, however, cannot imply causation; further study using prospective methods is needed to elucidate the relationship between postoperative hyperglycemia and patient outcomes.
Keywords: glycemic control, glucose, abdominal aortic aneurysm repair, infection
Introduction
Hospital hyperglycemia is common after surgical procedures.(1) Kiran et al. documented that even one postoperative hyperglycemic episode can be associated with increased morbidity and mortality.(2) It has been well documented that postoperative hyperglycemia is associated with myocardial infarction, surgical site infections, stroke, the need for re-operative interventions, increased length of stay, increased readmissions, and even death.(3, 4)
Multiple studies have reported that maintaining glucose levels between 80 and 180 mg/dL in the postoperative period enhances survival, improves outcomes, and decreases the incidence of wound complications and ischemic events.(5–7) Patients with diabetes have impaired immune status that may be minimized by better glycemic control. It has also been documented that patients with Type 2 diabetes can have better outcomes postoperatively, perhaps due to the protective effect of aggressive insulin therapy during the early postoperative period or the anti-inflammatory effect of insulin.(8, 9)
Few studies have evaluated outcomes associated with postoperative hyperglycemia in patients undergoing open and endovascular repair of an abdominal aortic aneurysm (AAA). The purpose of our study was to evaluate outcomes and complications associated with AAA repair in patients with postoperative hyperglycemia.
Methods
Data
We used the International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) diagnosis and procedure codes to identify patients who underwent open or endovascular repair for non-ruptured AAA from September 2008 to March 2014. Patients were identified from Cerner Health Facts®, a database comprised of electronic clinical records from hospital systems that have Cerner Corporation’s electronic health record. Hospitals that choose to participate can determine which data elements to include (e.g., encounter data, demographics, billing data, diagnosis and procedure codes, diagnostic tests, and medications). Health Facts® undergoes rigorous validity checks to ensure that terminology and units of measurement are consistent across institutions. Quality and benchmarking reports are provided to contributing institutions. The data are de-identified for compliance with the Health Insurance Portability and Accountability Act (HIPAA). Thus, identification of individual patients or hospitals is not possible. Compared with the National Inpatient Sample, patients in Health Facts are slightly younger and have slightly fewer in-hospital deaths, and hospitals with <99 beds are over-represented.(10) Health Facts data have been successfully used to study several health outcomes.(11–13) Patient consent was not obtained as this study used de-identified data; Health Sciences Institutional Review Board at the University of Missouri determined that the study was exempt.
Study Population
We included 2,478 patients who underwent endovascular (procedure code 39.71) or open (procedure codes 38.44 or 39.25) repair for a non-ruptured AAA (diagnosis code 441.4). We excluded patients who were less than 21 years old, had an emergent or urgent admission, had no post-operative medication or laboratory data in Health Facts®, or whose post-operative blood glucose levels were all below 80 mg/dL (hypoglycemic). See Appendix I for flowchart of exclusions.
Covariates
We included hospital characteristics (size and teaching facility), patient demographics (age, gender, and race) and acute/chronic problems the index admission (chronic heart disease, diabetes, hypertension, etc.). We used the Agency for Healthcare Research and Quality’s (AHRQ) Clinical Classifications Software to group ICD-9-CM diagnosis codes into clinically relevant conditions.(14) We also used ICD-9-CM diagnosis codes documented during the admission to calculate the Charlson Comorbidity Index, a measure associated with one-year mortality.(15)
We used the American Diabetes Association and American Association of Clinical Endocrinologists criteria to define suboptimal post-procedure glucose (hyperglycemia) values as serum glucose >180mg/dL without regard for preoperative serum glucose.(16, 17) We used the patient’s maximum post-operative blood glucose value, beginning on the calendar day following surgery and up to 7 days following the AAA procedure, to classify patients’ glucose control as optimal (all values between 80–180 mg/dL) or hyperglycemic (suboptimal – at least one value above 180 mg/dL).
Outcomes
We assessed several outcomes including infections, in-hospital mortality, and readmission within 30 days of discharge from the index admission. Hospital infections were indicated by the presence of a diagnosis code for sepsis, pneumonia, lower extremity cellulitis, surgical site infection, urinary tract infection, or another infection (see Appendix II for ICD-9 CM codes and descriptions). We also determined whether patients died in the hospital following the procedure.
Statistical Analysis
SAS version 9.4 (SAS Institute, Cary, NC) was used for all analyses. We used the chi-square statistic to evaluate bivariable relationships between post-operative hyperglycemia and patient characteristics, hospital characteristics, infections, acute and chronic problems, length of stay, and readmission. We calculated unadjusted relative risks (RR) and 95% confidence intervals (CI). Additionally, multivariable logistic regression was used to examine the association between postoperative hyperglycemia and infection, mortality, and readmission after adjusting for patient and hospital characteristics. We calculated odds ratios (OR) and 95% CIs. We 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 X2 test, with p-values > .05 indicating adequate fit.
Results
Overall
Of 2,478 patients, 407 (16.5%) had hyperglycemia while most (2,071, 83.5%) had optimal post-operative glucose control (Table I). Mean age of patients who had optimal glucose control was 69.7 years, while patients with suboptimal control were slightly younger, with a mean age of 67.4 years (p = .002). A majority of the study sample was male (72%) or Caucasian (86%). Sixty percent of the study patients underwent endovascular AAA repair. There was a lower rate of hyperglycemia among patients who had endovascular repair (9.9%) than those who underwent open procedures (26.2%, p < .0001). We used glucose readings up to 8 days after surgery. On average, there were 3.79 glucose readings for the full sample, with a mean postoperative glucose reading of 148. ENDO patients had an average of 2.29 readings during this period, with mean glucose level = 137. OPEN patients had significantly more glucose readings (mean = 6.02), as well as higher levels (mean = 165), likely resulting from their longer hospital stays (9.13 days vs. 3.16 days for open and endo, respectively). OPEN patients were also more likely to receive post-operative insulin and/or other diabetic medications than ENDO patients (42.9% vs. 25.1%, p <.0001).
Table I.
