Sage Journals: Discover world-class research

Abstract

Introduction: Periprosthetic joint infection after total knee arthroplasty (TKA) is a devastating complication and different risk factors, including hyperglycemia and increased glucose variability, have been suggested for this condition in previous literature. Our study aims to investigate the risk factors for postoperative hyperglycemia and glucose variability in patients undergoing unilateral primary TKA with continuous glucose monitoring (CGM) technology. Methods: Demographic factors, diabetic statuses, hemoglobin A1c (HbA1c) levels, and fructosamine levels of patients undergoing unilateral primary TKA were recorded preoperatively. These patients would either received 16 mg intravenous dexamethasone or no dexamethasone before their surgery as part of perioperative management. From the day of admission to postoperative day 4, CGM device monitored these patients’ blood glucose level, time of hyperglycemia, and glucose variability. Correlation and multivariate regression analysis were performed between the risk factors and the outcome measures, with multicollinearity between the predictor variables checked, to find out the most potent risk factors for postoperative hyperglycemia. Results: Eighty-three patients with primary unilateral TKA were analyzed. Diabetic, prediabetic, and nondiabetic patients were composed of 26.5%, 47.0%, and 26.5%, respectively. The average of HbA1c and fructosamine level was 6.0% and 249.3 mmol/L, respectively; 65.1% of the patients received 16 mg dexamethasone perioperatively. Multivariate regression analysis found that HbA1c, fructosamine, and dexamethasone injections are the significant risk factors for the percentage of time in hyperglycemia (R²= 0.374, p < 0.001) and average blood glucose level (R²= 0.493, p < 0.001), with HbA1c having the highest standardized coefficient, followed by fructosamine and dexamethasone injections. Although dexamethasone usage was significant in multivariate regression analysis on predicting glucose variability, the effect size of the model is very weak (R²= 0.05, p < 0.05). Conclusion: The most potent risk factor for hyperglycemic time and increased average blood glucose is HbA1c level, followed by fructosamine level and dexamethasone injections. No strong correlation was found between the factors in our study and patients’ glucose variability.

Keywords

Continuous glucose monitoring primary total knee arthroplasty hyperglycemia

Introduction

Periprosthetic joint infection (PJI) after total knee arthroplasty (TKA) is an uncommon yet devastating complication with severe morbidity and mortality. Different factors have been suggested to be associated with PJI and it is paramount for clinicians to be aware of them in order to minimize the risks of our patients developing this condition.

Out of these risk factors, postoperative hyperglycemia and glucose variability have been reported to be modifiable ones.^1–3 Another possible modifiable risk factor includes the use of steroid perioperatively. Although steroid injection is increasingly used to improve pain control and enhance recovery, the risk of hyperglycemia and PJI remains a concern.^4–6 Moreover, glycemic markers, such as hemoglobin A1c (HbA1c) and fructosamine, correlate with postoperative hyperglycemia and PJI, where the threshold for optimization, have been proposed with conflicting evidence.^7–11

There have been previous studies investigating the effects of these risk factors on perioperative hyperglycemia in patients undergoing TKA.^12,13 However, most of these studies rely on traditional intermittent self-monitoring of blood glucose level,^14,15 which has its limitations in reflecting the glucose profile after TKA. Recently, continuous glucose monitoring (CGM) technology has become a common alternative method for monitoring patients’ blood glucose levels. It enables continuous and instantaneous glucose level monitoring. It can also provide valuable information such as time in hyperglycemia, average blood glucose level, and glucose variability, which traditional glucose measurement methods cannot offer. Despite its more prevalent use and aforementioned advantages, the use of CGM in literature in investigating the risk factors of hyperglycemia after TKA are scarce.

Our study aims to investigate the risk factors for postoperative hyperglycemia and glucose variability in patients undergoing unilateral primary TKA, by looking at their average blood glucose level, the amount of hyperglycemic time, and the fluctuation of glucose level with the use of CGM technology.

