Controversy continues about the optimal timing of surgery for patients with bilateral symptomatic degenerative arthritis of the knee. Bilateral total knee replacement may be performed with the patient under a single anesthetic session either simultaneously with two surgical teams or sequentially with one team; it may be staggered and managed as separate procedures during a single hospitalization; or it may be staged in two hospitalizations, separated by a variable period of time1-7. For the patient, the physiologic insult from undergoing simultaneous or sequential procedures is greater than that from separate unilateral procedures8-10.
The purpose of this study was to compare the early complications associated with staggered bilateral total knee replacement and those associated with sequential procedures and staged procedures performed by the same surgeons. We hypothesized that, in a community setting, fewer complications would occur following the staggered procedures. We also investigated the differences in reimbursement to the hospital and surgeon among the three timing schedules.
Materials and Methods
Using a computerized patient database and medical records, we retrospectively identified 332 consecutive patients who underwent primary bilateral total knee arthroplasty (664 knees) from October 1997 through June 2001. All procedures were performed by one of two surgeons (D.D.G. and S.G.T.). The data were collected in a prospective fashion; the groups were assigned retrospectively. There were three groups of patients. The staggered group underwent bilateral total knee arthroplasty as separate procedures performed within four to seven days of each other during a single hospitalization. The sequential group underwent bilateral total knee arthroplasty performed by a single surgical team during a single anesthetic session. The staged group underwent two unilateral total knee arthroplasties performed during separate hospitalizations. Two hundred and forty-one patients, the majority of patients in this study, underwent staggered procedures spaced four to seven days apart (average, 4.5 days). Sixty-five patients underwent staged procedures spaced an average of 70.5 weeks apart (range, 1.6 to 270.9 weeks). Twenty-six patients underwent sequential bilateral replacement during a single anesthetic session.
The demographic data for the three groups are shown in Table I. A total of 319 patients (96%) had a diagnosis of primary osteoarthritis. Preoperative medical comorbidities were identified in 278 of the 332 patients. Hypertension was the most common medical comorbidity in all three groups and was present in 50% of all patients. Other common comorbidities included hypercholesterolemia (20%), obesity (15%), thyroid disorder (14%), diabetes mellitus (12%), and coronary artery disease (11%). There were significantly fewer medical comorbidities in the sequential group than in the staggered and staged groups (p = 0.012) (see Appendix). In the analysis of specific comorbidities, a significant difference was found with respect to the distribution of patients with hypertension. Patients with staged procedures had the highest rate of hypertension (63.1%); in comparison, the sequential group had a rate of 30.8% (p = 0.014).
During the study period, preoperative autologous blood donation was used by patients who requested it. Patients with substantial cardiac, pulmonary, or other serious comorbidities generally underwent staged procedures after receiving medical clearance from an internist.
Staggered or sequential replacements were generally performed on healthier patients. If the consulting internist concluded that the patient was not healthy enough to undergo staggered or sequential procedures, the patient received a unilateral procedure on the more symptomatic side and then underwent replacement of the contralateral side at a later date. Patients who were deemed to be healthy enough to undergo a bilateral procedure were then evaluated by an anesthesiologist. If the anesthesiologist concurred that the patient was a candidate for a bilateral procedure, he or she was offered a staggered procedure. If the patient requested a sequential procedure instead, the final decision was made in consultation with the surgeon and was based upon further evaluation of the potential medical risks (Fig. 1).
If medically stable after the first arthroplasty, patients in the sequential group then underwent the second arthroplasty. Patients in the staggered group who had an uncomplicated first knee replacement had the second knee replacement four to seven days later. If there was any problem intraoperatively or postoperatively, then the second arthroplasty was performed electively at a later hospitalization.
All patients underwent arthroplasty in a laminar airflow room and received perioperative antibiotics beginning thirty minutes prior to incision and continuing for forty-eight hours. The surgical team wore body-exhaust suits. All patients were admitted to an inpatient floor postoperatively. All patients received postoperative anticoagulation treatment with low-dose Coumadin (warfarin), low-molecular-weight heparin, or aspirin for four weeks. Patients were examined throughout the hospitalization for clinical evidence of deep-vein thrombosis, and additional workup was obtained if clinically indicated. Routine venous Doppler examination at the time of discharge was not obtained.
We recorded perioperative complications that occurred within the initial sixty postoperative days, blood transfusion requirements, length of acute inpatient hospitalization, and the need for placement in a rehabilitation hospital. A reimbursement analysis that included surgeon, anesthesia, and hospital charges was performed.
