Secondary Logo

Severe Intraoperative Hyperglycemia Is Independently Associated With Postoperative Composite Infection After Craniotomy

An Observational Study

Gruenbaum, Shaun E., MD*; Toscani, Laura, MD; Fomberstein, Kenneth M., MD; Ruskin, Keith J., MD§; Dai, Feng, PhD*; Qeva, Ega, MD; Rosa, Giovanni, MD; Meng, Lingzhong, MD*; Bilotta, Federico, MD, PhD

doi: 10.1213/ANE.0000000000001946
Neuroscience and Neuroanesthesiology: Original Clinical Research Report
Free

BACKGROUND: Postoperative infection after craniotomy carries an increased risk of morbidity and mortality. Identification and correction of the risk factors should be prioritized. The association of intraoperative hyperglycemia with postoperative infections in patients undergoing craniotomy is inadequately studied.

METHODS: A total of 224 patients were prospectively enrolled in 2 major medical centers to assess whether severe intraoperative hyperglycemia (SIH, blood glucose ≥180 mg/dL) is associated with an increased risk of postoperative infection in patients undergoing craniotomy. Arterial blood samples were drawn and analyzed immediately after anesthetic induction and again before tracheal extubation. The new onset of any type of infection within 7 days after craniotomy was determined.

RESULTS: The incidence of new postoperative composite infection was 10% (n = 22) within the first week after craniotomy. Weight, sex, American Society of Anesthesiologists score, preoperative and/or intraoperative steroid use, and diabetes mellitus were not associated with postoperative infection. SIH was independently associated with postoperative infection (odds ratio [95% confidence interval], 4.17 [1.50–11.56], P = .006) after fitting a multiple logistic regression model to adjust for emergency surgery, length of surgery, and age ≥65 years.

CONCLUSIONS: SIH is independently associated with postoperative new-onset composite infections in patients undergoing craniotomy. Whether prevention of SIH during craniotomy results in a reduced postoperative risk of infection is unknown and needs to be appraised by further study.

Published ahead of print February 8, 2017.

From the *Department of Anesthesiology, Yale University School of Medicine, New Haven, Connecticut; Department of Anesthesiology, Critical Care and Pain Medicine, Sapienza University of Rome, Rome, Italy; Department of Anesthesiology, New York Medical College, Valhalla, New York; and §Department of Anesthesia and Critical Care, The University of Chicago, Chicago, Illinois.

Published ahead of print February 8, 2017.

Accepted for publication January 6, 2017.

Funding: None.

The authors declare no conflicts of interest.

Reprints will not be available from the authors.

Address correspondence to Fedrico Bilotta, MD, PhD, Department of Anesthesiology, Critical Care and Pain Medicine, Sapienza University of Rome, Rome, Italy. Address e-mail to bilotta@tiscal.it.

Infection after craniotomy is a common and potentially fatal complication.1 Large epidemiologic studies have shown that up to 7% of patients develop a surgical site infection after neurosurgical procedures, and the incidence of extracranial infections (such as pulmonary, blood, and urinary) may be even higher.1 Infection increases the length of hospitalization and associated costs of medical care and carries an increased risk of morbidity and mortality.2,3 It is important to identify potentially modifiable factors that may contribute to postcraniotomy infection. Indeed, clinical pathways that aim to identify and correct the preventable contributing factors of perioperative complications (including infection of any type) have been advocated for ultimately improving patient outcomes.4

In recent years, glycemic control has gained much interest after diabetic patients were observed to be susceptible to both infectious and noninfectious perioperative complications.5 Hyperglycemia during neurosurgical procedures is common in both diabetic and nondiabetic patients, which may be the result of factors such as corticosteroid use and stress response in the perioperative setting.6,7 Previous studies in both diabetic and nondiabetic neurosurgical patients have shown that hyperglycemia is associated with poor outcomes including increased duration of hospital stay and higher mortality.6 There is, however, a paucity of data on whether intraoperative hyperglycemia is independently associated with all types of infectious complications after craniotomy.

