Preoperative A1c and Postoperative Infection in Elective Hand Surgery (2024)

Abstract

Diabetes mellitus affects 10.5% of the US population. Numerous studies have documented increased risk of complications for patients with diabetes after different types of surgery, including hand surgery. By aiming for a preoperative target hemoglobin A1c (A1c), the risk of surgical complications following elective hand surgery may be reduced for patients with diabetes. This literature review was conducted to evaluate the association between diabetes mellitus and surgical site infections and, more specifically, to determine whether there is any association between preoperative A1c level and postoperative infections in hand surgery. The risk for surgical site infections and wound complications appears to be higher for patients with insulin-dependent diabetes mellitus, but not necessarily for patients with noninsulin-dependent diabetes mellitus, when compared with patients without diabetes. The role of prophylactic antibiotics for patients with diabetes undergoing elective hand surgery was also considered. Prophylactic antibiotics have not been shown to be beneficial for healthy patients undergoing clean, elective hand surgery. However, preoperative antibiotics may have a protective role for some patients with poorly controlled hyperglycemia.

Keywords: surgery, specialty, infection, diagnosis, diabetes mellitus, antibiotic, hyperglycemia, A1c, hand surgery, elective, surgical site infection

Introduction

According to the Centers for Disease Control and Prevention (CDC), 10.5% of the US population had diabetes mellitus (DM) in 2018.¹ By the year 2030, the national prevalence of DM is anticipated to increase by 54%.² Hemoglobin A1c (A1c) represents average glycemia for approximately the prior 3 months. Historically, this has been the metric used most often in clinical trials that have shown the effects of glycemic control. The American Diabetes Association (ADA) recommends an A1c target of <7% for “many” nonpregnant adults with acknowledgment that targets for individuals should be patient centered and safely achievable without significant hypoglycemia.³ This guideline was established based on studies demonstrating significant reductions in the rates of development and progression of microvascular complications (eg, neuropathy, retinopathy, and nephropathy) for patients who met their A1c goal of 7%. The ADA notes that a less strict target of <8% may be appropriate for patients in whom the risks of treatment exceed the benefits or those with limited life expectancy.³

Numerous studies have documented the increased risk of complications and worse outcomes for patients with diabetes after different types of surgery, including hand surgery.^4-6 Among the most common complications are wound healing problems and surgical site infection (SSI). Surgical site infection after hand surgery can cause substantial morbidity. Subsequent stiffness and scarring from SSI can impair hand function. Seemingly minor postoperative wound issues can delay return to activities and work and more severe SSI can result in hospitalization, additional surgery, prolonged rehabilitation, increased health care expenditures, additional tissue loss or amputation, permanent disability, and even death from systemic complications (eg, sepsis). Antibiotic prophylaxis has been shown to effectively reduce the risk of SSI in many surgical specialties. We are unaware of any evidence-based guidelines regarding preoperative A1c target levels specifically for elective hand surgery. There also appears to be a lack of consensus in the hand surgery literature regarding the use of preoperative antibiotics for patients with diabetes undergoing elective hand surgery. This review was undertaken to summarize the current evidence with the aim of giving hand surgeons guidance for clinical practice.

Pathophysiology

While age, obesity, and/or malnutrition may correlate with DM and contribute to increased risk of infection, it is thought that DM alone can cause a relative immune deficiency and vasculopathy.⁷ There are multiple mechanisms in the setting of DM that can potentially impair the response to infection and wound healing. Some of the mechanisms are believed to be related to hyperglycemia causing an increase in the circulating inflammatory cytokines as well as suppression of some aspects of immune function, such as chemotaxis, neutrophil adherence, phagocytosis, generation of reactive oxygen species, and intracellular killing of bacteria.⁸ Furthermore, the degree of neutrophil dysfunction is related to the severity of hyperglycemia.⁹ This relative immunodeficiency increases susceptibility to infection, including SSI. Williamson et al showed that hyperglycemia can cause pseudohypoxia and mimic true hypoxia, leading to related complications.¹⁰ A common complication of DM is neuropathy, which impairs neural-mediated vasodilation and thereby impedes the wound healing response.¹¹

