Does Total Intravenous Anesthesia With Short-acting Spinal... : JAAOS - Journal of the American Academy of Orthopaedic Surgeons

• Updated on: 2018-05-29
• Read original article here

Results: Both total hip arthroplasty and total knee arthroplasty patients in the Updated protocol group had markedly reduced LOS compared with those in the Traditional protocol group (1.5 ± 0.1 days versus 2.4 ± 0.1 days; P < 0.05 and 1.4 ± 0.1 days versus 2.3 ± 0.1 days; P < 0.05). A higher proportion of patients in the Updated protocol group received at least 1 POD#0 PT session compared with those in the Traditional protocol group.

Dr. Incavo has received royalties from Innomed, Kyocera, OsteoRemedies, Smith & Nephew, Wright Medical Technology, and Zimmer; serves as a paid consultant to Zimmer; holds stock or stock options in Nimbic Systems and Medical/Orthopaedic publications editorial/governing board (Journal of Arthroplasty); and is a board member or committee member of the Knee Society. None of the following authors or any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Klavas, Dr. Karim, Dr. Lambert, Dr. Ferris, and Ms. Delgado.

Primary total hip arthroplasty (THA) and primary total knee arthroplasty (TKA) are two of the most common orthopaedic procedures currently performed in the United States, due in part to their effectiveness and patient satisfaction. By 2030, the demand for primary THA and primary TKA is projected to grow by 174% and 673%, respectively. Because of this anticipated growth, reducing hospital length of stay (LOS) to limit medical costs remains a focal point for most high-volume joint arthroplasty centers. Between 2002 and 2013, the mean hospital LOS decreased by approximately 1 day for TKA and 1.25 days for THA, whereas simultaneously the mean hospital cost increased by $7,849 for TKA and $7,858 for THA. This trend has prompted this past decade's considerable research efforts placed on defining rapid-recovery and fast-track clinical pathways to reduce hospital LOS. More emphasis is now being placed on the clinical variables associated with hospital LOS, such as mode of anesthesia, American Society of Anesthesiologists (ASA) score, surgical time, blood loss, postoperative analgesia, postoperative physical therapy (PT) protocol, and patient demographics. Previously, we have reported success with day-of-surgery mobilization postoperative day 0 (POD#0) at our institution.

Correlation between the mode of anesthesia and hospital LOS after primary total joint arthroplasty (TJA) is currently a topic of debate. The existing standard mode of anesthesia in joint replacement surgery comprises some combination of the following: general inhaled anesthesia and/or neuraxial anesthesia, regional nerve block, and local infiltration anesthesia administered intraoperatively for both TKA and THA. Memtsoudis et al. analyzed the anesthesia records of >380,000 patients who underwent primary TJA between 2006 and 2010. They found that 11% of procedures were performed under neuraxial anesthesia only, 14% under combined neuraxial-general anesthesia, and 75% under general anesthesia only. In an effort to build on these findings, we recently implemented a novel anesthesia protocol for primary THA and TKA that incorporates total intravenous anesthesia (TIVA) for general sedation and short-acting spinal blockade for neuraxial anesthesia.

The advent of intravenous delivery for anesthetic agents dates back to the year 1656 when the aristocrat Robert Boyle injected opium into the vein of a dog, noting a “brief period of anesthesia followed by a full recovery.” Large-scale use of intravenous anesthetics was implemented in the realm of modern anesthesia in the early 1990s when short-acting opioids such as remifentanil were first synthesized. These drugs possess a pharmacokinetic profile that allows them to be easily titrated to high levels of analgesia without causing prolonged respiratory depression. Total intravenous anesthesia, a form of general anesthesia that alleviates the need for long-acting sedating inhalational anesthetics, still requires ventilation with inhaled O. However, TIVA diverges from classic general anesthesia techniques by using a potent short-acting opioid such as remifentanil or fentanyl to provide analgesia, in combination with another drug, such as propofol, to provide hypnosis and amnesia. The major benefits of this drug combination are a completely attenuated response to noxious intraoperative stimuli and rapid postoperative emergence from sedation. Anesthesiologists who cover ambulatory surgery cases view TIVA as a clinically useful modality because they can safely and predictably awaken their patients from surgery without having to accommodate for prolonged side effects of anesthesia, such as postoperative nausea, vomiting, drowsiness, and fatigue.

