| With the rapid growth of population aging, life rhythm and life stress, chronic diseases have been a major health threat, increasing the economic and medical burden of the society. In recent years, management and research of chronic diseases have been widely concerned. Traditional Chinese medicine (TCM) plays an important role in this field, with its unique theories such as holistic conception, syndrome differentiation, preventive treatment and constitution theory. The therapeutic advantages of TCM, such as easy to use. inexpensive and effective diversified methods, made it more suitable for long term management of chronic diseases.The 12th Five-year Plan Period is the key period of chronic diseases management. State administration of traditional Chinese medicine set up a Special Fund for TCM-scientific Research on the prevention and treatment of chronic diseases. This project concentrated on the research of technical methods and programs on patient-oriented Chinese medicine prevention, treatment and rehabilitation. The results with scientific nature, authenticity and reliability, will provide a basis for decision-making in the application and promotion of Chinese medicine in chronic diseases. Quality assessment of TCM clinical research plays an important role in summary appraisal, supervision and encouragement, and quality promotion.Final quality assessment can provide an objective scientific measurement of research activities, about the implementation and goal attainment level. In view of the persistence of TCM clinical research on chronic diseases, there is a necessary to establish indicators of quality evaluation, which is suitable for TCM clinical research on chronic diseases. This dissertation was to establish a set of final quality evaluation indicators, to assess the whole quality of research activities in accordance with the scheduled programs, at the end of the project. Correspondingly, the ways of sample size estimation and indicator screening and evaluation, were also explored simultaneously.Objective1 To establish final quality evaluation indicators suitable for TCM clinical research on chronic diseases. Indicators were expected to be more specific, integrated and practicable.2 To explore the method of sample size estimation, which is required to construct the structural evaluation model. And explore the application of combination of Classical Test Theory (CTT) and Item Response Theory (IRT) in the establishment of final quality evaluation indicators for TCM clinical research on chronic diseases.MethodsIBase on the previous research on final quality evaluation indicators, we established the preliminary theoretical indicators model of final quality evaluation for TCM clinical research on chronic diseases, by ginger group discussion and Delphi method. Subjective weights of the indicators were determined by expert consultation.2 Using bootstrap to explore the sample size, which is required to construct the structural evaluation model.3 The reliability and validity of the indicators were evaluated based on the CTT. The reliability evaluation indicator was Cronbachfs alpha (Cronbach a coefficient). The construct validity was evaluated through Second-order Latent Variable Model (SLVM). With bootstrap method,95% confidence interval of each coefficient was calculated, testing the reasonableness of the individual factor (objective weight). The fitting condition of data and SLVM model was also analyzed. According to the evaluation result, indicators were adjusted.Using the GPCM model of IRT. each indicator performance was estimated individually by discrimination parameter. Discrimination parameter indicates the extent to which the indicator was related to the total quality level. Through the results of discrimination parameters, indicators were further screened and selected, ensuring that the selected subsets of indicators provided adequate information. Indicators selected by IRT. were evaluated by SLVM again, to ensure the good construct validity. Through the above screening and evaluation, the rest indicators can be confirmed as the final quality evaluation indicators for TCM clinical research on chronic diseases.4 The final quality scores of subjects were calculated and ranked by subjective weight weighted method. Case analysis was to evaluate the science and applicability of the indicators. At the end, measures and suggestions were proposed, to improve the implementation of quality control in TCM clinical research on chronic diseases.Results1 The preliminary established indicators of final quality evaluation for TCM clinical research on chronic diseases, consisted of 1 objective (final quality evaluation),6 essential factors (intervention application, compliance assessment, recording quality, quality control, subjects rights protection, and goal achievement) and 14 indicators (integrity and normalization of intervention recording, intervention application compliance, research method performance, compliance of subjects inclusion and exclusion, appearance rate, integrity, accuracy and normalization of case report form (CRF), recording and analysis of adverse events and serious adverse events, primary examination, secondary examination, ethical approval, ethics audit trail, signatures of informed consent, overall objective completion, and performance goals completion).2 The sample size of 150 was considered to be the minimum to obtain robust SLVM parameter estimation. Then each of 28 projects was given scores by 6 experts. The actual sample size was up to 168.3 Preliminary reliability analysis revealed that, Cronbach a coefficient of 14 indicators was 0.6. which was acceptable. Nevertheless, Cronbach a coefficient of recording quality and compliance assessment were extremely low, which indicated that internal consistency of the two essential factors was poor. The two essential factors were necessary from the professional perspective. Then scoring criteria of these indicators, belonging to the two essential factors, were adjusted according to the data characteristics. Reliability analysis of indicators after adjustment revealed that indicators after scoring criteria adjustment had good internal consistency.Results of the validity analysis revealed that, the weight coefficients were significantly non-zero, no matter between first order latent variables and corresponding observation variables, or between first and second order latent variables, or between second order latent variables and observation variables. Goodness-of-fit index (GFI), normed fit index (NFI). and comparative fit index (CFI) were above 0.6; especially GFI was 0.805. which was close to 0.9. In the empirical applications, the model fit index results were acceptable, which indicated good fit performance of data and models. The above results indicated that indicators after adjustment have good construct validity.Results of IRT analysis demonstrated that, the discrimination parameters of 14 indicators after scoring criteria adjustment, basically reached a good level. The overall design of the indicators was reasonable. Therefore, final quality evaluation indicators of TCM clinical research on chronic diseases can be ascertained.4 Case analysis showed consistency between indicators quantitative evaluation and qualitative evaluation. Thus, the indicators can be considered having good science and applicability. Quantitative evaluation revealed that the average score of final quality evaluation of 28 projects was 82.4, and 75% of which was higher than 80. The average scores of factors were above 80 except the quality control factor, which showed a relatively lower level. Measures and suggestions were proposed to improve the execution of quality control in future research, according to problem analyses in this empirical research.Conclusion1 Considering the key elements of quality control in TCM clinical research on chronic diseases, construction of quality evaluation indicators was necessary and significant.2 Bootstrap can be used to explore the sample size, required to construct the structural evaluation model. This may provide a method for the sample size estimation in future research on construction of quality evaluation indicators system.3 Classical test theory can evaluate the whole structure of indicators from a macro perspective, while item responding theory can not only overcome the theory limitations of CTT, but also evaluate indicators individually, from a micro perspective. Combination of the two methods will make indicators more reliability and validity. This may provide a methodology for future indicators research.4The constructed final quality evaluation indicators system can not only intuitively reflect the quality level of the implementation, it also make objective quantitative comparisons possible between different subjects, or different indicators. The evaluation results can also intuitively display the key factors affecting the quality level of subjects. That may provide objective evidence for overall quality control and quality assessment for future clinical research.5The implementation of all studies reached to a better level. At the same time, measures, such as establishment of quality control management system, innovations of various quality control management models, usage of quality control network platform, and improvement of quantitative evaluation system, can be further improved the quality control implementation level. |
PDF Full Text Request