This trial is a randomized, double-blinded controlled study, and was approved by the ethics committee of the Second Affiliated Hospital, Guangzhou University of Chinese Medicine (ethics number: ZF2020-031-01). The trial will be conducted in accordance with the Helsinki Declaration and monitored by the trial agency at the Guangdong Provincial Hospital of Chinese Medicine.
Recruitment and consent
Patients diagnosed with severe knee osteoarthritis (OA) and scheduled for total knee arthroplasty (TKA) will be recruited from the Department of Orthopedics, Second Affiliated Hospital, Guangzhou University of Chinese Medicine, with a target sample size of 234 subjects. All candidates will go through a standardized interview process and receive more information about the study and treatments. The participants’ written consents will be obtained. The purpose, procedures, and potential risks and benefits of the study will also be explained thoroughly to the participants. Participants are able to withdraw from the study at any time without consequence. The trial will be executed from October 2020 to July 2022, including enrollment and follow-up. (see Fig. 1).
All recruited patients will receive general anesthesia. Cefamandole will be administered intravenously 30 minutes before and after surgery. Low-molecular-weight heparin for thromboprophylaxis will be administered subcutaneously 8 hours postoperatively and then once daily until two weeks.
For multimodal pain management, preoperative oral doses of 400 mg celecoxib will be administered from day 1 before surgery. Postoperatively, all patients will receive 40 mg of parecoxib sodium twice intravenously until postoperative day (POD) 3, followed by oral doses of 200 mg celecoxib every 12 hours until discharge. Tramadol hydrochloride injection will be applied for rescue analgesia.
The same surgeon will perform all operations. A Depuy Attune PS prosthesis (Attune ®; DePuy Synthes, Warsaw, IN) will be inserted via a standard medial parapatellar approach in all patients.
Before wound closure, 80 ml of cocktail analgesic will be injected into the muscle and joint capsule for local infiltration analgesia, containing 300 mg 0.75% ropicacaine, 5 mg morphine, 50 mg flurbiprofen axetil, 0.4 ml 1:1000 epinephrine, and adequate saline.
Betamethasone will be mixed in the multimodal cocktail at 3 ml for high-dose betamethasone group (group H), 1 ml for normal-dose betamethasone group (group N), and a non-betamethasone group will serve as the control group (group C).
Randomization and blinding
Random assignment was performed before surgery by an independent assistant using a computer-generated, blocked random-allocation sequence with a 1:1:1 ratio. Computer-generated numbers will be located in a sealed envelope containing treatment information. After anesthesia, an envelope will be opened by an assisting nurse who is not involved in the study, and cocktail analgesics will be made according to the envelope. Milky white flurbiprofen axetil injection will be used as an ingredient of cocktail analgesics, which will guarantee that there are no differences in the appearance of the three groups; thus, the surgeons will be blinded. The outcome measurement will be performed by a blinded assessor who will not be involved in the randomization and blinding progress. The patients will be blinded due to general anesthesia.
Primary outcome measures
The primary outcomes of this trial will be the postoperative VAS pain score at rest and the strongest pain experienced during knee flexion. The VAS score will be evaluated by an independent assessor at preoperative day 1, postoperative 6, 24, 48, and 72 hours; day 5 and 14 after the surgery.
Secondary outcome measures
Postoperative 1-mintue walking ability
Physical activity is an important aspect of day-to-day life, and walking capacity is a measure of exercise tolerance requiring muscle strength. Walk tests measure the distance walked over a definite time period, with greater distances indicating better performance [19, 20]. The 1-minute walk test will be measured on preoperative days 1, POD2, 4, 7, 14, and 30.
Other rehabilitation-related outcomes
The circumference of the knee has been shown to have acceptable reliability for the determination of gross changes in knee swelling and inflammatory status in patients post-surgery . Knee swelling was measured using the circumference of the knee at the mid-portion of the patella with the knee in full extension. Measurements were then taken 10 cm proximal to the superior pole of the patella (thigh) and 10 cm distal to the inferior pole of the patella (calf). The skin is marked at these points, and the tape measure is placed circumferentially around the limb. The circumference at three measurement positions will be measured on preoperative day 1, POD 2, 4, 7, 14, and 30.
The Knee Society knee score will be applied to evaluate knee function on preoperative day 1, POD14, 30.
The presence of side effects during the postoperative period will be recorded, including nausea, vomiting, constipation, wound healing status, and other adverse events. Perioperative changes in serum C-reactive protein (CRP), interleukin 6 (IL-6), erythrocyte sedimentation rate (ESR), and blood glucose (Glu) will be noted on POD1, 7, 14, and 30.
Calculation of sample size was based on previous studies. Ikeuchi et al.  reported that POD3 VAS scores were 27 ± 16 mm in the steroid group and 43 ± 18 mm in the control group. Research in Kwon et al.  reported that POD3 VAS scores were 35 mm in the steroid group and 41 mm in the control group. We assumed that the mean difference in the visual analog scale (VAS) score between the high-dose betamethasone group and control group was 0.8, and the mean difference between the normal-dose betamethasone group and control group was 0.4. An estimated standard deviation (SD) of 1.5, significance level (α) of 0.05 (two-sided) and power of test (β) of 0.8000 will be adopted. Based on the former data, the sample size calculation will be 69 participants per group. Assuming a dropout rate of 10%, the final sample size will be 78 participants per group, and a total of 234 participants for three groups will be required.
Perioperative outcomes will be summarized using the mean ± SD for continuous variables and the ratio for categorical variables. All statistical analyses will be conducted using SPSS software for Windows 13.0. For continuous variables, analysis of variance (ANOVA) will be used to compare the respective differences in intragroup changes before and after surgery, and Dunnett’s t test will be conducted to compare between-group differences. For categorical variables, the chi-square test will be conducted to compare intragroup differences, and partitions of chi-square methods will be used to compare between-group differences. The significance level will be set to α = 0.05.