Patient, hospital, and procedural characteristics for patients with AAA who had lower extremity revascularization, by post-operative blood glucose levels [frequency (column %)].
| Post-operative Blood Glucose Levels | |||||||
|---|---|---|---|---|---|---|---|
| Total (N = 2478) |
Optimala (n = 2071) |
Suboptimalb (n = 407) |
p-valuec | ||||
| Patient characteristics: | |||||||
| Age (mean, SD) | 69.42 | (10.6) | 69.71 | (10.6) | 67.94 | (10.6) | .002 |
| 21–59 | 447 | (18.0) | 352 | (17.0) | 95 | (23.3) | |
| 60–69 | 747 | (30.1) | 628 | (30.3) | 119 | (29.2) | |
| 70–79 | 833 | (33.6) | 699 | (33.7) | 134 | (32.9) | |
| 80 or older | 451 | (18.2) | 392 | (18.9) | 59 | (14.5) | |
| Gender (male) | 1776 | (71.7) | 1497 | (72.2) | 279 | (68.5) | .12 |
| Race/ethnicity | .04 | ||||||
| African-American | 234 | (9.4) | 186 | (8.9) | 48 | (11.8) | |
| Caucasian | 2126 | (85.8) | 1793 | (86.6) | 333 | (81.8) | |
| Other/Unknown | 118 | (4.7) | 92 | (4.4) | 26 | (6.4) | |
| Hospital characteristics: | |||||||
| Bed size | .0008 | ||||||
| <200 | 119 | (4.8) | 108 | (5.2) | 11 | (2.7) | |
| 200–299 | 396 | (15.9) | 337 | (16.3) | 59 | (14.5) | |
| 300–499 | 705 | (28.4) | 610 | (29.4) | 95 | (23.3) | |
| 500 or more | 1258 | (50.8) | 1016 | (49.1) | 242 | (59.5) | |
| Teaching facility | 2076 | (83.8) | 1728 | (83.4) | 348 | (85.5) | .30 |
| Procedural/stay characteristics: | |||||||
| Procedure type | <.0001 | ||||||
| Endovascular | 1486 | (59.9) | 1339 | (64.6) | 147 | (36.1) | |
| Open | 992 | (40.0) | 732 | (35.3) | 260 | (63.9) | |
| Readmission (within 30 days) | 100 | (4.0) | 84 | (4.1) | 16 | (3.9) | .90 |
| In-hospital mortality | 58 | (2.3) | 24 | (1.2) | 34 | (8.4) | <.0001 |
| Length of stay, mean (SD) | 5.55 | (7.9) | 4.76 | (6.3) | 9.55 | (12.8) | <.0001 |
| > 10 days | 325 | (13.1) | 209 | (10.1) | 116 | (28.5) | <.0001 |
| Charlson Index, mean (SD) | 2.03 | (1.4) | 1.97 | (1.3) | 2.34 | (1.4) | <.0001 |
SD=standard deviation
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.
Most hospitals in the sample had 300 or more beds (79%) and the majority were academic centers (84%). All were located in urban areas. Patients with postoperative hyperglycemia had worse outcomes, experiencing longer mean stays (9.5 days) compared with those who had optimal glucose control (4.8 days, p < .0001), higher Charlson Index scores (2.3 vs. 2.0, respectively; p < .0001), and higher in-hospital mortality (8.4 vs. 1.2%, <.0001). Thirty-day readmission was 4% overall, and differed little between groups (4.1% optimal and 3.9% suboptimal, p = .90).
In bivariable analyses (Table II), patients who had post-operative hyperglycemia were 2.1 times more likely to acquire an infection (95% CI 1.70–2.60) than patients who had optimal glucose control post-operatively (18.2% vs. 7.9%, respectively). Diagnoses associated with post-operative hyperglycemia included: diabetes (RR 2.87, 95% CI 2.42–3.41); acute renal failure (RR 2.39, 95% CI 1.93–2.94); fluid and electrolyte disorders (RR 2.20, 95% CI 1.82–2.66); acute respiratory complications (RR 1.95, 95% CI 1.59–2.40); posthemorrhagic anemia (RR 1.95, 95% CI 1.54–2.47); coronary artery disease (RR 1.70, 95% CI 1.32–2.18); cardiac complications or MI (RR 1.65, 1.37–1.98); receipt of a blood transfusion (RR 1.43, 95% CI 1.13–1.80); chronic anemia (RR 1.34, 95% CI 1.04–1.72); or receipt of post-operative steroids (RR 1.60, 95% CI 1.26–2.04), post-operative insulin (RR 3.81, 95% CI 3.17–4.56), or other diabetes medications – e.g., glyburide, glipizide, glimepiride, metformin) (RR 2.65, 95% CI 2.13–3.28).
Table II.