Association Between DM and SSIs

The increased risk of SSI for patients with diabetes has been established for many different surgical specialties. A meta-analysis was performed by Martin et al to investigate the independent association between DM and SSI among different types of surgery.¹² The authors included 94 articles from US hospitals in their review. Among the categories of surgery were obstetrical and gynecological, colorectal, arthroplasty, breast, cardiac, spinal, and “other.” Studies included in the meta-analysis provided 866 427 procedures and 32 067 SSIs. Hemoglobin A1c levels were documented in some of the studies reviewed, but the number was insufficient to warrant a separate meta-analysis. An odds ratio (OR) of 1.53 (95% predictive interval [PI], 1.11-2.12) was determined to represent the overall effect size for the association between DM and SSI. The OR for the association between DM and SSI by surgery type was 1.16 for colorectal surgery (95% PI, 0.93-1.44), 1.26 for arthroplasty (95% PI, 1.01-1.66), 1.46 for other/multiple surgery types combined (95% PI, 1.07-2.00), 1.58 for breast (95% PI, 0.91-2.72), 1.61 for gynecological (95% PI, 1.15-2.24), 1.66 for spinal (95% PI, 1.10-2.32), and 2.03 for cardiac surgery (95% PI, 1.13-4.05). An overall OR of 1.88 (95% PI, 0.66-5.34) was found for the association between elevated blood glucose in the preoperative or intraoperative period and SSI.

Data in the National Surgical Quality Improvement Program (NSQIP) database at the Veterans Affairs Connecticut Healthcare System for 490 patients with diabetes who had undergone major noncardiac surgery were retrospectively reviewed by Dronge and colleagues.¹³ The SSI rate for patients with A1c <7 was 12%, while the SSI rate for those with A1c ≥7 was 20%. The investigators determined that a preoperative A1c level <7 was strongly associated with reduction in SSI. The type of treatment for DM did not affect SSI risk.

An increased risk of adverse outcomes for patients with diabetes has been observed in all orthopedic subspecialities, including sports medicine, foot and ankle, pediatrics, adult reconstruction, spine, and upper extremity (including hand) surgery.¹⁴ Goltsman et al reviewed 39 475 patients in the American College of Surgeons NSQIP database and looked at the relationship between DM and outcomes as well as wound-related complications (eg, SSI and wound dehiscence) after plastic surgery from 2007 to 2012.⁴ There were 1222 (3.10%) patients with insulin-dependent diabetes mellitus (IDDM) and 1915 (4.75%) patients with noninsulin-dependent diabetes mellitus (NIDDM). A variety of procedures, including lower extremity, breast, abdominal, craniofacial, and hand/upper extremity surgeries, were evaluated. Logistic regression analyses demonstrated that only patients with IDDM were more likely to have surgical complications, with an OR of 1.78 (P < .0001). Noninsulin-dependent diabetes mellitus did not appear to be significantly related to major surgical complications (P = .104). While NIDDM was not associated with wound complications (P = .112), the odds for patients with IDDM were 1.53 times (95% confidence interval [CI], 1.20-1.92; P < .001) that for patients without diabetes. Furthermore, patients with IDDM had higher likelihoods of wound dehiscence (OR, 2.05; 95% CI, 1.27-3.20; P < .0023) and SSI, both superficial and deep, (OR, 1.55; 95% CI, 1.18-2.02; P < .0012) than patients without diabetes. Noninsulin-dependent diabetes mellitus did not have a significant association with wound dehiscence (P = .8023) or SSI (P = .066). The odds of patients with IDDM, who underwent hand/upper extremity procedures, having surgical complications and wound infection were 4.64 times (95% CI, 2.24-8.59; P < .001) and 5.36 times (95% CI, 2.46-10.36; P = .0045) greater, respectively, compared with patients without diabetes. Patients with IDDM had a higher likelihood of SSI (OR, 4.95; 95% CI, 2.03-10.31; P < .001), but not wound dehiscence (P = .13), when compared with patients without diabetes. There was no association with hand/upper extremity surgical complications or wound complications for NIDDM (P = .25 and .87, respectively). The NSQIP database did not provide A1c values so the authors could not assess that variable.

Association Between DM and SSIs Following Hand Surgery

Numerous small studies (less than ~3000 cases/patients) examining SSI rates for common hand surgeries, such as partial palmar fasciectomy, carpal tunnel release (CTR), and trigger digit release (TDR), did not clearly show an increased SSI risk for patients with diabetes.^5,15-23 The overall complication rates were low, and the majority of these investigations were underpowered. Most studies examining the relationship between DM and hand surgery did not stratify patients with diabetes using A1c levels, and many did not distinguish between IDDM and NIDDM.

