Study design

This is a randomised, assessor-blinded, controlled trial of a 12-week multimodal, optimised non-surgical treatment (the MEDIC-treatment) with 12-month follow-up. Measurements will be taken at baseline, and after 12, 26 and 52 weeks. The study will conform to CONSORT guidelines for reporting parallel group randomised trials.34

Participants

Patients with a diagnosis of symptomatic and radiographic KOA considered ineligible for TKA will be included in this study.

We will recruit 100 patients meeting all of the following inclusion criteria:

1. Referred from primary care to an orthopaedic surgeon in a public hospital in The North Denmark Region for evaluation of the need for TKA;

2. Considered ineligible for a TKA by the surgeon;

3. Diagnosed with KOA using standing, weight-bearing knee radiographs (Kellgren-Lawrence score ≥1 on the original scale35 ,36);

4. Aged ≥18 years;

5. KOOS₄≤75 (the average score for four of the five Knee Injury and Osteoarthritis Outcome Score subscales covering pain, symptoms, activities of daily living and quality of life).37 ,38

The exclusion criteria are any of the following:

1. Previous ipsilateral knee arthroplasty;

2. Rheumatoid arthritis;

3. Mean pain the previous week >60 mm on a 100 mm Visual Analogue Scale (VAS);

4. Possible pregnancy or planning pregnancy;

5. Inability to comply with the protocol;

6. Inadequacy in written and spoken Danish.

Procedure

The overall structure of this study is outlined in figure 1. People in need of evaluation for TKA in The North Denmark Region are referred by their general practitioner to an orthopaedic surgeon at the outpatient clinics at Frederikshavn and Farsoe, Department of Orthopaedic Surgery, who specialises in TKAs. A standardised weight-bearing anteroposterior knee x-ray is obtained on arrival.8

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flowchart.

The orthopaedic surgeon will assess potential participants against inclusion criteria 1–4 and exclusion criteria 1–2 and a research health worker assigned to the project will assess them against inclusion criterion 5 and exclusion criteria 3–6. The research health worker will obtain informed written consent from patients who are eligible and willing to participate after they have received written and verbal information. After the baseline measures are obtained, patients who agree to participate in the RCT will be assigned to one of two treatments: (1) the MEDIC-treatment or (2) usual care.

Participants will be reassessed 3 months after randomisation (12-week follow-up) and again after 6 months (26 weeks) and 12 months (52 weeks). In addition, there will be long-term follow-ups 2, 5 and 10 years after randomisation. All current medication use, co-morbidities and co-interventions will be recorded at all follow-ups.

Patients declining to participate will be asked to fill out the Knee Injury and Osteoarthritis Outcome Score (KOOS) and report age and gender anonymously so as to improve the selection bias analysis.

Randomisation procedure and concealment of allocation

Before the initiation of the trial, the schedule for randomisation will be randomly generated in permuted blocks using a computer. To control for variation in patient characteristics between the two clinics, the randomisation will be stratified according to the clinic (Frederikshavn or Farsoe). The allocation numbers will be put in concealed, opaque C5 envelopes to conceal the outcomes of the randomisation. In blocks of eight, these envelopes will be placed in consecutively numbered opaque larger envelopes (seven larger envelopes in total for each clinic). A staff member, independent of this study, will prepare the envelopes. These will only be accessible by one research assistant at each of the respective clinics. A smaller envelope from the numbered larger envelopes will be opened by the research assistant following the informed consent and completion of the baseline measures, after which the allocation will be revealed to the participant. The smaller envelopes of the second larger envelope will be added, when only two smaller envelopes are left in the first of the numbered larger envelopes. The last two of the smaller envelopes will be added, when there are six smaller envelopes left in the sixth of the seven numbered larger envelopes at each clinic.

Blinding

The outcome assessor will be blinded to group allocation, unaffiliated with the treatment sites and not involved in providing the interventions. The participants, the project physiotherapist and the project dietician delivering the interventions cannot be blinded. The statistician performing the statistical analyses will be blinded.

