Effectiveness of Tuina manipulation in patients with knee osteoarthritis: study protocol for a randomized controlled trial

Background: Knee osteoarthritis (KOA) is a common musculoskeletal disorder. Previous studies reported that Tuina manipulation showed benicial effects in management of musculoskeletal disorders . However, there was no enough evidence to support the effectiveness of Tuina manipulation for KOA. The purpose of this study is to evaluate the effectiveness of Tuina manipulation on pain and dysfunction of patients with KOA compared with health care education. Methods/Design: This study is a single-centre, two-arm, open-label randomized controlled trial (RCT). A total of 170 eligible KOA patients will be randomly assigned to the Tuina manipulation group or the health care education group in a 1:1 ratio. In the Tuina manipulation group, patients will receive a 30 min treatment including pain point assessment and manual therapy. The health care education group will receive 45 min of lecture and discussion. All treatments will be conducted three times a week for four weeks. The primary outcome is the Knee Injury and Osteoarthritis Outcome Score (KOOS). Secondary outcomes include the McGill Pain Questionnaire (MPQ) and 36-item Short-Form Health Survey (SF-36). The results will show the evidence of the effect of Tuina manipulation for KOA compared with health care education. Discussion: The design and methodology of the trial is rigorous and allows the collection of valuable data to assess the effect of a specic Tuina regimen for the treatment of KOA. Therefore, the trial will provide a solid foundation for future clinical research on KOA and Tuina therapy.

