sign in sign up
<<
Home , Diabetes, Fructosamine, History of diabetes, Hyperglycemia, Prediabetes, Type 2 diabetes

1900

COPYRIGHT Ó 2017 BY THE JOURNAL OF BONE AND JOINT SURGERY,INCORPORATED

A commentary by James G. Wright, CM, MD, MPH, FRCS(C), FRCS(Ed), is linked to the online version of this article at jbjs.org. : A Simple and Inexpensive Test for Assessing Preoperative Glycemic Control

Noam Shohat, MD, Majd Tarabichi, MD, Eric H. Tischler, BA, Serge Jabbour, MD, and Javad Parvizi, MD, FRCS

Investigation performed at The Rothman Institute at Thomas Jefferson University, Philadelphia, Pennsylvania

Background: Although the medical community acknowledges the importance of preoperative glycemic control, the is inconclusive and the proper metric for assessment of glycemic control remains unclear. Serum fructosamine reflects the mean glycemic control in a shorter time period compared with glycated (HbA1c). Our aim was to examine its role in predicting adverse outcomes following total joint arthroplasty. Methods: Between 2012 and 2013, we screened all patients undergoing total joint arthroplasty preoperatively using serum HbA1c, fructosamine, and levels. On the basis of the recommendations of the American Association, 7% was chosen as the cutoff for HbA1c being indicative of poor glycemic control. This threshold correlated with a fructosamine level of 292 mmol/L. All patients were followed and total joint arthroplasty complications were evaluated. We were particularly interested in retrieving details on surgical-site (superficial and deep). Patients with fructosamine levels of ‡292 mmol/L were compared with those with fructosamine levels of <292 mmol/L. Compli- cations were evaluated in a univariate analysis followed by a stepwise analysis. Results: A total of 829 patients undergoing primary total joint arthroplasty were included in the present study. There were 119 patients (14.4%) with a and 308 patients (37.2%) with HbA1c levels in the prediabetic range. Overall, 51 patients had fructosamine levels of ‡292 mmol/L. Twenty patients (39.2%) had a fructosamine level of ‡292 mmol/L but did not have an HbA1c level of ‡7%. Patients with fructosamine levels of ‡292 mmol/L had a significantly higher risk for deep infection (adjusted odds ratio [OR], 6.2 [95% confidence interval (CI), 1.6 to 24.0]; p = 0.009), readmission (adjusted OR, 3.0 [95% CI, 1.1 to 8.1]; p = 0.03), and reoperation (adjusted OR, 3.4 [95% CI, 1.2 to 9.2]; p = 0.02). In the current study with the given sample size, HbA1c levels of ‡7% failed to show any significant correlation with deep infection (p = 0.14), readmission (p = 1.0), or reoperation (p = 0.7). Conclusions: Serum fructosamine is a simple and inexpensive test that appears to be a good predictor of adverse outcome in patients with known diabetes and those with unrecognized diabetes or . Our findings suggest that fructosamine can serve as an alternative to HbA1c in the setting of preoperative glycemic assessment. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

atients with diabetes have an increased risk for arthritis1, higher rates of myocardial , , superficial P which explains the high proportion of diabetes among complications, and periprosthetic joint infections2,6-8. Periop- patients undergoing total joint arthroplasty, at an esti- erative hyperglycemia has also been shown to be a common mated rate between 8% and 22%2-5. and independent for periprosthetic joint infection, Patients with diabetes, particularly those with poor gly- even among patients not diagnosed with diabetes9-11. A recent cemic control, have an increased risk for postoperative com- study demonstrated that >30% of patients undergoing total plications following total joint arthroplasty6. This includes joint arthroplasty have undiagnosed hyperglycemia12.

Disclosure: There was no external funding source for this study. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work and “yes” to indicate that the author had a patent and/or copyright, planned, pending, or issued, broadly relevant to this work (http://links.lww.com/JBJS/E454).

J Bone Joint Surg Am. 2017;99:1900-7 d http://dx.doi.org/10.2106/JBJS.17.00075 1901