In this multicenter and prospective single-cohort observational study, we assessed the association of severe intraoperative hyperglycemia (SIH) with the occurrence of composite infections after craniotomy. We hypothesized that the presence of SIH (blood glucose concentration [BGC] ≥180 mg/dL) is independently associated with postcraniotomy infection.

Back to Top | Article Outline

METHODS

Dr Bilotta registered this study before patient enrollment in http://www.clinicaltrials.gov (June 13, 2014, registration number NCT02165748). The study was conducted according to the recommendations set by the Helsinki committee, and it was approved by the Human Investigation Committees at the Sapienza University of Rome in Rome, Italy, and the Yale University School of Medicine in New Haven, Connecticut. This manuscript adheres to the applicable EQUATOR guidelines.

Back to Top | Article Outline

Patients

Adult patients who were ≥18 years of age and scheduled to undergo elective or emergency craniotomy were prospectively enrolled in the study between May 2013 and January 2016. Patients who were <18 years of age or with a known infection at the time of surgery were excluded. Written and verbal consents were obtained from the eligible patients or patients’ surrogates before surgery. All patients enrolled in the study received prophylactic antibiotics within 1 hour before surgical incision.

Back to Top | Article Outline

Intraoperative Blood Glucose Analysis

Intraoperative BGC was determined by extracting 1 mL of whole blood from an intra-arterial catheter at 2 time points during surgery: once immediately after induction of anesthesia and once at the end of surgery, but before tracheal extubation. BGCs were determined by arterial blood glucose analysis (Instrumentation Laboratory, GEM Premier 4000). SIH was defined as at least 1 arterial sample with a BGC ≥180 mg/dL. The prevalence of intraoperative hyperglycemia was defined as the number of hyperglycemic patients divided by the total number of patients enrolled in the study. The decision of how and when to initiate insulin therapy to manage SIH was made at the discretion of the anesthesia teams caring for the patients.

Back to Top | Article Outline

Potential Risk Factors and Infectious Complications

Patients’ demographics data including sex, age, and weight were collected as well as potential risk factors of infection to control for confounding variables, including American Society of Anesthesiologists classification, history of diabetes mellitus, recent corticosteroid use (within 6 weeks before surgery), intraoperative corticosteroid use, emergent versus elective surgery, and length of surgery. The new onset of composite infection within 7 days after surgery was determined by medical record review. Composite infection was defined as a combination of all types of newly diagnosed infection after surgery (ie, lung, wound, blood, etc). The diagnosis of specific infections was determined in accordance with the surveillance definitions and criteria established by Centers for Disease Control and Prevention, in which 14 major types of infection are organized according to body systems.8

Back to Top | Article Outline

Statistical Analysis

Patient demographics and clinical characteristics were summarized using mean (standard deviation) or median (interquartile range) for continuous variables and n (%) for categorical variables. A univariate analysis was performed to compare differences between patients with and without infection: a 2-sample ttest or Wilcoxon rank-sum test was used for continuous variables, and a χ2 or Fisher exact test was used for categorical variables. A multiple logistic regression analysis was then performed to assess the association between SIH (BGC ≥180 vs <180 mg/dL) and postoperative infection after adjusting for factors that were found statistically significant (P < .05) in the univariate analysis. All statistical analyses were performed using SAS version 9.4 (Cary, NC, Copyright 2013, SAS Institute Inc.), and a 2-sided P value < .05 was considered to be statistically significant.

On the basis of our clinical experience, we estimated that approximately 20% of patients might experience SIH (1:4 ratio of patients with SIH to normoglycemic patients). Although there are limited data on the incidence of SIH in patients undergoing craniotomy, previous studies have demonstrated approximately a 5% rate of postoperative infections in patients who receive antibiotic prophylaxis.9 Assuming the 5% rate of postoperative infection in the normoglycemic group and a 2-sided significance level of .05, it was determined that 30 patients in the SIH group and 120 patients in the normoglycemic group (1:4 ratio) would have a 80% power to detect an absolute rate difference of 18.4% between the 2 groups (ie, severe hyperglycemic and normoglycemic groups with infection rates of 23.4% and 5%, respectively). Power analyses were determined before patient enrollment.