weeks. The primary outcome is the Knee Injury and Osteoarthritis Outcome Score (KOOS). Secondary outcomes include the McGill Pain Questionnaire (MPQ) and 36-item Short-Form Health Survey (SF-36).
The results will show the evidence of the effect of Tuina manipulation for KOA compared with health care education.
Discussion: The design and methodology of the trial is rigorous and allows the collection of valuable data to assess the effect of a speci c Tuina regimen for the treatment of KOA. Therefore, the trial will provide a solid foundation for future clinical research on KOA and Tuina therapy.
Trial registration: This trial was registered with the Chinese Clinical Trial Registry on 24 February 2020 (ChiCTR2000030154), http://www.chictr.org.cn/index.aspx Background Knee osteoarthritis (KOA) is a disease characterized by joint pain, in ammation, stiffness and dysfunction [1]. KOA is a common musculoskeletal disorder that affects human health and brings a heavy nancial burden worldwide [2]. Speci c factors contributing to KOA include sex, advanced age, the use of postmenopausal hormone replacement therapy, educational attainment, and obesity [3]. Over 55 years old, the prevalence of pain and disability due to KOA is 10%, and a quarter of the affected individuals are severely disabled [4]. According to a survey, approximately 14% of American adults and 40% of Britons and Australians over the age of 65 suffer from KOA [5].
The pathophysiology of KOA has not been obviously established because it is di cult to determine the source of the pain. Studies have con rmed that cytokines such as tumour necrosis factor-α and proteases played important roles in the occurrence and development of osteoarthritis. The levels of TNF-α in the plasma and synovial uid of patients with osteoarthritis were signi cantly higher than those in unaffected people, and the level of TNF-α in the synovial uid was related to the degree of severity of the KOA lesions [6][7].
Asymmetrical narrowing of the joint space, osteosclerosis and/or cystic deformation at the lower end of the cartilage, and extra bone formation at the joint edge are the typical X-ray manifestations of KOA. MRI and CT scans can be used to achieve an early diagnosis of KOA, assist with the differential diagnosis and facilitate clinical research. MRI images of KOA are characterized by the thinning of cartilage thickness, tissue defects, meniscus injuries and degeneration, joint effusion, oedema and popliteal cysts. CT can show the stenosis of the affected joint space, subchondral bone sclerosis, cystic changes and osteophyte hyperplasia [8].
There are several different treatments than can be adopted to control KOA, such as intra-articular (IA) corticosteroid (CS) injection, non-steroidal anti-in ammatory drugs (NSAIDs), opioids, physical therapy (PT), arthroscopy, IA hyaluronic acid (HA) injection, surgery and exercise therapy [9][10][11][12]. However, NSAIDs may induce mucosal damage in the upper digestive tract and lower digestive tract [13]. Even with the use of gastric protectants, up to three-quarters of patients who use NSAIDs suffer from small bowel damage [14]. Opioids also carry the risk of addiction [15]. These treatments can achieve bene cial effects but are associated with recurrence and side effects. The indications for the use of each treatment depends on the stage of the disease and the speci c circumstances of the patient. When basic and medical treatments are not effective, patients can only undergo surgery. However, surgery is very expensive, which imposes a substantial economic burden on families and society. Therefore, to maximize the safety and minimize the cost of treatment, we need to nd a treatment with few side effects and a low cost.
Tuina is a type of physical therapy that relies on the strength of the practitioner's ngers, and it is a longstanding healing art [16]. When treating KOA with Tuina, the manipulation technique should be gentle and be applied for long enough to relieve pain to achieve optimal clinical e cacy.
As a type of physical therapy, manual therapy has been recommended by many countries as a non-drug therapy in management of musculoskeletal diseases. Recent clinical studies and systematic reviews have ithat manual therapy is safe and effective for the treatment of KOA [17][18][19]. Compared with NSAIDs, massage therapy can improve functional pain while incurring fewer costs. However, due to differences in treatment accuracy, the use of various treatment methods, insu cient sample sizes, low methodological quality, and the differences in the control groups, the e cacy is quite controversial.
Health care education is also an effective way of intervention, it shows positive effects for KOA patients [20][21][22]. The study reported short-term education lecture improved pain levels, physical function, and physical and mental quality of life of patients with KOA, as well as decreased BMI of them, which based on telephone calls one year after health education lecture [23].
Previous studies showed that Tuina manipulation may improve walking ability and muscle strength of patients with KOA [24,25]. Tuina manipulation also showed potential effects on joint dysfunction of patients with KOA. In order to provide reliable clinical evidence on pain and dysfunction of Tuina manipulation for patients with KOA, the randomized controlled trial is designed to investigate the effect on pain and dysfunction of Tuina manipulation for KOA compared with health care education. The primary outcome is the Knee Injury and Osteoarthritis Outcome Score (KOOS). Secondary outcomes include the McGill Pain Questionnaire (MPQ) and 36-item Short-Form Health Survey (SF-36).

Methods/design
Study design This is a multicentre, blinded RCT with two parallel groups: the Tuina manipulation group and the health care education group. The research protocol was approved by the Regional Ethics Review Committee of Yueyang Hospital of Integrated Traditional Chinese and Western Medicine a liated with Shanghai University of Traditional Chinese Medicine (Project No. 2019-097). It has been registered with the China Clinical Trial Registry (ChiCTR-2000030154). A total of 170 eligible patients with KOA will be randomly assigned to the two groups in a 1:1 ratio. All patients will be asked to provide written informed consent.
Outcome assessments and statistical analyses will be performed by independent investigators blinded to the patient assignments. The research design is illustrated in the ow chart in Figure 1, and the research schedule is shown in Figure 2.

Participants
Eligible participants will be patients diagnosed with KOA. Patients will be noti ed of the trial if knee pain or dysfunction affects their daily activities at work or elsewhere. If patients express willingness to participate, the clinical trial communicator will describe the trial to them and ask them if they would like to participate. If the patient wishes to participate, an interviewer will conduct a face-to-face interview in the reception room of the hospital. Patients who meet the inclusion criteria and do not meet the exclusion criteria can join after signing the consent form.

Recruitment
We will make this trial publicly available to potential participants in the following ways: 1) posting recruitment posters in the hospital's outpatient and ward corridors; 2) posting advertisements in community centres near the hospital; 3) posting information on the hospital's o cial Weibo and WeChat platforms, and 4) printing advertisements in newspapers.

