Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery

The Journal of Arthroplasty xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org

Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery Simon W. Young, FRACS a, John Mutu-Grigg, FRACS a, Christopher M. Frampton, PhD b, John Cullen, FRACS a a b

Department of Orthopaedics, North Shore Hospital, Auckland, New Zealand University of Otago, Christchurch, New Zealand

a r t i c l e

i n f o

Article history: Received 26 December 2013 Accepted 6 March 2014 Available online xxxx Keywords: total knee arthroplasty duration of surgery TKA

a b s t r a c t Longer operative times may be required in complex total knee arthroplasty (TKA), however little is known about outcomes in procedures performed rapidly. We analysed 58,009 primary TKAs from the New Zealand National Joint Registry. The mean surgical duration was 89 minutes, and 50% of procedures lasted between 60 and 89 minutes. There was no difference in adjusted revision rates for groups lasting between 40 and 120 minutes, however procedures lasting N 120 minutes had significantly higher revision rates. There was a higher revision rate in TKAs lasting b40 minutes (0.71 vs 0.48 revisions per 100 component years) but this was not statistically significant (P = 0.1). For primary TKAs lasting less than 120 minutes, further shortening operative time did not improve outcome, and very rapid procedures (b 40 minutes) may lead to an increased risk of revision. © 2014 Elsevier Inc. All rights reserved.

The speed at which an operation is performed has long been equated with a surgeon’s skill. Robert Liston, a Scottish surgeon in the 1830s, was famed for his ability to perform an above knee amputation in under 2.5 minutes [1]. Florence Nightingale’s Notes on Nursing states “the operator’s success will be in direct ratio to his quickness” [2]. While the advent of anaesthesia removed the main driver toward extremely rapid operations, economic factors mean surgeons remain under pressure to reduce operative time and maximise throughput, particularly in procedures such as in total knee arthroplasty (TKA) [3]. While in general surgeons aim to reduce operative time by improving efficiency [4,5], such pressures could potentially lead to compromises in surgical technique in order to maximize speed. Longer operative times in TKA are required in complex cases; therefore poorer functional outcomes or higher revision rates may be expected. However little is known about TKA outcomes in procedures performed very rapidly (b 40 minutes). A shorter operative time may reduce the risk of infection and anaesthetic risk, leading to improved outcomes. The aim of this study was to analyse functional outcome and revision rates by surgical duration in primary TKA.

Source of Funding: No funds were received by the authors of this study from any external source. The New Zealand Joint Registry if funded by the New Zealand Orthopaedic Association, individual orthopaedic surgeons, the Accident Compensation Corporation, the Ministry of Health, and the Southern Cross Hospitals Trust. Supplementary material available at www.arthroplastyjournal.org. The Conflict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2014.03.004. Reprint requests: Simon W. Young, FRACS, North Shore Hospital, Private Bag 93-503 Takapuna, Auckland City, 0740, New Zealand.

Methods A total of 61,438 primary TKA procedures performed between 1999 and 2012 from the New Zealand National Joint Registry were analyzed. Procedures were divided a-priori by surgical duration (skin to skin time) into those lasting b 40 minutes, 40–59, 60–89, 90–119, 120–179, and N180 minutes. National ethics committee permission was obtained and all patients signed a written consent to be included in the registry database. Implant survival rate was expressed as the number of revisions per 100 component-years. This is equivalent to the yearly rate of revision expressed as a percentage, and is calculated by dividing the number of prostheses revised by the observed component-years multiplied by 100. Functional outcomes were obtained from the New Zealand Joint Registry data. The Registry distributes the Oxford 12 questionnaire to a randomly selected sample of 28% of patients, which, with a 75% return rate, ensured a score for approximately 20% of primary knee arthroplasties. The Oxford Knee Score is a patient-generated outcome measure which consists of 12 questions on pain and function, and generates a score ranging from 0 points (worst) to 48 points (the best score) [6]. Questionnaires were sent at 6 months, 5 years, and 10 years post operatively. Statistical Analysis The effects of cementation, computer aided navigation, surgical approach and patella resurfacing on surgical duration were analysed using one-way ANOVA, with results summarised as means and 95% confidence intervals. The Oxford scores at 6 months, 5 years, and 10 years post operatively were compared between surgical duration groups using one-way ANOVA. Revision rates were compared

http://dx.doi.org/10.1016/j.arth.2014.03.004 0883-5403/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004

S.W. Young et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx

between surgical duration groups using Cox proportional hazards regression models. These models were run as adjusted analyses including as covariates cementation, age and the use of computer aided navigation and as unadjusted models without these covariates. The proportionality assumption for the cox regression models was tested for the comparisons among surgical duration groups by testing the interaction between loge(time) and surgical duration groups.

