1 Introduction

Total knee arthroplasty (TKA) continues to be a common solution for patients with knee osteoarthritis looking to decrease pain and improve functionality. As the general population ages, the incidence has steadily risen over the past 10–15 years. This may be attributed to the growing aging population in the United States, the knowledge that the outcome of TKA procedures can improve quality of life, newer techniques and equipment used for TKA procedures, and the rapid increased incidence of revision TKA procedures.¹

With the continued increase in TKA procedures has come an evolution in anesthetic procedures and practices over time. Unlike major abdominal, cardiac, or thoracic surgeries that require general anesthesia with an endotracheal tube or laryngeal mask airways, TKAs are now often performed with neuraxial techniques such as spinal anesthetic with concurrent intravenous sedation. Aimed at improving pain control while decreasing the time of effect of the anesthetic agent, these new techniques have also been shown to decrease blood loss in the operative field, decrease postoperative complications, and shorten hospital length of stay.^2–6 In concordance with this shift in analgesic modalities, the field of regional anesthesia has grown in the past two decades to provide new nerve block techniques that are often utilized to aid in postoperative pain control after a TKA, including enhanced needle and ultrasound technology, catheters, pumps, and longer lasting anesthetic agents.⁷

In addition to the continuing opioid epidemic, the relatively older age of patients undergoing TKA makes the development and utilization of a successful multimodal non-narcotic pain regimen vital. The many opioid-induced side effects that can occur, including nausea, vomiting, oversedation, constipated, and respiratory depression, lead to worening post-operative outcomes and increased hospital length of stay.⁸ Length of hospitalization and readmissions continues to remain a major predictor of total cost of hospital care for total joint arthroplasty.⁹

The Triple Aim refers to a theory designed to improve the U.S. healthcare system through the simultaneous pursuit of three goals: improving the patient experience, improving the health of populations, and reducing cost. Therefore, cost control must be achieved in the context of high value, high quality care and patient satisfaction. This study looks to explore the effect of different types of anesthesia in addition to a multimodal pain regimen on post-operative outcomes after undergoing TKA. The authors hope these results will help guide orthopaedic and anesthetic practitioners looking to improve a patient's clinical course after TKA while decreasing both patient complications and hospital costs.

2 Methods

Approval for this retrospective study was obtained by our hospital's institutional review board (IRB 2020-124). From January 2016 to December 2022, the charts of patients undergoing an elective single TKA (ICD-10 diagnosis code M17.11 and M17.12, principal procedure ICD-10 code 0SRC0J9 and 0SRD0J9) at our institution were reviewed. All patients were consented to be included in the study. Twenty-seven orthopedic surgeons performed TKA procedures in our hospital network during this time frame. Patients undergoing primary, isolated, and unilateral TKA procedures with either general or neuraxial anesthesia were included. Patients were grouped into three study arms: 1) those that received general anesthesia and no regional nerve block (GA); 2) general anesthesia with a local anesthetic adductor canal block (GA + ACB); and 3) spinal anesthesia with local anesthetic adductor canal block (SA + ACB). Patients who received femoral nerve blocks, pain catheters, or any other anesthesia types not listed above were excluded. In addition, patients undergoing revision procedures, bilateral procedures, concomitant procedures, robotic procedures or had a history of previous open surgeries or infections to the knee were excluded. Finally, any patient that underwent ACB utilizing liposomal bupivacaine was excluded. All patients underwent cemented TKA procedures using a medial parapatellar approach and posterior stabilized components.

For all patients, primary anesthetic type, regional nerve blocks, demographic information, hospital length of stay (LOS), readmission rates, and total hospital costs per patient (direct and indirect costs) were obtained. Direct costs included supplies, salaries, implants, medications, and hospital services provided by the orthopaedic, anesthesiology, physical and occupational therapy, and nursing departments (in addition to hospital room costs). Indirect costs included other hospital staff services, ancillary staff salaries, and any other hospital costs not included above. In addition, daily average opioid consumption during the postoperative period was also calculated. This was calculated by recording all intravenous and oral opioids taken from the immediate postoperative recovery period (PACU) until day of discharge. An opioid conversion chart was utilized to convert all opioids to an oral morphine equivalent (OME).