Inclusion criteria
Participants who meet all of the following criteria can be enrolled:

Exclusion criteria
Participants who meet any of the following conditions will be excluded: 1) A history of trauma and surgery in the knee or diseases such as cancer, tuberculosis, and osteomyelitis in the knee 2) Severe liver and kidney dysfunction, severe cardiovascular disease, diabetes and mental illness that could affect massage therapy 3) Inability to communicate or provide self-care due to mental or psychological disorders associated with severe mental illness or dementia 4) Physical pain caused by other diseases 5) Partial damage in the acute phase or local skin damage 6) Duration of knee joint pain or stiffness episodes less than 0.5 hours. 7) Pain attacks for less than 3 months or more than 2 years 8) Failure to adhere to treatment at the prescribed time 9) Use of massage or other related treatments within one week prior to entering the study 10) Participation in other clinical trials 11) Use of any other treatment (drug or non-drug) during the trial.

Dropout criteria
Participants who do not complete the clinical programme for the following reasons should be considered as having withdrawn from the study: 1) Patient withdrawal (poor e cacy or adverse reactions) 2) Lost to follow-up 3) Patient removal by a researcher (poor compliance, complications or serious adverse events).
The test communicator will have in-depth communication with a patient who wants to withdraw from the trial because of low effectiveness or a lack of time to promote retention.

Randomization
In this trial, the Yueyang Hospital Science and Technology Department will generate a random sequence using a random number generator (SPSS 21.0, SPSS Inc., Chicago, IL, USA). Random numbers will be placed in opaque envelopes, numbered sequentially, and sent to the therapists. The envelopes will be opened in numerical order to determine the assignment of participants during screening. After completing the baseline questionnaire, the therapist perform the physical examination and apply the exclusion criteria. Only after that has been completed will the envelope be opened, so that the study administrator, therapist, and the study participant are unaware of the group assignment until all baseline data have been collected.

Blinding
As an open-label clinical trial, both clinicians and patients will know which treatment is being administered. The clinical e cacy assessment will be conducted by the clinical assessor on the phone, and the assessor will conceal the treatment. During the data collection and analysis phase, clinical researchers, evaluators, and statisticians will not share research information with each other.

Interventions
Participants will receive 12 treatments within 4 weeks. The treatment will be carried out by a senior therapist who has studied acupuncture and massage and has been licensed for more than 10 years.
Participants will be asked to rest for 15 minutes before treatment. A constant room temperature of 23 to 25°C will be maintained to ensure that the patient is comfortable and relaxed during treatment. The intervention will be performed twice a week. Each treatment will last for 20 minutes. Knee function will be assessed at baseline and 2, 4, 8, and 12 weeks after the baseline assessment.

Tuina manipulation group
In this part of the study, the therapist will use a two-step approach to relieve pain around the knee and improve knee dysfunction. The therapist will release the soft tissue around the knee by inducing general relaxation while addressing speci c structural problems. The clinician will determine which approach will be most helpful for alleviating pain due to KOA and for promoting air transport (activation of blood circulation according to traditional Chinese medicine (TCM) theory). The speci c scheme used will be as follows.
The speci c method of massage implemented will be based on conventional acupoint tuina manipulation around the knee joint of the patient. The pain points will be identi ed based on the description provided by the patient and palpation of the area by the practitioner and then treated with Tuina manipulation, 20 minutes each time, 3 times a week for 4 weeks.
Step one: Check the pain points around the knee joint With the patient in a supine position, the examiner will identify pain points by holding the upper part of the patient's knee joint in one hand while sliding and pressing the thumb of the other hand along the inner knee, the outer knee eyes, any adductor nodules, the medial and lateral joint spaces, the femur, the tibia, and the periorbital and popliteal fossa.
To check the popliteal fossa, the patient will be placed in the prone position with the knee bent, and the practitioner will slide and press the thumb along the centre, the lower portion, the outside and the upper portion of the popliteal fossa. When inspecting the inner side of the apex patellae, the examiner will push the patella downwards with the left hand so that the apex patellae is lifted up and then press the facies medialis of the apex patellae with the ngertips of the right hand, with the thumb facing upward.
Step two: Tuina local manipulation The patient will be placed in the supine position. The doctor will stand on the affected side, apply Tuina manipulation to a regular acupuncture point around the patient's knee joint and then press and rub the patient's pain point. The pain point is generally considered to re ect the underlying condition and is often used for pain relief. The frequency of the press and rub manipulation should be 120-140 times/min. The doctors press the acupoints, including Futu (ST32), Liangqiu (ST34), Dubi (ST35), Xiguan (LR7), Neixiyan (EX-LE4), Xiyan (EX-LE5) and Ashi points, with the thumb to achieve Deqi sensation. Then, the patient will be placed in the prone position, and the doctor will repeat the above operation.