Back to Top | Article Outline

RESULTS

Patient and Surgery Characteristics

Table 1

Table 1

A total of 224 patients were enrolled in the study. Patient demographics and clinical characteristics categorized by infection or no infection are summarized in Table 1. Thirty-eight of these patients (17.0%) had at least 1 episode of SIH and 186 (83.0%) did not. SIH was demonstrated by only the first blood sample in 5 of 38 patients (13.2%), by only the second sample in 26 patients (68.4%), and in both samples in the remaining 7 patients (14.2%). An IV insulin bolus was administered to 7 of these 38 patients (18.4%). The indication for craniotomy included primary tumor (n = 144 [64.3%]), neurovascular lesions (n = 38 [17.0%]), metastatic tumor (n = 19 [8.5%]), temporal lobectomy for epilepsy (n = 12 [5.4%]), and traumatic brain injury (n = 11 [4.9%]). A total of 39 cases (17.4%) were emergent.

Back to Top | Article Outline

Association of Intraoperative Hyperglycemia With Postoperative Infections

Table 2

Table 2

A total of 22 of 224 (9.8%) patients developed an infection in the first week after craniotomy. Ten of 38 (26.3%) hyperglycemic patients developed an infection compared with 12 of 186 (6.5%) normoglycemic patients (P < .001; Table 1). Infections included pulmonary (n = 7 [31.8%]), urinary (n = 9 [40.9%]), wound (n = 4 [18.2%]), and blood (n = 2 [9.1%]). Factors associated with postoperative infection were emergency surgery (P = .001), length of surgery (P = .013), age ≥65 years (P = .043), and BGC ≥180 mg/dL (P < .001) based on univariate analysis and using a P value cutoff point of .05. Weight, sex, American Society of Anesthesiologists score, preoperative and/or intraoperative steroid use, and diabetes mellitus were not associated with postoperative infection (Table 1). Of note, all diagnoses of diabetes mellitus in this study were type II. SIH (BGC ≥180 mg/dL) was independently associated with postoperative infection (odds ratio [95% confidence interval], 4.17 [1.50–11.56], P = .006) after fitting a multiple logistic regression model to adjust for emergency surgery, length of surgery, and age ≥65 years (Table 2). As a sensitivity analysis, a nonparsimonious multiple logistic regression model was fit in which variables with P < .3 were adjusted in the univariate analysis that included sex, American Society of Anesthesiologists score, intraoperative steroid use, emergency surgery, length of surgery, and age ≥65 years. The positive association between SIH and postoperative infection remained unchanged (odds ratio [95% confidence interval], 4.12 [1.46–11.59], P = .007, data not shown).

Back to Top | Article Outline

DISCUSSION

This study demonstrated that SIH (BGC ≥180 mg/dL) is independently associated with postoperative composite infection in the first week after craniotomy after adjusting for several confounding variables including emergency surgery, length of surgery, and age ≥65 years. Importantly, however, it should be noted that the study design did not allow for sufficient adjustment for all confounding variables and might therefore be subject to bias. Weight, sex, American Society of Anesthesiologists score, preoperative and/or intraoperative corticosteroid use, and diabetes mellitus were not associated with postoperative infection. The overall incidence of postoperative composite infection was 10% in this study, in which 6.45% of normoglycemic patients and 26.32% of hyperglycemic patients developed an infection.

Postoperative infection is considered a major complication after craniotomy. Identification and correction of the modifiable risk factors associated with postcraniotomy infection may significantly impact patient outcomes. To date, there is limited and conflicting data on postoperative infection and its associated risk factors in neurosurgical patients. Importantly, most previous studies focused entirely on surgical site infections (negating extracranial infections altogether) were limited by either the retrospective study design or a relatively small sample size and have yielded largely inconsistent results. With the exception of antibiotic prophylaxis,2 there are no management guidelines that are known to reduce the risk of postoperative infection in patients undergoing craniotomy.

