| 1 ObjectiveTo optimize and revise the primary PRO scale with RA,analyze the correlation between RA-PRO and DAS28-ESR,HAQ and provide a reliable evaluation tool for the assessment and dynamic monitoring of RA though the cross-sectional and longitudinal study.2 Methods2.1 Inclusion and exclusion criteria2.1.1 Inclusion criteriaCross-sectional survey:① Meet the RA classification standard,which was revised in 1987 by the American College of Rheumatology(ARA);② Chinese residents aged 16 and over;③ Able to speak or read Chinese;④ Consent to sign informed consent.Longitudinal study:① Meet the RA classification standard,which was revised in 1987 by the American College of Rheumatology(ARA);②TCM syndrome differentiated by damp heat obstruction and blood stasis of collaterals syndrome;③Disease duration within 3 years;④DAS28 score>3.2;⑤Treatment without Methotrexate Tablets or Plaquenil;if treated by other DMARDS,only one type and stable dose for 3 mouths at least should be limited;⑥ Age between 18-65 years old.2.1.2 Exclusion criteriaCross-sectional survey:① The presence of cognitive or other impairment(such as vision,etc.)affects the participants to complete the self report;② Severe diseases,such as cardiovascular,brain,liver,lung,kidney and hematopoietic system,acute and chronic infectious diseases,malignant tumors,patients with mental disorders;③ Overlapped with other rheumatic diseases;④ Refused to participate in the study.Longitudinal study:① Skin irritation or skin ulceration;② Used Tripterygium wilfordii or glucocorticoid or biologics before and drug withdrawal within 4 weeks;③ Pregnancy,lactation and recent fertility planners;④ Severe diseases,such as cardiovascular,brain,liver,lung,kidney and hematopoietic system,acute and chronic infectious diseases,malignant tumors,patients with mental disorders;⑤ Overlapped with other rheumatic diseases such as systemic lupus erythematosus;⑥ Patients with retinopathy.2.2 Data sourcesCross-sectional survey study data came from the 1934 outpatient and inpatient cases visiting from February 2014 to August 2015 in line with the inclusion and exclusion criteria,excluding the collection of incomplete information in 66 cases.The effective data of 1868 cases were included from 23 medical institutions nationwide,which were all three grade A hospital.Chinese Rheumatoid Arthritis Data Center--Traditional Chinese Medicine(website:http://tcm.criponline.com)was set up.After double entering and consistency check,1868 cases of cross-sectional effective survey data were obtained.Longitudinal study was based on the 12th Five-Year National Science and technology support program "traditional Chinese medicine syndromes of rheumatoid arthritis and comprehensive treatment program" to extract partial data for anylasis.Data came from the outpatient cases visited 9 medical institutions nationwide including the Guang’an Men Hospital,the first affiliated hospital of Anhui University of traditional Chinese Medicine,the Chinese Academy of Medical Sciences Peking Union Medical College Hospital and so on.A total of 203 outpatient cases were included,17 cases with incomplete information excluded,186 cases were included in the study.The internet database was setup(website:http://124.205.181.142:8080/lfsgjy/index.jsp).After double entering and consistency check,186 cases of longitudinal effective survey data were obtained.2.3 Statistical Methods2.3.1 Descriptive AnalysisFrequency analysis of the general situation data and items of the patients was adopted and measurement data in accord with normal distribution using the mean±standard deviation(X±SD).2.3.2 Item OptimizationAnalyze the Klangbach coefficient and the factor load of each dimension and item,delete the item with lower coefficient or smaller factor load,and recalculate whether the Braunbach coefficient is raised to determine whether the item is deleted.2.3.3 Scale Weight and GradingItems are followed by four classification levels,assigned 0,1,2,3 respectively.The linear regression equation model was constructed to analyze the weight coefficient.According to the distribution of 25%quantile and 75%of the total score of PRO in different population,the degree of RA-PRO score was determined according to the classification of DAS28 disease activity.2.3.4 Scale EvaluationThe reliability test of the scale was measured by the Kronbach alpha coefficient and the Guttman folding half confidence method.The structural validity test was measured by the content validity and the structural validity method.The responsibility of the items was measured by the paired rank sum test and the response of the overall score was measured by the paired t test.2.3.5 PRO Correlation Analysis with DAS28-ESR and HAQThe correlation between the scale and DAS28-ESR and HAQ was analyzed by Pearson and Spearman correlation coefficients respectively.2.3.6 Variation tendency of PRO,DAS28 and HAQ in the longitudinal studyFor the correlation analysis between the two indicators of repeated measurements,the linear mixed effects(LME)model was adopted to analyze the variation tendency of PRO,DAS28 and HAQ in the longitudinal study.The above data analysis was achieved by SAS9.2 software.3 Results3.1 Scale RevisionBy using the Kronbach coefficient method and the factor analysis method to optimize the PRO