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

Although diabetes and perioperative hyperglycemia are joints replaced. We modified the Elixhauser index by considered by many as independent risk factors for compli- removing the diabetic element of scoring to improve the com- cations following total joint arthroplasty, there is inconsistency parison between groups. Data on dependent variables were with regard to the impact that preoperative glycemic control extracted from our institutional database of prospectively collected has on the outcome of total joint arthroplasty in patients with data. Deep infection was diagnosed on the basis of the Musculo- and without diabetes2,6,7,9,11,13-15. These discrepancies could be skeletal Infection Society (MSIS) criteria29.Superficial infection was attributed to a large extent to the heterogeneity of study designs defined as a local wound necessitating further and the variations in the definition of uncontrolled diabetes. treatment. The rate and the reason for readmission or reoperation, Although the American Diabetes Association (ADA) rec- as well as all “medical” complications, including cerebrovascular ommends hemoglobin A1c (HbA1c) as the measure for moni- accident, , , deep venous throm- toring glycemic control, the role of HbA1c in predicting the bosis, , mental status change, , outcomeofsurgicalproceduresinpatientswithdiabetesiscon- and gastrointestinal complications occurring within 90 days troversial. In the orthopaedic literature, the association between following total joint arthroplasty, were documented. The incidence HbA1c level and adverse outcome is unclear11,13,16-20. Further, the of periprosthetic fractures, stiffness, component loosening, and optimal preoperative HbA1c level in the surgical patients is un- instability (termed “mechanical” complications) was recorded. The known. Another major drawback with using HbA1c as a marker length of stay was also noted. for glycemic control is the delay that is associated with changes in the level of HbA1c with implementation of better glycemic con- Preoperative Glycemic Status trol. HbA1c levels are dependent on the life cycle of red blood cells Patients with HbA1c levels of ‡6.5% at preadmission testing who at 120 days and may take up to 3 months to reflect the changes in had no history of diabetes were considered as patients with newly the glycemic control21. diagnosed diabetes. Patients who had HbA1c levels between 5.7% However, fructosamine measures the level of glycated and 6.4% were categorized as prediabetic. Because patients may serum , mostly albumin22.Itreflects the mean glucose not have complied with our instructions and may not have fasted, levels over a short to intermediate time frame, 14 to 21 days we did not use the serum glucose level for diagnosis of diabetes. (reflecting the turnover of plasma proteins). Fructosamine Fructosamine was measured using quantitative spectro- levels show higher fluctuation than those of HbA1c, better photometry at a $19.77 charge. reflecting the variability and rapid changes of blood glucose The threshold for uncontrolled diabetes based on fruc- among patients with and without diabetes23,24, thus reflecting tosamine is not well established. We therefore examined the quicker response to treatment, and it may be a better marker percentile of uncontrolled diabetics on the basis of an HbA1c for poor glycemic control than HbA1c25-28. cutoff of 7% (as recommended by the ADA) to determine the The primary purpose of this study was to examine the as- fructosamine level being indicative of poor glycemic control. sociation between serum fructosamine levels and the risk for ad- We found 6% of patients (94th percentile) to have poor gly- verse outcomes following total joint arthroplasty among patients cemic control. We then examined what the 94th percentile with and without diabetes. The study also sought to compare value of fructosamine was and found it to be 292 mmol/L. fructosamine level with HbA1c in predicting outcomes. Statistical Analysis Materials and Methods Patients with fructosamine levels of ‡292 mmol/L (high fructos- etween September 2012 and July 2013, we screened all aminegroup)werecomparedwiththose with fructosamine levels Bpatients undergoing elective total hip arthroplasty or total of <292 mmol/L (low fructosamine group). Patients with HbA1c knee arthroplasty for glycemic control using HbA1c levels, levels of ‡7% were compared with those with HbA1c levels of fructosamine levels, and blood glucose levels. Blood samples <7%. To confirm the cutoff chosen for fructosamine (‡292 mmol/ were obtained at the preadmission testing 2 to 4 weeks before L), a receiver operating characteristic (ROC) curve analysis was the surgical procedure. Patients were instructed to fast starting performed and the area under the curve (AUC) was calculated for the night prior to these tests. On postoperative day 1, morning the risk of periprosthetic joint . The best fructosamine glucose levels were obtained for all participants as well. threshold correlating with periprosthetic joint infections was de- During this time period, 947 patients were screened and termined using the Youden index. Categorical variables were an- 118 patients undergoing revision arthroplasty or hip preser- alyzed with use of the chi-square test, and continuous variables vation surgical procedures were excluded. The present study were analyzed with use of the Student t test. Significance was set at includes 829 patients undergoing unilateral or bilateral pri- p < 0.05. Adjusted odds ratios (ORs) were determined for each mary total hip or knee arthroplasty. dependent variable separately (infection [superficial and deep], Following obtaining institutional review board approval, medical, mechanical, readmission, reoperation, overall) using the relevant and detailed data on all independent variables were stepwise logistic regression analysis with an entry of 0.2 and a extracted, including demographic characteristics (age, sex, race, removal of 0.15. Independent variables included age, sex, race, ), the Elixhauser comorbidity index, the American body mass index, Elixhauser comorbidity index, ASA classifica- Society of Anesthesiologists (ASA) classification, history of diabe- tion, type (knee or hip) and number (unilateral or bilateral) of tes, and type (knee or hip) and number (unilateral or bilateral) of joints replaced, and postoperative day-1 glucose levels. 1902

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

TABLE I Demographic Characteristics and History of Diabetes and Glycemic Control Based on Fructosamine Levels

Fructosamine <292 mmol/L (N = 778) Fructosamine ‡292 mmol/L (N = 51) P Value

Age* (yr) 63.7 ± 11.1 68.5 ± 10.1 0.002 Sex† 0.88 Male 325 (41.8%) 22 (43.1%) Female 453 (58.2%) 29 (56.9%) Body mass index* (kg/m2) 29.6 ± 5.5 29.9 ± 5.5 0.71 Elixhauser comorbidity index*‡ 1.49 ± 1.2 1.55 ± 0.8 0.63 Distribution 0.12 0† 162 (20.8%) 4 (7.8%) 1† 278 (35.7%) 24 (47.1%) 2† 205 (26.3%) 14 (27.5%) ‡3† 133 (17.1%) 9 (17.6%) Joint† 0.77 Knee 414 (53.2%) 26 (51.0%) Hip 364 (46.8%) 25 (49.0%) History of diabetes mellitus† <0.0001 Yes 86 (11.1%) 33 (64.7%) No 692 (88.9%) 18 (35.3%) HbA1c* (%) 5.7 ± 0.5 7.3 ± 1.2 <0.0001 Distribution <0.0001 ‡7%† 19 (2.4%) 31 (60.8%) <7%† 759 (97.6%) 20 (39.2%) Glucose levels* (mg/dL) Preoperative 99.2 ± 24.2 159.8 ± 57.7 <0.0001 Postoperative day 1 117.7 ± 22.1 155.3 ± 42.0 <0.0001