In critically ill neurosurgical patients, several studies have demonstrated the association of hyperglycemia with worsened neurologic outcomes and increased mortality.10–13 Perioperative hyperglycemia during neurosurgical procedures is common in both diabetic and nondiabetic patients6,7 and is associated with an increased risk of adverse events.6,14 In this study, the presence of even a single episode of SIH was associated with postoperative infection after craniotomy. Importantly, however, the presence of a single episode of SIH is likely indicative of more frequent episodes of hyperglycemia in these patients, which may be easily overlooked and underappreciated. SIH might result from the presence of underlying diabetes mellitus (and other glucose-intolerant conditions) that was not diagnosed and was therefore untreated. This is an important point, because previous studies have demonstrated that the presence of hyperglycemia in patients without diabetes may have a higher 1-year mortality rate than patients with diagnosed diabetes mellitus.15 Although hemoglobin A1C is probably the best screening tool to test for diabetes mellitus, it is infrequently measured in patients before craniotomy.

It has been suggested that hyperglycemia directly impairs the normal host defense mechanisms by a variety of mechanisms that include increasing vascular permeability and promoting edema, interfering with protein C3 of the complement cascade, and decreasing lymphocyte numbers and function.16 To date, only 1 study investigated the association between BGC and infection in patients undergoing craniotomy, and the study failed to demonstrate an association.17 However, that study was retrospective in design, only examined surgical site infection, and was limited to patients who underwent craniotomy for tumor resection only. In contrast, studies in patients undergoing cardiac surgery have demonstrated an association between perioperative hyperglycemia and all types of postoperative infection.16,18–20

A better understanding of the relationship between SIH and postoperative infection facilitates the clinical decision-making when taking care of patients undergoing craniotomy. Tight BGC control (80–110 mg/dL) carries a significant risk of hypoglycemia and increased mortality,5,21,22 and it is not recommended. Although the precise optimal range of BGC is unknown, some authors have recommended that a moderate target BGC (140–180 mg/dL) should be maintained in patients with neurologically critical conditions.23–25 However, there is currently no consensus for the target of BGC control during an intracranial procedure.6

Moreover, there are currently no guidelines that dictate the optimal method by which glucose should be managed in the perioperative period, and much of the management is clinician-dependent or institution-dependent. We recommend that BGC be intraoperatively maintained under 180 mg/dL with frequent BGC monitoring and continuous insulin infusion, which has been shown to reduce glucose variability compared with intravenous insulin boluses.26 Closed-loop glycemic control systems, in combination with continuous glucose monitoring, are an attractive experimental method by which computer algorithms are used to administer glucose and insulin to meet and maintain target BGC. Hybrid closed-loop insulin delivery systems, in which automated interprandial insulin administration is combined with user-delivered premeal insulin boluses, have also been recently investigated.27 Although some studies have demonstrated their accuracy and reliability, the cost-effectiveness and utility of such systems have yet to be determined.28

In our study, several risk factors of postoperative infection after craniotomy were examined for the purposes of controlling for confounding variables. We demonstrated that length of surgery and emergency surgery were associated, albeit nonindependently, with postoperative infection. Previous studies have demonstrated that a longer duration of surgery29,30 and emergency surgery29 were both independently associated with an increased risk of surgical site infection after neurosurgical procedures. In this study, the relatively small sample size may possibly account for the lack of an independent association between these variables.