3 Results

3.1

3.1 Demographics

Overall, 783 charts of patients undergoing TKA procedures were reviewed. Of these, 420 patients were excluded from our study. Sixty-seven patients had bilateral total knee arthroplasty procedures, 28 patients had a robotically-assisted total knee arthroplasty, and 103 patients had different primary anesthetics and different regional nerve blocks for postoperative pain control. An additional 136 patients were excluded because they received liposomal bupivacaine in their adductor canal blocks. Lastly, 29 patients were excluded secondary to lack of information regarding their pre-surgical knee x-rays and arthritis grade. Of the 420 patients, 63 patients were in the general anesthesia only cohort, 148 patients in the general anesthesia with adductor canal block cohort, and 209 patients in the spinal with adductor canal block cohort. (Fig. 1).

Close

Fig. 1 Inclusion and Exclusion of our Cohorts Undergoing Total Knee Arthroplasty. GA = General Anesthesia; ACB = Adductor Canal Block; SA=Spinal Anesthesia.

Export to PPT

All three of our cohorts were above 65 years of age on average and mostly female, non-Hispanic, and white. The average BMI of all three of our cohorts was above 30 (range 18.79–54.56). There was a significantly larger proportion of non-Hispanics and white patients in the spinal with adductor canal block group compared to the other two groups (Table 1). Most of our cohort had American Society of Anesthesiology (ASA) scores of 2 or 3, with the SA + ACB group having a significantly higher proportion of ASA 2 scores. Average Charles Comorbidity Index (CCI) scores were approximately 3 for all cohorts. Most patients did not have a documented history of chronic pain. Finally, most of our patients had Kellgren & Lawrence (KL) grades of 3 or 4 (Table 2).

Table 1 Demographic Information of our Sample Population. Percentages expressed as a proportion of the total cohort size.

		GA Only	GA/Adductor	Spinal/Adductor	p-value
Average Age		65.62 ± 10.02	67.07 ± 10.41	66.39 ± 10.30	0.62
Gender	Female	40 (63.5 %)	89 (60.0 %)	118 (56.7 %)	0.59
	Male	23 (36.5 %)	59 (40.0 %)	91 (43.3 %)
Ethnicity	Hispanic	5 (7.9 %)	11 (8.0 %)	4 (1.9 %)	0.04
	Non-Hispanic	55 (87.3 %)	137(92.0 %)	203 (96.7 %)
Race	White	54 (85.7 %)	134 (90.0 %)	200 (95.7 %)	0.02
	Black/Asian/Other	9 (14.3 %)	14 (10.0 %)	9 (4.3 %)
Marital Status	Married	38 (60.3 %)	87 (58.0 %)	124 (59.0 %)	0.88
	Single	5 (7.9 %)	18 (12.7 %)	26 (12.4 %)
	Other	20 (31.7 %)	43 (29.3 %)	59 (28.6 %)
Average BMI		34.68 ± 5.66	33.56 ± 7.04	32.72 ± 7.20	0.12
Total		63	148	209

Table 2 Medical History of our Sample Population.

		GA Only	GA/Adductor	Spinal/Adductor	p-value
American Society of Anesthesiologists (ASA) Score	1	0 (0.0 %)	0 (0.0 %)	2 (1.0 %)	<0.01
	2	23 (36.5 %)	51 (34.0 %)	116 (54.3 %)
	3	38 (60.3 %)	96 (63.3 %)	90 (42.9 %)
	4	1 (1.6 %)	1 (0.7 %)	1 (0.5 %)
Charles Comorbidity Index		2.7 ± 1.5	3.0 ± 1.5	2.9 ± 1.5	0.33
Chronic Pain	Yes	10 (15.9 %)	18 (12.7 %)	15 (7.6 %)	0.11
	No	53 (84.1 %)	130 (87.3 %)	194 (92.4 %)
Kellgren & Lawrence Grade	2	7 (11.1 %)	10 (6.7 %)	8 (3.8 %)	0.15
	3	29 (46.0 %)	63 (42.7 %)	106 (51.0 %)
	4	27 (42.9 %)	75 (50.7 %)	95 (45.2 %)
Total		63	148	209