*The values are given as the mean and the standard deviation. †The values are given as the number of patients, with the percentage in parentheses. ‡Excluding diabetic element of scoring.

Fructosamine (‡292 mmol/L) as a dichotomized variable was in- group (10.9%). Further, the mean HbA1c levels in the known cluded in all analyses, even if it did not enter the model. diabetic patients were 6.8% (range, 4.6% to 10.1%) compared with 6.8% (range, 6.5% to 7.8%) in the newly diagnosed diabetic Results patients; the mean fructosamine levels in the known diabetic total of 829 patients (347 male patients and 482 female patients were 277.1 mmol/L (range, 201 to 484 mmol/L) com- A patients) were included in the present study, undergoing pared with 279.0 mmol/L (range, 227 to 416 mmol/L) in the 440 total hip arthroplasties and 389 total knee arthroplasties. newly diagnosed diabetic patients. There were no significant The mean HbA1c level was 5.8% (range, 4.2% to 10.1%), and differences in outcomes between the dysglycemic patients (pa- the mean fructosamine level was 244.8 mmol/L (range, 178 tients with diabetics and ) and non-dysglycemic pa- to 484 mmol/L) (see Appendix). tients. Although the rate of superficial wound infection was There were 119 patients (14.4%) with a history of diabetes. higher in the dysglycemic group (2.5%) compared with the Twenty-one patients (2.5%) who had no history of diabetes were non-dysglycemic group (0.9%), the difference did not reach newly diagnosed. Thus, the overall prevalence of diabetes in the significance (OR, 2.7 [95% CI, 0.84 to 8.75]; p = 0.1). cohort was 16.9%. A total of 308 patients (37.2%) had HbA1c levels in the prediabetic range. There was no significant difference Characteristics of Patients with High Fructosamine (p = 0.14) with respect to diabetes control (OR, 0.37 [95% Overall, 51 patients (22 male patients and 29 female) had fruc- confidence interval (CI), 0.1 to 1.2]) between the known diabetic tosamine levels of ‡292 mmol/L, representing the 94th percentile group (38.7%) and the newly diagnosed diabetic group (19%); (Table I). The number of patients who had a fructosamine level of also, there was no significant difference (p = 0.71) with regard to ‡292 mmol/L but did not have an HbA1c level of ‡7% was 20 complicationrates(OR,1.35[95%CI,0.3to5.2])betweenthe (39.2%) (Fig. 1). The mean age was higher in patients with known diabetic group (14.3%) and the newly diagnosed diabetic fructosamine levels of ‡292 mmol/L at 68.5 years compared with 1903

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

Fig. 1 A scatterplot correlating HbA1c and fructosamine (R2 = 0.34). those with fructosamine levels of <292 mmol/L at 63.7 years (p = Fructosamine Predicts Adverse Outcomes 0.002). Interestingly, 18 patients (35.3%) in the high fruc- Patients with fructosamine levels of ‡292 mmol/L had a sig- tosamine group did not have diabetes. There were no sig- nificantly higher risk for developing adverse outcomes com- nificant differences in other between the pared with those with fructosamine levels of <292 mmol/L fructosamine groups. Fructosamine levels of ‡292 mmol/L (Table II). This included higher rates of infection, mainly correlated with the other glycemic control markers (Table I). periprosthetic joint infections (p = 0.02), readmissions (p = Postoperative day-1 glucose levels were not significantly asso- 0.04), and reoperations (p = 0.03). In a post hoc analysis, ciated with any of the adverse outcomes (see Appendix). fructosamine at a breakpoint of 293 mmol/L was confirmed to

TABLE II Adverse Outcomes in Patients with Low and High Fructosamine Levels

Fructosamine Group* Univariate Multivariate Low (N = 778) High (N = 51) OR† P Value Adjusted OR† P Value

Infection 21 (2.7%) 4 (7.8%) 3.1 (1.0 to 9.3) 0.06 3.2 (1.0 to 9.9) 0.04 Deep (periprosthetic joint 8 (1.0%) 3 (5.9%) 6.0 (1.5 to 23.4) 0.02 6.2 (1.6 to 24.0) 0.009 infection) Superficial 13 (1.7%) 1 (2.0%) 1.2 (0.1 to 9.2) 0.6 1.4 (0.2 to 11.3) 0.7 Medical‡ 45 (5.8%) 3 (5.9%) 1.0 (0.3 to 3.4) 1 1.1 (0.3 to 3.6) 0.9 Mechanical§ 20 (2.6%) 2 (3.9%) 1.5 (0.3 to 6.8) 0.6 1.7 (0.4 to 7.5) 0.5 Readmission 28 (3.6%) 5 (9.8%) 2.9 (1.0 to 7.9) 0.04 3.0 (1.1 to 8.1) 0.03 Reoperation 25 (3.2%) 5 (9.8%) 3.3 (1.2 to 8.9) 0.03 3.4 (1.2 to 9.2) 0.02 Overall 84 (10.8%) 8 (15.7%) 1.5 (0.7 to 3.4) 0.2 1.6 (0.7 to 3.5) 0.2