Unadjusted association of selected diagnoses during the index hospital encounter with post-operative blood glucose levels [frequency (column %)], overall
| Postoperative blood glucose level | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total (N = 2478) |
Optimala (n = 2071) |
Suboptimalb (n = 407) |
RR (95% CI) | p-valuec | |||||
| Acute problems | |||||||||
| Fluid and electrolyte disorders | 331 | (13.4) | 228 | (11.0) | 103 | (25.3) | 2.20 | (1.82 – 2.66) | <.0001 |
| Acute renal failure | 208 | (8.4) | 135 | (6.5) | 73 | (17.9) | 2.39 | (1.93 – 2.94) | <.0001 |
| Respiratory complications | 291 | (11.7) | 207 | (10.0) | 84 | (20.6) | 1.95 | (1.59 – 2.40) | <.0001 |
| Cardiac complications or MI | 66 | (2.7) | 34 | (1.6) | 32 | (7.9) | 1.65 | (1.37 – 1.98) | <.0001 |
| Chronic problems | |||||||||
| Anemia | 259 | (10.4) | 204 | (9.9) | 55 | (13.5) | 1.34 | (1.04 – 1.72) | .02 |
| Chronic heart disease | 998 | (40.3) | 818 | (39.5) | 180 | (44.2) | 1.18 | (0.98 – 1.41) | .07 |
| Coronary artery disease | 201 | (8.1) | 148 | (7.2) | 53 | (13.0) | 1.70 | (1.32 – 2.18) | <.0001 |
| Chronic kidney disease | 277 | (11.2) | 224 | (10.8) | 53 | (13.0) | 1.19 | (0.92 – 1.54) | .19 |
| Diabetes | 549 | (22.1) | 366 | (17.7) | 183 | (45.0) | 2.87 | (2.42 – 3.41) | <.0001 |
| Infections | |||||||||
| Any infectiond | 237 | (9.6) | 163 | (7.9) | 74 | (18.2) | 2.10 | (1.70 – 2.60) | <.0001 |
| LE cellulitis | 11 | (0.4) | 7 | (0.3) | 4 | (1.0) | 2.23 | (1.01 – 4.89) | .07 |
| Pneumonia | 97 | (3.9) | 63 | (3.0) | 34 | (8.4) | 2.24 | (1.68 – 2.98) | <.0001 |
| Sepsis | 47 | (1.9) | 24 | (1.2) | 23 | (5.7) | 3.10 | (2.28 – 4.21) | <.0001 |
| Surgical site infection | 28 | (1.1) | 16 | (0.8) | 12 | (3.0) | 2.66 | (1.72 – 4.12) | <.0001 |
| Urinary tract infection | 76 | (3.1) | 57 | (2.8) | 19 | (4.7) | 1.55 | (1.04 – 2.31) | .04 |
| Other Infection | 65 | (2.6) | 40 | (1.9) | 25 | (6.1) | 2.43 | (1.76 – 3.35) | <.0001 |
| Other complications | |||||||||
| Posthemorrhagic anemia | 191 | (7.7) | 134 | (6.5) | 57 | (14.0) | 1.95 | (1.54 – 2.47) | <.0001 |
| Blood transfusion | 309 | (12.5) | 240 | (11.6) | 69 | (16.9) | 1.43 | (1.13 – 1.80) | .002 |
| Post-operative medications | |||||||||
| Insulin | 752 | (30.3) | 498 | (24.1) | 254 | (62.4) | 3.81 | (3.17 – 4.56) | <.0001 |
| Other diabetic medicationse | 163 | (6.6) | 99 | (4.8) | 64 | (15.7) | 2.65 | (2.13 – 3.28) | <.0001 |
| Steroids | 233 | (9.4) | 175 | (8.5) | 58 | (14.3) | 1.60 | (1.26 – 2.04) | .0002 |
RR = relative risk; CI = confidence interval.
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 (LE cellulitis, pneumonia, sepsis, surgical site infection, urinary tract infection, and other infection).
glyburide, glipizide, glimepiride, metformin, or medications that include these agents in combination.
When stratifying by procedure type, hyperglycemia following endovascular AAA repair (Table III) was associated with: cardiac complications or MI (RR 5.14, 95% CI 3.26–8.11); posthemorrhagic anemia (RR 3.35, 95% CI 2.22–5.05); diabetes (RR 3.19, 95% CI 2.36–4.30); any infection (RR 3.18, 95% CI 2.17–4.64); acute renal failure (RR 2.68, 95% CI 1.75–4.10); acute respiratory conditions (RR 2.59, 95% CI 1.76–3.83); fluid and electrolyte disorders (RR 2.51, 95% CI 1.71–3.69); receipt of a blood transfusion (RR 2.35, 95% CI 1.63–3.47); coronary artery disease (RR 1.91, 95% CI 1.24–2.94); or receipt of post-operative steroids (RR 2.16, 95% CI 1.43–3.25), post-operative insulin (RR 3.79, 95% CI 2.81–5.12), or other diabetes medications (RR 3.88, 95% CI 2.78–5.41). Hyperglycemia following open AAA repair (Table IV) was associated with: acute respiratory conditions (RR 2.59, 95% CI 1.76–3.83); diabetes (RR 2.58, 95% CI 2.12–3.14); cardiac complications or MI (RR 1.93, 1.40–2.68); acute renal failure (RR 1.74, 95% CI 1.37–2.20); fluid and electrolyte disorders (RR 1.51, 95% CI 1.22–1.88); chronic heart disease (RR 1.32, 95% CI 1.07–1.62); coronary artery disease (RR 1.49, 95% CI 1.11–2.00); any infection (RR 1.29, 95% CI 1.02–1.64); or receipt of post-operative insulin (RR 2.92, 95% CI 2.33–3.67) or other diabetes medications (RR 2.10, 95% CI 1.60–2.75).
Table III.
Unadjusted association of selected diagnoses during the index hospital encounter with post-operative blood glucose levels [frequency (column %)] for patients who had endovascular procedures.
