| Background and objectiveBlinding method and patient's preference are two key problems in the clinical trials of non-drug therapy.To resolve these problems,many foreign researchers investigate the feasibility of partially randomized patient preference trials(PRPP).PRPP models has been widely used in non-drag therapy of clinical research.But there is no research investigating the feasibility of PRPP model in non-drug therapy of Traditional Chinese Medicine(TCM),and no completed PRPP research of TCM in China.To investigate the feasibility of a innovative clinical trial design(Partially Randomized Patients Preference model,PRPP)in non-drug therapy of traditional Chinese medicine(TCM).MethodsPart one:Bibliometrics of PRPP trialWe conducted literature search from 5 electronic databases(China Natinal KnowledgeInfrastructure,Chinese Dissertation Database,Chinese VIP information,PubMed and the Cochrane library)until 31st December 2017.Completed PRPP trials on any conditions were included.The design of data extraction table and extract data of included trials was conducted by Excel software.Qualitative research was conducted on PRPP trials which met the included criteria.The enumeration data were calculated by frequency analysis.Part two:Investigate the feasibility of PRPP model in non-drug therapy of TCMBased on a PRPP trial of acupuncture versus cupping therapy for fibromylgia,investigate the feasibility of PRPP model in non-drug therapy of TCM.This trial has been registered in Clinicaltrial.gov(ID:NCT01869712),and approved by Dongfang hospital affiliated to Beijing university of TCM(ID:2013052104-2).Pain visual scale(VAS)was used to calculate the sample size of the random part,and 1:1 estimated the sample size of the preference part.Participants were outpatients who were diagnosed with fibromylgia from acupuncture department in Dongfang hospital affiliated to Beijing university of TCM and the first person hospital in Dongcheng district.Patients diagnosed with fibromyalgia syndrome,VAS was equal or greater than 30,aged 20 to 60,and agreed to sign the informed consent were included.Combination of mental disorders and serious organic diseases(ie.tumors,blood diseases,organ failure,etc),other analgesics at the same time during the treatment,women in pregnancy and lactation,and patients who participated in other clinical trials were excluding.Participants who met the criteria were asked whether they had a strong preference for the treatment before informed consort.If they had,they received therapy according to their preferences.If had not,they wouldassigned by randomization.An opaque envelope was used to conceal the randomization and the statistical analyst and the outcome evaluators were blinded.Each participant was treated once a day,three times a week for five weeks.All outcomes were analyzed after 15 treatments except widespread pain score(WPI),syndrome score(SS),visual analogue scale(VAS)which measured 3 times after 5,10,15 treatments.The primary outcome was drop-out rate,patients' exception and patients' satisfaction.The second outcome was WPI,VAS,SS,short form 36 questionnaire(SF-36),Hamilton depression scale(HAMD)and fibromyalgia intensive questionnaire(FIQR).All the outcomes were compared among four groups(the randomized acupuncture group,the randomized cupping therapy,the preferred acupuncture group versus the preferred cupping therapy.X2 test and rank sum test was used for the primary outcome which were enumeration data.Single-factorvariance analysis and rank sum test were used comparing among four groups and t-test and rank sum test were used between two groups for the second outcome which were quantitative.Repeated measurements are used for multiple measurements.ResultsPart one:Bibliometrics of PRPP trial42 studies were included according the inclusion criteria.All included studies were published in English.Over time,PRPP trial has been increased,and it has reached the peak(6 trials)in 2009 according to the included studies.Pregnancy,childbirth and the puerperium disease(12 trials,28.57%),tumor(6 trials,14.29%),digestive system diseases(6 trials,14.29%)and mental and behavior disorders diseases(6 trials,14.29%)(i.e.depression)applied PRPP model mostly.The total sample size ranged from 31 to 5231,with an average sample size of 598.The sample size of the random part was 1 to 971 with an average sample size of 187 and the sample size of prefered part range was 20 to 4257,with an average sample size of 405.The ratio of randomization and preference group ranged form 0.01 to 4.65,and the average ratio is 0.89.Ratio varied by different disease.Calculation of the sample size of PRPP model is often similar with randomized controlled trials(17 trials,40.48%).4 trials(9.5%)trials only calculated the randomized part of PRPP trial and the preference part were recruited as outside trial,but both two part was analyzed finally.Except 2 trials(4.8%)which did not have the non-drug therapy treatment,all trial have non-drug therapy.The comparison of two different non-drug therapy(21 trials,50.8%),non-drug therapy versus drug therapy(11 trials,26.19%),and treatment given by differentroutes(5 trials,11.90%)(ie.local anesthesia versus general anesthesia)was the most common comparison type of interventions used in PRPP model.The aim of PRPP model was most used to compare the efficacy of intervention(33 trials,78.57%),patient preference(lltrials,26.19%)and patient satisfaction(6trials,14.29%)Part two:Investigate the feasibility of PRPP model in non-drug therapy of TCM126 participants was included and 21(16.66%)participants was dropped out which have not completed the second assessment after the 10th treatment.The preferred group was completed 8 months ahead of the randomized group.105 participants completed the trial totally(randomized acupuncture group:n=27,randomized cupping therapy group:n=28,prefered acupuncture group:n=31,prefered cupping therapy