Mean TKA Duration by Year of Operation 90

Minutes

2

85 80 75 70

Results The mean surgical duration for primary TKA was 89 minutes (range 24–421 minutes). Of 61,438 primary TKAs, 58,009 (94.4%) were performed for a diagnosis of primary osteoarthritis. The mean duration of TKA performed for osteoarthritis was 83.0 minutes, with 77% of procedures lasting between 60 and 119 minutes and 1.1% of procedures lasting b 40 minutes (Table 1). The 50 primary TKAs performed for a tumour diagnosis had the longest duration, with 84% of procedures lasting ≥ 120 minutes. There was a slight trend towards decreasing duration during the study period with a mean duration in 1999 of 85.4 minutes (95% CI 84.3–86.5) compared to a mean duration in 2012 of 82.3 minutes (95% CI 81.7–83.0, Fig. 1). Among TKAs performed for primary osteoarthritis, 4.3% were implanted without cement. The mean duration was 13 minutes shorter than when cemented implants were used (70.9 versus 83.9 minutes, P b 0.001, Table 2). TKAs performed with computer navigation were on average 8 minutes longer in duration than when conventional instrumentation was used (mean 90.2 minutes versus 82.3 minutes, P b 0.001). Similarly mean durations for procedures performed through a lateral (91.3 minutes) or ‘other’ approach (88.7 minutes, includes minimally invasive surgery) were longer in duration than a standard median parapatellar approach (82.6 minutes, P b 0.001 all comparisons). TKA procedures lasting 40–59, 60–89, or 90–119 minutes all had significantly (P b 0.01) lower revision rates than operations lasting N 120 minutes, after adjustment for diagnosis, age, navigation, and cementation (Tables 3a and 3b). Procedures lasting b40 minutes also had higher revision rates in comparison to procedures lasting 40–59 minutes (hazard ratio HR = 0.6, P = 0.03, Fig. 2), however after adjustment for the above factors the difference was no longer statistically significant (HR = 0.7, P = 0.1). At five years the mean oxford score in procedures lasting b 40 minutes was 38.6 (95% CI 36.6–40.6) and N180 minutes was 36.8 (95% CI 33.6–40.1), slightly lower than the most common TKA

1999

2001

2003

2005

2007

2009

2011

Operation Year Fig. 1. Mean procedure duration by operative year. Error bars represent 95% confidence intervals.

duration of 60–89 minutes (mean 40.2 points, 95% CI 39.9–40.5, Fig. 3). However, there was no statistically significant difference in functional scores at 6 months, 5, or 10 years between any of the procedure duration groups. Discussion This study found an operative time of N 120 minutes in TKA was an independent risk factor for higher revision rates after adjustment for age, diagnosis, computer navigation, and use of cement. Additionally, for primary TKAs lasting less than 120 minutes we found no evidence that further shortening operative time led to improved patient outcomes, and indeed there was a non statistically significant trend towards higher revision rates in procedures performed very rapidly (b40 minutes). There are a number of limitations to our study. Firstly, while registry data include a number of variables that influence operative duration, we could not control for other potential confounding factors. Body mass index (BMI) in particular is not recorded on the New Zealand registry. Elevated BMI has been associated with both increased operative duration and higher revision rates, although it is unclear which of these factors is causative [7–9]. However given the multiple variables that influence surgical duration, investigation of its effect on outcome is only possible with large, registry based studies such as this one. Secondly while the New Zealand joint registry provides data on both functional outcome and revision rates, it does not include data on other complications influenced by surgical duration such as cardiovascular, thromboembolic, or wound

Table 1 Patient Characteristics and Procedure Duration. Procedure Duration (Minutes)

