Efficacy, safety, and economic assessment of hominis placental pharmacopuncture for chronic temporomandibular disorder: a protocol for a multicentre randomised controlled trial

This study is a two-armed, parallel, multi-centre, randomised controlled study, and will be conducted at the Jaseng Hospital of Korean Medicine and Kyung Hee University Korean Medicine Hospital at Gangdong, located in South Korea. Prior to commencement, the study protocol has been approved by the institutional review board (IRB) at each centre (JASENG 2017–09–002-002, KHNMCOH 2019–08-002) and the Ministry of Food and Drug Safety (No. 31886). In addition, the study protocol has been registered at clinicaltrials.​gov (NCT04087005) and Clinical Research Information Service (CRIS) (KCT0004437), and the state of the research will be continually updated. Information about the research centres and the researchers can be found at the trial registration site.

Participants will be recruited by physician’s recommendation, posters placed in the hospitals, and adverts placed on the hospital homepages and the metro transport system. On the first visit, a researcher will explain the study to the participants and ask them to complete and submit an informed consent form (ICF). Next, an examiner will screen the participant. Participants will be enrolled according to the inclusion/exclusion criteria, will visit 15 times, and undergo surveys either through face-to-face, telephone, or online interviews, or by distribution and return. There will be 10 treatment sessions consisting of two sessions per week for 5 weeks. Thereafter, there will be follow-up visits at 6, 9, 13, and 25 weeks from baseline. On the second visit, participants will be randomly allocated to either the HPP group or the PT group and receive the corresponding treatment. All participant management and treatment processes will be performed according to the protocol. To check compliance, we will verify whether participants adhere to the treatment schedule specified in the protocol. For the HPP group, compliance will be assessed by checking the frequency and schedule of procedures; for the PT group, compliance will be assessed by checking the frequency and schedule of TENS procedures. A study flow chart is shown in Fig. 1.

An independent statistics expert will use nQuery Advisior 7.0 (or SAS 9.0 [SAS Institute Inc., Cary, NC, USA] or SPSS 21.0 [IBM Corp., Armonk, NY, USA]) to allocate participants into each group by block randomisation in a 1:1 ratio (41 persons/group). The results of random allocation will be sealed, such that they cannot be seen from outside, and sent to each centre, where they will be managed in a double-locked container. Before treatment, a random envelope will be opened in front of the enrolled participant, the random allocation numbers assigned to each group will be recorded in an electronic chart, and no changes will be allowed after allocation. Participant enrolment at each centre will proceed by competitive enrolment.

Because this study does not allow the participant or treatment provider to be blinded to their group, we will use a single-blinded design, where only the assessors are blinded. Participants will be assessed away from the treatment area, by an investigator who did not participate in treatment procedures and is blinded to the treatment group; efforts will be made to prevent the investigator from knowing the allocated group. The data analyst will not be blinded to two groups. The design is open label with only the outcome assessor being blinded, so unblinding will not occur in this study.

At both centres, treatment will be administered by a physician who has received prior training in a standardised protocol.

Although treatments that could affect TMD will be restricted during the treatment period (weeks 1–6), in cases where severe pain makes it unavoidable, participants may receive treatment after a subinvestigator has been informed; the treatment type and frequency will be recorded in detail in the case report form (CRF). In addition, treatments unrelated to TMD will be allowed on condition that a subinvestigator is informed of the relevant facts. No restrictions will be placed on other treatments during the follow-up period (after week 6), and regardless of allocation, both groups will receive education in the causes, prevention, treatment, and management of TMD, as well as self-stretching methods [20].

The specific participant schedule is shown in Table 1.

The trial may be stopped for a participant in the following circumstances:

Outcomes will be collected at different time points during a total of 25 weeks (Table 1), and the primary end point is 6 weeks after baseline.

This is an investigator-initiated clinical trial, and we have calculated the required sample size with reference to similar studies and our clinical experience. First, we set a significance level of α = 0.05 (two-tailed tests), a type 2 error (β) of 0.2, and a test power of 80%. Because there are no previous studies comparing pharmacopuncture with PT for the TMJ, we selected a moderate-to-high effect size for the comparison of mean VAS between the two groups based on our clinical experience, and used Cohen’s d = 0.65 [36]. Using these parameters, we calculated the sample size using G*Power 3.1.7. Based on an effect size = 0.65, the required sample size was estimated to be 39 persons/group (total 78 persons). However, as we plan to use an analysis of covariance (ANCOVA) correcting for the outcome at baseline for our main analysis, we calculated the sample size assuming the correlation between baseline and primary endpoint to be 0.4 [(1–0.4*0.4)*78], and found that the minimum required number of participants was a total of 65 persons [37]. Thus, to account for a dropout rate of 20%, we calculated that the number of participants we need to recruit is a total of 82 persons (41 persons/group).

Patients will be involved in recruitment and conduct of the study. We will keep in touch with the patients from recruitment to 6 months after treatment. The patients will be able to contact researchers at any time to get information they need.