Interventions

Baseline data

The radiographic severity of KOA will be assessed from the baseline x-ray using the Kellgren and Lawrence grading system.35 Furthermore, the following will be obtained by questionnaire: gender, age, nationality, height, alcohol intake, smoking habits, duration of KOA symptoms, previous injuries, treatment and use of medication regarding the affected knee, co-morbidities, physical activity and exercise levels, preferred treatment, previous arthroplasties, living arrangement, satisfaction with self-management of pain, education level and employment status, income, home help and the short version of the Hip/Knee Osteoarthritis Decision Quality Instrument (HK-DQI).45 After the randomisation, the participants will be asked to rate their belief in the effect of their received treatment on pain, function and quality of life.

Primary outcome measure

The primary outcome will be the change from baseline to 12 months in KOOS₄, with scores ranging from 0 (worst) to 100 (best; table 1).

Secondary outcome measures

A number of other patient-reported outcome measures will be taken (table 1): The five individual subscales of KOOS (the fifth scale being difficulty in sports and recreational activities),37 ,38 the EQ-5D-3L,46 and pain intensity measured on a 100 mm VAS with terminal descriptors of ‘no pain’ and ‘worst pain possible’ in the following situations: at rest, after 30 min. of walking, and worst pain and least pain in the previous 24 h. The participants will be asked to shade regions on a region-divided body chart where they have had pain during the previous 24 h. Furthermore, self-efficacy in relation to reduction in pain and increase in function and quality of life using a 100 mm VAS with terminal descriptors of ‘very unsure’ and ‘very sure’ will be used in this study.

Several objective measures will be assessed (table 1). The outcome assessor will be the same as in another trial involving KOA with the same objective measure42—someone who has undergone a period of supervised training in the use of the objective measures to optimise the reliability of the measurements. As measures of the functional performance of the participants, the Timed Up and Go47 and 20 mr walk test48 will be taken in this study. In addition, percentage change in weight from baseline to follow-up will be assessed. The measurement of weight will be performed barefooted on the same scales (seca 813, seca gmbh & co. kg., Hamburg, Germany) and at the same time of day.

Maximum isometric muscle strength will be measured in knee flexion and knee extension bilaterally in a make test using a handheld dynamometer (HHD), the Powertrack II Commander (JTech Medical Industries, Salt Lake City, Utah, USA). The procedure of this objective measure has been presented previously42 for both knee extension and knee flexion. The participant will be given a 30 s rest between each measurement.

To assess pressure pain thresholds (PPTs), a hand-held pressure algometer (Algometer Type II, Somedic AB, Hoerby, Sweden) with a 1 cm² probe will be used. The probe will be placed perpendicular to the skin and force applied at a constant rate of 30 kPa/s until the participant defines the pressure as pain and presses a button. PPTs will be assessed bilaterally at four sites in relation to bony landmarks in the peripatellar region: (1) 3 cm medial to the midpoint of the medial edge of the patella, (2) 2 cm proximal to the superior edge of the patella, (3) 3 cm lateral to the midpoint of the lateral edge of the patella and (4) at the centre of the patella. Furthermore, PPTs will be assessed at two control sites: (5) one on m. tibialis anterior (5 cm distal to the tibial tuberosity) and (6) one on m. extensor carpi radialis longus (5 cm distal to the lateral epicondyle of the humerus; figure 2). Before starting the measurement, the test is performed once or more on the dorsal aspects of the hand to make sure that the participant has understood the test procedure. A PPT will be obtained twice from each site and the mean of the two measurements will be used in the statistical analysis.49 ,50 The participant will be asked about the location and type of their knee pain using the interviewer-administered questionnaire Knee Pain Map, which has been found to be reliable for this purpose.51

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Pressure pain thresholds measurement sites.

The test setup for both isometric muscle strength and PPTs will be investigated in a test–retest reliability study on 20 participants.