| Postoperative blood glucose level | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total (N = 1486) |
Optimala (n = 1339) |
Suboptimalb (n = 147) |
RR (95% CI) | p-valuec | |||||
| Acute problems | |||||||||
| Fluid and electrolyte disorders | 112 | (7.5) | 87 | (6.5) | 25 | (17.0) | 2.51 | (1.71 – 3.69) | <.0001 |
| Acute renal failure | 78 | (5.2) | 59 | (4.4) | 19 | (12.9) | 2.68 | (1.75 – 4.10) | <.0001 |
| Respiratory complications | 104 | (7.0) | 80 | (5.9) | 24 | (16.3) | 2.59 | (1.76 – 3.83) | <.0001 |
| Cardiac complications or MI | 23 | (1.5) | 12 | (0.9) | 11 | (7.5) | 5.14 | (3.26 – 8.11) | <.0001 |
| Chronic problems | |||||||||
| Anemia | 139 | (9.3) | 122 | (9.1) | 17 | (11.5) | 1.27 | (0.79 – 2.04) | .33 |
| Chronic heart disease | 643 | (43.3) | 573 | (42.8) | 70 | (47.6) | 1.19 | (0.88 – 1.62) | .26 |
| Coronary artery disease | 113 | (7.6) | 93 | (6.9) | 20 | (13.6) | 1.91 | (1.24 – 2.94) | .003 |
| Chronic kidney disease | 179 | (12.0) | 152 | (11.3) | 27 | (18.4) | 1.64 | (1.12 – 2.42) | .01 |
| Diabetes | 316 | (21.3) | 248 | (18.5) | 68 | (46.3) | 3.19 | (2.36 – 4.30) | <.0001 |
| Infections | |||||||||
| Any infectiond | 86 | (5.8) | 62 | (4.6) | 24 | (16.3) | 3.18 | (2.17 – 4.64) | <.0001 |
| LE cellulitis | 2 | (0.1) | 1 | (0.1) | 1 | (0.7) | -- | -- | -- |
| Pneumonia | 24 | (1.6) | 17 | (1.3) | 7 | (4.7) | 3.05 | (1.60 – 5.79) | .001 |
| Sepsis | 9 | (0.6) | 4 | (0.3) | 5 | (3.4) | -- | -- | -- |
| Surgical site infection | 7 | (0.5) | 3 | (0.2) | 4 | (2.7) | -- | -- | -- |
| Urinary tract infection | 40 | (2.7) | 30 | (2.2) | 10 | (6.8) | 2.64 | (1.51 – 4.62) | .0001 |
| Other Infection | 29 | (1.9) | 17 | (1.3) | 12 | (8.2) | 4.47 | (2.81 – 7.09) | <.0001 |
| Other complications | |||||||||
| Posthemorrhagic anemia | 63 | (4.2) | 44 | (3.3) | 19 | (12.9) | 3.35 | (2.22 – 5.05) | <.0001 |
| Blood transfusion | 128 | (8.6) | 101 | (7.5) | 27 | (18.4) | 2.35 | (1.63 – 3.47) | <.0001 |
| Post-operative medications | |||||||||
| Insulin | 334 | (22.5) | 257 | (19.2) | 77 | (52.4) | 3.79 | (2.81 – 5.12) | <.0001 |
| Other diabetic medicationse | 103 | (6.9) | 70 | (5.2) | 33 | (22.4) | 3.88 | (2.78 – 5.41) | <.0001 |
| Steroids | 112 | (7.5) | 90 | (6.7) | 22 | (14.9) | 2.16 | (1.43 – 3.25) | .0003 |
RR = relative risk; CI = confidence interval.
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 (LE cellulitis, pneumonia, sepsis, surgical site infection, urinary tract infection, and other infection).
Glyburide, glipizide, glimepiride, metformin, or medications that include these agents in combination.
Table IV.
Unadjusted association of selected diagnoses during the index hospital encounter with post-operative blood glucose levels [frequency (column %)] for patients who had open procedures.
| Postoperative blood glucose level | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Total (N = 992) |
Optimala (n = 732) |
Suboptimalb (n = 260) |
RR (95% CI) | p-valuec | |||||
| Acute problems | |||||||||
| Fluid and electrolyte disorders | 219 | (22.0) | 141 | (19.2) | 78 | (30.0) | 1.51 | (1.22 – 1.88) | .0003 |
| Acute renal failure | 130 | (13.1) | 76 | (10.3) | 54 | (20.7) | 1.74 | (1.37 – 2.20) | <.0001 |
| Respiratory complications | 187 | (18.8) | 127 | (17.3) | 60 | (23.0) | 1.29 | (1.02 – 1.64) | .04 |
| Cardiac complications or MI | 43 | (4.3) | 22 | (3.0) | 21 | (8.1) | 1.93 | (1.40 – 2.68) | <.0001 |
| Chronic problems | |||||||||
| Anemia | 120 | (12.1) | 82 | (11.2) | 38 | (14.6) | 1.24 | (0.93 – 1.66) | .14 |
| Chronic heart disease | 355 | (35.7) | 245 | (33.4) | 110 | (42.3) | 1.32 | (1.07 – 1.62) | .01 |
| Coronary artery disease | 88 | (8.8) | 55 | (7.5) | 33 | (12.6) | 1.49 | (1.11 – 2.00) | .01 |
| Chronic kidney disease | 98 | (9.8) | 72 | (9.8) | 26 | (10.0) | 1.01 | (0.72 – 1.43) | .93 |
| Diabetes | 233 | (23.4) | 118 | (16.1) | 115 | (44.2) | 2.58 | (2.12 – 3.14) | <.0001 |
| Infections | |||||||||
| Any infectiond | 151 | (15.2) | 101 | (13.8) | 50 | (19.2) | 1.33 | (1.03 – 1.71) | .03 |
| LE cellulitis | 9 | (0.9) | 6 | (0.8) | 3 | (1.1) | -- | -- | -- |
| Pneumonia | 73 | (7.3) | 46 | (6.2) | 27 | (10.3) | 1.46 | (1.06 – 2.01) | .02 |
| Sepsis | 38 | (3.8) | 20 | (2.7) | 18 | (6.9) | 1.87 | (1.31 – 2.66) | .002 |
| Surgical site infection | 21 | (2.1) | 13 | (1.7) | 8 | (3.0) | 1.47 | (0.84 – 2.56) | .21 |
| Urinary tract infection | 36 | (3.6) | 27 | (3.6) | 9 | (3.4) | 0.95 | (0.54 – 1.69) | .86 |
| Other Infection | 36 | (3.6) | 23 | (3.1) | 13 | (5.0) | 1.40 | (0.89 – 2.19) | .16 |
| Other complications | |||||||||
| Posthemorrhagic anemia | 128 | (12.9) | 90 | (12.3) | 38 | (14.6) | 1.16 | (0.86 – 1.54) | .33 |
| Blood transfusion | 181 | (18.2) | 139 | (19.0) | 42 | (16.1) | 0.86 | (0.64 – 1.15) | .30 |
| Post-operative medications | |||||||||
| Insulin | 418 | (42.1) | 241 | (32.9) | 177 | (68.1) | 2.92 | (2.33 – 3.67) | <.0001 |
| Other diabetic medicationse | 60 | (6.0) | 29 | (3.9) | 31 | (11.9) | 2.10 | (1.60 – 2.75) | <.0001 |
| Steroids | 121 | (12.2) | 85 | (11.6) | 36 | (13.8) | 1.16 | (0.86 – 1.56) | .34 |
RR = relative risk; CI = confidence interval.
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 (LE cellulitis, pneumonia, sepsis, surgical site infection, urinary tract infection, and other infection).
glyburide, glipizide, glimepiride, metformin, or medications that include these agents in combination.