Some larger studies have identified an association between DM and SSI. The American College of Surgeons NSQIP database was assessed by Lipira et al for postoperative complications and related risk factors in hand surgery.²⁴ Notably, the NSQIP does not store data for some common hand surgery procedures, such as CTR, ulnar nerve release at the elbow, and certain types of arthrodesis. The investigators reviewed 10 646 patients between 2006 and 2011; 909 (8.5%) of these had DM. Univariate analysis revealed that DM had a significant association with increased risk of complications (relative risk, 2.41; 95% CI, 1.78-3.28; P < .001). Multivariate analysis, however, did not show a significant association between DM and increased complication risk. A1c was not included in the analysis.

Another retrospective review of 44 305 patients who underwent common hand surgery procedures in an ambulatory setting was performed using the California State Ambulatory Surgery database for 2010 and 2011.²⁵ There were 5362 patients with diabetes included. The investigators did not find an association between DM and increased risk for SSI resulting in an acute care visit (eg, hospitalizations, emergency department, or ambulatory surgical visits). A notable limitation is that postoperative visits to physician offices for SSI were not captured by the database.

Stepan et al studied risk factors and 30-day complication profiles for 52 727 patients in the NSQIP database, who underwent surgery from the distal humerus to the hand between 2005 and 2015.²⁶ The study included 297 Current Procedural Terminology codes. There were 4451 patients with diabetes (8.4%), including 2679 (5.1%) with NIDDM and 1772 (3.4%) with IDDM. The surgical complication rates were 0.87% for patients without diabetes, 0.86% for the NIDDM cohort, and 2.93% for the IDDM cohort. Patients with IDDM were found to be at greater risk of surgical site complications (OR, 2.35; 95% CI, 1.71-3.24), including deep SSI and superficial wound dehiscence, as well as superficial SSI (OR, 2.57; 95% CI, 1.75-33.76). There were no statistically significant differences identified when comparing the NIDDM group with the group without DM. However, because NSQIP categorizes patients who use diet alone to control DM as patients without diabetes, this could affect the actual difference in complication rates between patients with and without diabetes.

Diabetes mellitus was determined to be a significant risk factor for SSI in a study by Werner et al.²⁷ A group of 209 634 Medicare patients who underwent TDR within a national insurance database was assessed. The calculated SSI risk was 0.5% to 0.6%, and a significant statistical association with DM was reported. In a different study, Werner et al reviewed 454 987 Medicare patients in a national administrative database and identified 1466 postoperative infections following open CTR (0.32%).²⁸ Among the independent positive risk factors identified were male sex, younger age, tobacco use, alcohol use, obesity (body mass index of 30-40), morbid obesity (body mass index: >40), and comorbidities including DM, inflammatory arthritis, peripheral vascular disease, chronic liver disease, chronic lung disease, chronic kidney disease, and depression.

Data regarding A1c and its relationship to SSI are limited to a small number of studies. A retrospective chart review of 100 patients who had undergone TDR was performed by Blyth and Ross.²⁹ Eighteen of these patients had DM (6 IDDM). The mean A1c level for the group was 9.4% (range: 4.4-13.1), which was calculated by using an average of 8 different A1c levels for each patient. There were no wound complications or SSIs reported. Zwiebel and Becker retrospectively reviewed 658 CTR cases (260 with DM) and determined that 3.1% of patients with diabetes and 4.8% of patients without diabetes developed SSI.³⁰ After excluding patients who received antibiotic prophylaxis, they found that 3.8% of patients with diabetes and 4.9% of patients without diabetes developed SSI (P > .05). Moreover, their data disclosed that diabetic patients with A1c ≥7% did not have a significantly different SSI risk compared with patients with A1c <7% (P > .05).

After retrospectively reviewing 132 patients with diabetes who underwent CTR, Cunningham et al determined a superficial SSI incidence of 7.1% with A1c ≥7.8% versus 2.2%, although this was not significant.³¹

Another study by Werner and colleagues evaluated a group of 7958 patients with diabetes in a national private-payer insurance database.³² Patients were stratified into groups based on preoperative A1c levels, and the incidence of SSI after open CTR was calculated for each group. Through a receiver operating characteristic analysis, they determined that the inflection point for the curve representing 6-month SSI incidence occurred between A1c levels of 7.0 and 8.0 mg/dL. After adjusting for several comorbidities, they found that preoperative A1c levels greater than 8.0 mg/dL posed significantly higher risk for SSI. Zhuang and colleagues evaluated the records of 14 056 patients with diabetes who underwent some common elective upper extremity procedures within a large, insurance-based database.³³ They found that patients with A1c ≥7% had a significantly increased incidence (1.11% vs 0.72% ) of SSI within 30 days.

