Data Validity Of The Related Event In "Acupuncture And Clomiphene Citrate On Live Birth In Infertile Women With Polycystic Overy Syndrome:Randomized Controlled Trail"

Objective:To collect protocol deviation(violation)and drop-out cases in PCOSAct,and demonstrate definition,categorization and harm of such cases.To evaluate the compliance of researcher,research assistant and subject and discuss effects of different collecting methods on quality control;effective quality control can improve protocol compliance,decrease the occurence of protocol deviation(violation),monitor of protocol deviation can help to adjust the quality of local sites and maintain the uniformity,which can improve the research quality by increasing the validity of data;to report above cases in details and sum up experience and provide refference for implementation and quality control in other large RCT,and make some contribution to internationalization of TCM clinical trials.Methods:From the initiation of PCOSAct on July 2012 to completion of project on October 2015,we collect all protocol deviation(violation)and drop-out cases from:Research assistant’s reports to DCC and DSMB;In-site supervision documentions;Case report form verification documentions.Then we stastically analyzed the data.Results:1.Protocol deviation1.1 In PCOSAct,the averaged protocol deviations occured in each case is 1.49.1.2 Among protocol deviations,the 5 categorizations which contain the largest numbers of deviations are as follows:the subject didn’t complete protesterone and HCG tests every week;the subject received treatment before finishing all screening procedure;the subject lost medicine or bottle(which didn’t cause medication mistake);the subject received fetus protection medicine except dydrogesterone;the subject didn’t complete 2 times acupuncture therapies every week.The number of deviations in above 5 categorizationgs is 258,which compose 62.2%of all deviations.1.3 Deviations caused by researchers take up about 1/3 deviations,which mainly are incomleteness or expiration of baseline tests;Deviations caused by subject is distributed and mainly contains:the subject didn’t take P or HCH tests on time during treatment;the subject received fetus protection medicine except dydrogesterone when pregnancy.1.4 Deviations were reported by in-site supervision and reserch assistant report,which make up 42%and 5 8%of all deviations,respectively.Deviations found by in-site supervision mainly are caused by researchers,while most deviations found by research assistant report are caused by subjects(P<0.01).1.5 Until the first in-site supervision on December 2012,there were 123 deviations,about 29.6%of all.In-site supervision found 82 deviations,66.7%of all(82/123).Deviation’s occurence rate in initial stage is higher than those of midterm or later stage.In later stage,the number of deviationgs reported by researcher and research assistant became larger.1.6 Deviations mainly occured during baseling period(99/23.86%)and first treatment cycle(97/23.37%);during baseling period,most deviations were caused by researchers and found by in-site supervision;during treatment,devations were mainly cause by subjects and found by research assistant report.1.7 By analyzing the effect of hospital type on deviations,we found that deviations were all mainly caused by subjects in Chinese/We stern medicine hospital,affili at ed/non-affiliated,North/South hospitals;more deviations were found in Chinese medicine hospitals,affiliated hospitals,North hospitals than Western medicine hospitals,which is statisticly significant.2.Protocol ViolationIn PCOSAct,25 protocol violation cases occured,only 2.5%of all subjects.60%of them are violations of medication protocol,which is medicine taken more(3 cases)or less(15 cases)than 25%of protocol.3.Drop-outThe drop-out rate is 7.3%in PCOSact.The main reason for drop-out is subject’s unwilling to participate in the trial,which mainly occured in the first treatment cycle.Conclusions:1.Clinical trial is different from animal experiment,protocol complaince cannot be guaranteed during the procedure,especially in large,multi-center,long-duration clincial trials;there are limited reports of deviations all over the world becasue of its slight effect on evaluation of intervention.Monitor of protocol deviation is a good measurement to improve data quality.2.Effective quality control can improve protocol compliance,decrease the occurence of protocol deviation(violation),adjust the quality of local sites,maintain the uniformity,and improve the validity of data.3.Protocol violation and drop-out rate in PCOSAct is relative low,which guaranteed the date effectivness.On one hand,implementation of this project was under direction of international experts who have rich experience in quality control.On the other hand,in-site supervision by project group and other quality control measurements were used flexibly.