1 Introduction

For patients undergoing unilateral total knee arthroplasty (TKA) for the treatment of osteoarthritis, 37.2% undergo contralateral TKA within 10 years1. The additional procedure, commonly referred to as a staged bilateral TKA, has previously been shown to increase the risk of perioperative complications due to multiple hospitalizations and prolonged cumulative anesthesia exposure2. With as many as 19% of patients having severe, bilateral osteoarthritis at the time of initial presentation3, a simultaneous bilateral total knee arthroplasty (simBTKA) is an attractive option and addresses concerns associated with staged procedures. However, previous studies evaluating simBTKA have reported increased rates of mortality and vascular complications, such as deep vein thrombosis (DVT) and pulmonary embolism, compared to unilateral or staged bilateral TKA3–5. Although these complications are often attributed to the prolonged, single anesthesia exposure, other contributing factors, such as obesity, are not fully understood.

Despite the national pandemic of obesity, only a few studies have examined post-operative complications following simBTKA specifically in obese patients. In a two-team simBTKA approach, perioperative complications for the morbidly obese were found to be comparable to unilateral TKA patients6. However, these results are not generalizable, as the surgical and anesthesia times are not representative of single-surgeon simBTKA. In contrast, an evaluation of a national database found a higher incidence of thromboembolic events following simBTKA (2.48%) compared to unilateral TKA (1.4%), with obesity contributing significantly to the increased risk7. Although these previous studies provide insight to the added complication risk for simBTKA, in the absence of direct complication comparison between specific obesity classifications, the influence of obesity is still uncertain and the best practice for the severely and morbidly obese presenting with bilateral knee osteoarthritis is still unknown. Therefore, the purpose of this study was to compare perioperative complications between defined obesity classifications in patients having undergone simBTKA.

2 Materials and methods

This institutional review board approved, retrospective review included an unselected, consecutive cohort of 313 patients who were identified as having undergone simBTKA between October 2013 and November 2019. All patients met the standard clinical and radiographic indications for bilateral TKA and all procedures were performed at a multi-specialty community hospital by a single, experienced fellowship trained orthopedic surgeon. Standard surgical procedures were performed on each knee and were completed in succession with a standard protocol for completing the left knee first. Following wound closure and application of dressings, the right knee was prepped and draped, with new sterile sets of instruments used as though performing a completely new case. The transition time between the end of surgery on the left knee and starting surgery on the right knee was approximately 30 min.

Prior to surgery, all patients received appropriate antibiotics and 500 mg–1000mg of acetaminophen, unless contraindicated. All patients received an ultrasound guided adductor canal nerve block prior to induction performed by an experienced high volume anesthesiologist containing bupivacaine 0.5% (20 cc), epinephrine (100 mcg) and Clonidine (1 mcg/kg). All patients received a pericapsular injection during surgery consisting of bupivacaine 0.15% (1 cc/kg) and toradol (30 mg) with half the maximum dose per patient administered in each knee. Surgery was performed under general anesthesia and a tourniquet for all knees. All patients received 1 g of intravenous tranexamic acid prior to incision and before closure of the arthrotomy. The capsule was closed using a barbed suture and the subcutaneous layer was closed with interrupted braided suture. Final wound closure was completed using either staples or a zipper method (ZipLine® Medical, Silicon Valley, CA). Dressings consisted of gauze and six inch elastic bandage over the incision area only. Intermittent mechanical foot compression was used for all patients until discharge and DVT chemoprophylaxis included aspirin (325 mg) for six week unless contraindicated. Patients who required chronic anticoagulation for other reasons, such as presence of a cardiovascular stent or artificial valve, were allowed to restart preoperative anticoagulants on post-operative day one. Patients with a history of thromboembolic events were given a low molecular weight heparin, including 40 mg subcutaneously for 10 days following surgery or 10 mg of rivaroxaban for 21 days following surgery.

Immediate full, unrestricted weight bearing was allowed as tolerated. Discharge criteria included the ability to walk 50 feet with an assistive device, adequate pain and nausea control, hemodynamic stability, and, if stairs were present in the home environment, successful navigation of one flight of stairs. If patients could not prove safe, independent function, the patient was kept overnight and discharged when safe function was demonstrated. Transfer to an acute care rehabilitation facility or skilled nursing facility was arranged if the patient could not demonstrate progression toward safe, independent post-operative function within 1–2 days.

Demographics were collected for each patient at the time of surgery, including age, body mass index (BMI), and American Society of Anesthesiologists’ (ASA) classification, as determined by a core group of experienced anesthesiologists. Outcome measures collected included adverse events during or immediately following surgery, transfusions required, hospital length of stay and discharge disposition. Perioperative complications were defined as any wound or systemic complication arising within 6 months following surgery. Complications were recorded from emergency room visits, readmissions, or clinic follow-up visit.

Data were separated by BMI category (BMI<30: non-obese; BMI = 30–35: obese); BMI>35: severely obese). Descriptive statistics for all outcome variables, including mean, standard deviation and ranges, were determined for each group. Normality for all continuous variables was evaluated with the Kolmogorov-Smirnov test. For all continuous variables, an analyses of variance (parametric) or Kruskal-Wallis test (non-parametric) was performed to determine differences between BMI categories. For all significant main effects, Bonferroni post-hoc tests were performed to determine the location of the difference. Categorical variables were evaluated using a Chi-square tests. For transfusions required, univariate logistic regression analyses were performed for age, gender, BMI, ASA, and surgical times. Regression results were reported as odds ratios (OR) and 95% confidence intervals (CI). All statistical tests were performed using SPSS v25, with significance level of p < 0.05.