Postural balance will be assessed using an instrumented force platform (Good Balance, Metitur Oy, Jyvaskyla, Finland), measuring the centre of pressure (COP) excursion body sway of the participants (100Hz). Participants will be asked to stand barefooted in a comfortable position with their feet positioned side-by-side (about a shoulder width apart). Further, they will be given the standardised cue ‘Stand as still as possible’ with their arms folded across their chest while focusing their eyes on a visual target positioned 3 m away while being tested. Four different sensory conditions will be applied to explore the contribution of different conditions to the postural control in these patients: (1) standing on a firm surface with eyes open, (2) standing on a firm surface with eyes closed, (3) standing on a soft surface (foam) with eyes open and (4) standing on a soft surface (foam) with eyes closed. Each condition will last 1 min and be repeated three times in a random order. During all measurements, an experienced experimenter will be standing next to the patient in case they lose their balance. Between each trial, participants will have the option of a rest if needed. Bipedal static COP measures have previously been proven to be a reliable tool for investigating postural balance.52

Other measures

A number of other measures will be obtained in this study (table 1). In the group allocated to MEDIC-treatment, compliance with exercise will be monitored by the physiotherapist during the intervention period as the total number of exercise sessions completed out of the planned 24 sessions (2 sessions a week for 12 weeks). Good compliance is defined as participation in 75% or more of the exercise sessions, medium compliance as participation in 50–74% of the sessions and poor compliance as participation in less than 50% of the sessions. The participants in the group allocated to the MEDIC-treatment will be requested to record their weekly exercise until the long-term follow-up 2 years after randomisation to investigate the long-term compliance. Use of medication in the group allocated to the MEDIC-treatment will be recorded in a medication diary, which will be examined as part of the follow-up. At each follow-up, all participants will be asked to report their compliance with what they have learned in this study using a five-point scale (never, every month, every week, everyday, all the time). All participants will also be asked to rate their satisfaction with the treatment to date on a five-point Likert scale at each follow-up.

Adverse and seriously adverse events will be registered in two ways and divided into index knee or sites other than index knee. The project physiotherapist will record any adverse events that the participant experiences or tells them about. At all follow-ups, the assessor will use open-probe questioning to assess adverse events in all participants.

Information on direct healthcare costs and direct non-healthcare costs will be collected retrospectively and at all follow-ups. Direct healthcare costs will include cost of the MEDIC-treatment and compliance with the treatment. These elements will be valued using published Danish prices for medical costs. Direct non-healthcare costs will include sick pay (if relevant), change in home help, number of days lost from work and shorter working hours.

Sample size

It is expected that the group allocated to MEDIC-treatment will improve 10 points more than the group allocated to usual care based on the primary outcome KOOS₄ at the main endpoint after 12 months. With a common between-subject SD of 14, sample size calculations show that 41 participants in each group are required to detect a statistical difference (power of 90% and significance level at 0.05 (two-sided)). Therefore, a total of 100 participants will be included to allow for crossovers and missing data (drop-out rate will be set to 20%). The minimal clinically important difference between patients having optimised non-surgical treatment in patients not considered eligible for TKR is not known. Some studies have applied an improvement of 15% as a cut-off to determine number needed to treat (NNT).53 We will closely follow the ongoing discussion within this area and apply a cut-off supported by current knowledge at the time of analysis.

Statistical analysis

The primary outcome measure will be the KOOS₄ score at the 12-month follow-up. The statistical analysis will follow an intention-to-treat approach and be based on a Generalised Estimating Equations regression model for the KOOS₄ score at all follow-ups to take the repeated measurements on the patients into account. The following aspects will be incorporated in the model: the effects of treatment, follow-up time, treatment-by-follow-up time interaction, and KOOS₄ score at baseline. Secondary analyses will assess heterogeneity between sites and a within-group analysis will be done to investigate if treatment compliance is associated with the change in KOOS₄. Furthermore, an analysis of NNT will be performed. NNT estimates the number of people who would need to go through the MEDIC-treatment for one person to have a clinically meaningful improvement in KOOS₄ from baseline to the follow-ups.

Ethical considerations

The protocol is designed to conform to the principles of the Declaration of Helsinki and has been approved by the local Ethics Committee of The North Denmark Region (N-20110085). The participants in this study will be allocated to either usual care or the MEDIC-treatment, which means that the treatment they receive will be either equivalent to, or superior to, the treatment that they would receive if they did not participate in this study.

Timelines and dissemination plans

Approval from The Danish Data Protection Agency was given in January 2012 while ethics approval was obtained from The North Denmark Region in February 2012. Recruitment and training of the involved project physiotherapist and dietician were undertaken in July and August 2011 and recruitment of participants started in April 2012.

All participants are expected to have completed the 12-month follow-up by April 2014. The statistical analysis will start immediately after the data monitoring is completed. Publications will be ready for submission in the summer of 2014.