Controversy remains regarding the relative safety of simultaneous bilateral TKA performed with the patient under 1 anesthetic [3–12]. Unfortunately, published studies describe a variety of staged, sequential, and 2-team simultaneous approaches, which prevents valid comparison between studies. To complicate the topic further, many conflicting reports exist concerning the effect of obesity on the risk of short- and long-term complications in patients undergoing unilateral primary TKA [15–22]. We have attempted to carefully assess outcomes in obese patients undergoing simultaneous, 2-team bilateral TKA and compare them with a matched cohort of obese patients undergoing unilateral TKA.
We believe that we are the first study comparing these 2 cohorts undergoing 2-team, simultaneous bilateral TKA and unilateral TKA. The study by Benjamin et al comparing obese and nonobese patients undergoing unilateral and bilateral procedures is not equivalent to our study, as the bilateral procedures were 1-team, sequential procedures . However, similar to our study, they were able to show a nonsignificant difference in wound and systemic complications between unilateral and bilateral obese groups, despite the approximate doubling of surgical time needed for sequential procedures as compared to unilateral procedures.
Wound problems have been among the most frequently cited complications of TKA in the obese population. Wilson et al were among the first to correlate obesity and wound infections in TKA . Winiarsky et al found that TKA in obese patients, while commonly successful, is associated with increased rates of infection, wound complications, and medial collateral ligament avulsions . In our series, there was a low overall rate of wound complications and infections, and there was no increase in this type of morbidity in patients undergoing the simultaneous bilateral procedure.
The use of blood transfusion is not without risk and is of concern to both the surgeon and patient . A significant variation in reported blood transfusion rates for simultaneous and sequential bilateral total knee arthroplasty exists; rates reported have ranged from 17% to 91% [3, 6, 7, 25–28]. This is likely due to differences in the criteria of reporting blood transfusion rates and blood loss, as well as the varying approaches used to manage acute blood loss. The rates of blood transfusion in our study decreased over time in both groups as policy was changed so that patients were treated symptomatically rather than automatically receiving a transfusion if hemoglobin levels dropped below particular levels (approximately 8 g/dL). Rates of autogenic preoperative donation also decreased over time in our study population, which may have lowered our transfusion rate over time. However, rates of transfusion in bilateral TKA patients in published studies have shown a universal increase in blood loss and transfusion rates, as would be expected with twice the surgical insult [3, 6–8, 11, 25–28]. Lane et al observed that longer surgical duration in TKA is associated with higher crystalloid replacement, leading to a dilutional component of anemia . Our bilateral group did have a significantly increased (P < .01) crystalloid replacement of approximately 10% over our unilateral group, which may also have contributed to a greater need for transfusion in patients undergoing the bilateral procedure. Our bilateral cohort did have a significantly higher (P < .01) transfusion rate without a significantly larger increase in postoperative hemoglobin levels (P = 0.23); this may be due to the increased crystalloid replacement, unseen postoperative blood loss, or some other unknown factor.
Greater emphasis is being placed on the cost benefit of various surgical procedures. Despite the fact that TKA has been shown to be an effective and cost-beneficial procedure, much attention continues to be paid to cost-cutting procedures [29–31]. The data contained in our study may have important consequences in this regard. The length of stay in the bilateral group was significantly longer than the unilateral group (3.72 vs. 3.30 days, P < .01), but staging the procedure for a bilateral situation would roughly double the unilateral time, causing significant increases in hospital inpatient stay costs. Operative time, which was also significantly increased in the bilateral group, would similarly be increased if a staged procedure would take place, leading to increased operating room expenses. However, 3 other important health system cost variables would not show a significant decrease in simultaneous 2-team TKA: the percentage of patients requiring transfusion, mean number of transfused units of packed red blood cells, and percentage of patients going to an extended-care facility at discharge. On the other hand, doubling the percentage of patients requiring extended-care facility treatment at discharge in the unilateral group to simulate staged procedures is only a very rough estimate, as level of deconditioning or decreased function as a result of a recent contralateral TKA is not taken into account, and may actually increase the use of extended-care facilities. This rough estimate prevents any conclusion regarding this statement in our study population. Reuben et al retrospectively compared the cost of unilateral vs. 1-team sequential bilateral TKA and noted a 36% cost reduction in the sequential bilateral total knee group as compared to a staged procdure . Similarly, Brotherton et al determined that the overall hospital bill may be more than 50% greater when a staged TKA is performed rather than a sequential bilateral TKA .
Our study does have several inherent weaknesses. Its retrospective nature, as well as inability of randomization, could influence results by introducing bias. The small sample size of our groups could also introduce a type-II statistical error. However, because of the relatively low mortality of patients undergoing this procedure, an extremely larger and possibly impractical number of patients would have to be included to avoid such an error if evaluating mortality . Although all of the patients were defined as obese by virtue of a BMI >30, there is likely a stratification of risk as patients reach more morbid levels of obesity, such as the patient with a BMI of 61.4 in our series. Additionally, although this series provides valuable outcome data, a comparison to non-obese patients at the same institution may generate beneficial data, as well. Another potential weakness is the fact that investigation into deep venous thrombosis was done only if the physicians had clinical suspicion in the perioperative period or as on an outpatient basis. Nonclinical deep venous thrombosis may have been missed and therefore the potential of bias from missing these nonsymptomatic thromboses is introduced.