Methods

We performed a prospective cohort study on patients undergoing unilateral primary TKA from 2020 to 2022. Patients’ demographic factors were collected. Diabetic status was based on medical history and HbA1c level. According to the American Diabetes Association, HbA1c ≥ 6.5% and ≥5.7% were considered diabetic and prediabetic, respectively.¹⁶ HbA1c and fructosamine levels were checked one month before surgery. Patients received either 16 mg or no dexamethasone as decided by the anesthetist prior to their surgery. Potential risk factors for hyperglycemia were analyzed, including body mass index (BMI), diabetic status, perioperative dexamethasone injection, HbA1c, and fructosamine level. Patients who were having poorly controlled Diabetes mellitus requiring regular insulin injections, patients who were immunocompromised or with contraindications of using steroids (e.g., hepatitis carrier) were not given dexamethasone after anesthetists’ assessment. Patients who were scheduled for one-staged bilateral TKA or refused to use CGM were excluded from this study. Patients who were having HbA1c > 7.5% were postponed for surgery and were not recruited in our study as well. Patients who were newly diagnosed for diabetes mellitus but with HbA1c ≤7.5% were not excluded from surgery or from using dexamethasone. Since the use of fructosamine level is still at an investigative stage as a correlation measurement for perioperative hyperglycemia, we did not use fructosamine as a parameter for the decision of the use of dexamethasone or postponing surgery.

The intravenous fluids and calorie intake during the in-hospital stay of the patients in this study were standardized according to their body weight. We divided the patients into three different groups according to their weights (49 kg or below, 49–59 kg, and 60 kg or above) and limited their calories intake to 1200 kcal, 1500 kcal, and 1800 kcal, respectively, which is about 25–30 kcal per kilogram of their body weight. Normal saline was the only type of intravenous fluid given to our patients in this study in order to standardize their intravenous fluid intake.

We used CGM to monitor the glucose level from before the operation to at least postoperative day 4. Outcome measures include the percentage of time in hyperglycemia (percentage of time with blood glucose level >10.0 mmol/L), increased average glucose level (>10.0 mmol/L), and increased glucose variability (percent coefficient of variation >36%) during the monitoring period. The CGM system we have used in this study is the FreeStyle Libre System (Abbott, IL, USA), which is a CGM system that has been frequently used in other clinical studies on blood glucose monitoring.^17–19 A CGM sensor was applied to each of the patients’ upper arm on the day of admission until at least postoperative day 4. The blood glucose level and the blood glucose variability of these patients were continuously recorded by the sensors during this period.

According to a sliding scale, insulin was injected whenever the patient's blood glucose level was ≥10.1 mmol/L due to ethical considerations to prevent the development of dangerously high hyperglycemia.

Statistical analysis

IBM SPSS Statistics Version 28.0.1.0 was used for statistical analysis for this study. Unpaired Student t-test and correlation and multivariate regression analysis were performed between the risk factors and the outcome measures. Multicollinearity between the predictor variables was checked. Continuous variables are presented in means and standard deviations (SDs), while qualitative variables are presented in percentages. P-value ≤0.05 was considered statistically significant.

Results

A total of 83 patients were analyzed in our study; 82% of the patients were females. The mean age and BMI were 69.1 years old (SD 6.6) and 28.8 kg/m² (SD 3.8), respectively. The average of HbA1c and fructosamine level was 6.0% (SD 0.5) and 249.3 mmol/L (SD 27), respectively. Diabetic, prediabetic, and nondiabetic patients were composed of 26.5%, 47.0%, and 26.5%, respectively; 65.1% received 16 mg dexamethasone perioperatively. Patients’ demographics and their baseline characteristics are shown in Table 1.

Table 1.

Patients’ demographics and baseline characteristics.

