By using the hospital’s electronic medical records and appointment system, patients who are referred to the orthopedic surgery outpatient clinic at a tertiary hospital in Singapore will be screened based on the eligibility criteria prior to their actual consultation visit. During the clinic visit, potential patients will be invited to participate in the study if they meet all the inclusion and exclusion criteria. A research assistant will share details of the study with patients who are interested. Written informed consent will be obtained for those who subsequently agree to join the trial. The reasons for rejection to participate will be recorded. Each consenting participant will be asked to complete the baseline measures. Patients will then be randomized into the control arm or the intervention arm. Considering previous studies [15,16], patients who have received TCM treatment within 1 week prior to the trial will be subject to a 2-week washout period, counting from their last TCM treatment date. They will be requested to stop any form of acupuncture before beginning the trial. For this group of patients, randomization and baseline measures will be performed after the wash-out period. Participants are allowed to take any analgesics as needed throughout the study duration, and this will be recorded. All medications (prescription and over the counter), vitamin and mineral supplements, and herbs taken by the participant will be documented. Data collection will be done at baseline and weeks 6, 12, and 26. The research assistant will follow up with the participants regularly to ensure study completion. Figures 1 and 2 illustrate the study schedule of enrollment, interventions, and outcome measurements according to the SPIRIT guidelines and the CONSORT flowchart, respectively.

Patients who consent to participate will be randomized in a 1:1 allocation ratio to intervention or standard care using a stratified, permuted, block randomization method (block sizes of 4, 6, and 8). Stratification is based on gender with a female to male ratio of 2:1. The unbalance ratio was based on the World Health Organization’s estimate of the prevalence of symptomatic osteoarthritis for people aged 60 years or older, among whom the prevalence of women is almost double that of men (male vs female: 9.6% vs 18%) [17]. The allocation sequence is generated by an independent statistician a priori using Stata (version 16.1; StataCorp) and will be kept concealed from the study team. Randomization will be done using the REDCap (Research Electronic Data Capture; Vanderbilt University) randomization module based on the allocation sequence; allocation will be locked once assigned. Randomization and intervention allocation will only be performed by the study team after the patient is counselled fully about the study and provides informed consent.

After reviewing international literature and engaging with local TCM experts, this clinical intervention was designed in accordance with local practice. The intervention involves acupuncture with far-infrared heat therapy in addition to standard care, which will be administered by qualified TCM practitioners from Tan Tock Seng Hospital’s Complementary Integrative Medicine Clinic (CIMC) and Singapore Chung Hwa Medical Institution (SCHMI), who have full registration with the Traditional Chinese Medicine Practitioners’ Board and a minimum of 3 years of acupuncture practice. Before the commencement of the trial, all the acupuncturists from the participating institutions will undergo training on acupoint locations and needling methods to ensure standardization of acupuncture techniques in accordance to the acupuncture protocol that was developed by consensus among all the senior acupuncturists participating in this trial. The treatment protocol and rationale are elaborated below.

Our study aims to evaluate the therapeutic effectiveness of acupuncture combined with far-infrared heat therapy in addition to standard care, compared to standard care alone. To achieve this purpose, we have not included sham acupuncture as an inactive control in our study design. There are controversies about the validity of sham acupuncture. A study found that sham acupuncture may produce comparable effects on biomarkers as verum acupuncture, which suggests that sham acupuncture, being an inactive intervention akin to a placebo, should be reevaluated [18]. Additionally, heat therapy will yield beneficial outcomes in the treatment, regardless of whether verum or sham acupuncture is used.

Although the 2007 Singapore Ministry of Health clinical guidelines for KOA recommend electroacupuncture as an adjunct therapy, anecdotal input from our local subject matter experts suggests that some patients are afraid of electrical stimulation and that it is generally less well tolerated due to the pain of electrical stimulation. Studies have also shown that even though electroacupuncture has advantages compared to standard care, acupuncture with heat therapy, a more tolerable option, possesses similar efficacy as electroacupuncture when combined with the usual pharmacotherapies [19]. Therefore, we chose to evaluate acupuncture with heat therapy to establish its clinical effectiveness as an adjunctive therapy in the management of KOA, with long-term implementability a key consideration.

Upon careful review of the evidence-based Guidelines of Clinical Practice with Acupuncture and Moxibustion for KOA [20] and the intervention protocols of several acupuncture RCTs that examined patients with KOA (Multimedia Appendix 1, Figure S1 [10, 21-29]), acupoints that were previously reported to be effective in pain relief and improving knee flexibility were identified.

Guided by international evidence in consultation with the local TCM collaborators, a consensus on a set of proximal and distal acupoints based on TCM meridian differentiation was reached (Multimedia Appendix 1, Figure S2 [30,31]).