group:n=19).114 participants were included in analysis.Except SF-36 questioner and FIQR,the baseline of each group was no significant difference.The SF-36 baseline of the prefered acupuncture group was 10.43 points higher than the randomized cupping therapy group,and the FIQR baseline score of the randomized cupping group was 9.30 points higher than that of the prefered acupuncture group.2.1 Primary outcome:21 participants were droupped out(the randomized group:n=3,the randomized cupping therapy:n=2,the preferred acupuncture group:n=8,the preferred cupping therapy group:n=8).The total drop out rate of the four groups(X2=1.85,P=0.625),the randomized group versus the prefrred group(X2=0.23,P=0.632)and the acupuncture group versus cupping theapy group(X2= 1.44,P=0.23)had not siginificant difference.The droup out rate of the preferred group was higher than the randomized group at the initial stage of the trial(X2=5.09,P=0.024),but after 5 treatment(X2=3.20,P=0.074)and 10 treatment(X2=0.00,P=1.00)the drop out rate of the preferred group and randomized had no significant difference.The drop-out rate of acupuncture versus cupping therapy was no siginificant difference in each time point of measurement.In the four group,71(56.8%)participantswere expected to relieve symptoms generally,and 39(31.2%)were expected to relieve symptoms totally and 15(12%)werre expected symptoms slightly relieved.Expectations was no statistical difference between the four groups(f=2.943,p=0.83),the randomzied group versus the prefered group(f=0.289,p=0.90),and the acupuncutre group versus the cupping therapy(f=2.043,p=0.38).In the four groups,43(40.95%)participants very satisfied,37(35.24%)participants saticfied;16(15.24%)participants satisfied slightly,3(2.9%)participants the degree of satisfaction is equal to the degree of dissatisfaction,3(2.9%)participants dissatisfied,2(1.90%)participants dissaticfied slightly,and 1(0.9%)participants which from the randomized acupuncture group very dissatisfied.Satidfiction was no statistical difference between the four groups(f=f=15.476,p=0.63),the randomzied group versus the prefered group(f=5.66,p=0.46),and the acupuncutre group versus the cupping therapy(f=5.791,p=0.456).2.2 Secondly outcome:The results of repeated measurements show that the data of each group are decreasing.There were no statistically significant differences among four groups in WPI(f=0.94,p=0.43),SS(f=0.49,p=0.69),and VAS(f=0.893,p=0.447).Similar results were found in randomized versus preferred group and acupuncture versus cupping therapy group in WPI,SS and VAS.Difference between tha Acupuncture and cupping therapy have similar effect on symptoms relieve.Difference between the end of treatement and the baseline in VAS showed there was no sigificant difference between acupuncture(-30.90±22.70)and cupping therapy(-36.47±21.09)(t=-0.195,p=0.846)in relieveing pain.Similar results was found btetween randomized group and the prefered group(t=-0.195,p=0.846)with the randomized group decreased by 33.799±2.97 and the prefered group decreased by 32.98±2.90.There were no siginificant difference were found in other outcomes(i.e WPI,SS,SF-36,HAMD,and FIQR)among four groups,the randomized group versus the prefrred group and the acupuncture group versus the cupping therapy group.ConclusionPart one:Bibliometrics of PRPP trialOver time,PRPP model used in clinical trials has increased in recent years according to the included studies.This model is widely used in the field of pregnancy,childbirth and puerperal diseases,tumors,digestive systems,and psychiatric and behavioral disorders currently.Calculation of the sample size of PRPP model is often similar with randomized controlled trials.The ratio of different diseases was different between the randomized group and the preference group and the average ratio was 0.89.The comparison of two different non-drug therapy,non-drug therapy versus drug therapy,and treatment given by differentroutes(ie.local anesthesia versus general anesthesia)was the most common comparison type of interventions used in PRPP model.The aim of PRPP model was most used to compare the efficacy of intervention,patient preference and patient satisfaction.The stardards of reporting in PRPP model needs to be made in future.According to the included studies,the original intention(patients' preference and not having perfect blinging method)of PRPP model is similar with non-drug therapy in traditional Chinese medicine(TCM).PRPP model is considered to apply to non-drug therapy in TCM in theory.Part two:Investigate the feasibility of PRPP model in non-drug therapy of TCMDue to appropriate ratio between each groupin non-drug therapy of TCM which can accepted by most of Chinese,the PRPP modle was suitable for compare different kinds of non-drug therapy of TCM.PRPP model was good for recruitment.The randomized part of new model can keep the internal validity of the clinical trials,and at the same time,the non-randomized part which taking the patient's preference into consideration can increase the external validity of clinical trials.But the new model will need larger sample size than the traditional model(randomized clinical trial,RCT).The baseline comparison among four groups indicates that RCT has a certain representation in the sample size.In the early stage of PRPP trial,supeivise and management of preferrced group should strengthen to prevent the drop-out of participants.Qualitative interview of PRPP model should be conducted to investigate the feasibility in non-drug therapy of TCM in further.Patients with fibromyalgia had high expectations and satisfaction with acupuncture and cupping therapy.Acupuncture and cupping therapy have the same effect for fibromyalgia.Cost-effectiveness studies should be considered in the future studies to establish evidence for best decision-making in acupuncture versus cupping therapy for fibromylgia. |
PDF Full Text Request