Total intravenous anesthesia techniques are generally viewed as safe. The need for gas delivery systems and free radical scavenger equipment is alleviated. Furthermore, TIVA drugs are less toxic than inhalational drugs and pose a lower risk of the development of malignant hyperthermia. When we compared the safety of TIVA with that of volatile inhaled anesthetics, we found that previous studies have yielded no differences. Total intravenous anesthesia maintained comparable hemodynamic parameters with fewer episodes of bradycardia in patients who underwent coronary artery bypass graft procedures. In addition, postcardiac surgery measurements of the liver and renal function were found to be similar in patients who received either TIVA or inhaled anesthetics.

The goal of implementing an updated protocol is to achieve safe, rapid recovery while maintaining excellent pain control. Improving postoperative pain control facilitates earlier mobilization and accelerated PT, two factors that have proved to reduce the mean LOS after primary TJA. The purpose of this study was to (1) investigate how hospital LOS after primary TJA was affected when patients received a novel mode of anesthesia consisting of TIVA combined with a short-acting spinal anesthetic and (2) determine whether the updated anesthesia protocol facilitated quicker postoperative mobilization while maintaining patient safety both during their hospital stay and after discharge. We hypothesized that this updated protocol would result in improved postoperative mobilization with a clinically significant reduction in hospital LOS and no significant increase in 30-day complication rates.

After institutional review board approval was obtained, we retrospectively reviewed the use of the updated anesthesia protocol identified as the “Updated protocol group” for primary THA and TKA patients compared with the control group, consisting of primary THA and TKA patients with a standard anesthesia protocol identified as the “Traditional protocol group.” The senior orthopaedic surgeon performed the procedures on these patients between January 2014 and December 2014. Data collected for the Updated protocol group were obtained through a review of the charts of all primary THA and TKA surgeries performed by the same surgeon over the course of a 6-month period from June 2016 to December 2016 immediately after the initiation of the updated anesthesia protocol. All surgeries took place at a single tertiary care center with operating rooms (ORs) and ancillary staff dedicated to orthopaedic surgical procedures. The inclusion criterion was any adult patient who underwent primary elective unilateral TJA of the hip or knee during the study period. The exclusion criterion was any patient who underwent revision joint arthroplasty, bilateral joint arthroplasty during the same hospital stay, unicompartmental joint arthroplasty, or THA performed for femoral neck fractures. After inclusion and exclusion criteria were applied, a total of 235 consecutive patients (116 THA and 119 TKA), who underwent primary elective joint arthroplasty during the Traditional protocol study period, and a total of 118 consecutive patients (59 THA and 59 TKA), who underwent primary elective joint arthroplasty during the updated protocol study period, were included. All patients who met the study criteria were included for analysis, despite age or comorbidities.

Medical hospital charts were reviewed for patient demographics, intraoperative data, postoperative course, progress notes, nursing notes, and physical therapist notes. Patient records were also reviewed for complications and readmissions for a period up to 30 days after discharge. Follow-up in the office for wound assessment, radiographs, and range of motion assessment was documented with a progress note at 2 to 3 weeks postoperatively. All patients who met the inclusion criteria were followed up.

Hospital LOS was measured from the time of admission in the preoperative holding area on the day of surgery to the time of discharge from the hospital, as documented by the charge nurse in the acute care orthopaedic unit. Physical therapy sessions for patient ambulation were conducted on the day of surgery (POD#0) and each subsequent day until discharge. These sessions remained consistent across both study groups in THA and TKA patients alike, with up to three daily sessions directed at early mobilization, range of motion exercises, and strength-building exercises. Physical therapy was performed after patient consent was obtained.

