| [Background] The World Health Organization (WHO) has claimed that the lack of transparency was the main reason for corruption and low efficiency in medicine sector. Transparency, as a new way of regulation has been adopted for health and medical regulation in many developed countries. In China, attempts concerning transparency have been made in medicine sector. How patients would use the drug use information has been identified as the key element of the efficiency of this transparency system which still lacks investigation. Whether this transparency system would change the behavior of patients is still unknown.[Purposes] (1) To establish a concept model concerning how patients use drug use information which exploring the mechanism between information availability, information perception and information reaction. (2) To test the path between these 3 elements, and to identify the key path in this mechanism. (3) To test the ultimate effect of this mechanism from the angle of doctors'patient volume, i.e. whether the transparency of drug use information would lead to the change of patients' selection of doctors.[Methods] Literature review and focus group discussion were used to clarify the theoretical concept of this transparency mechanism research, and to establish the theoretical model of how patients use drug use information; Laboratory Studies with controlled experimental design were used to identify the best way of intervention which would benefit the patients most. Logistic regression was then used to analyze the data; B city in A province was chosen as the study site, matched-pair cluster-randomization was used. All of the 20 primary healthcare institutions in B city were clustered as 10 pairs. The two institutions in every pair were randomly assigned to the intervention group or control group. The institutions in intervention group then were subjected to transparency intervention. A convenience sampling method was used, the sampled patients were investigated with a structured questionnaire. Mediating Role Analysis and Structural equation modeling were employed as the main analysis method to fit the data, the mathematical model of how patients use drug use information was then established. To evaluate the effect of transparency intervention on patients'ultimate selection behavior, a pre-post quasi-experimental design was used. The panel date of both intervention and control group was obtained. The doctors' patient volume was the predict variable, after proper data processing, Difference-in-difference model was then used to analyze the aimed effect.[Results] (1) Literature analysis of 37 articles on transparency definition revealed that relevance, clarity, timeless and easy to access of information are the most frequent mentioned character of information quality related to information consumers'use of information. The defined dimensions used to evaluate the degree of user embedding of information included:information's perceived value in achieving users'goals; its compatibility with decision-making routines; and its comprehensibility.(2) Laboratory experiment showed that the research objects in group 1 (numeric, sorted, added information) have the most accuracy in selection of the best (74.43%) and the worst doctors (90.11%) concerning drug use. Without background information, the research objects in group 3 have the lowest accuracy (59.77%) in choosing the best doctor in drug use. Without sorting the information, the research objects in group 5 have the lowest accuracy (58.89%) in choosing the worst doctor in drug use. Logistic analysis showed that:when information were framed with numbers, proper sorting and with added background information, information users have the most probability of choosing the right doctor (OR=1.45 for choosing the best and 2.87* for choosing the worst compared to the baseline group).(3) The mechanism of how patients use drug use information analysis:566 valid questionnaire were obtained (300 for intervention group,266 for control group), Cronbach's a was 0.83 for the overall questionnaire, the Cronbach's a for every identified components were 0.919(information availability),0.93(perceived usefulness), 0.80(perceived comprehension),0.67(perceived compatibility). Factor analysis extracted five common factors, total explained 75.56% of the cumulative variance, and the item distribution in five common factors was completely consistent with theoretical model, indicating good construct validity. Independent samples T test showed that the intervention group has a higher factor score than control group on all 5 components. But this different was only statistically significant on two components namely information availability and information comprehension. Mediating Role Analysis showed that perceived comprehension was the mediating factor for the relationship between information availability and perceived usefulness, perceived usefulness was incomplete mediating factor for the relationship between perceived usefulness and information use tendency. Six statistically significant paths were identified through Structural equation modeling analysis, these paths were:1) perceived compatibility can promote better understanding of information(standard regression coefficients.146, P=0.009); 2)perceived compatibility can promote perceived usefulness(standard regression coefficients.117, P=0.017);3) perceived comprehension can promote information availability(standard regression coefficient=0.387, p<0.001); 4) information availability can promote perceived usefulness(standard regression coefficient=0.285, p<0.001); 5)perceived comprehension can promote perceived usefulness(standard regression coefficient=0.224, p<0.001); 6)perceived usefulness can promote the tendency to use these information(standard regression coefficient=0.653,p<0.001).(4) DID analysis included 295 doctors (139 for intervention group,156 for control group), the transparency intervention has made a positive effect on the relationship between the ranking of doctors'drug use quality and the patients'volume. The intervention has made a more intense relationship between whether a doctor ranks in the first 1/3 on antibiotic use and whether a doctor's patient volume ranks in the first 1/3(OR=3.86, p=0.039,95%CI[1.07,13.86]). The job title of doctors positively influence their patients volume, as the job title being higher graded, the possibility that their patients volume being ranked in the first 1/3 gets higher(OR= 1.61,1.50,1.48respectivly in 3 model, P <0.001).Income also has a positive impact on patients volume.[Conclusions] (1) The identified elements of drug us mechanism of patients contained: 1) information availability; 2) perceived usefulness; 3) perceived comprehension; 4) perceived compatibility. The theoretical model of the mechanism was:information availability-information preciseness-use tendency-behavior.(2) When information were framed with numbers, proper ordering and with background information, information users have the most probability of choosing the right doctor.(3) All pre elements except perceived usefulness have an indirect impact on the tendency to use transparent information, i.e. perceived compatibility, perceived comprehension and information availability can have an indirect impact on the tendency to use transparent information through perceived usefulness. The relationship between perceived compatibility, perceived comprehension and availability are as follows: perceived compatibility can promote perceived comprehension, perceived comprehension can then promote information availability.(4) The transparency intervention has a positive impact on patients' selection behavior, the transparency intervention has made a positive effect on the relationship between the ranking of doctors'drug use quality and their patients' volume, i.e. the transparency intervention has made patients more likely to choose doctors who have a better ranking on antibiotic use.[Innovations and Limitations] this research theoretically was based on transparent action cycle, taking the specialty of drug use sector and the current transparency situation into account. Quasi-experimental design was employed to explore the mechanism between information availability, information perceives and information reaction, mathematical model were then established. This research has made a theoretical contribution to drug use regulation. The theory innovation of this study was that a concept model of how patients use drug use information was established and then the mathematical model was also explored. This has made an important improvement to transparent regulation theory. Quasi-experimental design were also used for intervention evaluation, this strong design has also made the result very powerful.Due to the restriction of research condition, we have only sampled the primary health care institutions, the tier-2 hospitals and tier-3 hospitals is beyond our study which may bare a different conclusion; this is the first research theoretically based on transparent action cycle and applied it to the patients perceive process for information, there is a lack of comparison on the Mathematical path; relatively, the education level of patients in primary health institutions is low, this may has a negative impact on the intervention effect. During the mathematical analysis, there may be a lot of uninvestigated influence factors, as to these uninvestigated factors, further research would be needed. |
PDF Full Text Request