3 Results

Overall, 248 patients were included in data analysis, including 139 non-obese patients, 61 obese patients and 48 severely obese patients. 65 patients did not have a six month follow-up at the time of data collection and were therefore excluded from data analysis. Patient demographics and perioperative variables are presented in Table 1. There were no differences between BMI groups for gender (p = 0.543). A significant main effect for age was present between the BMI groups, with non-obese patients being significantly older than patients in the obese (p = 0.006) and severely obese (p < 0.001) categories. A significant main effect for surgical time was also present between the BMI groups, with non-obese (p < 0.001) and obese patients (p = 0.002) having significantly lower surgical times than severely obese. Similarly, a significant main effect for tourniquet time was present between BMI groups, with non-obese (p < 0.001) and obese patients (p = 0.027) having significantly lower tourniquet times than severely obese. As BMI increased, patients were more likely to have a higher ASA score (p = 0.044) but there were no differences in length of stay (p = 0.663), discharge disposition (p = 0.165) or length of follow-up (p = 0.979) between BMI groups.

Post-operative complications are presented in Table 2. Out of the 248 patients, 37 patients required transfusion but the incidence was not different between BMI groups (p = 0.288). There were no significant differences in post-operative complications within six months, with the severely obese group sustaining only one (1.0%) deep infection, one (1.0%) DVT and one (1.0%) manipulation. Two patients sustained a fracture; one tibial tuberosity avulsion fracture while walking six weeks post-operative and one sustained during a manipulation, with both being treated non-operatively. No difference was seen between BMI groups for fractures (p = 0.455). Three knees required revision at one month, with one failing due to aseptic loosening and two revised due to deep infections. Twenty-one knees underwent a manipulation between six and ten weeks following simBTKA, with no significant difference in the frequency between BMI groups. Univariate regression results are presented in Table 3. Age was the only predictor of requiring a transfusion (OR: 1.089, CI = 1.036–1.145, p = 0.001).

4 Discussion

Previous research has evaluated perioperative complications following simBTKA but the consensus for the safety of this procedure remains unclear for obese patients.^3,5,8 Specifically, the risk of complications has been reported to be higher following simBTKA compared to unilateral or staged procedures with BMI found to be a contributing factor.^3,5 The results of the current study do not support these previous findings, as there were no significant differences in early post-operative complications between obesity classifications. With only two severely obese patients (4.2%) sustaining a wound or systemic complication, the results of the current study suggest that BMI over 35 should not be an automatic exclusionary criteria for patients seeking simBTKA for treatment of bilateral osteoarthritis.

Obesity contributes to a host of comorbidities, with previous research reporting a positive correlation between BMI and ASA classification.⁶ The presence of additional comorbidities, measured by ASA of three or greater, is commonly a contraindication for simBTKA,^2,9 which limits the representation of the severely obese patients in research. Therefore, the results of the current study add to the limited breadth of available literature evaluating severely and morbidly obese patients following simBTKA. As expected, the current study is consistent with previous reports, finding 70.8% of severely obese patients having an ASA ≥3. However, this high percentage did not translate to an increased incidence of thromboembolic events, with only one DVT (1.0%) seen in the severely obese group, compared to two pulmonary embolisms (1.4%) and four DVTs (1.4%) in the non-obese group. These low incidences of thromboembolic events are not only similar to previous literature reporting thromboembolic events in 2.48% of simBTKA and 1.4% of unilateral TKA,⁷ but also suggest severely obese patients with ASA ≥3 should not be automatically excluded from considering simBTKA.

These results should be considered with caution, as previous research has reported surgical and tourniquet times to be positively correlated with BMI.⁶ Consistent with previous research,¹⁰ the added surgical difficulty for the severely obese group was evidenced in the current study by significantly increased surgical and tourniquet times with increasing BMI categories. Despite this increase, average tourniquet times reported for the severely obese group was only 41.34 min per knee, which was much lower than previously reported tourniquet times of 117.6 and 116.0 min.^6,11 Previous research indicates an association between increased surgical and tourniquet times with increased risk of complications following TKA,¹² therefore, the low tourniquet times in the current study may explain the low overall incidence of complications. Based on these results, surgeons should continue to individually determine if simBTKA can be safely and efficiently performed in their severely obese patients.

Consistent with previous literature,^13–15 the rate of transfusion was not significantly different between obesity groups. On the contrary, transfusion rate seemed to decrease as BMI increased, although this difference was not statistically significant (p = 0.288). Upon further analysis, age was the only significant contributor to the risk of transfusion, with the risk increasing with age. Therefore, the significantly younger age of the severely obese group, coupled with higher blood volumes and increased coagulability,¹⁶ may help explain the lack of difference in transfusion requirements between BMI groups.

Our study has several limitations. All procedures were performed by a single, fellowship-trained, high volume surgeon at a community medical center with nearly 10 years of fast track type arthroplasty service experience and highly experienced anesthesiologists. Surgical technique during this time period was consistent across all BMI groups, however, the results of this study may not be generalizable to lower volume settings with less experienced anesthesia services. Secondly, the results of this study may not be comparable to two team simultaneous bilateral TKA performances. Finally, the small sample size of the severely obese group may limited the ability to find statistical significance. Additionally, patients with a BMI>40, of which there were 13 in the current study, could not be separated from the BMI>35 group due to the small sample size. A greater sample size would be preferable, particularly in the higher BMI groups, to more accurately reflect the risk of (morbid) obesity on perioperative complication rates.

5 Conclusion

There were no differences in wound or systemic complications following simBTKA across obesity classifications in the current study. These results suggest that higher BMI and associated higher ASA classifications only resulted in significantly longer tourniquet and operative times. Advanced age, not obesity, was the only variable associated with increased risk for requiring transfusions. Based on these findings, patients with BMI >35 and ASA ≥3 should not automatically be excluded from consideration for simBTKA.