Major complications were defined as death, myocardial infarction, pulmonary embolism, cerebrovascular accident, deep infection, or a return to the operating room for any reason. Minor complications included urinary retention, urinary tract infection, deep-vein thrombosis, pneumonia, superficial infection, early knee manipulation for poor motion, atrial fibrillation, and admission to the hospital without monitoring in the intensive care unit.
The Fisher exact test was used for comparisons between nominal data, such as gender, diagnosis, comorbid medical conditions, inpatient rehabilitation stay requirements, and complications. A Wilcoxon rank-sum test was performed for comparisons of nonparametric ordinal data. A least squares means test was used to compare the mean ages among the three groups. The Student t test was used to compare age at the time of surgery and complications. Odds ratios were determined to compare minor complications among the groups. The Kruskal-Wallis test was used to compare reimbursement among the three groups of patients. When the test for overall group difference was significant, post hoc pairwise multiple comparisons based on the Kruskal-Wallis rank sums were performed.
Asignificant difference was found in the gender distribution among the three groups, with the sequential group having more male patients (54%) and the staged group having more female patients (78%) (p = 0.006). A significant difference was detected in the mean ages for the staged group (67.2 years) and the sequential group (59.3 years) (p = 0.0002). Comparisons of age between the sequential and staggered groups and between the staged and staggered groups showed no significant difference (p = 0.07 and p = 0.4, respectively). No significant difference was found among the three groups with respect to the distribution of the diagnosis (p = 0.16).
The mean length of acute inpatient hospitalization was five days (range, four to nineteen days) in the sequential group, nine days (range, seven to twenty days) in the staggered group, and nine days (range, four to twenty-one days), when both hospitalizations were summed, in the staged group. The sequential group had a significantly shorter inpatient hospitalization than the other two groups (p = 0.0001).
An allogenic or autologous blood transfusion was required in 165 patients (50%). A total of 140 (58%) of the 241 patients in the staggered group and fourteen (54%) of the twenty-six patients in the sequential group required a transfusion. In contrast, only eleven (17%) of the sixty-five patients in the group that had staged procedures required a transfusion at either operation (p = 0.0001). The median number of units of blood transfused was two (range, one to six units) in the sequential group, one (range, one to six units) in the staggered group, and one (range, one to two units) in the staged group.
A greater proportion of patients in the sequential group (31%; eight of twenty-six) and the staggered group (33.6%; eighty-one of 241) required skilled rehabilitation services compared with that in the staged group (17%; eleven of sixty-five) (p = 0.029).
The overall rate of complications was 13% (thirty-two patients) in the group managed with staggered knee replacements, 35% (nine patients) in the sequential group, and 31% (twenty patients) in the staged group. The staggered group had approximately 2.5 times fewer complications than the other groups; the difference was significant (p = 0.0009). Major complications were rare and occurred only in the staggered group and the staged group (Table II). There were no major complications in the sequential group. The only death was a patient in the staggered group who had an early postoperative fatal cardiac arrhythmia. With the numbers available, no significant difference was detected with respect to major complication rates between the sequential group and the staggered group (p = 1.0), the sequential group and the staged group (p = 1.0), or the staged group and the staggered group (p = 0.085).
Minor complications occurred in 12% of the patients in the staggered group, 35% of those in the sequential group, and 25% of those in the staged group (see Appendix). The odds of a patient undergoing sequential knee replacements having a minor complication was 3.92 times (95% confidence interval, 1.40, 10.41) greater than the odds of a minor complication occurring in a patient in the staggered group (p = 0.009). The odds of a patient in the staged group having a minor complication was 2.64 times (95% confidence interval, 1.23, 5.56) greater than the odds of a minor complication occurring in a patient in the staggered group (p = 0.012). No significant difference was detected with respect to the odds of a minor complication occurring in the sequential group compared with that in the staged group (p = 0.59). The readmission rate for treatment of a deep-vein thrombosis without pulmonary embolism was 1% for the staggered group, 3% for the staged group, and 0% for the sequential group.
The age at the time of surgery did not correlate with the prevalence of major or minor complications. The mean age (and standard deviation) was 66.1 ± 9.6 years for the patients with complications and 65.9 ± 9.0 years for those without complications (p = 0.885).
At least one medical comorbidity was identified in fifty-four (89%) of sixty-one patients with either a major or minor complication and in 224 (82.7%) of 271 patients without a complication (see Appendix). Seven of the eight patients who had a major complication and 271 (83.6%) of 324 patients who did not have a major complication were found to have a comorbid condition (see Appendix).