Role of Prophylactic Antibiotics for Patients With Diabetes Undergoing Elective Hand Surgery

Another question is whether antibiotics can influence the SSI rate in elective soft tissue hand surgery, particularly for patients with DM. Antibiotic use is not benign. Possible repercussions from prophylactic antibiotics range from minor adverse reactions, such as a skin rash, to major allergic reactions, such as life-threatening anaphylaxis. Other serious sequelae of antibiotic use include development of antibiotic-resistant strains of organisms and Clostridium difficile enterotoxemia. In addition, there is financial cost to the health care system incurred by unnecessary administration of antibiotics.

A prospective series of 236 consecutive patients who underwent elective and emergency hand surgery was analyzed by Platt and Page with the aim of developing guidelines for perioperative antibiotic administration.³⁴ Antibiotic use was at the discretion of the attending surgeon. They found that 4 (8%) of the 48 patients who underwent elective hand surgery and received prophylactic antibiotics experienced SSI. This was compared with 8 (12%) of the 64 patients who did not receive antibiotics. The difference was not significant (P = .55). They also determined that odds of infection for emergency hand surgery without antibiotics were 8.5 times greater (95% CI, 1.5-49; P = .014) for clean wounds and 13.4 times greater (95% CI, 2.5-72; P = .002) for dirty wounds. The authors recommended antibiotic prophylaxis for elective hand surgery only in the following situations: patients with major comorbidities (eg, DM, rheumatoid arthritis, chronic steroids), percutaneous wires, implants (eg, metal, plastic), and procedures lasting greater than 2 hours.

Vasconselos and colleagues conducted a retrospective review of 346 patients (50 had DM) who underwent clean, elective, outpatient hand surgery.³⁵ Preoperative antibiotics were given at the discretion of the attending surgeon to 180 (52%) of the total number and to 27 (54%) of the patients with diabetes. There were 4 SSI cases (1% infection rate), and 1 of these was a patient with diabetes. The infection rate was not statistically different for patients with diabetes, and prophylactic antibiotics did not have a statistically significant effect on the development of SSI, even for patients with diabetes.

A prospective multicenter study by Bӓcker and colleagues included 434 patients who had elective soft tissue hand surgery.³⁶ Prophylactic antibiotics were given to 257 (59.2%) of the patients. There were 44 patients with diabetes (18 of whom received antibiotics). There were 1 SSI (0.006%), 1 wound dehiscence, and 1 suture abscess in the group that received antibiotics. There were 1 SSI (0.003%) and 4 suture abscesses in the group that did not receive antibiotics. There was no significant difference between the groups regarding infection rate or comorbidities.

A study by Ko et al identified a trend for greater protective effect of prophylactic antibiotics in patients with DM (OR = 0.46) compared with those without DM (OR = 1.36).³⁷ They assessed 469 CTR procedures by chart review and implemented a propensity score-matched model. The DM cohort consisted of 163 patients. Preoperative antibiotics were given to 219 (46.7%) of the total number and to 110 (67%) of the patients with diabetes. The SSI rate was 3.1% for patients with diabetes versus 6.5% for patients without diabetes. The difference was not statistically significant. The SSI rate for patients who received antibiotics was 5.0% versus 5.6% for those who did not. The investigators stratified patients with diabetes based on A1c levels. The SSI rate was 1.9% for diabetic patients with A1c >7 who received preoperative antibiotics. In contrast, the SSI rate was 5.7% for the same group of patients with diabetes who did not receive antibiotics; however, the difference did not reach statistical significance. The authors concluded that there was no significant benefit for preoperative antibiotics in healthy patients who undergo CTR, but they posited that the trend identified warrants further investigation.

In their evaluation of 3003 patients (546 patients with diabetes) treated by CTR, Harness et al found that antibiotic prophylaxis did not appear to decrease the risk of SSI, even for patients with diabetes.¹⁶ The authors acknowledged that prophylactic antibiotics may be appropriate for poorly controlled patients with diabetes. They also noted that the study had insufficient power to detect a difference in the SSI rate between groups because of the overall low rate of SSI (0.36%).