We will use e-CRFs via the Internet-based Clinical Research and Trial (iCReaT) management system, which is operated by the Korea Centers for Disease Control and Prevention. An investigator from the coordinating centre will distribute a standard operating procedure (SOP) to be used in trial processes such as CRF completion and data input and will provide instruction for assessors and investigators at each centre. CRF data will use double data entry, and data entry will be performed by each participating centre and the coordinating centre. After cross-checking the input data, access to the data will be blocked from all investigators except the data coordinator.

As an RCT of chronic TMD patients, this study will compare the difference within and between groups for data on efficacy, safety, and economics. The main analysis will be the ITT analysis, and a PP analysis will also be performed. The demographic results, such as sex, age, occupation, and disease history, and the treatment expectancy will be assessed by group; continuous variables will be presented as the mean ± standard deviation (SD) and categorical variables will be presented as n (%). Differences will be tested using chi-square tests and independent t tests. Efficacy outcomes are the changes in continuous outcomes (VAS, NRS, K-BDI, JFLS, EQ5D-5 L, SF-12) at each time point relative to baseline. The primary outcome is the change in VAS from baseline to the primary end point (week 6) in each group. For safety outcomes, we will compare the incidence of AEs between groups; after graphing all the reported ARs in participants who have received treatment at least once, we will calculate the proportion of patients who have experienced AEs in each group, and compare these using chi-square tests and Fisher’s exact test. To compare the HPP and usual care groups, for continuous variables, we will perform an ANCOVA adjusted for the baseline outcome. For categorical variables, we will use chi-square tests of Fisher’s exact test, and to analyse within-group differences between time points, we will use paired t tests. Additionally, we will perform an ANCOVA adjusted for baseline outcomes and covariant factors that show a statistically significant difference between the treatment groups at baseline, and we will perform a repeated measures analysis of variance (RM ANOVA) to test differences in the changes across each visit. We will perform subgroup analyses according to RDC/TMD diagnosis (Group Ia, Ib), mouth opening, and K-BDI score. Missing values will be handled in a blinded state, mainly using multiple imputation (MI), last observation carried forward (LOCF) will additionally be used, and sensitivity analyses will be performed. All statistical analysis will be performed using SAS version 9.4 statistical package (SAS Institute, Cary, NC, USA), and statistical significance will be defined as p < 0.05.

We will perform an economic analysis to investigate the cost-effectiveness of HPP compared to PT. The economic analysis can be performed from an insurer’s perspective, healthcare system perspective, and societal perspective; we have selected a societal perspective [38]. We will perform a cost-utility analysis including medical costs and non-medical costs, and costs associated with disease-related loss of production. For the final analysis results, we will present the incremental cost-effectiveness ratio (ICER) using the major economic assessment index of costs per quality-adjusted life year (QALY) gained [39]. ICER is an index that indicates the additional costs required per unit increase in effect, compared to an alternative. In addition, we will separately present the total effect (efficacy) and total costs of each treatment. For effect units, we will use QALY, which will be calculated using the quality of life derived by EQ-5D-5 L and using the area under the curve method with receiver operating characteristic (ROC) curves. For cost units we will convert all costs to 2019 South Korean Won (KRW), and apply 5% discount in accordance with the economic analysis guidelines of the Health Insurance Review and Assessment Service [40]. The treatment costs incurred by the clinical trial will be calculated by combining the treatment frequency and the unit costs; the unit costs will be calculated using the health insurance costs and the centre’s costs data. The analysis period is set to the entire clinical trial period (6 months), including the follow-up period. If estimates for the subsequent period are required, we will perform secondary analyses, such as extrapolating the costs and effects through a regression model or performing decision modelling analysis. In the baseline analysis, we will use representative values (mean, etc.) for the parameters used in the trial. To examine the uncertainty of the parameters used in the model, we will perform a sensitivity analysis; specifically, we will perform a probabilistic sensitivity analysis in which we estimate the prior distribution of the parameters, randomly extract the parameter values from these distributions, and use these results to perform a cost-effectiveness analysis.

Monitoring will be performed by the monitoring coordinator at the coordinating centre. During monitoring, CRFs will be compared with evidentiary documentation, and participants’ safety data will be reviewed to verify the completeness of safety and research data. The monitoring will be conducted once every 2 months at two sites. The Trial Steering Group will meet to review conduct throughout the trial period once every 2 months. The Data Monitoring Committee was not organised as no severe AE regarding HPP or PT previously reported are expected to occur. The interim analysis will not be done in this study.

The principal investigator will decide the termination of the study; at any time during the trial period, if a participant is exposed to clear, unexpected risk or unacceptable risk, or if at least 25% of all participants experience severe AEs thought to be related to JHG002, the trial can be terminated. Severe AEs that might justify terminating the trial include local infection and severe progressive neurological deficits [41, 42].

In the event of direct harm related to this study, participants will receive the appropriate medical treatment as determined by a subinvestigator and will be compensated according to the terms of the trial insurance, which will be designated beforehand.

The results of this trial will be registered on clinicaltrials.​gov and CRIS, and will be published in an academic journal. After the publication of this article, important changes to the study protocol and any other changes will be updated regularly on the trial registration site.

The current protocol is version 2.3 (2 February 2020). First participant was enrolled on 17 October, 2019. We anticipate that the recruitment will be completed by October 2020.