For the Traditional protocol group, a combination of general anesthesia and long-acting neuraxial anesthesia was used for both THA and TKA patients. General anesthesia consisted of inspired sevoflurane gas with O via endotracheal tube airway. Neuraxial anesthesia consisted of 0.2 mg of morphine administered via one-time spinal injection into the thecal sac. Peripheral nerve blocks for TKA patients were not performed. Foley catheters were placed preoperatively after intubation for all patients in the Traditional protocol group to avoid urinary retention with long-acting spinal blockade. All Foley catheters were removed on POD#1, unless medically indicated for prolonged use.

For the Updated protocol group, a combination of TIVA and short-acting spinal anesthesia was used. Total intravenous anesthesia consisted of a propofol intravenous bolus using weight-based dosing, followed by continuous intravenous propofol infusion, with a combination of fentanyl, remifentanil, or hydromorphone for narcotic analgesia. In all cases, the amounts of these drugs were titrated at the discretion of the anesthesiologist to maintain an adequate depth of anesthesia. Ventilation was maintained with inhaled O via either laryngeal mask airway or endotracheal intubation. Neuraxial anesthesia consisted of 0.75% bupivacaine administered via one-time injection into the thecal sac by the attending anesthesiologist. Foley catheters were not used in the Updated protocol group, unless indicated for a medical reason. Unless contraindicated, all TKA recipients also received an ultrasonography-guided saphenous nerve block with 0.2% ropivacaine, 100 μg of clonidine, and 5 mg of dexamethasone at the adductor canal (Table 1). There were no contraindications for saphenous nerve blockade. However, the success of every nerve blockade was judged preoperatively by both delivering anesthesiologist and senior orthopaedic surgeon.

In addition to the respective anesthesia protocols just described, 10 mg/kg of intravenous tranexamic acid was administered in all cases to help limit blood loss, and 10 mg of intravenous dexamethasone was administered intraoperatively to minimize postoperative nausea. All THA procedures in the study were performed with conventional instruments and uncemented components through a minimally invasive posterior approach, direct anterior approach, or anterolateral approach. All TKA procedures were performed with a standard medial parapatellar approach, conventional instruments, and a tourniquet. The patella was resurfaced, and cement fixation was used in all patients. Drains were not used for any THA or TKA cases. Before closure in all cases, a locally injected anesthetic cocktail was used to assist with postoperative analgesia (Table 1).

Postoperative day 0 PT was conducted for all patients in the study. For the Traditional protocol group, one PT/ambulatory session was offered on the day of surgery. For the Updated protocol group, day-of-surgery ambulation was offered for three sessions to emphasize rapid mobilization. Patients did not undergo formal PT before surgery. The PT protocol for both groups consisted of subjective assessment of patient condition, lower limb–strengthening exercises, and maximum gait training with the use of an assistive walking device as needed. Posterior hip movement precautions were taken for any patient who underwent posterolateral approach THA to protect against postoperative dislocation. Continuous passive motion was not used in any TKA patient in this study.

Patients were cleared for discharge by a multidisciplinary team after they met the following criteria: absence of comorbid medical conditions necessitating inpatient management, intact surgical incisions with no evidence of surrounding erythema without discharge, intact lower extremity neurovascular examination, adequate pain control with oral analgesics, and all PT goals achieved (independence with bed mobility, independent transfers, standing balance, and a minimum gait distance of 150 ft). Patients who were unable to be discharged home were transferred to skilled nursing facilities or home health care. Follow-up appointments were arranged for 2 to 3 weeks postoperatively.

Of the 235 patients who met the inclusion criteria and underwent primary elective joint arthroplasty by traditional anesthesia methods, 7 patients (4 THA and 3 TKA) were excluded from analysis in the Traditional protocol group because of complications that prolonged hospital LOS. Of the 118 patients who met the inclusion criteria and underwent primary elective joint arthroplasty by updated anesthesia methods, 2 patients (1 THA and 1 TKA) were excluded from analysis in the Updated protocol group because of complications that prolonged hospital LOS (Table 2).