*The fructosamine levels were denoted as low (<292 mmol/L) or high (‡292 mmol/L). The values are given as the number of patients, with the percentage in parentheses. †The values are given as the OR, adjusted or not, with the 95% CI in parentheses. ‡This category included cere- brovascular accident, myocardial infarction, arrhythmia, deep venous thrombosis, urinary tract infection, mental status change, pneumonia, and gastrointestinal complications occurring within 90 days following total joint arthroplasty. §This category included periprosthetic fractures, stiffness, component loosening, and instability. 1904

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

Fig. 2 Bar graph showing the comparison between patients with fructosamine levels of ‡292 mmol/L with those with HbA1c levels of ‡7% in predicting adverse outcomes following total joint arthroplasty. The asterisk indicates a significant change compared with controls. PJI = periprosthetic joint infection.

be the best threshold for predicting periprosthetic joint in- compared with patients with HbA1c levels of <7% (2.8%) but fections (Youden index, 0.11; cutpoint, 0.03). In a stepwise the difference was not significant (p = 0.18). Also, the rates of logistic regression analysis, fructosamine was the only variable periprosthetic joint infection were higher in patients with to enter the model assessing for periprosthetic joint infections HbA1c levels of ‡7% (4.0%) compared with patients with (adjusted OR, 6.2 [95% CI, 1.6 to 24.0]; p = 0.009), read- HbA1c levels of <7% (1.2%), but this was also not significant mission (adjusted OR, 3.0 [95% CI, 1.1 to 8.1]; p = 0.03), and (p = 0.14). Moreover, the rate of readmission was similar (p = reoperation (adjusted OR, 3.4 [95% CI, 1.2 to 9.2]; p = 0.02). 1.0) in patients with HbA1c levels of ‡7% (4%) compared There were also higher rates of mechanical and overall com- with patients with HbA1c levels of <7% (4%) Also, the rate of plications in the elevated fructosamine group, but these did reoperation was similar (p = 0.7) in patients with HbA1c not reach significance. levels of ‡7% (4%) compared with patients with HbA1c There was no notable difference (p = 0.39) with respect to levels of <7% (3.6%). In addition, the overall complications mean length of hospital stay (95% CI for the difference, 20.46 to rates were similar (p = 1.0) in patients with HbA1c levels of 0.18) between the high fructosamine level group (2.16 days) and ‡7% (10%) compared with patients with HbA1c levels of the low fructosamine level group (2.01 days). There was also no <7% (11.2%). significant difference (p = 1.0) with regard to medical compli- cations between the high fructosamine level group (5.9%) and Discussion the low fructosamine level group (5.8%). There was no signifi- o our knowledge, this is the first study that examines the cant difference (p = 0.6) with regard to cardiovascular compli- Trole of serum fructosamine, which can be measured by a cations between the high fructosamine group (0%) and the low simple and inexpensive test, in predicting outcome following fructosamine group (2.1%). There was also no significant dif- primary total joint arthroplasty in patients with diabetes and ference (p = 1.0) with regard to the rate of deep venous throm- prediabetes. The study demonstrated that elevated fructos- bosis or pulmonary between the high fructosamine amine is an independent risk factor for infections following group (0%) and the low fructosamine group (0.4%). total joint arthroplasty, an association that remained strong even when adjusting for covariates. Our findings suggest that Fructosamine Compared with HbA1c fructosamine can serve as a preoperative marker for glycemic We compared the group of patients with fructosamine levels of control and is a strong predictor of adverse overall outcome, ‡292 mmol/L with those with HbA1c levels of ‡7% in pre- and in particular infections, for patients with diabetes who are dicting adverse outcomes following total joint arthroplasty undergoing total joint arthroplasty. The associations of fruc- (Fig. 2). Unlike fructosamine, HbA1c levels of ‡7% failed to tosamine with adverse outcomes were higher than those ob- show any significant association with adverse outcomes in the served for HbA1c. These results suggest that fructosamine may current study with the given sample size. The rates of infec- serve as an alternative to HbA1c in preoperative glycemic control tion were higher in patients with HbA1c levels of ‡7% (6.0%) . 1905

