Purpose
To evaluate the association of diabetes and perioperative hemoglobin A1C (HgA1C) value with postoperative wound healing complications following carpal tunnel release (CTR) and trigger finger release (TFR).
Methods
A retrospective review of diabetic patients who underwent CTR and/or TFR between 2014 and 2018 was performed. Hemoglobin A1C value within 90 days of surgery was recorded for all diabetic patients. A nondiabetic comparison group was selected from within the same study period in an approximately 1:1 procedural ratio, although direct matching was not performed. A chart review was used to examine postoperative wound healing complications, such as wound infection, wound dehiscence, or delayed wound healing.
Results
Two hundred sixty-two diabetic patients and 259 nondiabetic patients underwent 335 and 337 CTR and/or TFR procedures, respectively. There were 36 wound complications in the diabetic group and 9 complications in the nondiabetic group. Logistic regression analysis demonstrated an increased association of wound healing complications with diabetic patients compared to nondiabetic patients. Additionally, an increased association was demonstrated among diabetic patients with an HgA1C value above 6.5% compared with those with an HgA1C value below 6.5%.
Conclusions
Compared with nondiabetic controls, diabetic patients have increased associated risk of postoperative wound healing complications following CTR and/or TFR. This increased association was further demonstrated among diabetic patients with elevated perioperative HgA1C values.
Type of study/level of evidence
Prognostic IV.
Key words
The prevalence of diabetes mellitus (DM) has been estimated to be 12%–14% in the United States among those aged 20 years and older, and its incidence continues to increase.
1
,2
Surgical outcomes in diabetic patients are reported in many surgical specialties, including hand surgery. The literature published in the field of arthroplasty, sports medicine, as well as foot and ankle has presented evidence that DM is associated with a higher rate of adverse events when compared with a nondiabetic cohort.Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2020. Accessed August 28, 2020. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf
3
, 4
, 5
, 6
, 7
Within the field of hand surgery, the association of DM and postoperative wound healing complications has not been demonstrated to the same degree. Stepan et al8
conducted the largest study on the effect DM may have on patients undergoing hand surgery. They demonstrated that patients who have insulin-dependent DM are at a greater risk of postoperative complications, including infection, compared with nondiabetic patients across several different upper-extremity procedures. However, carpal tunnel release (CTR) or trigger finger release (TFR) was not assessed in this study, and HgA1C values were not collected. The effect DM may have on postoperative wound healing complications in minor hand procedures, such as carpal tunnel surgery, is less clear, with many studies failing to demonstrate an effect.9
, 10
, 11
, 12
The main criticism regarding these prior studies is their small sample size. Additionally, some lacked a nondiabetic group to demonstrate a baseline rate of complication.Glycosylated hemoglobin (HgA1C) is a marker of long-term glucose control, and its value has been examined in relation to postoperative outcomes. Recent studies in adult reconstruction and foot and ankle surgery have evaluated the association of HgA1C with adverse events, including surgical site infection (SSI).
4
,5
,7
Within the field of hand surgery, diabetic patients are at increased risk of infection, and this risk is exacerbated with poor glycemic control.13
The rate of infection in patients undergoing hand surgery is generally low. Harness et al
9
showed that the rate of infection following CTR was 11/3,003 in all patients and 3/546 among diabetic patients. SSI following hand surgery in all patients has been estimated to occur in 1.7/1,000 patients, with 10% of postoperative patients with SSI ultimately requiring a secondary procedure.14
Diabetic patients who develop upper-extremity infections experience a greater number of polymicrobial infections, require more frequent serial debridement, and undergo a greater number of amputations for definitive management.13
Altogether, we find the current body of literature conflicting in its overall findings and thus, lacking the ability to answer an important question: are diabetic patients at increased risk of developing a wound healing complication following a CTR or TFR? Examining this relationship was the primary aim of this study. The secondary aim was to determine if diabetic patients with poor perioperative glycemic control have an increased risk of wound healing complications following these procedures compared with diabetic patients with good perioperative glycemic control.