Within each protocol group, data were analyzed using both combined and separate observations from THA and TKA procedures. Age, body mass index (BMI), LOS, surgery time, time spent in the postanesthesia care unit (PACU), estimated blood loss (EBL), and the number of PT sessions conducted were compared between groups (Traditional versus Updated protocol) using an independent two-tailed t-test. Chi-square analysis was used to compare the percentage of patients within each group who received PT on the day of surgery and the ASA Physical Status Classification (1 to 4). Type I error was set at α = 0.05 for all analyses. For all statistically significant differences (P < 0.05), the effect size was calculated using either a Cohen D statistic (t-test analysis) or a phi statistic (chi-square analysis). To quantify the size of any observed differences between the two protocol groups, Cohen D and phi statistic were interpreted as follows: <0.1, negligible effect size; 0.1 to 0.3, small effect size; 0.3 to 0.5, moderate effect size; 0.5 to 0.7, large effect size; and >0.7, very large effect size. All data were analyzed using SPSS software for Windows (version 20; IBM Statistics).

For all primary TJA recipients included in the analysis of this study, there were no significant differences between the two groups with regard to age, BMI, and ASA status. However, surgery time (79.8 ± 1.4 minutes versus 69.4 ± 1.4 minutes; P < 0.01), EBL (175.4 ± 10.5 mL versus 112.4 ± 10.4 mL; P < 0.01), and length of time in the PACU (133.5 ± 4.8 minutes versus 115.8 ± 4.2 minutes; P < 0.05) were each significantly higher for the Traditional protocol group than those for the Updated protocol group (Supplemental Table 3, Supplemental Digital Content 1, http://links.lww.com/JAAOS/A90). The Cohen D statistic was 0.58 (large effect) for surgery time, 0.46 (moderate effect) for EBL, and 0.30 for time spent in the PACU (moderate effect). A similar proportion of patients attended at least four or more PT sessions before discharge in both groups: 116/228 (50.8%) in the Traditional group and 60/116 (51.7%) in the Updated group. For the total number of PT sessions attended before discharge, only the subset of THA patients showed any difference (P < 0.05): mean 3.6 ± 0.1 sessions before discharge in the Traditional protocol group versus 4.3 ± 0.2 sessions before discharge in the Updated protocol group. The effect size for this difference was 0.44 (moderate). Overall, there were a significantly greater proportion of patients in the Updated protocol group who received POD#0 PT (87.1%) compared with the Traditional protocol group (66.7%) (P < 0.05). This difference had an effect size of 0.22 (small) (Supplemental Table 3, Supplemental Digital Content 1, http://links.lww.com/JAAOS/A90).

With regard to the hospital LOS of all patients who underwent either THA or TKA in this study, the Updated protocol group had a mean LOS of 1.4 ± 0.1 days. This was markedly lower than in the Traditional protocol group, which had a mean LOS of 2.3 ± 0.1 days (P < 0.01). The subset of THA patients in the Updated protocol group had a mean hospital LOS of 1.5 ± 0.1 days compared with 2.4 ± 0.1 days for THA patients in the Traditional protocol group (P < 0.01). Total knee arthroplasty patients in the Updated protocol group had a mean hospital LOS of 1.4 ± 0.1 days compared with 2.3 ± 0.1 days for TKA patients in the Traditional protocol group (P < 0.01) (Supplemental Table 3, Supplemental Digital Content 1, http://links.lww.com/JAAOS/A90). The effect size as measured by the Cohen D statistic was considered very large when comparing the differences in the LOS of all primary joint recipients (0.90), THA alone (0.86), and TKA alone (0.98).

The percentage of patients remaining in the hospital after POD#1 was 91.2% for the Traditional protocol group and 31.9% for the Updated protocol group. By POD#2, 32.2% of patients in the Traditional protocol group remained hospitalized compared with 8.6% of patients in the Updated protocol group. By POD#4, 2.6% of patients in the Traditional protocol group remained hospitalized, whereas all patients from the Updated protocol group were discharged (Figure 1).