Multivariable Models
Multivariable logistic regression was used to determine whether hyperglycemia following AAA repair was associated with infection (Table V) or in-hospital death (Table VI). We found no association between post-operative glucose levels and 30-day readmission in unadjusted analyses (p = .90), and therefore did not develop a model for readmission. Models included patient and hospital characteristics, the Charlson Comorbidity Index, procedure type (endovascular or open), associated comorbidities/conditions, and post-operative medications (including steroids, insulin, and other diabetes medications – glyburide, glipizide, glimepiride, metformin and combinations including one of these agents). An interaction term between diabetes and hyperglycemia was tested in each of the models. As no significant interactions were found, no interactions are included in the results.
Table V.
Multivariable logistic regression models for risk factors for infections following AAA repair.
|
Full Sample (n = 237) |
Endovascular
Only (n = 86) |
Open
Only (n = 151) |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | (95% CI) | p-value | OR | (95% CI) | p-value | OR | (95% CI) | p-value | |
| Procedure type (open) | 1.78 | (1.26 – 2.53) | .001 | ||||||
| Post-operative hyperglycemia | 1.40 | (0.97 – 2.02) | .06 | 1.85 | (0.98 – 3.51) | .054 | 1.29 | (0.82 – 2.01) | .27 |
| Age (years) | 0.99 | (0.97 – 1.00) | .20 | 1.00 | (0.97 – 1.02) | .74 | 0.99 | (0.98 – 1.01) | .35 |
| Gender (female) | 1.13 | (0.82 – 1.56) | .43 | 0.97 | (0.54 – 1.74) | .91 | 1.26 | (0.84 – 1.87) | .26 |
| Race (reference = Caucasian) | |||||||||
| African-American | 1.39 | (0.87 – 2.20) | .15 | 1.40 | (0.62 – 3.15) | .41 | 1.39 | (0.78 – 2.48) | .26 |
| Other | 1.73 | (0.89 – 3.37) | .10 | 1.16 | (0.37 – 3.59) | .79 | 2.04 | (0.84 – 4.98) | .11 |
| Charlson Index (reference = 0) | |||||||||
| 1 | 0.85 | (0.38 – 1.88) | .69 | 0.36 | (0.11 – 1.16) | .08 | 1.48 | (0.48 – 4.54) | .49 |
| 2 | 1.56 | (0.70 – 3.47) | .26 | 0.53 | (0.16 – 1.71) | .28 | 3.26 | (1.05 – 10.1) | .04 |
| 3+ | 1.99 | (0.86 – 4.59) | .10 | 1.04 | (0.32 – 3.38) | .95 | 3.07 | (0.92 – 10.2) | .06 |
| Hospital bed size (reference = <200) | |||||||||
| 200–299 | 1.90 | (0.73 – 4.93) | .18 | 1.15 | (0.27 – 4.79) | .85 | 1.97 | (0.53 – 7.30) | .31 |
| 300–499 | 1.62 | (0.64 – 4.10) | .30 | 0.75 | (0.18 – 3.06) | .68 | 2.95 | (0.81 – 10.6) | .10 |
| 500 or more | 1.20 | (0.47 – 3.04) | .69 | 0.59 | (0.14 – 2.44) | .46 | 2.29 | (0.63 – 8.30) | .20 |
| Teaching facility | 1.74 | (1.00 – 3.00) | .04 | 2.64 | (1.11 – 6.29) | .02 | 0.93 | (0.41 – 2.11) | .85 |
| Comorbidities | |||||||||
| Diabetes | 0.57 | (0.37 – 0.88) | .01 | 0.95 | (0.49 – 1.85) | .88 | 0.42 | (0.24 – 0.75) | .003 |
| Fluid and electrolyte disorders | 1.37 | (0.94 – 1.99) | .10 | 2.39 | (1.21 – 4.71) | .01 | |||
| Renal Failure | 2.74 | (1.80 – 4.16) | <.0001 | 3.25 | (1.55 – 6.83) | .001 | 2.60 | (1.54 – 4.38) | .0003 |
| Respiratory problems | 1.66 | (1.13 – 2.44) | .009 | 1.78 | (1.13 – 2.82) | .01 | |||
| Cardiac complications or MI | 2.05 | (1.10 – 3.83) | .02 | 2.39 | (0.78 – 7.36) | .12 | 2.15 | (1.00 – 4.63) | .04 |
| Chronic heart disease | 0.64 | (0.46 – 0.89) | .008 | 0.53 | (0.34 – 0.82) | .004 | |||
| Chronic heart failure | 1.46 | (0.93 – 2.30) | .09 | ||||||
| Posthemorrhagic anemia | 1.77 | (1.15 – 2.72) | .009 | 2.96 | (1.27 – 6.88) | .01 | |||
| Post-operative medications | |||||||||
| Steroids | 2.39 | (1.62 – 3.54) | <.0001 | 3.07 | (1.61 – 5.85) | .0007 | 2.01 | (1.22 – 3.30) | .006 |
| Insulin | 1.57 | (1.12 – 2.20) | .007 | 1.82 | (1.04 – 3.18) | .03 | 1.54 | (1.01 – 2.34) | .04 |
| Other diabetic medicationsa | 0.71 | (0.34 – 1.48) | .37 | 0.67 | (0.23 – 1.93) | .46 | 0.61 | (0.21 – 1.78) | .36 |
OR = odds ratio; 95% CI = 95% confidence interval.
Glyburide, glipizide, glimepiride, metformin, or medications that include these agents in combination.
Table VI.
