Design and setting
This study is a single-centered, randomized, double-blind, and placebo-controlled trial to evaluate GFW's efficacy and safety for PD. The survey designed will comply with the Declaration of Helsinki (2013 version) and Consolidated Standards of Reporting Trials guidelines Extension for Chinese Herbal Medicine Formulas 2017 (CONSORT CHM Formula) [1]. Figure 1 provides the flow diagram. This study's protocol is prepared based on the Standardized Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines [21]. The Ethical Committee of The Second People’s Hospital of Yichang has approved the protocol (Approval No.2020001), which was registered on the Chinese Clinical Trials Registry in 2020 (Registration No. ChiCTR2000034118.).
The trial will be conducted at the Second People’s Hospital of Yichang. Participants will be required to complete a screening test, and eligible patients will enter a baseline period of one week, during which they will be asked to complete a dysmenorrhea diary and record-related symptom scores. Subjects meeting the including criteria will be randomly allocated to either the GFW group or the placebo control group. Participants in both groups will receive treatment for three menstrual cycles and three months of follow-up through telephone calls or social media. Patient enrollment will start in late October 2020 and is expected to end in December 2021 ( Flowchart of the study design see Fig. 1 and the assessment schedule see table.1).

Participants
Inclusion criteria
Participants meeting the following criteria will be included: (1) nulliparous women of 18 to 30 years old; (2) meet the diagnostic criteria of PD by Chinese Obstetrics And Gynecology(3rd edition); (3) accord with the TCM diagnosis of “heat-burning blood-stasis syndrome”; (4) pain symptoms last from 6 months to 15 years; (5) a VAS score of menstrual pain ≥ 40 mm at least for three consecutive menstrual cycles; (6) signature of an informed consent form.
Exclusion criteria
Participants would be excluded if they meet any of the following criteria: (1) unable to complete or comply with the study; (2) imaging examination suspects secondary dysmenorrhea; (3) pregnant during the period of the trial; (4) unable to cooperate with the questionnaire due to intellectual or mental disorder; (5) history or presence of clinically relevant severe cardiovascular and cerebrovascular diseases, liver and kidney dysfunction, mental illness, or life-threatening conditions; (6) treatment with analgesics, antidepressants, or psychotropic drugs for PD in the past month; (7) are participating in the other clinical trials of drugs.
Potentially eligible participants will be screened by the research staff through physical examination and clinical tests. All participants will be required to sign an informed consent form before enrollment. Patients’ information will be kept confidential, and the patient has the right to withdraw it at any time without prejudice.
Interventions
Participants in the GFW group receive GFW (herb granule, 30gper packet) treatment. It has the effect of clearing heat and dispelling dampness, blood circulation and removing blood stasis and clearing away heat and blood. It consists of five natural herbs: Guizhi (Cinnamon Twig) 6g, Fuling (Poria Cocos) 6g, Mudanpi (Cortex Moutan) 6g, Chishao (Radix Paeoniae Rubra) 6g, Taoren (Peach Kernel) 6g [22].
Participants in the control group receive GFW simulated preparation composed of Maiya (Hordei Fructus Germinatus), edible and medicinal. The dosage form, color, appearance, and packaging of the GFW fake preparation are consistent with GFW in the treatment group [23].
Both GFW formula and GFW simulated preparation are granules provided by Beijing Tcmages Pharmaceutical Co. Ltd. (Beijing, China) and distributed by the Second People’s Hospital of Yichang. The production process will be subject to quality control to ensure that the packaging has consistent purity, microbial content, weight, and other physical characteristics.
Patients in the two groups will be given oral medication every 12 hours a day, 30g each time, from 7 days before menstruation to 3 days after menstruation for three menstrual cycles. Participants were allowed to use NSAIDs such as ibuprofen if required, and the outcome assessment is not scheduled in the next 48 hours [24]. The use of all drugs should be documented in the case report form (CRF). Medicine will be kept sealed in a cool, dry, and secure locked place, and patients will be asked to record menstrual pain and acute medication use in a dysmenorrhea diary throughout the trial. Besides, if patients have intolerable symptoms, the attending doctor will treat the patient according to the actual situation and estimate whether adverse events are related to the study drugs; if so, they will decide to terminate the experiment.