There were no hospital readmissions for postoperative complications within 30 days in the Traditional protocol group. The Updated protocol group had 1 TKA patient who, 2 weeks after discharge, presented with a fall at home with wound dehiscence and was treated with irrigation and débridement in the OR and 6 weeks of intravenous antibiotics. The Traditional protocol group had two patients who sustained a mechanical fall in the hospital before discharge, whereas the Updated protocol group had no patients who sustained a mechanical fall before discharge. Of note, 92.5% of patients in the Traditional protocol group were discharged to home compared with 95.8% of patients in the Updated protocol group. The remaining patients were discharged to a supervised setting.

Factors that enable expeditious discharge from the hospital after primary elective total joint surgery include adequate postoperative analgesia, short-lived side effects of anesthetics, patient motivation, and clinical pathways that promote early and aggressive rehabilitation by physical therapists. However, because hospital LOS has gradually shortened in recent years, the concern surrounding patient safety and increased risk of complications has heightened. A major risk associated with aggressive PT is hip dislocation after THA. Fortunately, no hip dislocations were reported in our cohort of patients. Other minor postoperative risks include prolonged swelling, bleeding from the surgical site, urinary tract infections, and surgical site infections. Major postoperative risks are pulmonary embolism, myocardial infarction, and deep surgical wound infection necessitating return to the OR.

Minor complications that occur after discharge are not always reported by the patient; therefore, their true incidence in our study population after discharge is unknown. Nevertheless, we are able to report that there were two major complications necessitating return to the OR and/or a hospital readmission within 30 days: 1 patient in the Traditional protocol group who was taken back to the OR on the same day of surgery for evacuation of a hematoma (Table 2) and 1 patient readmitted to the hospital for wound dehiscence after a fall, necessitating irrigation and débridement and intravenous antibiotics. This equates to a major complication rate of 0.56% and a readmission rate of 0.29%. Data from the American College of Surgeon's National Surgical Quality Improvement Program on elective primary TJA patients show that the 30-day readmission rate after elective primary TJA is 4.6% for TKA and 4.2% for THA, which is much higher than the rate reported in the current study. One possible reason for this discord is that the current study only screened for readmission to the same institution and may have missed patients who were readmitted elsewhere. Similarly, data from the American College of Surgeon's National Surgical Quality Improvement Program show that hospital discharge within 2 days from surgery does not increase the risk of major complications and hospital readmission. Patient demographics (BMI, ASA score, and medical comorbidities) and perioperative variables (surgery time and blood loss) were attributed to an increased risk of major complications.

Three perioperative variables measured in this study (surgery time, estimated blood loss [EBL], and length of time spent in the PACU) were found to be significantly lower in the Updated protocol group compared with the Traditional protocol group. The authors think that shorter surgery times may be due to bias in favor of the Updated protocol group, whose patients underwent surgery over a period of time 2 years later than the Traditional protocol group. Such an advantage in chronicity could potentially grant the Updated protocol group with the benefit of more intraoperative experience and familiarity. The observed difference in EBL can be explained by institutional changes in the way blood loss is reported in the electronic medical record after surgery. For the Traditional protocol, the anesthesiologist alone estimates blood loss, whereas for the Updated protocol, the anesthesiologist and attending surgeon estimate blood loss after discussing with one another. Nevertheless, increased surgical time and increased blood loss have been previously described to positively correlate with prolonged hospital LOS. Our analysis did not control for perioperative variables when analyzing hospital LOS between the two study groups.

The results of our study show that TIVA used in conjunction with short-acting nonopioid spinal anesthetics facilitated earlier postoperative mobilization in a safe manner and led to markedly reduced LOS in the hospital, validating the study hypothesis. The incidence of complications and hospital readmission in the cohort of patients in the current study who received TIVA was much lower than what is reported in the literature, attesting to the safety of this perioperative protocol. Furthermore, because the Updated protocol replaced a long-acting spinal anesthetic with a short-acting spinal anesthetic, we were able to discontinue the routine use of Foley catheters.