Multivariable logistic regression models for risk factors for in-hospital mortality following AAA repair.
|
Full Sample (n = 58) |
Endovascular
Only (n = 20) |
Open
Only (n = 38) |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | (95% CI) | p-value | OR | (95% CI) | p-value | OR | (95% CI) | p-value | |
| Procedure type (open) | 1.61 | (0.79 – 3.29) | .18 | ||||||
| Post-operative hyperglycemia | 3.48 | (1.78 – 6.80) | .0003 | 7.54 | (1.95 – 29.1) | .003 | 3.05 | (1.29 – 7.21) | .01 |
| Diabetes | 0.40 | (0.15 – 1.02) | .055 | 0.21 | (0.03 – 1.74) | .14 | 0.43 | (0.13 – 1.39) | .15 |
| Age (years) | 1.07 | (1.03 – 1.11) | <.0001 | 1.15 | (1.05 – 1.25) | .001 | 1.06 | (1.01 – 1.10) | .007 |
| Gender (female) | 1.02 | (0.52 – 2.00) | .95 | 0.31 | (0.07 – 1.31) | .11 | 1.45 | (0.62 – 3.38) | .39 |
| Charlson Index (reference = 0) | |||||||||
| 1 | 0.33 | (0.09 – 1.16) | .08 | 0.47 | (0.04 – 5.78) | .55 | 0.26 | (0.05 – 1.27) | .09 |
| 2 | 0.20 | (0.05 – 0.81) | .02 | 0.16 | (0.01 – 2.96) | .21 | 0.17 | (0.03 – 0.95) | .04 |
| 3+ | 0.34 | (0.08 – 1.46) | .14 | 0.36 | (0.02 – 6.61) | 49 | 0.30 | (0.05 – 1.98) | .21 |
| Hospital bed size (reference = <200) | |||||||||
| 200–299 | 0.65 | (0.11 – 3.77) | .63 | 0.35 | (0.01 – 10.0) | .54 | 0.60 | (0.05 – 6.97) | .68 |
| 300–499 | 0.38 | (0.07 – 2.06) | .26 | 0.20 | (0.01 – 6.33) | .35 | 0.21 | (0.02 – 2.14) | .18 |
| 500 or more | 0.48 | (0.10 – 2.45) | .37 | 0.36 | (0.01 – 9.34) | .54 | 0.28 | (0.03 – 2.67) | .27 |
| Teaching facility | 4.91 | (0.93 – 25.8) | .06 | 2.89 | (0.24 – 35.3) | .40 | 15.09 | (1.22 – 187.2) | .03 |
| Comorbidities/Problems | |||||||||
| Renal Failure | 5.93 | (2.79 – 12.6) | <.0001 | 5.38 | (1.10 – 26.3) | .03 | 7.09 | (2.66 – 18.9) | <.0001 |
| Respiratory Problems | 3.50 | (1.76 – 6.97) | .0004 | 4.15 | (1.75 – 9.82) | .001 | |||
| Cardiac/MI | 4.69 | (2.00 – 10.9) | .0004 | 6.63 | (2.23 – 19.7) | .0007 | |||
| Chronic Heart Failure | 7.92 | (1.66 – 37.7) | .009 | ||||||
| Posthemorrhagic anemia | 13.00 | (2.43 – 69.4) | .002 | ||||||
| Post-operative medications | |||||||||
| Insulin | 3.05 | (1.50 – 6.19) | .002 | 4.17 | (1.13 – 15.4) | .03 | 2.75 | (1.11 – 6.82) | .02 |
| Other diabetic medicationsa | 0.41 | (0.06 – 2.66) | .35 | 0.11 | (0.00 – 5.45) | .26 | 0.49 | (0.05 – 4.92) | .54 |
OR = odds ratio; 95% CI = 95% confidence interval.
Glyburide, glipizide, glimepiride, metformin, or medications that include these agents in combination.
Almost ten percent of the sample had an infectious complication (n = 237, 9.5%). Although not significant, patients with post-operative hyperglycemia who underwent any type of AAA repair had a higher odds of developing an infection (OR 1.40, 95% CI 0.97–2.02, p = .06) than patients with optimal post-operative glucose control after adjusting for patient, procedure, and hospital characteristics (Table V). Hyperglycemic patients who underwent an open AAA repair had higher odds of infection than patients who had endovascular repair (OR 1.78, 95% CI 1.26–2.53, p=.001). Additional factors associated with infection after any AAA repair included acute renal failure (OR 2.74, 95% CI 1.80–4.16), acute respiratory problems (OR 1.66, 95% CI 1.13–2.44), MI or other cardiac complications (OR 2.05, 95% CI 1.10–3.83), posthemorrhagic anemia (OR 1.77, 95% CI 1.15–2.72), receipt of post-operative steroids (OR 2.39, 95% CI 1.62–3.54), and post-operative receipt of insulin (OR 1.57, 95% CI 1.12–2.20). Having a diabetes diagnosis was associated with lower odds of infection (OR 0.57, 95% CI 0.37–0.88), while receiving other diabetes medications was not associated (p=.37). Model discrimination was good (c-statistic = 0.80), and calibration was adequate (p = .84).
After an endovascular AAA repair was performed, the association between infectious complications and post-operative hyperglycemia was not statistically significant (OR 1.85, 95% CI 0.98–3.51, p = .054; Table V). Several factors were associated with infectious complications in endovascular AAA repair patients: acute renal failure (OR 3.25, 95% CI 1.55–6.83), receipt of post-operative steroids (OR 3.07, 95% CI 1.61–5.85), posthemorrhagic anemia (OR 2.96, 95% CI 1.27–6.88), fluid and electrolyte disorders (OR 2.39, 95% CI 1.21–4.71), and post-operative receipt of insulin for type I diabetes (OR 1.82, 95% CI 1.04–3.18). Model discrimination was good (c-statistic = 0.84) and model calibration was adequate (p = .79).
After open AAA repair, patients with post-operative hyperglycemia did not have greater odds of infection (OR 1.29, 95% CI 0.82–2.01, p = .27) compared with patients who had optimal glucose control (Table V). However, acute renal failure (OR 2.60, 95% CI 1.54–4.38), receipt of post-operative steroids (OR 2.01, 95% CI 1.22–3.30), respiratory problems (OR 1.78, 95% CI 1.13–2.82), and post-operative receipt of insulin (OR 1.54, 95% CI 1.01–2.34) were associated with greater odds of infection, while a diabetes diagnosis was associated with lower odds (OR 0.42, 95% CI 0.24–0.75). Model discrimination was moderate (c = 0.76) and calibration was good (p = .22).