Materials and Methods
We retrospectively reviewed the charts of patients from our tertiary care institution who underwent open CTR or TFR over a 4-year period (2014–2018). Access to and the collection of patient health information prior to 2014 was limited because of an institutional change of electronic medical records (EMRs) at that time. The procedures were performed by 7 different fellowship-trained hand surgeons. All the procedures were performed on an outpatient basis in an operating room setting using a variety of anesthetic techniques, ranging from wide-awake with local anesthesia to general anesthesia. As part of routine practice at our institution, including for diabetic patients, preoperative antibiotics were not administered for simple hand procedures. All the procedures were open releases, with no endoscopic procedures included in either cohort. The experimental cohort consisted of patients who had a pre-existing diagnosis of DM at the time of surgery and, additionally, had an HgA1C value documented within 90 days of the surgery. We did not differentiate patients based on type 1 or type 2 DM, and we did not collect data on the medical management of their DM, regardless of whether the patients were insulin-dependent. The exclusion criteria were any revision surgery, concurrent bone/fracture surgery (ie, a CTR with a distal radius fracture treated with open reduction and internal fixation), autoimmune disorder, human immunodeficiency virus/acquired immunodeficiency syndrome, hypothyroidism, malignancy, systemic immunosuppression/steroid use, or chemotherapy. Institutional review board approval was obtained.
Nondiabetic patients were selected in an approximately 1:1 TFR/CTR ratio. To accomplish this, all nondiabetic patients who underwent CTR or TFR and did not meet the exclusion criteria were included sequentially from October 2016 to December 2017. This method of comparison group selection was chosen to ensure that we captured a baseline rate of complication within our nondiabetic cohort. Given the limitations in the number of control patients within the study period, true patient-to-patient age and sex, body mass index (BMI), and smoking status matching was not performed.
Patients who underwent multiple procedures during separate and distinct surgical encounters over the study period were included as separate encounters (eg, a patient who underwent a right CTR in 2016 and left CTR in 2017). Any patient who underwent multiple procedures during a single visit (CTR and TFR or multiple TFRs) was included as 1 encounter. For a patient to be included, the procedures had to have been separated by at least 3 months, and a separate HgA1C value had to be available for all diabetic patients.
Hemoglobin A1C values were collected from all diabetic patients’ EMRs or outside shared health records. The smoking status and BMI, when available in the patients’ EMR, were assessed as documented, although these variables were not matched for diabetic and nondiabetic patients. Patients were considered active, former, or never smokers. Wound healing complications were assessed at the first documented postoperative visit, which was approximately 2 weeks after the surgery date. Wound healing complication was defined as: delayed wound healing, continued serous drainage, wound dehiscence, or surgical site infection following the definitions set by the CDC.
15
This was assessed using either the visit diagnosis and/or within the text of the visit’s documentation. Complications occurring after the 2-week postoperative visit were not regularly assessed in this study because most patients did not have further follow up scheduled. Among patients who were followed up beyond 2 weeks, none were identified as having late wound healing complications.Statistical analysis
The Student t test and chi square test were used to evaluate differences between continuous and categorical variable groups, respectively. The dependent variable, wound healing complication, was considered a yes/no binary variable. An estimate of the incidence of diabetic and nondiabetic wound healing complications (4% vs 0.5%, respectively) was used for a power analysis calculation, estimating n = 281 per group. The predictors of complications were identified a priori and included patient sex, age, perioperative smoking status, BMI, perioperative HgA1C value, and diabetic status. To determine an HgA1C cutoff point, a series of sequential univariate logistic regression models was used. The patients were separated into 2 groups starting at an HgA1C value above or below 6.0% and increased in 0.1 increments to determine an upper limit of difference that was statistically significant between the groups. To determine the independent predictors of wound healing complications, while adjusting for multiple repeated measures, a random-effects, generalized linear mixed model was run using binary logistic regression, with deidentified patient number as the random effect. Statistical significance was set at P = .05.