scale.The "body feeling" dimension of the a coefficient increased significantly after deleting the "10 fear of the wind cool cool",so item 10 was deleted.In this paper,we use the multiple linear regression method and combine the clinical significance of each item of PRO,taking the weighting factors into account,fit the linear regression model of PRO total score and each item.The model fitting statistics F=2381.87,P<0.0001,with statistically significance,the revised PRO total score is calculated as follows:PRO Score ＝1.43×pro1 + 0.93×pro2 + 0.40×pro3 + 1.09× pro4 + 0.11×pro56 + 0.18×pro78+0.24×pro9 + 0.04×pro11+ 1.17×pro12 + 0.19× pro1314 + 0.02× pro1516According to the distribution of 25%and 75%of the total score of PRO in different groups,we divided the criteria of RA-PRO into no activity(RA-PRO<3.61),mild activity(3.61≤RA-PRO<5.54),moderate activity(5.54≤RA-PRO≤7.79),severe activity(RA-PRO>7.79).3.2 Reliability,validity,responsiveness measurement of the revised PRO scaleThe overall Cronbacha value was 0.756 and the fold reliability was 0.733,which indicated that the internal consistency and stability of the scale was good.Three factors were extracted by factor analysis,which reflected the three aspects of physical,social and psychological status with the internal cross-logical relationship between the items.It can be considered that this scale has a good structure validity and surface validity in RA population’s disease measurement.The 12 weeks and 24 weeks information of each item were compared with the baseline data respectively by the rank sum test.It was found that there were significant differences between the items except item "12 appetite" and the total score of the RA-PRO scale after 3mounths and 6mounths treatment(P<0.001)and the responsibility is good.There was no significant difference of the item“12 appetite" between the baseline and the 12 weeks(P>0.05),but there was significant difference between the baseline and the 24 weeks(P<0.05).3.3 Correlation between RA-PRO and DAS28-ESR,HAQ and other evaluation tools3.3.1 Correlation between RA-PRO and DAS28-ESRThe correlation coefficient between RA-PRO score and DAS28-ESR score was 0.640,and the RA-PRO rank level and DAS2-ESR score is 0.523,indicating that the correlation between them was good(P<0.05).The correlation coefficient between the dimension of somatesthesia and DAS28-ESR was the highest(P<0.05).3.3.2 Correlation between RA-PRO and HAQThe correlation between RA-PRO score and HAQ score was calculated by the Person correlation coefficient(r = 0.581),indicating there was a certain correlation(P<0.05).In the dimension correlation analysis of four groups,the dimension of somatic function and HAQ were highly consistent(P<0.05).To further compare the correlations between the three tools,the results show that RA-PRO was more relevant than HAQ to DAS28.3.3.3 Correlation between RA-PRO and ESRThe correlation coefficient between RA-PRO score and ESR is 0.339,indicating that the correlation between them was Medium down(P<0.05).There was a weak correlation between the RA-PRO score and ESR only in the group of high disease activity.There was no correlation between the dimension of psychology and DAS28-ESR(P<0.05).3.3.4 Correlation between RA-PRO and GHThe correlation coefficient between RA-PRO score and GH was 0.542,and the DAS2-ESR was 0.670,HAQ was 0.450(P<0.05),respectively.In the high disease activity group,the correlation coefficient between the PRO score and GH was the highest(P<0.05).The correlation coefficient between the dimension of somatesthesia and GH was the highest(P<0.05).3.4 Correlation between revised RA-PRO scale and DAS28,HAQ dynamic trendThe dynamic correlation coefficients of RA-PRO and DAS28 and HAQ were 0.4622 and 0.4464,respectively.The correlations between their dynamic trend was pretty good(P<0.0001),and the dynamic trend correlation between RA-PRO scale and DAS28 was better.4 ConclusionBased on the previous PRO scale,the revised RA-PRO was a RA evaluation scale which reflected Chinese characteristics and TCM thinking characteristics.This scale improved and perfected the traditional Chinese medicine’s evaluation system from the "disease" level,and to a certain extent,reflected the advantages and efficacy of traditional Chinese medicine treating RA.From the perspective of PRO,it could be used as a supplement to the existing RA evaluation tools,from the perspective of Chinese medicine efficacy evaluation it could be used as a reliable method for the evaluation of RA efficacy of Chinese medicine treatment.Specific conclusions are as follows:(1)There was a good correlation between the revised RA-PRO and DAS28 in the assessment of RA,which could reflect more disease information in the evaluation of RA and could reflect more "hidden" therapeutic benefits.(2)The revised RA-PRO has a good ability to assess the joint function of the patient.The correlation coefficient between RA-PRO score and DAS28 was higher than HAQ,indicating PRO was more sensitive to reflect the condition of RA.(3)The revised RA-PRO has a good ability of dynamic assessment and has considerable effectiveness of DAS28.It could reflect more disease information and had a certain application value and advantage in disease monitoring and RA follow-up researches. |
PDF Full Text Request