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

Although elevated HbA1c values (‡7%) have been as- levels. These findings suggest that perioperative hyperglycemia sociated with increased perioperative complications and may be a more reliable predictor than the mean glycemic control in patients undergoing nonorthopaedic surgical procedures30-32, over the 2 to 3 months prior to the surgical procedure4,11,37.An- most studies examining its role in total joint arthroplasty other explanation could be that fructosamine detects fluctuation have failed to show correlation with periprosthetic joint in- and rapid variations of glucose and may detect short-term hy- fections and other adverse outcomes (see Appendix). In con- perglycemic events better than HbA1c23,24,38. Short-term fluctua- trast, most studies comparing patients with diabetes and tions in glucose levels may have a larger effect on inflammatory complications with those with diabetes without complications as cytokine levels than continuous hyperglycemia39. Furthermore, a measure for uncontrolled diabetes did findtheformertobea acute short-term hyperglycemia has been shown to affect all predictor for adverse outcomes6,33-35. Marchant et al.6 categorized major components of innate immunity and impairs the ability the glycemic control based on the International Classification of the host to combat infection40. Finally, continuous glucose of , Ninth Revision (ICD-9) coding. Using the Nation- measurements have been shown to be more tightly correlated wide Inpatient Sample (NIS) database (1988-2005), the authors with glycated albumin compared with HbA1c41,42.Maetal.43 demonstrated a higher risk for complications in those with pre- reported glycated albumin to be more closely correlated with operative uncontrolled diabetes, with an adjusted OR for death of coronary than HbA1c in a high-risk Chinese 3.23 (95% CI, 1.87 to 5.57) and an adjusted OR for infection of population. Fructosamine (mostly composed of albumin) 2.31(95%CI,1.42to3.75).However,Adamsetal.14, using the could therefore be a more reliable marker in assessing glyce- Kaiser Permanente Registry (2001-2009), failed to show an as- mic control. sociation between HbA1c and eventual outcome among patients Importantly, 35.3% of patients with fructosamine levels of with diabetes undergoing total knee arthroplasty, when using the ‡292 mmol/L in our study did not have diabetes. Previous studies HbA1c threshold of 7% to categorize controlled diabetes com- on fructosamine showed a strong association with microvascular pared with uncontrolled diabetes. Nevertheless, HbA1c continues complications, coronary disease, ischemic stroke, and heart to be used as the marker for glycemic control in the preoperative failure in patients with and without diabetes44-47.Jamsen¨ et al.9 setting, usually with <7% being the threshold for good glycemic showed an association between preoperative glucose levels and control. The limited predictive value of HbA1c puts into question the risk for infection in patients not known to have diabetes. its utility in the preoperative risk assessment for arthroplasty. We Patients with glucose levels of ‡124 mg/dL had higher infection add to these studies and report no significant correlation between rates (1.15%) than patients with glucose levels of <124 mg/dL uncontrolled diabetes, as determined by HbA1c levels of ‡7%, (0.28%) (p = 0.002). However, these differences did not reach and adverse outcome in patients undergoing total joint arthro- significance after a multivariate analysis. plasty, with the given sample size. However, we did find fruc- The current study had several limitations. The study did tosamine levels to be predictive of adverse outcomes when using not control for factors that might influence fructosamine levels a threshold equivalent to HbA1c levels of ‡7%. such as losing states (e.g., ), re- Another limitation of using HbA1c as a preoperative duced protein production (i.e., hepatic ), and high screening tool was the delay in response to treatment and levels of glycated immunoglobulins (specifically IgA) or bili- subsequent delay in the surgical procedure (2 to 3 months) that rubin. The sample size in the study may have been too small to could ensue with the use of a threshold for HbA1c as a marker show an association between HbA1c and the adverse outcomes, for glycemic control. This results in unnecessary suffering for a potential for the presence of type-II statistical error. The study the patients and may very well discourage patients from un- involved a single baseline measurement of fructosamine. It is dergoing the surgical procedure36. Fructosamine holds better possible that multiple measurements of fructosamine, particu- promise as its level changes within 14 to 21 days compared with larly in the postoperative period, could have provided better 120 days for HbA1c. Being an inexpensive and readily available insight into glycemic control in general and the examination test, it could easily be ordered during the preadmission period of the association between glycemic control and eventual out- in patients undergoing surgical procedures. Patients with high come of total joint arthroplasty. Moreover, although our findings levels of fructosamine could be put on a strict glycemic control suggest a value of 292 mmol/L as a threshold for complications, protocol and could be reassessed after a short period of 2 weeks, mainly periprosthetic joint infections, this value may vary be- avoiding further delay in the surgical procedure. tween different laboratories48. Finally, although an association There could be a number of reasons to explain the su- between elevated fructosamine and adverse outcome was dem- periority of fructosamine over HbA1c in predicting outcome onstrated, the study did not examine whether correcting the following total joint arthroplasty, in particular, the risk of in- fructosamine level did indeed lead to a reduction of complica- fections. First, fructosamine reflects the mean glucose levels tions. Thus, there is a need for a prospective clinical trial to closer to the time of the surgical procedure: 2 to 3 weeks examine the role of strict fructosamine control on the outcome compared with 2 to 3 months. Preoperative glucose levels taken of total joint arthroplasty. close to the time of the surgical procedure have been shown to Despite the aforementioned limitations, the current correlate with adverse outcomes including infection and death study demonstrates that fructosamine may be a viable and while HbA1c failed to do so9,10,18. We found a positive linear valuable marker for glycemic control in patients undergoing correlation between fructosamine and preoperative glucose surgical procedures. With the emergence of recent guidelines 1906