Results
Between June 2014 and March 2018, there were 386 patients with DM who underwent 465 CTR or TFR procedures. Further refinement for the available HgA1C values and exclusion criteria narrowed this to 262 diabetic patients undergoing 335 surgical procedures. The control group consisted of 259 nondiabetic patients undergoing 337 surgical procedures. There were 61 patients in the diabetic group and 66 patients in the control group who underwent multiple procedures (Table 1, Table 2).
Table 1Patient Demographic Data
| Variable | Diabetic Group | Nondiabetic Group |
|---|---|---|
| Sex | ||
| Males | 98 | 98 |
| Females | 164 | 161 |
| Mean age (y) | 58.4 | 58.8 |
| Number of CTRs | 206 | 215 |
| Number of TFRs | 129 | 122 |
| Mean BMI (kg/m2) | 34.0 | 30.6 |
| Active smokers | 33 | 25 |
| Mean HgA1C value (%) | 7.4 | |
| Median HgA1C value (%) | 7.2 | |
| HgA1C value days from the time of surgery (%) | 39.6 |
Table 2Number of Wound Complications Based on the Group and Surgical Procedure
| Wound Complications | Diabetic Group | Nondiabetic Group |
|---|---|---|
| Infections | 24 | 4 |
| TFR | 10 | 2 |
| CTR | 14 | 2 |
| Delayed wound healing | 12 | 5 |
| TFR | 2 | 0 |
| CTR | 10 | 5 |
| Total | 36 | 9 |
There were 36 wound healing complications in our diabetic group, with 24 infections requiring antibiotic therapy, 9 wounds with dehiscence, and 3 wounds with serous drainage. There were 9 wound healing complications in the nondiabetic group, with 4 infections requiring antibiotic therapy, 2 wounds with delayed healing, 2 wounds with dehiscence, and 1 wound with serous drainage. On a per-procedure basis, this equated to a 10.7% incidence of complications in the diabetic group and 2.7% incidence in the nondiabetic group. One patient in the control group required admission for intravenous antibiotic therapy. The remaining 27 patients with infections resolved with an outpatient course of oral antibiotic monotherapy. There were no patients who required returning to the operating room in either group (Table 3).
Table 3Detailed Representation of Complications With Their Subsequent Treatment
| HgA1C Value (%) | Surgery | Complication | Treatment | Tobacco Smoking | |
|---|---|---|---|---|---|
| Diabetic procedures | |||||
| 1 | 5.9 | CTR | Cellulitis | Cephalexin | Never |
| 2 | 11.5 | CTR | Cellulitis | Cephalexin | Former |
| 3 | 7.2 | TFR | Purulence | Doxycycline | Never |
| 4 | 7.1 | CTR | Dehiscence | Local wound care | Never |
| 5 | 7.9 | CTR | Purulence | Cephalexin | Never |
| 6 | 8.3 | CTR | Purulence | Augmentin | Never |
| 7 | 9.1 | CTR | Purulence | Augmentin | Never |
| 8 | 6.2 | CTR | Serous drainage | Local wound care | Former |
| 9 | 9.7 | CTR | Cellulitis | Cephalexin | Never |
| 10 | 9.4 | CTR | Dehiscence | Local wound care | Never |
| 11 | 11.2 | CTR | Purulence | Clindamycin | Never |
| 12 | 6.9 | CTR | Dehiscence | Local wound care | Never |
| 13 | 14.5 | CTR | Purulence | Cephalexin | Never |
| 14 | 7.4 | TFR | Purulence | Trimethoprim/Sulfamethoxazole | Never |
| 15 | 8 | CTR | Dehiscence | Former | |
| 16 | 7.4 | TFR | Purulence | Trimethoprim/Sulfamethoxazole | Never |
| 17 | 8.3 | CTR | Cellulitis | Cephalexin | Never |