Mean age Mean ASA score Diagnosis (n) Osteoarthritis Rheumatoid arthritis Other inflammatory Tumour Postfracture Postligament disruption Avascular necrosis

20–39

40–59

60–89

90–119

120–179

N180

n = 672

n = 8044

n = 30515

n = 16786

n = 5128

n = 293

68.5 2.1

68.7 2.1

68.2 2.1

67.6 2.2

66.5 2.2

61.8 2.2

1.1% 643 0.8% 13 0.9% 5 0% 0 0.5% 3 0.3% 1 2.4% 5

13% 7718 9.5% 154 9.2% 51 2.0% 1 7.8% 51 7.4% 29 16.1% 34

8% 4649 12% 194 14% 76 32% 16 20% 128 17% 67 10.4% 22

0.4% 220 0.8% 13 0% 0 52% 26 4.4% 29 1.5% 6 0.0% 0

50% 29084 45% 729 41% 228 6.0% 3 37% 245 44% 173 41.7% 88

27% 15695 32% 512 35% 192 8.0% 4 31% 201 29% 114 29.4% 62

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004

S.W. Young et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx Table 2 Operative Factors and Mean Procedure Duration in TKA Performed for Osteoarthritis Only. % of Mean Procedure Procedures Duration

n Cement Cemented

52057 89.7%

83.9

Uncemented

2501

4.3%

70.9

Hybrid

3445

5.9%

79.3

Navigation Conventional 52789 91.0%

82.3

Computer aided Approach Medial

5214

9.0%

90.2

53046 96.0%

82.6

Lateral

846

1.5%

91.3

Other

1362

2.5%

88.7

17694 30.5%

85.7

Unresurfaced 40312 69.5%

81.8

Patella Resurfaced

Surgeon experience Basic trainee 1274 Advanced trainee Consultant

2.2%

86.3

5858 10.1%

95.8

50635 87.7%

81.4

95% CI

P Value

83.6– 84.1 69.9– 71.8 78.5– 80.1

P b 0.001 all comparisons

82.1– 82.5 89.5– 90.9

P b 0.001

82.3– 82.7 89.6– 92.9 87.4– 90.0

P b 0.001 all comparisons

85.4– 86.1 81.6– 82.1

P b 0.001

84.9– 87.6 95.2– 96.4 81.2– 81.6

P b 0.001 all comparisons

Table 3a Revision Rates by Procedure Duration.

b40 40–59 60–89 90–119 120–179 ≥180

n 672 8044 30515 16786 5128 293

Revision Rate Per 100 Component-Years (95% CI)

Adjusted P Value⁎

0.71 0.45 0.48 0.50 0.65 1.13

0.92 0.001 b0.001 0.006 -

(0.46–1.1) (0.39–0.52) (0.44–0.51) (0.45–0.55) (0.56–0.75) (0.66–1.8)

Table 3b Revision Rates by Procedure Duration for OA Only. Procedure Duration (Minutes) b40 40–59 60–89 90–119 120–179 ≥180

n 643 7718 29081 15692 4649 220

Revision Rate Per 100 Component-Years (95% CI)

Adjusted P Value⁎

0.66 0.44 0.48 0.49 0.63 0.77

0.88 0.001 0.002 0.019 -

(0.41–1.0) (0.38–0.52) (0.44–0.52) (0.45–0.55) (0.53–0.73) (0.36–1.5)

⁎ Adjusted for age, cementation, image guided and surgeon experience, for comparison to procedures lasting N120 minutes.

complications. Finally, while the large number of patients in this study increases our ability to identify associations, proving causation is more difficult. Clinical examination findings and radiographs were not available, and may be able to identify factors leading to higher revision rates such as component malposition or ligamentous imbalance. Our findings support those of two previous large studies on surgical duration in arthroplasty, which also found both prolonged and very short operative times to be associated with higher revision rates. Smabrekke analysed operative time in 31,745 primary total hip arthroplasties reported to the Norwegian registry [10]. The mean operating time was 96 minutes, and operative times N 150 minutes were associated with a two-fold increase in revision rates. Additionally, cemented THAs performed in less than 51 minutes had a significantly greater risk of aseptic loosening (RR 1.6, P = 0.03). Ong reviewed Medicare claims data from 1997 to 2004 and found a median TKA duration (measured by anaesthetic billing time) of 148 minutes. TKA procedures lasting more than 240 minutes were significantly associated with higher revision rates when compared to