In-hospital mortality was relatively rare (n = 58, 2.3%). For patients undergoing any AAA repair procedure (Table VI), there was a strong association of post-operative hyperglycemia with in-hospital mortality (OR 3.48, 95% CI 1.78–6.80). The association between diabetes diagnosis and lower mortality was not statistically significant (OR 0.40, 95% CI 0.15–1.02, p=.055). The following variables were also associated with mortality: acute renal failure (OR 5.93, 95% CI 2.79–12.6), cardiac complications or MI (OR 4.69, 95% CI 2.00–10.9), respiratory problems (OR 3.50, 95% CI 1.76–6.97), receipt of post-operative insulin (OR 3.05, 95% CI 1.50–6.19), and being older (OR 1.07, 95% CI 1.03–1.11). Model discrimination was good (c = 0.95) and calibration was adequate (p = .86). When the procedure types were modeled separately (Table VI), patients with post-operative hyperglycemia had higher odds of dying following both procedures than those with optimal post-operative glucose control (EVAR: OR 7.54, 95% CI 1.95–29.1; open: OR 3.05, 95% CI 1.29–7.21). Post-operative insulin receipt was associated with higher odds of death, while diabetes diagnosis was not.
Discussion
We found that postoperative hyperglycemia was common among patients who underwent AAA repair, affecting one of every six patients. After adjusting for hospital and patient characteristics, including diabetes diagnosis and medications, hyperglycemia remained significantly associated with in-hospital mortality. The association with infections was just shy of statistical significance for all patients (p=.06) and for patients who underwent endovascular repair (p=.054). Receiving insulin in the post-operative period was also associated with increased odds of infection or mortality, while receipt of other diabetes medications was not. Even after controlling for insulin administration and post-operative hyperglycemia, having diabetes was associated with lower chances of infection or in-hospital mortality. This analysis is not implying causation, but highlights that hyperglycemia in the postoperative period is associated with increased complications. Therefore, those patients receiving insulin likely had hyperglycemia leading to inferior outcomes. The finding that diabetic patients did not have increased infection or hospital mortality suggests that the non-diabetic patients that develops hyperglycemia may be at greatest risk and warrants further investigation using prospective methods.
Hyperglycemia is relatively common in the postoperative period. It is directly related to insulin resistance as a consequence of the hypermetabolic stress response, which creates imbalance between endogenous hepatic glucose production and peripheral glucose uptake (insulin-mediated).(18) Even though this response was initially referred as an adaptive mechanism to critical illness, today there is supportive literature with evidence indicating that acute postoperative hyperglycemia can be a potential cause for mortality and severe morbidity.(19)
Hyperglycemia after surgery has been associated with poor outcomes in the general surgery patients, non-cardiac surgery procedures, and patients who undergo vascular lower extremity procedures.(12, 20–22) In this analysis, poor glycemic control was associated with longer hospitalizations, an increased number of nosocomial infections, and increased hospital mortality. It has been recently shown prospectively in a single center series that for open AAA, a postoperative insulin protocol posed low risk to patients and was associated with reduced length of stay.(23)
In this analysis, patients undergoing an open AAA repair had higher infection rates compared to patients undergoing EVAR (15.2% vs. 5.8%, p=.0002). In our analysis, factors associated with developing an infection were open AAA repair, postoperative hyperglycemia, renal failure, respiratory complications, myocardial infarction, posthemorrhagic anemia, and use of postoperative steroids or insulin administration.
Overall, patients with suboptimal glucose control had a mortality rate of 8.4% compared 1.2% for patients with optimal post-operative glucose control. Subset analysis determined that patients with postoperative hyperglycemia following an EVAR procedure had seven times the odds of hospital mortality compared to those with optimized glucose management, while those undergoing open repair had almost four times the odds. Given the shorter length of stay (3.2 vs 9.1 days, respectively) and fewer serum glucose tests (2.3 vs. 6.0, respectively) following EVAR than open procedures, there may be less opportunity to identify poor glucose control. The administration of insulin in the postoperative period was associated with increased rates infection for open as well as endovascular AAA repair. In multivariable models, the association between hyperglycemia and infections did not reach significant in the overall population and for patients who had endovascular AAA repair. The use of insulin suggests that patients with hyperglycemia were more likely to develop and infection, which was found to be significant for all groups, studied using multivariable logistic regression models. Patients who received EVAR, however, had higher odds of infection if they had postoperative hyperglycemia. Interestingly, following open procedures, diabetes was not associated with increased infection rates. It has been suggested that, in general, insulin treatment is more consistently administered in diabetic patients with perioperative hyperglycemia.(24–26) We found a weaker association between postoperative hyperglycemia and infections for open AAA repair than EVAR, perhaps due to increased blood glucose measures and more frequent treatment with insulin among open patients. In contrast, a recent single center prospective study using a standardized insulin infusion protocol after open AAA repair did not find reduced complications, including surgical site infection, with optimal perioperative glycemic control. This study concluded that postoperative hyperglycemia is common after bypass and AAA repair and can be effectively managed with an insulin infusion protocol.(23) The protocol was found to be low risk and was associated with reduced LOS and cost. The current analysis did not look at the introduction of an insulin protocol, but did find that the administration of insulin in the postoperative period was associated with infectious complications and mortality. Reasons for this are unclear, but may be that hyperglycemia is a marker for patients who are sicker leading to hyperglycemia. As well, this is not implying causation, but highlights that hyperglycemia in the postoperative period is associated with increased complications.
Frisch et al. reported that mortality and hyperglycemia were highly correlated, concluding that perioperative hyperglycemia was associated with increased LOS, hospital complications, and mortality after noncardiac general surgery.(27) van Kyijk et al. recommended that non-diabetic vascular surgery patients should be tested for glucose regulation disorders before surgery and the authors reported that patients with no history of diabetes but with impaired glucose tolerance had an increased risk of cardiovascular complications and mortality in the postoperative period.(25) After controlling for insulin administration and post-operative hyperglycemia, a diagnosis of diabetes had decreased odds of infection and overall mortality.