Complete reimbursement data were available for 253 patients (76%) who had the surgical procedures performed after December 1997. These included twenty-one patients (81%) in the sequential arthroplasty group, 186 (77%) in the staggered arthroplasty group, and forty-six (71%) in the staged arthroplasty group. All reimbursement data were converted to 2001 dollars with use of the Consumer Price Index conversion factors made available by the United States Bureau of Labor Statistics. Hospital reimbursement differed significantly among the groups. A pairwise comparison of the groups showed a significantly greater mean hospital reimbursement for the staged group ($18,958.39) compared with the sequential group ($14,291.85) (p < 0.0001) and the staggered group ($15,506.42) (p = 0.010). No significant difference in reimbursement was detected between the sequential and staggered groups (p = 1.0) (see Appendix).
Although no significant difference was found with respect to the mean surgeon reimbursement among the groups, the data demonstrated a trend toward higher surgeon reimbursement for staged (two-admission) procedures compared with bilateral procedures in a single hospitalization (p = 0.073) (see Appendix).
The timing sequences for bilateral total knee replacement have been well described in the literature1-5,11-14. Sequential bilateral knee arthroplasty under a single anesthetic session has been associated with an increased risk of mortality compared with unilateral total knee arthroplasty8. Blood loss is greater in bilateral replacement15,16. Fat and marrow embolism syndrome has also been reported in sequential bilateral total knee replacement9,10,17. Although Ritter et al. reported low complication rates for patients undergoing sequential total knee replacement13, their patients were treated in a highly specialized center for total joint replacement with resources that may be unavailable to many orthopaedic surgeons performing knee replacement surgery in a community setting. Our study examined the three timing sequences of total knee arthroplasty when performed in a community hospital.
One of the most important concerns regarding the timing of the surgical procedures is whether the method of replacement increases the mortality rate. In our study, one death occurred in the staggered group for a prevalence of 0.41%. This mortality rate is comparable with the rate of 0.49% reported by Parvizi et al.8 in a study of 2679 patients who had bilateral total knee arthroplasty.
Many authors have shown that blood loss and the transfusion rate are greater in bilateral knee replacement than in unilateral replacement1,2,15-17. In our patients who were managed with bilateral knee replacement during the same hospitalization, either staggered or sequential, we noted that the transfusion rate was approximately three times greater than that for the patients who underwent staged procedures.
Numerous studies have noted that sequential procedures often reduce the number of inpatient hospital days after the procedure1,3,4,12,17. The hospital stay for our patients with sequential arthroplasties was four days shorter than that for either the staggered group or the staged group.
Some reports have noted that the need for postoperative rehabilitation increased after bilateral total knee arthroplasty compared with that after unilateral total knee replacement2, whereas other studies have demonstrated that there was less need1. Our patients who had bilateral knee replacement during a single hospitalization, either staggered or sequential, required nearly twice the rate of inpatient postoperative rehabilitation compared with those undergoing staged arthroplasty. This increased need for inpatient rehabilitation after a bilateral replacement during a single hospitalization in some respects mitigates the reported cost advantages of sequential replacement.
The data for our patients demonstrated a significant difference in hospital reimbursement for the sequential and staggered groups compared with the staged group. Greater hospital reimbursement was associated with two hospital admissions, while the number of inpatient hospital days had a minimal impact on reimbursement, which was probably related to the diagnosis-related group reimbursement rates.
A weakness of this study was selection bias in retrospectively constructing the three groups. The sequential group was healthier, with significantly fewer medical comorbidities than either the staggered group or the staged group. In addition, a sequential procedure was performed only at the request of the patient and was always at the surgeon's discretion; hence, the number of patients in this group was small. Nevertheless, on the basis of medical criteria alone, it would have been expected that the sequential group would have fewer overall complications. However, this group had the highest overall complication rate of all the groups, confounding the bias introduced into this study by the patient selection process. Nevertheless, the sequential group did not have any major complications, and, while no significant difference was found with respect to the rate of major complications among the groups, our inability to detect a difference may have been the result of the small number of patients who received sequential knee replacements.
Minor complications occurred most frequently in the sequential and the staged groups. This may have been due to the greater surgical insult to the patients undergoing simultaneous total knee arthroplasty and to the overall worse general health of the group of patients who had staged procedures.
Other studies have attempted to identify risk factors that preclude or stratify risk in certain patients who are to undergo sequential bilateral knee replacement8. In our study, medical comorbid conditions were present in the vast majority of patients. However, the presence of a medical comorbidity (or the number of comorbidities) did not predict whether a particular patient would have a perioperative complication. Likewise, no particular comorbidity was found to be associated with a minor or major complication. Increased age was not a factor in predicting complications despite a large number of patients (132) who were more than seventy years old. Of these patients, 120 underwent either staggered or staged procedures, indicating that age was an important determinant in excluding a patient as a candidate for a bilateral arthroplasty during a single anesthetic session.