Total patient number	83
Gender	15 males, 68 females
Average BMI (kg/m²)	28.8 ± 3.8
Average age (years)	69.1 ± 6.6
Diabetic status (Number of patients (% in study))	Diabetic 22 (26.5%), pre-diabetic 39 (47%), non-diabetic 22 (26.5%)

Univariate analysis between each risk factor and outcomes found that HbA1c, fructosamine, and dexamethasone injection are statistically significant as a risk factor on the percentage of time in hyperglycemia (p < 0.05) and average blood glucose level (p < 0.05). Also, it has been shown that dexamethasone injection is a statistically significant risk factor for glucose variability in our patients (p < 0.05). The mean values, SDs, and p-values of the outcome measurements are shown at Tables 2–6.

Table 2.

Means, standard deviations, and p-values (BMI ≥30 kg/m² vs patients with BMI <30 kg/m²).

	Average percentage of time of hyperglycemia (%)	Average blood glucose level (mmol/L)	Average glucose variability (%)
Patients with BMI ≥30 kg/m²	13.14 ± 10.06	7.44 ± 1.05	27.44 ± 6.40
Patients with BMI <30 kg/m²	11.47 ± 10.40	7.23 ± 1.08	28.90 ± 7.79
p-Value	0.49	0.39	0.39

Table 3.

Means, standard deviations, and p-values (HbA1c ≥ 5.7% vs HbA1c < 5.7%).

	Average percentage of time of hyperglycemia (%)	Average blood glucose level (mmol/L)	Average glucose variability (%)
Patients with HbA1c ≥ 5.7%	14.60 ± 10.77	7.54 ± 1.13	29.30 ± 7.81
Patients with HbA1c < 5.7%	5.35 ± 4.00	6.68 ± 0.51	26.08 ± 5.45
p-value	0.0001	0.0007	0.069

Table 4.

Means, standard deviations, and p-values (Fructosamine ≥ 285 µmol/L vs Fructosamine < 285 µmol/L).

	Average percentage of time of hyperglycemia (%)	Average blood glucose level (mmol/L)	Average glucose variability (%)
Patients with fructosamine ≥ 285 umol/L	24.33 ± 11.69	8.52 ± 0.94	27.97 ± 6.16
Patients with fructosamine < 285 umol/L	11.21 ± 9.56	7.23 ± 1.01	28.45 ± 7.51
p-value	0.002	0.003	0.88

Table 5.

Means, standard deviations, and p-values (Known history of diabetes vs Without known history of diabetes).

	Average percentage of time of hyperglycemia (%)	Average blood glucose level (mmol/L)	Average glucose variability (%)
Patients with known history of diabetes	12.16 ± 10.13	7.32 ± 1.06	27.69 ± 6.08
Patients without known history of diabetes	11.90 ± 10.52	7.27 ± 1.09	29.22 ± 8.56
p-value	0.91	0.83	0.35

Table 6.

Means, standard deviations, and p-values (received dexamethasone vs did not receive dexamethasone).

	Average percentage of time of hyperglycemia (%)	Average blood glucose level (mmol/L)	Average glucose variability (%)
Patients who received dexamethasome	13.72 ± 10.20	7.50 ± 1.08	29.58 ± 7.69
Patients who did not receive dexamethasone	8.90 ± 9.76	6.92 ± 0.96	26.21 ± 6.20
p-value	0.040	0.018	0.046

Multivariate regression analysis found that HbA1c, fructosamine, and dexamethasone injection are the significant risk factors for the percentage of time in hyperglycemia (R²= 0.374, p < 0.001) and average blood glucose level (R²= 0.493, p < 0.001).

For the analysis for the time in hyperglycemia, HbA1c had the highest standardized coefficient (0.469, p < 0.001), followed by fructosamine (0.277, p = .01) and dexamethasone injections (0.194, p = 0.02) (Table 7).

Table 7.

Multivariate regression analysis for time of hyperglycemia (percentage of time of blood glucose level >10 mmol/L).