Kotagal et al. reported that only forty percent of nondiabetic patients with blood glucose levels of 180–250 mg/dL and 55 percent of those with blood glucose above 250mg/dL received insulin therapy in the early postoperative period.(24) Insulin treatment may have a different effect on non-diabetic patients undergoing surgery, with subsequent and more pronounced inflammation and oxidative stress, worsened atherosclerosis, and a hypercoagulable state.(24, 26, 28) It has been suggested that chronic exposure to hyperglycemia, may allow patients with diabetes to better tolerate the blood glucose fluctuation.(24)
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 events. As well, as the assignment of diagnosis codes can be inexact and for these reasons, we did not separate Type I and Type II diabetes in this analysis. We did not have data on operative time or estimated blood loss, which could have biased the results, but have included whether the patients received a blood transfusion in the hospital. As well, almost fifty percent of all of the encounters in the data set had blood glucose measurements, which may be associated with selection bias based on who received laboratory evaluations during her hospital stay. Furthermore, preoperative glucose values were not available for the majority of patients in this data set, making it impossible to evaluate change in serum glucose. We further assume that if point-of-care testing was performed, these data were contained within the EMR; if these data were not entered into the EMR, this could add bias to the analysis. No outpatient facilities are contained within the data set. In addition, we are unable to determine readmissions to hospitals in different health systems. As Cerner Corporations’ Health Facts is a proprietary database comprised of electronic clinical records from hospitals and hospital systems that use Cerner’s electronic health record, the patient population may not be completely reflective of the US population.
In conclusion, this analysis has revealed that postoperative hyperglycemia occurs often after both open AAA repair and EVAR. After adjustment for age, gender, comorbidities and diabetes medications, hyperglycemia remained significantly associated with death after AAA repair. Hyperglycemia was linked with hospital mortality and was correlated with mortality including cardiac and respiratory complications. A diagnosis of diabetes by itself was not associated with increased risk of infection, in-hospital morbidity, or mortality. Postoperative hyperglycemia as a clinical marker was associated found to be with inferior outcomes, specifically for non-diabetic patients. Patients undergoing AAA repair, particularly EVAR, should receive glycemic monitoring in the postoperative period to allow for treatment of hyperglycemia, which may decrease risk of postoperative complications and mortality. We suggest further investigation using prospective methods to more fully explore the association of hyperglycemia with postoperative outcomes and potential confounders.
Acknowledgements:
Research reported in this publication was supported by the Agency for Healthcare Research and Quality (R24HS022140). The content is solely the responsibility of the authors and does not necessarily represent the official views of the Agency for Healthcare Research and Quality.
Footnotes
Presented at the 41st Annual Midwestern Vascular Surgical Society, Plenary Session, Chicago, Illinois, September 7–9, 2017
References
- 1.Kwon S, Thompson R, Dellinger P, Yanez D, Farrohki E, Flum D. Importance of perioperative glycemic control in general surgery: a report from the Surgical Care and Outcomes Assessment Program. Ann Surg 2013;257(1):8–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kiran RP, Turina M, Hammel J, Fazio V. The clinical significance of an elevated postoperative glucose value in nondiabetic patients after colorectal surgery: evidence for the need for tight glucose control? Ann Surg 2013;258(4):599–604. [DOI] [PubMed] [Google Scholar]
- 3.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]
- 4.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. [DOI] [PubMed] [Google Scholar]
- 5.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]
- 6.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]
- 7.Krinsley JS. Effect of an intensive glucose management protocol on the mortality of critically ill adult patients. Mayo Clin Proc 2004;79(8):992–1000. [DOI] [PubMed] [Google Scholar]
- 8.Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitabchi AE. Hyperglycemia: an independent marker of in-hospital mortality in patients with undiagnosed diabetes. J Clin Endocrinol Metab 2002;87(3):978–82. [DOI] [PubMed] [Google Scholar]
- 9.Vriesendorp TM, Morelis QJ, Devries JH, Legemate DA, Hoekstra JB. Early post-operative glucose levels are an independent risk factor for infection after peripheral vascular surgery. A retrospective study. Eur J Vasc Endovasc Surg 2004;28(5):520–5. [DOI] [PubMed] [Google Scholar]
- 10.DeShazo JP, Hoffman MA. A comparison of a multistate inpatient EHR database to the HCUP Nationwide Inpatient Sample. BMC Health Services Research 2015;15:384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.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]
- 12.Vogel TR, Smith JB, Kruse RL. The association of postoperative glycemic control and lower extremity procedure outcomes. J Vasc Surg 2017;66(4):1123–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.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 Int Med 2012;172(3):246–53. [DOI] [PubMed] [Google Scholar]
- 14.Nandan AR, Bohnen JD, Chang DC, Yeh DD, Lee J, Velmahos GC, et al. The impact of major intraoperative adverse events on hospital readmissions. Am J Surg 2017;213(1):10–7. [DOI] [PubMed] [Google Scholar]
- 15.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]
- 16.Moghissi ES, Korytkowski MT, DiNardo M, Einhorn D, Hellman R, Hirsch IB, et al. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Endocr Pract 2009;15(4):353–69. [DOI] [PubMed] [Google Scholar]
- 17.Umpierrez GE, Hellman R, Korytkowski MT, Kosiborod M, Maynard GA, Montori VM, et al. Management of hyperglycemia in hospitalized patients in non-critical care setting: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 2012;97(1):16–38. [DOI] [PubMed] [Google Scholar]
- 18.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]
- 19.Duncan AE. Hyperglycemia and perioperative glucose management. Curr Pharm Des 2012;18(38):6195–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.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]
- 21.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]
- 22.Jacobi J, Bircher N, Krinsley J, Agus M, Braithwaite SS, Deutschman C, et al. Guidelines for the use of an insulin infusion for the management of hyperglycemia in critically ill patients. Crit Care Med 2012;40(12):3251–76. [DOI] [PubMed] [Google Scholar]
- 23.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]
- 24.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]
- 25.van Kuijk JP, Dunkelgrun M, Schreiner F, Flu WJ, Galal W, van Domburg RT, et al. Preoperative oral glucose tolerance testing in vascular surgery patients: long-term cardiovascular outcome. Am Heart J 2009;157(5):919–25. [DOI] [PubMed] [Google Scholar]
- 26.Fiorillo C, Rosa F, Quero G, Menghi R, Doglietto GB, Alfieri S. Postoperative hyperglycemia in nondiabetic patients after gastric surgery for cancer: perioperative outcomes. Gastric Cancer 2016. [DOI] [PubMed] [Google Scholar]
- 27.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]
- 28.NICE-SUGAR Study Investigators, Finfer S, Chittock DR, Su SY, Blair D, Foster D, et al. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360(13):1283–97. [DOI] [PubMed] [Google Scholar]