3.2

3.2 Hospital outcomes

Patients in the SA + ACB group had a significantly shorter LOS of 2.4 ± 1.0 days compared to both the GA + ACB (2.79 ± 1.3 days) and GA (2.9 ± 1.7 days) groups (p < 0.01) when performing multivariate regression analysis, even though the median LOS was 2 days in all three groups. (Fig. 2). Additionally, patients with a KL grade of 4 had longer LOS (2.7 days) compared to those with KL grade 2 (2.4 days) and 3 (2.5 days) (p < 0.01). Finally, there were no readmissions in the GA group, while 3 patients (3.3 %) in the GA + ACB and 1 patient (0.5 %) in the SA + ACB groups were readmitted within 30 days, notably for cellulitis of operative leg and various cardiac complications (symptomatic bradycardia, atrial fibrillation with rapid ventricular rate, and cardiovascular accident.

Close

Fig. 2 Hospital Length of Stay. Length of stay in days for each anesthesia type.

Export to PPT

3.3

3.3 Opioid use

While most patients in all three groups had minimal opioid consumption (0–40 OME), the SA + ACB group had the lowest number of patients with moderate opioid use (41–60 OME) and excessive opioid use (>90 OME) (Fig. 3). The SA + ACB group had a statistically lower daily average OME requirement (36.5 OME) compared to the GA (40 OME) and GA + ACB (47 OME) groups (p < 0.01). This held true in the multivariate regression analysis accounting for confounding demographic factors as well (p < 0.01). Finally, patients with a KL grade 4 (45.5 OME) had a significantly lower amount of average daily OME compared to patients with KL grade 2 (60.6 OME) and 3 (51.5 OME) as well (p < 0.01).

Close

Fig. 3 Daily Average Opioid Requirement After Surgery. GA = General Anesthesia group; GA + ACB=General Anesthesia with adductor canal block group; SA + ACB=Spinal Anesthesia with adductor canal block group.

Export to PPT

3.4

3.4 Cost analysis

On average, patients in the GA group had the highest average total cost of care of $15,006.38. Patients in the SA + ACB group had the lowest average total cost of $11,683.91. This was also statistically cheaper than the other two groups (p < 0.01). Direct costs ($8015.86) and indirect costs ($3668.04) were also lowest in the SA + ACB group (Fig. 4).

Close

Fig. 4 Average cost per patient by anesthesia type.

Export to PPT

4 Discussion

With the continuing opioid epidemic as well as focus on decreasing total hospital costs, orthopaedic surgeons and anesthesiologists have been tasked with efficiently and cost-effectively controlling postoperative pain in the face of an increasing number of TKA procedures. As we implemented a multimodal pain regimen, our study found that SA + ACB was more successful in decreasing hospital LOS and opioid usage compared to GA and GA + ACB, all while decreasing total hospital costs. This confirms previous reports of the increased complications associated with general anesthesia and the benefits of utilizing spinal anesthesia.^10–12 Surgeons and anesthesiologists involved in the care of these patients should consider this anesthesia type when aiming to improve pain control and postoperative outcomes.

While not accounting for outpatient opioid use, our inpatient opioid usage was lowest in the SA + ACB group. Our results point to this anesthesia type eliminating a higher proportion of excessive opioid users, even though most patients in all three groups had low opioid usage. This was associated with lower hospital costs and shorter LOS. The decrease in need for pain medication and earlier improved pain control logically led to shorter LOS (by cutting out longer stays due to uncontrolled pain) and lower hospital costs. Pain control after TKA is especially hard to achieve, with many studies analyzing the best way to maximize analgesic efficacy. Our study confirms that a multimodal pain regimen should include spinal anesthesia and peripheral nerve blocks in addition to the non-narcotic medications.¹³ In addition, while an analysis on risk factors for increased LOS was not in the scope of this study, patient factors such as increased age, rheumatoid arthritis, higher ASA grades, higher CCI, higher BMI, and lower hemoglobin levels, among other factors, have been shown to increase LOS and increase complication rates.^14–16