	Standardized coefficient	p-Value
BMI	−0.078	0.375
HbA1c	0.469	<0.001
Fructosamine	0.277	0.01
Diabetic status	0.098	0.244
Use of dexamethasone	0.194	0.02

For the analysis of high average blood glucose level, once again HbA1c had the highest standardized coefficient (0.397, p < 0.001), followed by fructosamine (0.320, p = 0.003) and dexamethasone injections (0.261, p = 0.002) (Table 8).

Table 8.

Multivariate regression analysis for high average blood glucose level (average blood glucose level >10 mmol/L).

	Standardized coefficient	p-value
BMI	−0.041	0.641
HbA1c	0.397	<0.001
Fructosamine	0.320	0.003
Diabetic status	0.120	0.154
Use of dexamethasone	0.261	0.002

For the analysis for glucose variability, although dexamethasone usage was significant in multivariate regression analysis on predicting glucose variability (standardized coefficient 0.230, p = 0.044) (Table 9), the effect size of the model is very weak (R²= 0.05, p < 0.05). All independent variables did not show any multicollinearity with tolerance >0.25 and variance inflation factor <4.

Table 9.

Multivariate regression analysis for glucose variability (percentage of coefficient of variation >36%).

	Standardized coefficient	p-Value
BMI	0.022	0.845
HbA1c	−0.102	0.505
Fructosamine	0.118	0.412
Diabetic status	0.161	0.163
Use of dexamethasone	0.230	0.044

Discussion

Postoperative hyperglycemia is a risk factor for complications in orthopedics and other types of surgeries.^1,2,20–22 The surgical stress after major surgeries triggers a catabolic response and the release of hormones, such as cortisol and glucagon, which in turn leads to hyperglycemia and increase insulin resistance. This transient hyperglycemia creates an optimal environment for bacterial growth while impairing our immune system's ability to combat infection.^23,24 This explains the clinical finding of increased PJI and wound complications after arthroplasty with postoperative hyperglycemia.^2,10 Therefore, we should identify, screen, and optimize the risk factors for postoperative hyperglycemia before joint replacement surgeries.

Our study found that patients’ preoperative HbA1c level was the most robust predictor for the time in hyperglycemia and high average blood glucose, followed by fructosamine and steroid injections. On the contrary, the patient's medical history of diabetic status was not a significant risk factor after multivariate regression analysis. Our findings are supported by other studies demonstrating hyperglycemia is more critical than diabetic status in predicting postoperative complications.^20,25 Also, postoperative hyperglycemia can develop in diabetic and nondiabetics.²⁶ Hence, screening and optimizing the glycemic markers, especially HbA1c, is essential in all patients before arthroplasty to help minimize their risks of developing infection. Due to the relatively small sample size of our study, we were not able to conclude a cutoff value of HbA1c as a contraindication to TKA. However, despite excluding patients with HbA1c level >7.5%, in our study HbA1c still appeared to be a predictor of hyperglycemia. From this result, it appears that a cutoff HbA1c level lower than 7.5% may be beneficial to guide us in assessing the hyperglycemic risk of patients undergoing TKA; however, a study with a larger sample size will be required for further confirmation. Even though steroid injection is one of the risk factors for hyperglycemia, its standardized coefficient is the lowest out of the three risk factors. This may suggest that the correlation between the use of it and perioperative hyperglycemia may not be as strong as conventionally presumed. Thus, consideration of a more lenient use of this medication should be explored in the future especially when many studies have shown that steroids are effective in enhancing recovery which is particularly useful in fast-track arthroplasty.^27,28 Therefore, the risk and benefits of its use should be weighed and discussed for individual patients.

Shohat et al. reviewed point-of-care capillary and serum blood glucose levels in arthroplasty patients and reported that glucose variability in the postoperative period is associated with an increased rate of PJI.³ The traditional methods of glucose measurements might not capture the peaks and troughs of the glucose level, thus, underestimating its variability. On the other hand, CGM can accurately reflect the variation in glucose over time. Unfortunately, the risk factors analyzed in this study only had a weak effect size in predicting glucose variability. This indicates glucose variability is more volatile and more predictors are needed to explain its fluctuation.