| 18 | 11.4 | CTR | Purulence | Cephalexin | Never |
| 19 | 6.6 | TFR | Dehiscence | Local wound care | Never |
| 20 | 5.8 | CTR | Dehiscence | Local wound care | Never |
| 21 | 6.3 | CTR | Purulence | Doxycycline | Never |
| 22 | 9 | TFR | Cellulitis | Cephalexin | Never |
| 23 | 7.1 | CTR | Cellulitis | Cephalexin | Former |
| 24 | 6.6 | CTR | Serous drainage | Local wound care | Never |
| 25 | 6.6 | CTR | Cellulitis | Cephalexin | Current |
| 26 | 6 | CTR | Cellulitis | Cephalexin | Never |
| 27 | 7.1 | TFR | Cellulitis | Doxycycline | Never |
| 28 | 7.9 | CTR | Serous drainage | Local wound care | Never |
| 29 | 7.1 | CTR | Dehiscence | Local wound care | Never |
| 30 | 7.3 | TFR | Cellulitis | Cephalexin | Never |
| 31 | 7.9 | TFR | Purulence | Cephalexin | Never |
| 32 | 8.2 | TFR | Cellulitis | Augmentin | Never |
| 33 | 5.9 | CTR | Dehiscence | Local wound care | Never |
| 34 | 8.3 | TFR | Cellulitis | Doxycycline | Never |
| 35 | 6.6 | TFR | Dehiscence | Local wound care | Never |
| 36 | 8.3 | TFR | Purulence | Trimethoprim/Sulfamethoxazole | Never |
| Nondiabetic procedures | |||||
| 1 | CTR | Purulence | Trimethoprim/Sulfamethoxazole | Current | |
| 2 | CTR | Purulence | Cephalexin | Never | |
| 3 | TFR | Purulence | Cephalexin | Never | |
| 4 | CTR | Delayed healing | Local wound care | Former | |
| 5 | CTR | Serous drainage | Local wound care | Never | |
| 6 | CTR | Dehiscence | Local wound care | Never | |
| 7 | CTR | Delayed healing | Local wound care | Current | |
| 8 | TFR | Cellulitis | Cephalexin | Never | |
| 9 | CTR | Dehiscence | Local wound care | Former | |
∗ Staged bilateral surgery performed in 2 patients at different time points.
Diabetic patients were at an increased risk of postsurgical complications compared with nondiabetic patients (odds ratio [OR] = 2.66, 95% confidence interval [CI] 1.42–5.0).
Among diabetics, there were 229 encounters with an HgA1C value above 6.5%, with 30 wound complication occurrences, and 107 encounters with an HgA1C value below 6.5%, with 6 wound complication occurrences. The median HgA1C value was 7.2%, with a mean of 39.6 days between the date of surgery and HgA1C laboratory draw.
Among diabetic patients, those with an HgA1C value greater than 6.5% were at increased risk of developing complications compared with those with an HgA1C value below 6.5% (OR = 2.61, 95% CI 1.05–6.54).
There were 33 active smokers within the diabetic group and 25 in the nondiabetic group. The possible effect smoking had on wound healing complications was determined in the whole study (OR = 1.23, 95% CI 0.40–3.76) and diabetic groups (OR = 4.26, 95% CI 0.56–32.26).
The BMI within 1 year of surgery was available for 300 diabetic patients and 212 nondiabetic patients. The mean BMI across the total group was 32.6 kg/m2 (range: 17.9–64.0, SD: 7.6). Mixed-model logistic regression was performed on the entire group of 512 patients with the available BMI data (OR = 1.04, 95% CI 1.01–1.08) and among only the 300 diabetic patients (OR = 1.05, 95% CI 1.00–1.09). This OR represented an increased risk of wound healing complications for each 1.0-incremental increase in the BMI.