In total, 14 needle insertions will be made on a fixed set of proximal and distal acupoints for each patient per session (Figure 3).

Distal acupoints will be applied on the opposite elbow [32] to increase the analgesic efficacy according to contralateral collateral needling (缪刺) from Huang Di Nei Jing (Yellow Emperor's Inner Classic in English). It is a therapy where the right side is acupunctured if the left side is diseased and vice versa; the upper side is acupunctured if the below is diseased and vice versa. Studies have shown that contralateral collateral needling unblocks collaterals and is potentially more effective than conventional acupuncture for some conditions, including hemiplegia following acute ischemic stroke, cervical shoulder pain, insomnia, and KOA [33-35]. From a pragmatic perspective, patients with bilateral knee pain will only receive acupuncture on the most painful knee and the opposite elbow. This practice is supported with evidence that suggests unilateral acupuncture is equally as effective as bilateral acupuncture in increasing function and reducing the pain associated with KOA [36].

The patient will be treated in a supine position with pillows under both knees. After a standard disinfection procedure, a 30 × 40–mm disposable stainless steel acupuncture needle (YangTzeKing Impex S Pte Ltd) will be inserted according to the location defined; the direction of insertion and the safe depth range are specified in Multimedia Appendix 1, Figure S2 [21, 30,31]. Basic manipulations such as lifting and thrusting, twirling, and rotating are applied until the needling sensation (de-qi) is obtained. After de-qi is obtained, uniform reinforcing-reducing (平补平泻) manipulations are applied. The needles will be retained for 30 minutes (except for weizhong [BL40], where the needle is removed immediately after the needling sensation is obtained), whereupon 2 manipulations will be applied at 10-minute intervals before needle removal. During needle retention, far-infrared heat therapy will be applied at the lateral, medial, and anterior sides of the knee for 10 minutes each for a total of 30 minutes while maintaining a 30-cm safety distance (Figure 4). The treated region will be observed and checked for redness at 10-minute intervals to prevent burn injury. A total of 12 acupuncture sessions will be administered twice a week for 6 weeks in addition to standard care. A minimum of 50% attendance (6 of 12 sessions) is required to qualify as completing the study treatment. There are no make-up sessions should the patient miss the scheduled appointment. Total numbers and dates of attended and missed sessions will be recorded for analysis purposes. The study intervention will be conducted at TTSH, CIMC, or SCHMI. The patients are allowed to indicate their preference for the treatment location at the start of the intervention and must attend at the selected location throughout the 6-week intervention. Each TCM session takes approximately 40 minutes, involving acupuncture and heat therapy. Safety precautions will also be exercised to ensure the safety of the patients. Generally, acupuncture is safe and has a low risk of adverse events [37]. However, since the most common adverse events are subcutaneous hematomas and hemorrhages at the site of needling [37], the TCM practitioner will take extra precaution if patients indicate during informed consent that they are taking antiplatelets or anticoagulants. All adverse events that occur during the treatment will be documented.

Standard care generally constitutes simple lifestyle advice, analgesia, and a referral to the outpatient physiotherapist. The physiotherapist conducts an assessment and recommends a variety of lifestyle modifications and exercise therapy. Any other treatment or the involvement of allied health professionals, such as dieticians, will take place at the discretion of the managing physician. The patients in the standard care group will be discouraged from seeking acupuncture treatment for 6 to 8 weeks after randomization. However, if they do so, it will be noted and accounted for in the analysis.

The study intervention may be discontinued under the following circumstances: (1) a serious adverse event arises from the intervention, (2) a participant opts for total knee replacement, (3) a participant is unable to continue the study (eg, because of the sudden onset of any medical condition that prohibits continuing the study), and (4) a collaborator withdraws due to unforeseen circumstances. The participants may also withdraw voluntarily from the study intervention for any reason.

The choice of outcome measures is based on the OARSI guidelines for lifestyle diet and exercise clinical trials in osteoarthritis [38]. The recommended core outcomes include pain, physical function, and global patient assessment. Additional outcomes include health-related QOL and global physician assessment. Outcome measures will be collected at baseline (pretreatment), week 6, week 12, and week 26. The primary end point in this study is the Knee Osteoarthritis Outcome Score (KOOS-12) at week 6. KOOS-12 is a specialized tool designed to evaluate patients’ knee condition, and it assesses the short- and long-term effects of knee injuries [39]. It reduces respondent burden by 70% from the original KOOS while providing scale scores for knee-specific pain, function, and QOL, along with a summary measure of overall knee impact [40]. The KOOS score has been validated in Singapore [39,41].

Baseline measures such as demographics, socioeconomic status, medical comorbidities, knee symptoms, radiographic severity, height, and weight will be collected.