A similar conclusion was reached by Tosti et al.³⁸ There was no difference in the rate of SSI between those who did and did not receive prophylactic antibiotics in a multicenter retrospective review of 600 consecutive patients who had elective soft tissue hand surgery. The overall SSI rate was 0.66%, with all being superficial. The SSI rate for the group that received preoperative antibiotics was 0.47%, and it was 0.77% for the group that did not. The difference was not significant (P = 1.00). The authors noted that antibiotics were not randomly administered and that the low SSI rate precluded multivariate analysis for individual risk factors.

Bykowski et al retrospectively reviewed the records of 8850 patients who had clean, elective outpatient hand surgery to evaluate whether prophylactic antibiotics were efficacious for decreasing SSI risk after surgery.⁶ Of these patients, 968 had DM. The number of patients who received preoperative antibiotics was 2755 (31%) of the total and 354 (37%) of patients with diabetes. They found an overall SSI rate of 0.35%, with no significant difference between those who received antibiotics (0.54%) and those who did not (0.26%). The authors noted that patients with DM experienced suture granulomas and wound dehiscence (7 and 4 times, respectively) more frequently than patients without diabetes. Diabetes mellitus was identified as a significant predictor of SSI (OR = 2.8; 95% CI, 1.2-6.5; P = .019) on multivariate regression analysis. A subgroup evaluation for the protective effect of prophylactic antibiotic administration in patients with diabetes revealed no significant difference.

An administrative insurance claims database was retrospectively assessed by Li and colleagues to determine whether prophylactic antibiotics reduced the risk of SSI following soft tissue hand surgery.³⁹ There were 516 986 patients (99 033 patients with diabetes) included, and 58 201 (12 041 patients with diabetes) of them received preoperative antibiotics. The researchers calculated propensity scores to create cohorts that were matched based on possible SSI risk factors (eg, age, procedure type, steroid use, tobacco use, and comorbidities). They used multivariable logistic regression both before and after the propensity score matching to test whether preoperative antibiotics were related to a reduction in SSI risk within 30 postoperative days. The diabetic matched cohort consisted of 12 041 patients who received antibiotics and 11 718 who did not. The authors found no difference in SSI risk between patients who received preoperative antibiotics and those who did not (OR, 1.03; 95% CI, 0.93-1.13; P = .585).

A confounding factor is the definition of SSI. Variability in reported infection rates partially reflects the absence of a standardized definition for SSI across studies. It is difficult to precisely define infection, and so the criteria for diagnosis vary among investigators. Some swelling and erythema can be normal following surgery. A suture abscess may be sterile, and a small wound dehiscence may be colonized but this does not necessarily represent infection. The CDC has defined SSI and made a distinction between superficial and deep infections.⁴⁰ However, the CDC guidelines were developed after some of the studies presented here, and many investigators have not implemented the definitions in their research. Furthermore, Platt and Page pointed out that there is not a deep layer of fascia in the hand that surgeons routinely repair.³⁴

Summary

There is a paucity of data regarding the relationship between preoperative A1c and SSI after elective hand surgery. The overall hand surgery SSI risk is low. Infection rates for elective hand surgery have variously been reported to range from 0% to 10%, with the majority of them being <2% (Table 1). It is difficult to achieve the statistical power required to detect some differences, such as the influence of DM and the treatment effect of prophylactic antibiotics, because large sample sizes are needed, given such low SSI rates. A few studies reviewed disclosed a significant association between DM and SSI after hand surgery.^6,27,28,33 Two studies demonstrated a significant relationship for IDDM and SSI but not for NIDDM.^4,26 One study identified a significant association between preoperative A1c >8 and SSI, and another showed significance for A1c ≥7.^32,33 None of the studies reviewed showed a significant protective effect for prophylactic antibiotics in patients with or without diabetes, although 1 study did reveal a trend toward a protective effect for diabetic patients with an A1c >7.³⁷ It may be that perioperative glycemic control has a greater effect on the risk of complications than the presence of DM. The use of insulin may also be an independent risk factor for SSI, as concluded by Goltsman and colleagues.⁴ Future studies with large enough sample sizes to achieve statistical power are needed to develop meaningful guidelines that will minimize SSI risk and optimize outcomes.

Footnotes

References