Advancements overtime in orthopaedic techniques and technology have been met by improvements in anesthetic and peripheral nerve block techniques. Spinal anesthesia has been commonly favored compared to general anesthesia as it may decrease intubation-related complications and blood loss with hypotensive anesthesia while minimizing the recovery time after.^17–19 In addition, an increase in fellowship-trained anesthesiologists in regional anesthesia along with the use of ultrasound guidance has led to improved accuracy and efficacy of blocks as well. Furthermore, nerve block anesthetic agents have evolved over time, with multiple long-acting agents existing that help improve the longevity of these blocks.^20,21 Finally, adductor canal blocks are preferred to femoral nerve blocks after TKA due to decreased deactivation of the quadriceps and the ability to start physical therapy earlier postoperatively because of this.²²

Patients with higher arthritis grades had an expected longer length of stay. However, they surprisingly had a lower amount of average daily opioid use compared to patients with lower grades. In previous reports, patients with milder radiographic arthritis tend to have higher postoperative pain, more chronic pain, and lower satisfactory rates after TKA compared to those with more severe radiographic arthritis.^23,24 Patients should be counseled on this fact before undergoing TKA procedures, so they have a better understanding of what to expect. The culture of staying in the hospital for several days after TKA has changed over the years to become outpatient procedures. During the last few years of our study when more SA + ACB were performed, the expectation was to be discharged on postoperative day one.

As more emphasis has been placed recently on quicker postoperative recoveries after TKA procedures, preoperative Surgical Home (PSH) clinics were introduced to provide a patient-centric approach to the complexities of perioperative care. Guidance for the patient and their families starts from the decision for surgery to the post-discharge phase.^25–27 There is strong emphasis on process standardization, and the use of evidence-based clinical care pathways, such as Clinical Pathways and Enhanced Recovery after Surgery (ERAS) protocols. Such pathways and protocols have been touted as an efficient and cost-effective use of health resources.²⁸ While we did not track compliance to ERAS protocols in our patients, they were an integral part of peri-operative care in all of our patients that cannot be overlooked.

Finally, despite our study being underpowered, the significant difference in opioid usage and our other secondary outcomes among anesthesia types point to the superiority and efficacy of spinal anesthesia with an adductor block in reducing opioid usage and improving postoperative outcomes after TKA.

Limitations in our study are those mostly inherent to a retrospective chart review, including limited patient sample size and the possibility of charting errors. The hospital network initiated their use of EPIC in January 2016, and so our clinical analytics department held patient data as far back as 2016 only. As a result, the GA only cohort size was smaller. In our hospital, performing GA during this time frame only would also only result as a consequence of patient refusal or medical contraindication to both spinal anesthesia and regional nerve block. This includes attempted spinal without success, history of spinal fusion, or recent anticoagulation that does not permit spinal anesthesia. The surgical variability and postoperative rehabilitation protocols between our 27 orthopaedic surgeons and the variability in effectiveness of peripheral nerve blocks performed by anesthesiologists was not accounted for. Finally, the reason for longer LOS was not documented, and this could be due to medical reasons not directly related to the TKA. However, we believe accounting for patient ASA scores helps mitigate this bias.

Ethical statement

IRB Approval was obtained for this study by our institution's IRB (IRB 2020-124).

Funding sources

B Braun Grant.

Patient consent

All patients consented to having their de-identified data included in this study.

IRB approval

IRB approval was obtained both from the St. Luke's University Health Network IRB Committee and the Temple University School of Medicine IRB Committee.

CRediT authorship contribution statement

Ajith Malige: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. Michael DeRogatis: Conceptualization, Data curation, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – review & editing. Allincia Michaud: Data curation, Formal analysis, Investigation, Project administration, Resources, Software, Validation, Visualization, Writing – review & editing. Michael Usewick: Data curation, Formal analysis, Investigation, Project administration, Resources, Software, Validation, Visualization, Writing – review & editing. Anna Ng-Pellegrino: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – review & editing.