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL by the World Health Organization and the U.S. Centers for associations between postoperative day-1 glucose levels and Disease Control and Prevention49, emphasizing the importance adverse postoperative outcomes through univariate and mul- of glycemic control, fructosamine may a critical role in tivariate analysis; and a review of studies assessing the impact of glycemic control in patients and measuring the glycemic control (controlled compared with uncontrolled) effect of treatment strategies. One of the critical advantages of using HbA1c as a dichotomous variable are available with the fructosamine, compared with HbA1c, is the shorter half-life of online version of this article as a data supplement at jbjs.org the glycated proteins that may reflect the effect of treatment (http://links.lww.com/JBJS/E455). n within 2 or 3 weeks compared with , which could take up to 120 days. On the basis of the findings of this study, we propose that all patients undergoing total joint arthroplasty should be screened for glycemic control, as a Noam Shohat, MD1,2 portion of patients in this cohort had unknown prediabetes Majd Tarabichi, MD1 and diabetes, and fructosamine could serve as the screening Eric H. Tischler, BA1 Serge Jabbour, MD3 marker of choice. Although the exact threshold for fructos- 1 amine that is indicative of strict glycemic control could not be Javad Parvizi, MD, FRCS determined, the association between elevated fructosamine 1The Rothman Institute at Thomas Jefferson University, levels and adverse outcome is strong and should provide an Philadelphia, Pennsylvania impetus toward normalization of fructosamine levels prior to subjecting the patients to an elective surgical intervention. 2Tel Aviv University, Tel Aviv, Israel Future studies should explore whether correction of fructos- 3 amine reduces the risk for complications. Department of , Thomas Jefferson University, Philadelphia, Pennsylvania

Appendix E-mail address for J. Parvizi: [email protected] Tables showing demographic characteristics, glycemic status, and preoperative data on the study participants; ORCID iD for J. Parvizi: 0000-0002-6985-5870

References

1. Louati K, Vidal C, Berenbaum F, Sellam J. Association between diabetes mellitus 13. Iorio R, Williams KM, Marcantonio AJ, Specht LM, Tilzey JF, Healy WL. Diabetes and : systematic literature review and meta-analysis. RMD Open. 2015 mellitus, hemoglobin A1C, and the incidence of total joint arthroplasty infection. Jun 2;1(1):e000077. J Arthroplasty. 2012 May;27(5):726-9.e1. Epub 2011 Nov 4. 2. Bolognesi MP, Marchant MH Jr, Viens NA, Cook C, Pietrobon R, Vail TP. The 14. Adams AL, Paxton EW, Wang JQ, Johnson ES, Bayliss EA, Ferrara A, Nakasato C, impact of diabetes on perioperative patient outcomes after total hip and total knee Bini SA, Namba RS. Surgical outcomes of total knee replacement according to arthroplasty in the . J Arthroplasty. 2008 Sep;23(6)(Suppl 1):92-8. diabetes status and glycemic control, 2001 to 2009. J Bone Joint Surg Am. 2013 3. Singh JA, Sloan J. Higher comorbidity, poor functional status and higher health Mar 20;95(6):481-7. care utilization in veterans with prevalent total knee arthroplasty or total hip ar- 15. Hogan C, Bucknell AL, King KB. The effect of diabetes mellitus on total joint throplasty. Clin Rheumatol. 2009 Sep;28(9):1025-33. Epub 2009 Jun 11. arthroplasty outcomes. JBJS Rev. 2016 Feb 16;4(2):01874474-201602000- 4. Stryker LS, Abdel MP, Morrey ME, Morrow MM, Kor DJ, Morrey BF. Elevated 00003. postoperative blood glucose and preoperative hemoglobin A1C are associated with 16. Harris AH, Bowe TR, Gupta S, Ellerbe LS, Giori NJ. Hemoglobin A1C as a marker increased wound complications following total joint arthroplasty. J Bone Joint Surg for surgical risk in diabetic patients undergoing total joint arthroplasty. J Arthroplasty. Am. 2013 May 1;95(9):808-14. 