Discussion
In this retrospective study, we report wound healing complications among diabetic patients undergoing 2 of the most common hand, soft-tissue surgeries: CTR and TFR. Compared with nondiabetics, diabetic patients have a statistically significant increase in the odds of developing a wound healing complication following this surgery. We acknowledge that this finding disagrees with that of previous studies. The prospective study by Mondelli et al
10
concluded that there are no differences in the complication rate between diabetics and nondiabetics, but their diabetic cohort had only 24 patients and may have been underpowered to detect a difference between the groups. Similarly, Thomsen et al12
reported no complications in either cohort but had a diabetic cohort of only 35 patients. Perioperative wound healing complications were not a primary outcome measure for either of these studies. Harness et al9
noted no significant increase in the rate of SSIs in diabetics in a multicenter retrospective review of CTR. Their overall infection rate was low at 0.34%, but their method of screening relied on accurate SSI coding via International Classification of Diseases, ninth revision, codes, and infections may not have been captured in the dataset if they were not correctly coded in the EMR. Our study reviewed the text of each chart of the postoperative encounters to evaluate wound healing problems. Additionally, we evaluated wound healing problems other than SSI, such as dehiscence and delayed wound healing, because we felt that this was clinically relevant to the postoperative period. The goal of this study design was to have adequate power to detect a difference between the cohorts and have more stringent chart review beyond the International Classification of Diseases, ninth revision, codes. This may account for the difference in our result compared with that of previous studies.We found that diabetic patients with perioperative HgA1C values above 6.5% have increased odds of wound healing complications. This was slightly below the cutoff point noted by Werner et al,
16
who concluded that a perioperative HgA1C value between 7% and 8% could serve as a threshold/cutoff point for increased risk of SSI. However, their study relied on an insurance database that estimated a lower frequency of wound healing complications following CTR/TFR compared with our data. Additionally, their study did not include noninfectious wound healing complications. An HgA1C value of 6.5% is below the level of 7.0%, which most diabetes and endocrine societies have historically used to classify their patients as well controlled, although these recommendations have recently changed. The 2019 recommendations of the American Diabetes Association call for more rigorous glycemic targets, specifically an HgA1C value less than 6.5%, in patients with “little comorbidity and long life expectancy… if they can achieve it safely without hypoglycemia or significant therapeutic burden.”17
We do not enforce a strict HgA1C cutoff of 6.5% in all patients, but we do counsel diabetic patients on potential additional surgical risks. This study did not assess the effect that insulin dependence may have had on our results, which was previously demonstrated to be an independent risk factor for postoperative complication, compared with nondiabetic patients undergoing more complex upper-extremity surgery.8
Additionally, we chose not to differentiate patients based on type 1 or type 2 DM. Instead, we chose to focus on perioperative HgA1C values because we believed that this provides an objective quantification of the quality of an individual patient’s disease management, and it is a modifiable risk factor in preoperative diabetic patients.According to London et al,
18
there appears to be a dose-dependent effect of BMI among obese patients such that increasing-obesity heightens the risk of complications following hand, wrist, and elbow surgery. With only a 5% (range: 0%–9%) increase in the odds of a complication for each 1.0-point increase in the BMI of diabetic patients, there is minimal impact in our study. There may have been a confounding effect between increasing BMI and HgA1C value that we were unable to assess with the current study design.Tobacco smoking and its associated complications were not a primary outcome of our study, and we did not demonstrate a significant effect in either group; however, in another study, smoking has been found to be an independent risk factor for postoperative complications.
19
With such a low prevalence of smokers, our study was not adequately powered to assess the impact of smoking on wound healing complications, and we included this information to demonstrate that it did not have a large influence in our study group.This study has limitations that should be recognized. Although age, sex, BMI, and smoking status were assessed in diabetic and nondiabetic patients, direct patient-to-patient matching between the groups was not performed, which prevented the nondiabetic cohort from acting as a true control group. Limiting the chart review to 2 weeks after the surgery may have resulted in the underestimation of our infection rate if there were late-presenting infections; however, we believed that majority of the acute postoperative wound healing problems would have been recognized at the time of initial follow up. This is a retrospective cohort study performed at a single institution, which may have limited its external validity. Some possible confounding variables not evaluated include nutritional status, renal insufficiency, peripheral neuropathy, insulin dependence, and type 1 versus type 2 DM, many of which have previously been shown to be associated with increased postoperative complications following hand and wrist operations.