Health care education group
In the health care education group, patients with KOA will participate in a 45-minute health care educational seminar three times a week for 4 weeks. The programme will include a 30-minute lecture and a 15-minute discussion. The content will involve the diagnosis of KOA, healthy diet, pain management, physical and mental health, and healthy lifestyle factors.
Each participant will be given an introduction about the underlying mechanisms and predisposing factors of KOA. Interventionists will draw up speci cation of daily routines for participants, for example: Keep the knees warm; prevent obesity; maintain a bene cial posture and abandon bad exercise posture; avoid fatigue, excessive exercise or weight lifting; choose appropriate exercise; maintain a good mental state, and inform them that chronic pain has a long process.
Allowance for concurrent treatment of patients All KOA treatments are prohibited during the trial, including oral muscle relaxants, anaesthetics, analgesics, surgery, drug injections, acupuncture and physical therapy. They may receive any treatment that is not related to knee pain. Any changes to concurrent treatments will be recorded at each visit.

Primary outcome measurements KOOS
The knee injury and osteoarthritis outcome score (KOOS) has been widely used in research and has been shown to be e cient and reliable when use to measure knee pain [26]. We used the KOOS (0-100 scale, 0 for extreme knee problems, 100 for no problems) as our primary outcome at baseline and at the end of the intervention to measure the clinical results. The KOOS scale is an extension of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), a self-administered 42-question survey covering ve patient-related domains: pain, other disease-speci c symptoms, daily functional activities, motor and recreational functions, and knee-related quality of life. The scale is an effective, reliable, and responsive method of measuring results with good test reliability and structural validity for total joint replacement. We used other KOOS subscale scores as the primary outcome. We will measure the outcome at these ve time points (baseline, two, four, eight and twelve weeks) and compare. In this trial, we will compare the mean and standard deviation of the two groups at each time point. then, the change in values from baseline until the time point will be included in the comparison.

MPQ
The McGill pain questionnaire (MPQ) is a very common scale that measures and evaluates various pain [27][28]. The MPQ contains 4 groups of 20 pain descriptions. Each group of words is arranged in increasing order with regard to the severity of pain. Dimensions one to ten are sensory, 11 to 15 are affective components of pain, 16 are pain evaluations, and 17 to 20 are miscellaneous. The test subject selects a word to indicate the degree of pain in each group (they must select a word).

SF-36
The SF-36 questionnaire assesses the association between health-related quality of life and various factors [29]. eight other aspects. For each domain, the score ranges from 0 to 100, with higher scores re ecting a better quality of life. The SF-36 will be evaluated using repeated longitudinal analysis. Same as primary outcome, secondary outcome will be measured at these ve different time points (baseline, two, four, eight and twelve weeks).the method of aggregation and the metric for comparison is the same as KOOS.