2013 Sep;28(8)(Suppl):25-9. Epub 2013 Jul 30. 5. King KB, Findley TW, Williams AE, Bucknell AL. Veterans with diabetes receive 17. Hwang JS, Kim SJ, Bamne AB, Na YG, Kim TK. Do glycemic markers predict arthroplasty more frequently and at a younger age. Clin Orthop Relat Res. 2013 occurrence of complications after total knee arthroplasty in patients with diabetes? Sep;471(9):3049-54. Epub 2013 May 7. Clin Orthop Relat Res. 2015 May;473(5):1726-31. Epub 2014 Nov 18. 6. Marchant MH Jr, Viens NA, Cook C, Vail TP, Bolognesi MP. The impact of glycemic 18. Maradit Kremers H, Lewallen LW, Mabry TM, Berry DJ, Berbari EF, Osmon DR. control and diabetes mellitus on perioperative outcomes after total joint arthro- Diabetes mellitus, hyperglycemia, hemoglobin A1C and the risk of prosthetic plasty. J Bone Joint Surg Am. 2009 Jul;91(7):1621-9. joint infections in total hip and knee arthroplasty. J Arthroplasty. 2015 7. Meding JB, Reddleman K, Keating ME, Klay A, Ritter MA, Faris PM, Berend ME. Mar;30(3):439-43. Epub 2014 Oct 15. Total knee replacement in patients with diabetes mellitus. Clin Orthop Relat Res. 19. Han HS, Kang SB. Relations between long-term glycemic control and postop- 2003 Nov;416:208-16. erative wound and infectious complications after total knee arthroplasty in type 2 8. Fisher DA, Dierckman B, Watts MR, Davis K. Looks good but feels bad: factors diabetics. Clin Orthop Surg. 2013 Jun;5(2):118-23. Epub 2013 May 15. that contribute to poor results after total knee arthroplasty. J Arthroplasty. 2007 20. Lavernia CJ, Heiner AD, Villa JM, Alcerro JC, Rossi MD. Preoperative glycemic Sep;22(6)(Suppl 2):39-42. Epub 2007 Jul 26. control on total joint arthroplasty patient-perceived outcomes and hospital costs. 9. J¨amsen E, Nevalainen P, Kalliovalkama J, Moilanen T. Preoperative hyperglyce- J Arthroplasty. 2017;32(1):6-10. Epub 2016 Jul 15. mia predicts infected total knee replacement. Eur J Intern Med. 2010 Jun;21 21. Gallagher EJ, Le Roith D, Bloomgarden Z. Review of hemoglobin A(1c) in the (3):196-201. Epub 2010 Mar 15. management of diabetes. J Diabetes. 2009 Mar;1(1):9-17. Epub 2009 Jan 27. 10. Mraovic B, Suh D, Jacovides C, Parvizi J. Perioperative hyperglycemia and 22. Armbruster DA. Fructosamine: structure, analysis, and clinical usefulness. Clin postoperative infection after lower limb arthroplasty. J Diabetes Sci Technol. 2011 Chem. 1987 Dec;33(12):2153-63. Mar 1;5(2):412-8. 23. Garlick RL, Mazer JS. The principal site of nonenzymatic of human 11. Chrastil J, Anderson MB, Stevens V, Anand R, Peters CL, Pelt CE. Is hemoglobin serum albumin in vivo. J Biol Chem. 1983 May 25;258(10):6142-6. A1c or perioperative hyperglycemia predictive of periprosthetic joint infection or 24. Rondeau P, Bourdon E. The of albumin: structural and functional im- death following primary total joint arthroplasty? J Arthroplasty. 2015 Jul;30 pacts. Biochimie. 2011 Apr;93(4):645-58. Epub 2010 Dec 16. (7):1197-202. Epub 2015 Jan 31. 25. Ceriello A. Postprandial hyperglycemia and diabetes complications: is it time to 12. Capozzi JD, Lepkowsky ER, Callari MM, Jordan ET, Koenig JA, Sirounian GH. The treat? Diabetes. 2005 Jan;54(1):1-7. prevalence of diabetes mellitus and routine hemoglobin A1c screening in elective 26. Malmstr¨om H, Walldius G, Grill V, Jungner I, Gudbj¨ornsdottir S, Hammar N. total joint arthroplasty patients. J Arthroplasty. 2016 Jun 23. Fructosamine is a useful indicator of hyperglycaemia and glucose control in clinical 1907