Correlation between aggressive postoperative rehabilitation (including day-of-surgery ambulation) and decreased hospital LOS is an evolving concept in the subspecialty of joint replacement surgery. This retrospective review was conducted on patients who underwent uncomplicated primary joint replacement after a modified early ambulation program was implemented on the acute care orthopaedic unit at our institution. As part of this enhanced recovery program, PT protocols were altered such that three daily walking sessions were offered. Both groups in the current study had the benefit of this aggressive mobilization regimen to their advantage, confirmed by an insignificant difference in the total number of PT sessions received before discharge (3.7 ± 0.1 versus 4.0 ± 0.2; P > 0.05) (Supplemental Table 3, Supplemental Digital Content 1, http://links.lww.com/JAAOS/A90). On a similar note, neither group trended toward a higher likelihood of being discharged to a skilled nursing facility or long-term acute care facility, further supporting the advantages provided by rapid mobilization.

Based on the results of this study, the authors propose that TIVA with a short-acting nonopioid spinal anesthetic and no routine Foley catheter use markedly increases a patient's likelihood of ambulating on the day of surgery in an enhanced recovery program. The current study advances the discussion surrounding the efficacy of TIVA and short-acting spinal anesthesia on clinical outcomes in primary joint arthroplasty, which has limited reports in the literature. Two randomized controlled trials conducted by Harsten et al at a Swedish fast-track joint arthroplasty center looked at recovery times after TJA in patients receiving solely TIVA versus solely spinal blockade. In patients who received only TIVA, they reported a median time to meet discharge criteria of 26 hours in their THA cohort and a median time to meet discharge criteria of 46 hours in their TKA cohort. The current study reports a mean hospital LOS of 40.9 hours in THA patients and 38.2 hours in TKA patients. Our methods differ in that we did not specify “meeting discharge criteria” as the end point when we measured hospital stay. Instead, we used the time of physical discharge from the hospital, thus potentially increasing the mean LOS compared with other studies.

Our finding of a reduced hospital LOS by approximately 22 hours (0.9 day) holds clinical significance in addition to statistical significance. Although it would have been inappropriate and may have introduced study bias to preemptively declare a cutoff for reduction in LOS as clinically significant, the authors retrospectively propose that any reduction in LOS >12 hours should be interpreted as clinically beneficial. For example, a difference in this magnitude could reflect one of two scenarios: (1) an early morning discharge on POD#2 as opposed to evening or (2) an evening discharge on POD#2 as opposed to an additional night in the hospital before discharging the morning of POD#3. In either instance, the reduction in LOS portends to reduced utilization of hospital resources, decreased overall costs, and increased patient satisfaction.

The study is limited by its retrospective nature, relatively small sample size, and asymmetry between study group sizes. A retrospective study design made it difficult to collect data on postoperative nausea, vomiting, and visual analog scale pain scores because these values were inconsistently recorded in the patients' medical chart; however, it is our observation that postoperative nausea and vomiting were greatly reduced in patients who did not receive inhaled anesthetics. The statistical analysis did not control for certain perioperative factors that may have influenced study results. Furthermore, the analysis did not take into account how the surgical approach affected LOS, how the day of surgery affected LOS, or how the time of surgery start may affect LOS. Moving forward, now that our institution has transitioned to the Updated protocol for 100% of elective primary TJA procedures, reimplementing a Traditional protocol to perform a randomized controlled trial comparing the two groups would not be in the best interest of our patients. The data presented show clear improvement in the postoperative course with direct patient benefits at no additional risk to patient safety. The strengths of this study include its adequate statistical power, measure of effect size, lack of difference in age, sex, BMI, or ASA status, and reproducible methods for potential use in future investigations.

These results provide useful information for any individual surgeon or group of surgeons who regularly perform primary elective TJA and wish to decrease their patients' postoperative LOS. With decreased LOS comes improved cost of care. Because greater importance is placed on value-based care within bundled-care settings, improvements such as these will be necessary. Our results indicate that TIVA and short-acting nonopioid spinal anesthetics are a promising route toward achieving decreased hospital LOS after TJA.