Procedure Duration (Minutes)

3

⁎ Adjusted for age, cementation, image guided and surgeon experience, for comparison to procedures lasting N120 minutes.

those lasting 120–50 minutes (adjusted odds ratio OR = 1.3, P = 0.01). TKA procedures shorter than 90 minutes also had significantly higher revision rates compared to those lasting 120–150 minutes (adjusted OR = 1.47; P = 0.008). An association between prolonged operative time and poorer outcomes may be explained by a number of factors. The duration of the procedure may be an indicator of surgical complexity or an inexperienced or slow working surgical team [10,11]. Obesity is also associated with prolonged operative times [7–9] and higher complication rates such as deep infection [7,12–14]. Prolonged operative times have been associated with TKA complications such as mortality, infection [13,15], venous thromboembolism, and postoperative neurologic dysfunction. In a retrospective study of 1001 primary and revision TKA patients, Horlocker found significantly higher rates of peroneal or tibial nerve palsy with longer surgical duration and tourniquet time [16]. In a small study of 60 patients undergoing revision hip arthroplasty, Kessler calculated that each additional minute of operating time led to a 3% increase in perioperative complications including periprosthetic fracture, dislocation, thromboembolism, bleeding, and wound damage [17]. The effect of extremely short operative duration is less clear. Our lower cutoff of 40 minutes is arbitrary, and only 1.1% of TKA procedures performed for a diagnosis of osteoarthritis were within this category. It is conceivable that after a certain point, operative duration can be shortened further only by omitting key steps or compromising surgical technique, thus risking a poorer patient outcome. However the lowest revision rate in this study was seen in TKA procedures lasting between 40 and 59 minutes, suggesting optimising surgical efficiency remains a desirable goal. The issue is pertinent as limited healthcare resources and increasing demand mean surgeons remain under pressure to maximise throughput. When Centers for Medicare and Medicaid Services began a transition from a charge-based system to a resource-based relative value scale for Medicare reimbursements, they anticipated that physicians would increase surgical volume in response [11]. Our data suggest that surgeons attempting to reduce operative time in order to complete more cases should do so with caution, and ensure patient outcome is not compromised for economic reasons. In conclusion, we report an association between an operative time of N120 minutes and higher revision rates in TKA surgery. For primary TKAs lasting less than 120 minutes we found no evidence that further shortening operative time led to improved patient outcomes. Additionally, we report a non-statistically significant trend toward an increased risk of revision in very rapid procedures (b 40 minutes) that warrants consideration in future studies.

References 1. Ellis H. Operations that Made History. Cambridge, UK: Cambridge University Press; 1996. 2. Nightingale F. Notes on Nursing: What it Is, and what it is Not. New York: D. Appleton & Company; 1860. 3. Booth RE. Minimizing operating time: does speed kill? Orthopedics 2001;24(9):853.

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004

4

S.W. Young et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx

Revision Rates Primary TKA 1.75

Revisions per 100 component years

1.50

A

1.25 1.00 0.75 0.50 0.25 0.00 <40

40-59

60-89

90-119

120-179

>180

Operation time (Mins)

Revision Rates Primary TKA for OA only

Revisions per 100 component years

1.75 1.50 1.25 1.00 0.75 0.50 0.25

B

0.00 <40

40-59

60-89

90-119

120-179

>180

Operation time (Mins) Fig. 2. Revisions per 100 component-years by procedure duration. Error bars represent 95% confidence intervals.