Safety evaluation
Treatment safety will be assessed by researchers monitoring patient symptoms before and after treatment. In this trial, the following events will be de ned as adverse events (AEs): (1) severe pain; (2) syncope; (3) deterioration in walking function; (4) hospitalization; and (5) life-threatening changes, all of which will be carefully recorded in the case report form. In the case of any AEs, whether or not they are related to the treatment, treatment will be stopped immediately. The patient will be treated, and the AEs will be reported to the responsible unit and the ethics committee to determine if the patient should withdraw from treatment.

Follow-up
To assess the short-term e cacy, long-term e cacy, and safety of the interventions, we will track participants for 2 months after the trial. During the 2 months of unsupervised follow-up, no participants will receive special treatment other than regular knee care. At weeks 8 and 12, the evaluators will investigate the recurrence of knee pain over the telephone. Patients will also be able inform the assessor of their clinical signs and mood changes by email, text message or WeChat at the relevant time. During the follow-up period after the trial, any health issues will be assessed, and the committee will be noti ed.

Data collection and monitoring
The assessor will record the clinical results in the case report form (CRF), including the assessment of the treatment effect, the changes in physiological function and details of AEs. The completed CRFs will be reviewed by a four-person steering committee composed of the heads of the four research centres. Then, the CRFs will be submitted to the data stewards, who are independent of the research team and do not know the group allocation of the participants. The stewards will enter the data into an Excel database. All data administrators will have data analysis quali cations and uni ed training. To ensure the accuracy of the data, the two data administrators will independently enter the information and check the accuracy of the entered data. If there is a problem with the information on a CRF, the data administrator will complete the requisite form, and then the data administrator can modify the data according to any revisions provided by the steering committee. After con rming that the database is correct, the steering committee, data stewards, and statisticians will lock the database.

Health economics
All costs associated with this test will be recorded. These costs mainly include medical costs for the direct treatment of KOA, such as inpatient bed fees, outpatient registration fees, Tuina treatment fees, daily care and testing fees, and in addition, the cost of treatment for any AE will be recorded and included in the health cost evaluations.

Sample size calculation
The expected e cacy rate of Tuina manipulation is 70%, and the e cacy rate of health care education based on the literature is 45%. The sample size was calculated with a signi cance level of 0.05 and power of 0.90, with consideration of a maximum dropout tolerance of 10%. Finally, 170 patients are needed for the trial, with 85 patients in each group.

Data analysis
Demographic and baseline data will be analysed with standard descriptive statistics. Between-group differences will be tested using repeated measures analysis of variance. The acceptable signi cance level for all analyses will be P<0.05. The complete analysis set, including dropouts, will be analyzed through an intention-to-treat (ITT) population analysis. Statisticians who are independent of the research team and are blinded to group allocation will perform the data analysis. Data analysis will be performed with SPSS software (for Microsoft Windows® SPSS, Inc., SPSS 21.0 KO, Chicago, Illinois, USA).

Quality control
To improve the reliability of the results of this study, we designed quality control procedures with regard to the following aspects: (1) baseline homogeneity: this study limits the age of participants to between 40 and 65 years, and we will assess people with similar living habits; (2) Tuina manipulation: in this study, all researchers will be trained to fully understand the design and protocol of the trial, the use of the CRF, and the manipulation technique. To avoid operational differences, two experienced doctors who will have received specialized standardized training before the trial will perform all treatments in accordance with the standard operating procedures. In addition, during the treatment, except for that needed to identify pain points, communication between the therapist and the patient is prohibited; (3) Pain points: considering the importance of pain point identi cation for accurate treatment evaluation, the pain threshold at the tender points will be measured after each treatment to evaluate whether there is a signi cant difference in stimulus and pain perception after each treatment, which will help rule out the bias of using the pain points; (4) data collection and completeness: Yueyang Hospital of Integrated Traditional Chinese and Western Medicine Data Monitoring Team will be responsible for controlling bias and identifying problems in the project. Detailed records of changes will be kept. Meanwhile, a quali ed clinical trial expert will monitor the therapists after every course of treatment, and regular board meetings will be held to ensure the quality of the study process.