8
,20
Given the relatively low number of occurrences of wound healing complications in our study, we did not feel that our cohort was an appropriate size to address all of these possible comorbid conditions as separate independent variables. Additionally, patients who underwent multiple procedures during the study’s duration were included as separate encounters, which could have served to amplify any underlying confounding effects. We performed a generalized linear mixed-model analysis to control for this; however, this is imperfect, and the effect that these confounding variables had on the data presented here likely persists to some degree. Lastly, this study only assessed wound healing and did not comment on the short- or long-term patient-reported outcomes of CTR or TFR.Diabetes mellitus is associated with a significantly increased risk of wound healing problems after soft-tissue surgery of the hand (CTR and TFR). Furthermore, an elevated perioperative HgA1C value can help identify patients at a higher risk of surgical complications. Surgical candidates with DM should be appropriately counseled prior to surgery.
References
- Prevalence of and trends in diabetes among adults in the United States, 1988-2012.JAMA. 2015; 314: 1021-1029
Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2020. Accessed August 28, 2020. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf
- The impact of glycemic control and diabetes mellitus on perioperative outcomes after total joint arthroplasty.J Bone Joint Surg. 2009; 91: 1621-1629
- Diabetes mellitus, hyperglycemia, hemoglobin A1C and the risk of prosthetic joint infections in total hip and knee arthroplasty.J Arthroplasty. 2015; 30: 439-443
- Is hemoglobin A1C or perioperative hyperglycemia predictive of periprosthetic joint infection or death following primary total joint arthroplasty?.J Arthroplasty. 2015; 30: 1197-1202
- Factors associated with infection following anterior cruciate ligament reconstruction.J Bone Joint Surg. 2015; 97: 450-454
- Trends in postoperative infection rates and their relationship to glycosylated hemoglobin levels in diabetic patients undergoing foot and ankle surgery.J Foot Ankle Surg. 2014; 53: 307-311
- Insulin dependence is associated with increased risk of complications after upper extremity surgery in diabetic patients.J Hand Surg. 2018; 43: 745-754
- Rate of infection after carpal tunnel release surgery and the effect of antibiotic prophylaxis.J Hand Surg. 2010; 35: 189-196
- Outcome of surgical release among diabetics with carpal tunnel syndrome.Arch Phys Med Rehabil. 2004; 85: 7-13
- A comparison of outcomes of carpal tunnel release in diabetic and non-diabetic patients.J Hand Surg. 2013; 38: 485-488
- Clinical outcomes of surgical release among diabetic patients with carpal tunnel syndrome: prospective follow-up with matched controls.J Hand Surg. 2009; 34: 1177-1187
- Quantifying the effect of diabetes on surgical hand and forearm infections.J Hand Surg. 2018; 43: 105-114
- Surgical site infection in hand surgery.Int Orthop. 2015; 39: 2191-2198
- Surgical site infection event.(Accessed March 8, 2020.)
- The association of perioperative glycemic control with postoperative surgical site infection following open carpal tunnel release in patients with diabetes.Hand. 2019; 14: 324-328
- 6. Glycemic targets: standards of medical care in diabetes-2019.Diabetes Care. 2019; 42: S61-S70
- The impact of obesity on complications of elbow, forearm, and hand surgeries.J Hand Surg. 2014; 39: 1578-1584
- The impact of smoking on early postoperative complications in hand surgery.J Hand Surg. 2021; 46: 336.e1-336.e11
- The effect of malnutrition on postoperative complications following surgery for distal radius fractures.J Hand Surg. 2019; 44: 742-750
Article info
Publication history
Published online: July 02, 2021
Accepted:
May 12,
2021
Received:
October 7,
2019
Footnotes
J. N. Lawton serves as a consultant for Innomed (Savannah, GA) unrelated to the subject of this article. No benefits in any form have been received or will be received by the other authors related directly or indirectly to the subject of this article.
Identification
Copyright
© 2021 by the American Society for Surgery of the Hand. All rights reserved.