Outcomes and follow-up
Primary outcome
The primary outcome will be menstrual pain intensity measured by the visual analog scale (VAS) scores. Participants will be asked to indicate a perception of pain intensity (most commonly) scored from 10 to 100 (0, no pain; 100, maximum) [25].
Secondary outcomes
The secondary outcomes include the Cox Menstrual Symptom Scale (CMSS), the Self-rating Depression Scale (SDS), the Self-rating Anxiety Scale (SAS), the 12-item Short-Form health survey (SF-12), and the dysmenorrhea diarymenstrual pain.
Safety assessments
Safety is assessed by adverse events and serious adverse events.
The primary and secondary outcomes will be recorded at baseline, treatment period and the follow-up periods. Safety assessments will be detected at the treatment period and the follow-up periods.
The assessments will be performed in a separate space in the outpatient department. All of the outcome assessors will be trained in conducting interviews and performing measurements before the study. The following measures will be taken to optimize participants' retention: free examination and treatment will be provided during the survey; appointments will be arranged flexibly to suit the patients' schedule [26].
The participants will be contacted monthly by telephone or social media to monitor their intake of the granules. If someone has taken less than 80% of the medicine, they would be considered non-compliant [27].
Randomization and allocation concealment
Eligible participants will be randomly assigned to either the treatment group or the control group at a 1:1 ratio after the screening visit [22]. Random numbers will be generated by a researcher who is not practicing throughout the study, using the SPSS Statistics for Windowa, Version 22.0 (SPSSInc., Chicago, III., USA). For allocation concealment, the name of each group (A or B), prepared by another researcher (not involved in the study), will be written on a card and placed in opaque envelopes labeled with sequential numbers, according to the sample size. Then the envelopes are sequentially opened for every new participant to determine the group. The envelopes will be placed numerically in a safe place until the study is completed.
Blinding
In this double-blind trial, none of the participants, clinical practitioners, the outcome evaluators, the data manager, and statistician will be aware of group assignments until the end of the study period. The placebo granule will be matched to the GFW herb granule in color, taste, smell, and outer packaging. Elimination of blindness will only be considered in cases of serious medical emergencies.
Sample size
Assuming a difference of 1cm on the VAS as the minimum clinically relevant difference and SDs was 1.67, consistent with those observed in previous similar studies overall 2-tailed 0.05 level significance 90% power [29–30]. The ratio between the two groups at 1:1, a minimum sample size of 116 patients (58 patients in each group) is needed. A minimum sample size of 128 patients (64 patients per arm) at the baseline will be required to allow for a maximum dropout tolerance of 10%.
Data collection and management
Data will be collected from the dysmenorrhea daily of symptoms recorded by participants, the CRF tables, and TCM syndrome scores recorded by the trial team [31]. All patients will be invited to answer the data administrator's self-administered electronic questionnaire [32]. All filled questionnaires will be checked for completeness by data collectors, and questionnaires with missing items will be returned to the patient in time to maximize the richness of the information and minimize recall bias [32, 33]. Regarding the recruitment rate, the study screening log will record the reasons for unqualified and non-recruitment patients. Retention and reasons for the withdrawal of participants throughout the trial will be documented [34]. The outcome measures will be described according to Fig. 1.
Qualified sites, and well-trained assessors are crucial factors that ensure the quality of a clinical trial [35, 36]. An independent data monitoring committee will be composed of two expert clinicians with experience in conducting and monitoring clinical trials [37] and they will check the data routinely. Any incomplete data will be coded as unknown, missing, or not applicable. Confidentiality of patient data will be maintained throughout the study. Trial data will be accessible only to the investigator team [38].
Data analysis
At the end of all measurements' data collection, statistical analyses will be performed by SPSS Statistics for Windowa, Version 22.0 (SPSSInc., Chicago, III., USA) [39]. we will descriptively summarise patient characteristics, outcome variables and the adverse events. Continuous variables, such as age and disease course, will be presented as mean ± standard deviation (SD) and analysed by independent-sample T test. Quantitative variables such as baseline variables, sociodemographic data will be analyzed using χ2 tests or Mann-Whitney U tests. The repeated measurement variables, such as the primary outcome and second outcome, are analyzed by Repeated ANOVA or generalized estimation equations (GEE).
Both the full analysis set and per-protocol analysis set will be used simultaneously to evaluate the effect and safety of GFW for PD. The last-observation-carried-forward method will be used for missing data [40]. Statistical significance will be set at P < 0.05.