Additionally, after considering the progression and severity of KOA as defined in TCM clinical guidelines published in China (Multimedia Appendix 1, Figure S3), 2 evaluation forms have been designed. The Knee Osteoarthritis TCM Evaluation Form (Multimedia Appendix 1, Figure S4) was specifically created to integrate the KOOS-12 knee survey with questions based on TCM syndrome differentiation for KOA. This will enable the investigators to assess the efficacy of the treatment by analyzing syndrome differentiation. This form will be administered at week 1 (start of treatment, before the first acupuncture session), week 4 (midtreatment, before the seventh acupuncture session) and week 6 (end of treatment, after the 12th or last acupuncture session). The Knee Osteoarthritis Visual Analog Scale and Range of Motion Evaluation Form (Multimedia Appendix 1, Figure S5) will be administered before and after each of the 12 biweekly acupuncture sessions over 6 weeks.

Clinical data (glycated hemoglobin, blood pressure, and other chronic disease parameters) and operational and cost data (health care use) will also be obtained through the health care administrative databases for economic evaluation.

Except for the TCM evaluation forms (Multimedia Appendix 1, Figures S4 and S5), the remaining outcome measures will be measured by blinded outcome assessors. All outcome assessors will receive training prior to study initiation to ensure good inter- and intraobserver reliability, particularly for the functional performance testing. Patients will be instructed not to reveal their allocation to the outcome assessors. Outcome assessment will be conducted at the TTSH physiotherapy clinic. Unblinding will be carefully considered and justified in each case of medical emergency or clinical treatment decisions that require the intervention to be known, following the study protocol and ethical guidelines.

The sample size needed to find a 10-point difference for KOOS-12 (the primary outcome) between the intervention and control arms with a power of 90%, a P value of .05 (2-sided) and an SD of 14 is n=42. The 10-point difference is based on the minimal clinical important difference for KOOS [42]. The anticipated baseline and SD are based on studies done in similar populations. Accounting for a 20% drop-out rate, a total of 100 patients will be needed for the study.

Results will be analyzed by the intention-to-treat principle. However, data will also be analyzed by a per-protocol approach to account for protocol violations, such as patients who were deemed not compliant to treatment or patients who underwent a surgical procedure of the knee due to treatment failure during the study. Descriptive frequency analysis will be used for baseline characteristics. For continuous variables, the mean and SD will be reported and for categorical variables, the frequencies and percentages will be reported. Between-group comparisons of change from baseline to the 26-week follow-up in the primary and secondary continuous outcomes will be analyzed using a generalized linear mixed model (GLMM). Testing for normality of distributions of outcomes will be based both on the Shapiro-Wilks test and a visual analysis of the histogram plot. Categorical secondary outcomes will be analyzed using the ordinal logistic regression function under the GLMM. A 2-sided P value less than .05 will be considered statistically significant. An analysis will be done on the nature of the missing data to determine if the data are missing at random or a systemic bias is present resulting from the missing data.

All data will be monitored by the principal investigator and the study team independently of the study sponsor. Data quality measures include queries to identify outliers and missing data. A unique identifier will be assigned to each patient after enrollment to ensure patient confidentiality. Data will be collected and stored on the REDCap system, which is a widely used and secure web application for clinical data management in research. REDCap is password protected, and access rights to the study database are only granted to authorized people. Physical study files will be stored under lock and key.

A safety monitoring review will be done quarterly. Data monitored include serious adverse events and research-related events such as pain, burn, or any other adverse events that may arise due to the conduct of study activities. Safety monitoring reviews will be done by the principal investigator and the co–principal investigator of the study team. The study may be stopped by the principal investigator if serious safety concerns emerge. In the event that a critical adverse event occurs and a decision must be made to terminate the study, interim analyses will be conducted by an independent statistician.

Adverse events and other unanticipated effects of the intervention will be spontaneously reported by the acupuncturists. All reporting will be collected and assessed by the study team. Unanticipated problems involving risk to participants or other events will be reported to the institutional review board.

Only authorized study personnel will have access to the secured cabinet(s) or room(s) where hard-copy data will be stored. The electronic data will be stored on a password-protected desktop computer, laptop, and external HDD. The databases will not include subject identifiers, and the data relating subject identifiers to subject identification codes will be stored separately.The research data will only be accessible to the study team (including principal investigators, co-investigators, research coordinators, and statisticians). The anonymized and password-protected sharing of research data will be allowed with external institutions.

Enrolled participants are covered by indemnity for unexpected negligent harm due to the trial procedure.

We anticipate that it will take between 3 and 4 months to compile the final results for the appropriate journal. The study results will be made available to the participating physicians and the medical and TCM practitioner community via academic publication.