Preoperative or intraoperative steroid administration was not associated with postoperative infection in this study. Several studies have previously examined the use of corticosteroids and postoperative infection in neurosurgical and nonneurosurgical procedures with conflicting results. In 1 observational study in the general surgery population, steroids were associated with a significantly increased risk of infection and mortality.31 However, a recent multicenter, randomized, double-blind, placebo-controlled trial in patients undergoing cardiac surgery demonstrated that steroid administration resulted in a decreased incidence of postoperative infection, duration of postoperative mechanical ventilation, and lengths of intensive care unit and hospital stays.32 The few studies examining the association between steroid use and infection in the neurosurgical population have yielded similar conflicting results.33–36

Our study showed that age ≥65 years was associated with postoperative infection (albeit not independently), whereas diabetes mellitus was not. Previous studies that examined these associations have yielded inconsistent results. Increasing age was previously shown to be independently associated with an increased risk of surgical site infection after neurosurgical37 and nonneurosurgical procedures.38 In contrast, a different study showed that patients >50 years had a lower risk of developing surgical site infection after neurosurgical procedures compared with patients <50 years old.30 Likewise, some studies showed an increased risk of infection in diabetic patients undergoing neurosurgical37,39 and nonneurosurgical38 procedures, whereas others failed to show this association.1 However, it should be noted that the presence or absence of diabetes mellitus in this study was determined via retrospective chart review of the medical records after craniotomy, and it is possible that diabetes mellitus was underdiagnosed in this study population. This is particularly true of patients who underwent emergency surgery, in whom an extensive preoperative workup was not performed. However, as discussed previously, hemoglobin A1C levels are infrequently measured in patients before elective craniotomy, and therefore the presence of diabetes mellitus may also be missed in these patients.

Our study had several limitations. The relatively small sample size precludes adequate subgroup analyses. It is currently unknown whether specific subsets of patients (ie, patients with metastatic tumor) are more susceptible to increases in BGC and are at an increased risk of postoperative infection. Perhaps most importantly, because there was a relatively small number of postoperative infections relative to the number of confounding variables, it was not possible to adequately adjust for confounding variables. As such, the association between SIH and postoperative infection is subjected to potential bias. It should be noted that a parsimonious model was not favored with such few significant factors, in which adjusting for confounding might be insufficient. Because the infection outcome is rare, this study is limited in its ability to adjust for more covariates given the general rule of 5 to 10 events per predictor variable in a logistic regression analysis.40,41 Finally, we are not able to determine the causal effect of SIH on postoperative infection as a result of the single-cohort and nonrandomized design of this study.

The findings in this study raise many questions that should be investigated in future studies. First, it is not known for certain why some patients developed SIH and others did not and whether the increased rate of infection was directly attributable to the presence of SIH. Moreover, it is unknown whether patients with hyperglycemia at induction of anesthesia had worse outcomes compared with patients who only developed hyperglycemia at the end of surgery. It is also unknown whether the duration or severity of hyperglycemia affects outcomes. These questions clearly highlight the need for a larger follow-up prospective study in which patients are randomized to receive either improved intraoperative glucose control or standard therapy and postoperative infection rates are determined.

In summary, we demonstrated that a single episode of SIH (BGC ≥180 mg/dL) is independently associated with postoperative infection in the first 7 days after craniotomy. Although the study design and analysis did not allow for sufficient adjustment of confounding variables, and might therefore be subject to bias, this study suggests that BGC monitoring should be more routinely used in patients undergoing craniotomy. Whether the practice of maintaining BGC under 180 mg/dL during craniotomy would reduce the risk of postoperative infection needs to be appraised by further study.

Back to Top | Article Outline

ACKNOWLEDGMENTS

S.G. and F.D. are supported by CTSA Grant Number UL1 TR000142 from the National Center for Advancing Translational Science (NCATS), a component of the National Institutes of Health (NIH). F.D. is also supported by National Institutes of Health (NIH)/NIAID R01AI14780. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NIH.

Back to Top | Article Outline

DISCLOSURES

Name: Shaun E. Gruenbaum, MD.

Contribution: This author helped conceive the study, collect and analyze the data, and write and edit the manuscript.

Name: Laura Toscani, MD.

Contribution: This author helped collect and analyze the data, and write and edit the manuscript.

Name: Kenneth M. Fomberstein, MD.

Contribution: This author helped collect and analyze the data, interpret the results, and edit the manuscript.

Name: Keith J. Ruskin, MD.

Contribution: This author helped analyze the data, interpret the results, and edit the manuscript.

Name: Feng Dai, PhD.