Although we standardized the intravenous fluid and diet, controlling extra intake and snacks between meals is challenging. Moreover, insulin injections are given to hyperglycemic patients per protocol as it is unethical to leave them untreated, thus underestimating the actual duration of hyperglycemia. In addition, patients with complex medical comorbidities and a higher risk of postoperative glucose disturbance will likely opt not for steroid injection by the anesthetist, which will further confound our results. Nevertheless, although these limiting factors can potentially affect our results, we believe that they also reflected real-life clinical scenarios where similar treatments would have been provided accordingly. As patients with HbA1c > 7.5% were not recruited in our study, it means our study was mainly focused on patients who were supposed to have relatively good control of their diabetes, and our results may not be directly applicable for patients with poorly controlled diabetes. However, since the variation of glucose control is still commonly observed in patients with relatively well-controlled diabetes, we believe our results are still useful in providing insight into perioperative hyperglycemia risk of this group of patients.

Despite the limitations, our study does shed some light on the risk factors for postoperative hyperglycemia after TKA and we hope it can provide evidence for clinicians to consider when managing this group of patients. However, the sample size of our study is relatively small to see the effect of controlling the risk factors identified in the current study, including HbA1c, fructosamine, and steroid injections. Also, in our study, we were not able to draw any conclusion on a target threshold for these risk factors for patient optimization, and this is once again likely due to the limited size of our study. In view of these limitations, a larger scale of study is welcomed to further examine the relationships between these risk factors and postoperative hyperglycemia. Furthermore, in our study, we have only been looking at patients undergoing unilateral primary TKA. Nowadays bilateral one-staged TKAs or revision surgery are becoming more common, so our results may not be directly applicable to those patients. Future studies on the risk factors for postoperative hyperglycemia for these patients will be useful.

Conclusion

Our study found that the most potent risk factor for hyperglycemic time and increased average blood glucose is HbA1c, followed by fructosamine and dexamethasone injections. With our results, we suggest that universal screening of these glycemic markers, that is, HbA1c and fructosamine, is beneficial for patients undergoing primary TKA, even in nondiabetic patients. However, the factors analyzed in our study cannot adequately explain the difference in glucose variability in our patients, and further study in this regard is welcomed.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

L H Brian Leung

W K Vincent Chan

Michelle H Luk

References

Kheir

Tan

Kheir

, et al. Postoperative blood glucose levels predict infection after total joint arthroplasty. J Bone Joint Surg Am 2018; 100: 1423–1431.

Mraovic

Suh

Jacovides

, et al. Perioperative hyperglycemia and postoperative infection after lower limb arthroplasty. J Diabetes Sci Technol 2011; 5: 412–418.

Shohat

Restrepo

Allierezaie

, et al. Increased postoperative glucose variability is associated with adverse outcomes following total joint arthroplasty. J Bone Joint Surg Am 2018; 100: 1110–1117.

Allen

Jedrzynski

Michelson

, et al. The effect of dexamethasone on postoperative blood glucose in patients with type 2 diabetes mellitus undergoing total joint arthroplasty. J Arthroplasty 2020; 35: 671–674.

Herbst

Telford

Hunting

, et al. The effects of perioperative dexamethasone on glycemic control and postoperative outcomes. Endocr Pract 2020; 26: 218–225.

Lindberg-Larsen

Kehlet

Bagger

, et al. Preoperative high-dose methylprednisolone and glycemic control early after total hip and knee arthroplasty: a randomized, double-blind, placebo-controlled trial. Anesth Analg 2018; 127: 906–913.

Godshaw

Ojard

Adams

, et al. Preoperative glycemic control predicts perioperative Serum glucose levels in patients undergoing total joint arthroplasty. J Arthroplasty 2018; 33: S76–s80.