T HE J OURNAL OF B ONE &JOINT SURGERY d JBJS. ORG SERUM FRUCTOSAMINE:ASIMPLE AND I NEXPENSIVE T EST FOR VOLUME 99-A d N UMBER 22 d N OVEMBER 15, 2017 ASSESSING P REOPERATIVE GLYCEMIC CONTROL

and epidemiological studies—cross-sectional and longitudinal experience from 39. Esposito K, Nappo F, Marfella R, Giugliano G, Giugliano F, Ciotola M, Quagliaro the AMORIS cohort. PLoS One. 2014 Oct 29;9(10):e111463. L, Ceriello A, Giugliano D. Inflammatory cytokine concentrations are acutely 27. Moura BP, Amorim PR, Silva BP, Franceschini SC, Reis JS, Marins JC. Effect of a increased by hyperglycemia in humans: role of oxidative . Circulation. 2002 short-term program on glycemic control measured by fructosamine test in Oct 15;106(16):2067-72. patients. Diabetol Metab Syndr. 2014 Feb 11;6(1):16. 40. Turina M, Fry DE, Polk HC Jr. Acute hyperglycemia and the innate immune 28. Shima K, Komatsu M, Noma Y, Miya K. Glycated albumin (GA) is more advan- system: clinical, cellular, and molecular aspects. Crit Care Med. 2005 Jul;33 tageous than hemoglobin A1c for evaluating the efficacy of in achieving (7):1624-33. glycemic control in patients with type 2 diabetes. Intern Med. 2014;53(8):829-35. 41. Ogawa A, Hayashi A, Kishihara E, Yoshino S, Takeuchi A, Shichiri M. New Epub 2014 Apr 15. indices for predicting glycaemic variability. PLoS One. 2012;7(9):e46517. Epub 29. Parvizi J. New definition for periprosthetic joint infection. Am J Orthop (Belle 2012 Sep 27. Mead NJ). 2011 Dec;40(12):614-5. 42. Suwa T, Ohta A, Matsui T, Koganei R, Kato H, Kawata T, Sada Y, Ishii S, Kondo A, 30. Dronge AS, Perkal MF, Kancir S, Concato J, Aslan M, Rosenthal RA. Long-term Murakami K, Katabami T, Tanaka Y. Relationship between clinical markers of gly- glycemic control and postoperative infectious complications. Arch Surg. 2006 cemia and glucose excursion evaluated by continuous glucose monitoring (CGM). Apr;141(4):375-80; discussion 380. Endocr J. 2010;57(2):135-40. Epub 2009 Nov 19. 31. Halkos ME, Lattouf OM, Puskas JD, Kilgo P, Cooper WA, Morris CD, Guyton RA, 43. Ma X, Hu X, Zhou J, Hao Y, Luo Y, Lu Z, Bao Y, Jia W. Glycated albumin is more Thourani VH. Elevated preoperative hemoglobin A1c level is associated with reduced closely correlated with than 1,5-anhydroglucitol and glycated long-term survival after coronary artery bypass surgery. Ann Thorac Surg. 2008 hemoglobin A1c. Cardiovasc Diabetol. 2015 Feb 7;14:16. Nov;86(5):1431-7. 44. Nathan DM, McGee P, Steffes MW, Lachin JM, DCCT/EDIC Research Group. 32. Halkos ME, Puskas JD, Lattouf OM, Kilgo P, Kerendi F, Song HK, Guyton RA, Relationship of glycated albumin to blood glucose and HbA1c values and to reti- Thourani VH. Elevated preoperative hemoglobin A1c level is predictive of adverse nopathy, nephropathy, and cardiovascular outcomes in the DCCT/EDIC study. Dia- events after coronary artery bypass surgery. J Thorac Cardiovasc Surg. 2008 betes. 2014 Jan;63(1):282-90. Epub 2013 Aug 29. Sep;136(3):631-40. 45. Parrinello CM, Selvin E. Beyond HbA1c and glucose: the role of nontraditional 33. Maoz G, Phillips M, Bosco J, Slover J, Stachel A, Inneh I, Iorio R. The Otto Aufranc glycemic markers in diabetes diagnosis, prognosis, and management. Curr Diab Award: modifiable versus nonmodifiable risk factors for infection after hip arthro- Rep. 2014;14(11):548. plasty. Clin Orthop Relat Res. 2015 Feb;473(2):453-9. 46. Selvin E, Rawlings AM, Grams M, Klein R, Sharrett AR, Steffes M, Coresh J. 34. Pedersen AB, Mehnert F, Johnsen SP, Sørensen HT. Risk of revision of a total Fructosamine and glycated albumin for risk stratification and prediction of incident hip replacement in patients with diabetes mellitus: a population-based follow up diabetes and microvascular complications: a prospective cohort analysis of the study. J Bone Joint Surg Br. 2010 Jul;92(7):929-34. Risk in Communities (ARIC) study. Lancet Diabetes Endocrinol. 35. Singh JA, Lewallen DG. Diabetes: a risk factor for poor functional outcome after 2014 Apr;2(4):279-88. Epub 2014 Jan 15. total knee arthroplasty. PLoS One. 2013 Nov 13;8(11):e78991. 47. Selvin E, Rawlings AM, Lutsey PL, Maruthur N, Pankow JS, Steffes M, Coresh J. 36. Giori NJ, Ellerbe LS, Bowe T, Gupta S, Harris AH. Many diabetic total joint Fructosamine and glycated albumin and the risk of cardiovascular outcomes and arthroplasty candidates are unable to achieve a preoperative hemoglobin A1c goal of death. Circulation. 2015 Jul 28;132(4):269-77. Epub 2015 May 28. 7% or less. J Bone Joint Surg Am. 2014 Mar 19;96(6):500-4. 48. Hill RP. The effect of calibration on the between-laboratory variation of serum 37. Koenig RJ, Peterson CM, Jones RL, Saudek C, Lehrman M, Cerami A. Correla- fructosamine. Ann Clin Biochem. 1988 Jul;25(Pt 4):435-9. tion of glucose regulation and hemoglobin AIc in diabetes mellitus. N Engl J Med. 49. Berr´ıos-Torres SI, Umscheid CA, Bratzler DW, Leas B, Stone EC, Kelz RR, Reinke 1976 Aug 19;295(8):417-20. CE, Morgan S, Solomkin JS, Mazuski JE, Dellinger EP, Itani KMF, Berbari EF, Segreti 38. Lee EY, Lee BW, Kim D, Lee YH, Kim KJ, Kang ES, Cha BS, Lee EJ, Lee HC. J, Parvizi J, Blanchard J, Allen G, Kluytmans JAJW, Donlan R, Schecter WP; Health- Glycated albumin is a useful glycation index for monitoring fluctuating and poorly care Infection Control Practices Advisory Committee.Centers for Disease Control and controlled type 2 diabetic patients. Acta Diabetol. 2011 Jun;48(2):167-72. Epub Prevention Guideline for the Prevention of Surgical Site Infection, 2017. JAMA Surg. 2010 Dec 10. 2017 May 3. [Epub ahead of print.]