Abstract
Introduction
We designed this study to compare 30-day complications and length of hospital stay between patients undergoing total knee arthroplasty (TKA) with general anesthesia, to those undergoing TKA with spinal, epidural anesthesia, or Monitored Anesthesia Care (MAC, a combination of local anesthesia with sedation and analgesia provided by an anesthesiologist) with or without regional nerve blocks.
Methods
We identified patients ≥18 years undergoing TKA between the years of 2006 and 2017 from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP). We collected patient demographics, anesthesia type, 30-day complications, length of operation and hospital stay from the database. We used multivariable regression to compare complications and length of stay (LOS) between anesthesia types, while adjusting for relevant covariables
Results
We identified 265,325 TKA patients. Of these, 91 (0.03%) underwent epidural anesthesia with a nerve block, while 1,855 (0.70%) underwent epidural anesthesia with no block, 12,800 (4.82%) underwent MAC with a block, 25,643 (9.66%) underwent MAC with no block, 13,575 (5.12%) underwent spinal anesthesia with a block, 80,803 (30.45%) underwent spinal anesthesia with no block, and 130,558 (49.21%) underwent general anesthesia.
The rate of complications was not associated with the presence of a block, while the unadjusted LOS was significantly lower with the use of a block in patients treated with spinal anesthesia (2.54±2.07 vs. 2.84±2.25), epidural anesthesia (2.87±1.81 vs. 3.88± 4.67), and MAC (2.51±2.14 vs. 2.68±2.11), p<0.0001.
The unadjusted rate of major complications was significantly lower in patients who underwent spinal anesthesia (2.10%), and MAC (1.91%) compared to general anesthesia (2.31%), p<0.0001. Similarly, the unadjusted rate of minor complications was significantly lower for patients treated with spinal anesthesia(1.86%) and MAC (1.78%) compared to general anesthesia (2.11%), p<0.0001.
The unadjusted LOS was significantly longer in patients treated with epidural (3.83±4.58), compared to general (2.94±3.64) anesthesia, p<0.0001. In contrast, the unadjusted LOS was significantly lower for patients treated with spinal anesthesia (2.80±2.23), and MAC (2.62±2.12) compared to general anesthesia, p<0.0001.
Following covariable adjustment, spinal anesthesia and MAC were associated with a 0.93 (0.87–0.98), and 0.84 (0.78–0.91), odds of major complications compared to general anesthesia. Similarly, spinal anesthesia and MAC were associated with a 0.92 (0.87–0.98) and 0.89 (0.82–0.97) odds of minor complications compared to general anesthesia.
Following covariable adjustment, epidural anesthesia increased the LOS by 0.25 (0.27–0.28) days compared to general, while spinal anesthesia and MAC decreased the LOS by 0.04 (95%CI 0.05–0.04), and 0.10 (0.11–0.09) days, compared to general.
In patients treated with spinal anesthesia, epidural anesthesia, and MAC, the use of a block was independently associated with a decreased LOS by 0.10 (0.12–0.90), 0.24 (0.39–0.09), and 0.07 (0.08–0.05).
Conclusion
Patients who undergo TKA with spinal anesthetic, and MAC appear to have superior outcomes compared to those who undergo TKA with general anesthesia. In addition, the use of a regional nerve block appears to be independently associated with a shorter LOS in patients who undergo TKA with neuraxial (spinal and epidural) anesthetic, and MAC.
For figures, tables, or references, please contact authors directly.