Discussion
Knee pain is one of the most challenging public health problems in the world. Given its prevalence, there is a need to identify effective methods of treating knee joint disease. Tuina is an important aspect of Chinese medicine and has contributed to the health of the Chinese people for more than a thousand years. A large body of scienti c evidence supports this treatment for knee pain, but its e cacy has not been determined. Therefore, it is important to provide strong evidence of the effectiveness of massage therapy for the treatment of KOA.
This trial is a comparative study of the e cacy and safety of massage (intervention) and health care education (control) for pain relief and functional recovery in KOA patients. For high-quality RCTs, proper control is critical. Both pain intensity and motion function are subjective, and we will use validated scales and questionnaires to assess these clinical outcomes. The KOOS will be used as the primary outcome, as it measures knee pain and the impairment of daily function. The secondary outcomes are the MPQ, which measures the intensity of pain in general and the emotional effect of pain, and the SF-36, which measures health-related changes in quality of life.
The A-Shi point in Chinese medicine refers to a point on the surface of the body. When it is lightly pressed, it will reproduce the speci c pain being treated. Its position indicates the exact location where the Qi and blood are blocked. Manipulation of the A-Shi point promotes the free movement of the Qi and improves blood circulation in the area, eliminating obstacles and relieving pain. Studies have shown that one mechanism by which Tuina acts is by activating the piezo electric channel as a physical stimulus, after which the membrane potential changes due to the ow of charged ions and generates an analgesic effect [30]. Strength training around the joint muscles and low-intensity aerobic exercise can relieve pain. Maintaining joint mobility and improving joint function can control the progression of KOA. Tuina is a treatment that uses manipulation by a doctor to achieve muscle stimulation and improve joint function.
The RCT design and methodological rigor of the trial will allow the collection of valuable high-quality data that can be used to assess the e cacy of Tuina manipulation and health care education for the treatment of KOA. The trial will provide a solid foundation for the clinical treatment of KOA and future research on Tuina and massage therapy.

Study limitation
One of the primary methodological di culties inherent to studies evaluating physical interventions is blinding of participants. In this study, Tuina manipulation and health care education are different forms of physical therapy, which hardly be blinded for the participants and therapists.Therefore, the quality of the blinding method should be monitored to control the questionnaire measurement credibility, which is composed of administrator, evaluators and the data analysts. Tasks of Shanghai Health Planning Committee (2018LP040). The funders had no role in the design of the study; analysis, collection, and interpretation of the data; or the writing of and decision to publish the manuscript. All this funding applies to this study.
Authors' contributions ZL and GG participated in the protocol writing and design of the trial, and LK drafted the manuscript. QZ will perform the Tuina treatments. WS recruited and screened quali ed participants in the outpatient department, and MF was responsible for the study. YC is responsible for coordinating the treatment of patient ZW, who participated in the design of the trial and helped prepare the manuscript. SX participated in evaluating treatment e cacy and collecting clinical data. XZ participated in the strict revision of the manuscript. SZ participated in evaluating treatment e cacy and collecting clinical data. CY participated in the strict proofreading of the manuscript. All authors have read and discussed the nal manuscript, and all approved the manuscript for publication.

Ethics approval and consent to participate
This study will be conducted in accordance with the guidelines and principles of the Declaration of Helsinki. The protocol was approved by the Ethics Review Committee of Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, which is a liated with Shanghai University of Traditional Chinese Medicine (reference 2019-097) and is registered with the Chinese Clinical Trial Registry (ID: ChiCTR-2000030154). Informed consent from all study participants will be obtained before clinical trial assessors begin collecting any data. All participants will provide their written consent. No one except investigators will have access to the nal data.

Availability of data and material
The datasets analysed in the whole study can be obtained from the corresponding authors upon reasonable request.