4. Stiehl JB, Jackson S, Szabo A. Multi-factorial analysis of time efficiency in total knee arthroplasty. Comput Aided Surg 2009;14(1–3):58. http://dx.doi.org/10.3109/ 10929080903030996. 5. Smith MP, Sandberg WS, Foss J, et al. High-throughput operating room system for joint arthroplasties durably outperforms routine processes. Anesthesiology 2008;109(1):25. http://dx.doi.org/10.1097/ALN.0b013e31817881c7. 6. Murray DW, Fitzpatrick R, Rogers K, et al. The use of the Oxford hip and knee scores. J Bone Joint Surg Br 2007;89(8):1010. http://dx.doi.org/10.1302/0301620X.89B8.19424. 7. Namba RS, Paxton L, Fithian DC, et al. Obesity and perioperative morbidity in total hip and total knee arthroplasty patients. J Arthroplasty 2005;20(7 Suppl 3):46. http://dx.doi.org/10.1016/j.arth.2005.04.023. 8. Silber JH, Rosenbaum PR, Zhang X, et al. Influence of patient and hospital characteristics on anesthesia time in medicare patients undergoing general and orthopedic surgery. Anesthesiology 2007;106(2):356.

9. Raphael IJ, Parmar M, Mehrganpour N, et al. Obesity and operative time in primary total joint arthroplasty. J Knee Surg 2013;26(2):95. http://dx.doi.org/10.1055/s-00331333663. 10. Småbrekke A, Espehaug B, Havelin LI, et al. Operating time and survival of primary total hip replacements: an analysis of 31,745 primary cemented and uncemented total hip replacements from local hospitals reported to the Norwegian Arthroplasty Register 1987–2001. Acta Orthop Scand 2004;75(5):524. http://dx.doi.org/ 10.1080/00016470410001376. 11. Ong KL, Lau E, Manley M, et al. Effect of procedure duration on total hip arthroplasty and total knee arthroplasty survivorship in the United States Medicare population. J Arthroplasty 2008;23(6 Suppl 1):127. http://dx.doi.org/10.1016/j.arth.2008.04.022. 12. Jibodh SR, Gurkan I, Wenz JF. In-hospital outcome and resource use in hip arthroplasty: influence of body mass. Orthopedics 2004;27(6):594. 13. Peersman G, Laskin R, Davis J, et al. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res 2001;392:15.

5 Year Functional Outcomes 42 41

Oxford Score

40 39 38 37 36 35 <40

40-59

60-89

90-119

120-179

>180

Operation time (Mins) Fig. 3. Mean Oxford Score at 5 year duration grouped by procedure length. Error bars represent 95% confidence intervals.

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004

S.W. Young et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx 14. Namba RS, Inacio MCS, Paxton EW. Risk factors associated with deep surgical site infections after primary total knee arthroplasty: an analysis of 56,216 knees. J Bone Joint Surg 2013;95(9):775. http://dx.doi.org/10.2106/JBJS.L.00211. 15. Willis-Owen CA, Konyves A, Martin DK. Factors affecting the incidence of infection in hip and knee replacement: an analysis of 5277 cases. J Bone Joint Surg Br 2010;92 (8):1128. http://dx.doi.org/10.1302/0301-620X.92B8.24333.

5

16. Horlocker TT, Hebl JR, Gali B, et al. Anesthetic, patient, and surgical risk factors for neurologic complications after prolonged total tourniquet time during total knee arthroplasty. Anesth Analg 2006;102(3):950. http://dx.doi.org/10.1213/01. ane.0000194875.05587.7e. 17. Kessler S, Kinkel S, Käfer W, et al. Influence of operation duration on perioperative morbidity in revision total hip arthroplasty. Acta Orthop Belg 2003;69(4):328.

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004

5.e1

S.W. Young et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx

Appendix Reasons for Revision by Initial Surgery Duration (Minutes)

Reason for revision (n) Loosening femoral component Loosening tibial component Pain Deep infection Patella complications

20–39

40–59

60–89

90–119

N120

Overall

n = 25 8.0%

n = 205 15.6%

n = 850 22.9%

n = 491 18.7%

n = 216 22.0%

n = 1787 10.2%

2 0.0%

32 22.9%

80 18.7%

45 22.0%

23 23.2%

182 20.4%

0 44.0 % 11 16.0% 4 32.0% 8

47 26.8% 55 18.1% 37 16.6% 34

159 27.5% 234 23.4% 199 20.9% 178

108 25.9% 127 21.8% 107 21.2% 194

50 24.1% 52 24.1% 52 18.6% 39

364 26.8% 479 22.3% 399 20.3% 363

Please cite this article as: Young SW, et al, Does Speed Matter? Revision Rates and Functional Outcomes in TKA in Relation to Duration of Surgery, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.004