Contribution: This author helped analyze the statistics, interpret the results, and write and edit the manuscript.

Name: Ega Qeva, MD.

Contribution: This author helped analyze the data, interpret the results, and edit the manuscript.

Name: Giovanni Rosa, MD.

Contribution: This author helped analyze the data, interpret the results, and edit the manuscript.

Name: Lingzhong Meng, MD.

Contribution: This author helped analyze the data, interpret the results, and edit the manuscript.

Name: Federico Bilotta, MD, PhD.

Contribution: This author helped conceive the study, collect and analyze the data, and write and edit the manuscript.

This manuscript was handled by: Gregory J. Crosby, MD.

Back to Top | Article Outline

REFERENCES

1. McClelland S 3rd, Hall WA. Postoperative central nervous system infection: incidence and associated factors in 2111 neurosurgical procedures. Clin Infect Dis. 2007;45:55–59.
2. Dashti SR, Baharvahdat H, Spetzler RF, et alOperative intracranial infection following craniotomy. Neurosurg Focus. 2008;24:E10.
3. van Vught LA, Wiewel MA, Klein Klouwenberg PM, et alMolecular Diagnosis and Risk Stratification of Sepsis Consortium. Admission hyperglycemia in critically ill sepsis patients: association with outcome and host response. Crit Care Med. 2016;44:1338–1346.
4. Smith KA, Matthews TW, Dubé M, Spence G, Dort JC. Changing practice and improving care using a low-risk tracheotomy clinical pathway. JAMA Otolaryngol Head Neck Surg. 2014;140:630–634.
5. Russo N. Perioperative glycemic control. Anesthesiol Clin. 2012;30:445–466.
6. Godoy DA, Di Napoli M, Biestro A, Lenhardt R. Perioperative glucose control in neurosurgical patients. Anesthesiol Res Pract. 2012;2012:690362.
7. Lukins MB, Manninen PH. Hyperglycemia in patients administered dexamethasone for craniotomy. Anesth Analg. 2005;100:1129–1133.
9. Korinek AM, Golmard JL, Elcheick A, et alRisk factors for neurosurgical site infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on 4,578 patients. Br J Neurosurg. 2005;19:155–162.
10. Schlenk F, Vajkoczy P, Sarrafzadeh A. Inpatient hyperglycemia following aneurysmal subarachnoid hemorrhage: relation to cerebral metabolism and outcome. Neurocrit Care. 2009;11:56–63.
11. Mowery NT, Gunter OL, Guillamondegui O, et alStress insulin resistance is a marker for mortality in traumatic brain injury. J Trauma. 2009;66:145–151.
12. Griesdale DE, Tremblay MH, McEwen J, Chittock DR. Glucose control and mortality in patients with severe traumatic brain injury. Neurocrit Care. 2009;11:311–316.
13. McGirt MJ, Chaichana KL, Gathinji M, et alPersistent outpatient hyperglycemia is independently associated with decreased survival after primary resection of malignant brain astrocytomas. Neurosurgery. 2008;63:286–291.
14. Bilotta F, Rosa G. Glucose management in the neurosurgical patient: are we yet any closer? Curr Opin Anaesthesiol. 2010;23:539–543.
15. Abdelmalak BB, Knittel J, Abdelmalak JB, et alPreoperative blood glucose concentrations and postoperative outcomes after elective non-cardiac surgery: an observational study. Br J Anaesth. 2014;112:79–88.
16. Golden SH, Peart-Vigilance C, Kao WH, Brancati FL. Perioperative glycemic control and the risk of infectious complications in a cohort of adults with diabetes. Diabetes Care. 1999;22:1408–1414.
17. Hardy SJ, Nowacki AS, Bertin M, Weil RJ. Absence of an association between glucose levels and surgical site infections in patients undergoing craniotomies for brain tumors. J Neurosurg. 2010;113:161–166.