In conclusion, patients who had bilateral knee replacement with the procedures performed in close succession, either sequentially or staggered, required more blood transfusions and were more likely to require an additional inpatient rehabilitation. The sequential group had the shortest mean hospital stay but the greatest overall rate of complications. Minor complications occurred less frequently when patients had staggered bilateral knee replacement, but with the numbers available, no difference was detected among the groups with regard to major complications. No risk factors that could be used to predict the occurrence of complications were identified. Bias that was introduced by the patient selection process reinforces our ultimate conclusion that staggered bilateral total knee replacement is a safe and effective method of bilateral knee replacement and is associated with a low overall complication rate.
Tables presenting medical comorbidities, complications, and reimbursement comparisons for the three study groups are available with the electronic versions of this article, on our web site at jbjs.org (go to the article citation and click on “Supplementary Material”) and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM). ▪
In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from DePuy. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. A commercial entity (DePuy) paid or directed, or agreed to pay or direct, benefits to a research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
Investigation performed at the Department of Orthopaedics, University of Iowa Hospitals and Clinics, Iowa City, and Des Moines Methodist Hospital, Des Moines, Iowa
1. , Sculco TP, Ranawat CS, Behr C, Tarrentino S. One-stage versus 2-stage bilateral total knee arthroplasty. Clin Orthop. 1994;309: 94-101.
2. , Hozack WJ, Shah S, Rothman RH, Booth RE Jr, Eng K, Smith P. Simultaneous bilateral versus unilateral total knee arthroplasty. Outcomes analysis. Clin Orthop. 1997;345: 106-12.
3. , Fisher RL. Bilateral total knee arthroplasties. Comparison of simultaneous (two-team), sequential, and staged knee replacements. Clin Orthop. 1985;199: 220-5.
4. , Mamlin LA, Melfi CA, Katz BP, Freund DA, Arthur DS. Outcome implications for the timing of bilateral total knee arthroplasties. Clin Orthop. 1997;345: 99-105.
5. , Jessup DE, Clelland C. Simultaneous bilateral total knee replacement versus unilateral replacement. Am J Orthop. 1996;25: 292-5.
6. , Singri P, Khanna S, Gradisar IA. Medical and financial aspects of same-day bilateral total knee arthroplasties. Biomed Sci Instrum. 1997;33: 429-34.
7. , Meyers SJ, Cox DD, Elliott M, Watson M, Shim SD. Cost comparison between bilateral simultaneous, staged, and unilateral total joint arthroplasty. J Arthroplasty. 1998;13: 172-9.
8. , Sullivan TA, Trousdale RT, Lewallen DG. Thirty-day mortality after total knee arthroplasty. J Bone Joint Surg Am. 2001;83: 1157-61.
9. , Merkel C, Mellman MF, Klein I. Fat emboli in bilateral total knee arthroplasty. Predictive factors for neurologic manifestations. Clin Orthop. 1989; 248: 112-9.
10. , Trousdale RT, Ilstrup DM. Complications after concomitant bilateral total knee arthroplasty in elderly patients. Mayo Clinic Proc. 1997;72: 799-805.
11. , Adams RA, Ilstrup DM, Bryan RS. Complications and mortality associated with bilateral or unilateral total knee arthroplasty. J Bone Joint Surg Am. 1987;69: 484-8.
12. , Chen SH. Simultaneous bilateral total knee arthroplasty in a single procedure. Int Orthop. 1998;22: 390-3.
13. , Harty LD, Davis KE, Meding JB, Berend M. Simultaneous bilateral, staged bilateral, and unilateral total knee arthroplasty. A survival analysis. J Bone Joint Surg Am. 2003;85: 1532-7.
14. , Binazzi R, Insall JN, Nordstrom TJ, Pellicci PM, Goulet JA. Successive bilateral total knee replacement. J Bone Joint Surg Am. 1985;67: 573-6.
15. , Freeman BJ, Pullyblank A, Newman JH. Blood loss in sequential bilateral total knee arthroplasty. J Arthroplasty. 1998;13: 77-9.
16. , Meding JB, Faris PM, Ritter MA. Predictors of transfusion risk in elective knee surgery. Clin Orthop. 1998;357: 50-9.
17. , Forrest CJ, Benjamin JB. Safety and efficacy of bilateral total knee arthroplasty. J Arthroplasty. 1997;12: 497-502.