Thever

Teo

BJX

Tan

HCA

. Predictive value of common serum glycaemic markers on periprosthetic joint infection following total joint arthroplasty: a review of the literature. J Orthop 2020; 22: 278–281.

Shohat

Tarabichi

Tan

, et al. 2019 John Insall Award: fructosamine is a better glycaemic marker compared with glycated haemoglobin (HbA1C) in predicting adverse outcomes following total knee arthroplasty: a prospective multicentre study. Bone Joint J 2019; 101-B: 3–9.

10.

Yang

Sun

, et al. Is hemoglobin A1c and perioperative hyperglycemia predictive of periprosthetic joint infection following total joint arthroplasty?: a systematic review and meta-analysis. Medicine (Baltimore) 2017; 96: e8805.

11.

Citak

Toussaint

Abdelaziz

, et al. Elevated HbA1c is not a risk factor for wound complications following total joint arthroplasty: a prospective study. Hip Int 2020; 30: 19–25.

12.

Jämsen

Nevalainen

Eskelinen

, et al. Risk factors for perioperative hyperglycemia in primary hip and knee replacements. Acta Orthop 2015; 86: 175–182.

13.

Godshaw

Warren

Nammour

, et al. Insulin-Dependent diabetic patients are at increased risk of postoperative hyperglycemia when undergoing total joint arthroplasty. J Arthroplasty 2020; 35: 2375–2379.

14.

Park

Chang

Kim

, et al. Effects of intravenous dexamethasone on glycemic control in patients with type 2 diabetes Mellitus after total knee arthroplasty. J Arthroplasty 2021; 36: 3909–3914.

15.

Nurok

Cheng

Romeo

, et al. Dexamethasone and perioperative blood glucose in patients undergoing total joint arthroplasty: a retrospective study. J Clin Anesth 2017; 37: 116–122.

16.

American Diabetes Association. https://www.diabetes.org/diabetes/a1c/diagnosis.

17.

Zueg Zueger

Schuler

Stettler

, et al. Assessment of three frequently used blood glucose monitoring devices in clinical routine. Swiss Med Wkly 2012; 142: w13631.

18.

Krakauer

Botero

Lavalle-González

, et al. A review of flash glucose monitoring in type 2 diabetes. Diabetol Metab Syndr 2021; 13: 42.

19.

Slattery

Choudhary

. Clinical use of continuous glucose monitoring in adults with type 1 diabetes. Diabetes Technol Ther 2017; 19: S55–S61.

20.

Rahimi-Nedjat

Sagheb

Pabst

, et al. Diabetes and hyperglycemia as risk factors for postoperative outcome in maxillofacial surgery. J Surg Res 2017; 217: 170–176.

21.

Jones

Cain

Mitchell

, et al. Hyperglycemia predicts mortality after CABG: postoperative hyperglycemia predicts dramatic increases in mortality after coronary artery bypass graft surgery. J Diabetes Complications 2008; 22: 365–370.

22.

Akiboye

Rayman

. Management of hyperglycemia and diabetes in orthopedic surgery. Curr Diab Rep 2017; 17: 13.

23.

Turina

Fry

Polk

HC Jr

. Acute hyperglycemia and the innate immune system: clinical, cellular, and molecular aspects. Crit Care Med 2005; 33: 1624–1633.

24.

Xiu

Stanojcic

Diao

, et al. Stress hyperglycemia, insulin treatment, and innate immune cells. Int J Endocrinol 2014; 2014: 486403.

25.

Kwon

Thompson

Dellinger

, et al. Importance of perioperative glycemic control in general surgery: a report from the surgical care and outcomes assessment program. Ann Surg 2013; 257: 8–14.

26.

Mannion

Rather

Manifold

, et al. Postoperative hyperglycemia in patients with and without diabetes after major joint replacement: the impact of an enhanced glucose management program. JB JS Open Access 2021; 6: 3.

27.

Chan

VWK