18. Gandhi GY, Nuttall GA, Abel MD, et alIntraoperative hyperglycemia and perioperative outcomes in cardiac surgery patients. Mayo Clin Proc. 2005;80:862–866.
19. Khaodhiar L, McCowen K, Bistrian B. Perioperative hyperglycemia, infection or risk? Curr Opin Clin Nutr Metab Care. 1999;2:79–82.
20. Rogers SO Jr, Zinner MJ. The role of perioperative hyperglycemia in postoperative infections. Adv Surg. 2009;43:103–109.
21. Kreisel SH, Berschin UM, Hammes HP, et alPragmatic management of hyperglycaemia in acute ischaemic stroke: safety and feasibility of intensive intravenous insulin treatment. Cerebrovasc Dis. 2009;27:167–175.
22. Coester A, Neumann CR, Schmidt MI. Intensive insulin therapy in severe traumatic brain injury: a randomized trial. J Trauma. 2010;68:904–911.
23. Kramer AH, Roberts DJ, Zygun DA. Optimal glycemic control in neurocritical care patients: a systematic review and meta-analysis. Crit Care. 2012;16:R203.
24. Bilotta F, Rosa G. Optimal glycemic control in neurocritical care patients. Crit Care. 2012;16:163.
25. Evans CH, Lee J, Ruhlman MK. Optimal glucose management in the perioperative period. Surg Clin North Am. 2015;95:337–354.
26. Duncan AE. Hyperglycemia and perioperative glucose management. Curr Pharm Des. 2012;18:6195–6203.
27. Bergenstal RM, Garg S, Weinzimer SA, et alSafety of a hybrid closed-loop insulin delivery system in patients with type 1 diabetes. JAMA. 2016;316:1407–1408.
28. Gruenbaum SE, Bilotta F. Use of continuous glucose monitoring in the ICU: a review. Int J Intensive Care. 2014;21:25–29.
29. Korinek AM. Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. The French Study Group of Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Service Epidémiologie Hygiène et Prévention. Neurosurgery. 1997;41:1073–1079.
30. Valentini LG, Casali C, Chatenoud L, Chiaffarino F, Uberti-Foppa C, Broggi G. Surgical site infections after elective neurosurgery: a survey of 1747 patients. Neurosurgery. 2008;62:88–95.
31. Ismael H, Horst M, Farooq M, Jordon J, Patton JH, Rubinfeld IS. Adverse effects of preoperative steroid use on surgical outcomes. Am J Surg. 2011;201:305–308.
32. Dieleman JM, Nierich AP, Rosseel PM, et alDexamethasone for Cardiac Surgery (DECS) Study Group. Intraoperative high-dose dexamethasone for cardiac surgery: a randomized controlled trial. JAMA. 2012;308:1761–1767.
33. Marshall LF, King J, Langfitt TW. The complications of high-dose corticosteroid therapy in neurosurgical patients: a prospective study. Ann Neurol. 1977;1:201–203.
34. Merkler AE, Saini V, Kamel H, Stieg PE. Preoperative steroid use and the risk of infectious complications after neurosurgery. Neurohospitalist. 2014;4:80–85.
35. Cooper PR, Moody S, Clark WK, et alDexamethasone and severe head injury. A prospective double-blind study. J Neurosurg. 1979;51:307–316.
36. Dearden NM, Gibson JS, McDowall DG, Gibson RM, Cameron MM. Effect of high-dose dexamethasone on outcome from severe head injury. J Neurosurg. 1986;64:81–88.
37. Erman T, Demirhindi H, Göçer AI, Tuna M, Ildan F, Boyar B. Risk factors for surgical site infections in neurosurgery patients with antibiotic prophylaxis. Surg Neurol. 2005;63:107–112.
38. Korol E, Johnston K, Waser N, et alA systematic review of risk factors associated with surgical site infections among surgical patients. PLoS One. 2013;8:e83743.
39. Takanari K, Araki Y, Okamoto S, et alOperative wound-related complications after cranial revascularization surgeries. J Neurosurg. 2015;123:1145–1150.
40. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379.
41. Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol. 2007;165:710–718.
© 2017 International Anesthesia Research Society