How public reporting of prescription quality ... - Wiley Online Library

bs_bs_banner

Journal of Evaluation in Clinical Practice ISSN 1365-2753

How public reporting of prescription quality indicators influence prescribing practices? A survey of general practitioners Xin Du PhD candidate,1 Xinping Zhang PhD,2 Yuqing Tang PhD candidate,1 Xiaopeng Zhang Master’s degree candidate,1 Lijun Wang Master’s degree candidate1 and Xuan Wang PhD candidate1 1 Postgraduate, 2Professor, School of Medicine and Health Management, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei Province, China

Keywords attitude, behavioural intention, information accessibility, perceived risk, perceived value, public reporting Correspondence Professor Xinping Zhang School of Medicine and Health Management Tongji Medical College Huazhong University of Science and Technology No.13 Hangkong Road 430030 Wuhan Hubei Province China E-mail: [email protected] Accepted for publication: 27 May 2015 doi:10.1111/jep.12410

Abstract Rationale, aims and objectives Public reporting of performance data is one of the most popular topics in health care field. The aim of this study was to investigate the transparency mechanism, that is, how public reporting influenced general practitioners’ (GPs) prescribing practices. Method GPs who had the license to prescribe medicine of all 10 primary care institutions were surveyed. Data were collected by an instrument, which exhibited satisfactory reliability and validity (Cronbach’s alpha > 0.7; average variance extracted > 0.5; composite reliability > 0.7). Data analysis was conducted by structural equation model. Results The results showed that GPs’ perceived value (GP’s overall assessment of the worth of the public reporting) and attitude (the psychological reaction to public reporting) had a significantly direct effect on behavioural intention (r = 0.28; r = 0.36), and were affected by information accessibility (r = 0.63; r = 0.32). Attitude had a significant effect on perceived value(r = 0.45). Perceived risk (the perceptions of the possible loss due to public reporting, e.g. decreasing their income) did not have a significant relationship with information accessibility, attitude and behavioural intention(r = −0.09; r = 0.01; r = −0.07). Conclusion The information accessibility, perceived value and attitude have strong effects on prescribing practices of GPs, whereas perceived risk did not play a role in influencing the prescribing practices. Policymakers need to improve the accessibility of prescription quality indicators and pay attention to the perceived values and attitudes of GPs. Policymakers also need to strengthen the risk education of GPs and attach incentives to transparent regulation.

Introduction Transparency as a popular policy issue Transparency has become a popular topic and has often been cited as a response to the accountability concerns of global actors [1]. Many governments and organizations, such as the USA and the European Union, have designed public systems to minimize corruption, protect civil rights, improve public services, and reduce financial, health and safety risks [2]. The Anti-Corruption and Civil Rights Commission stressed that transparency and anticorruption was the pathway to lead South Korea preeminent in the Asia-Pacific [3]. In the field of health care, public reporting of performance data has become a response to calls for increasing transparency [4]. In

the USA, almost all states have implemented numerous reporting programmes for hospitals. The Center for Medicare and Medicaid Services has initiated public disclosure programmes for hospitals, manages care plans, nursing homes and home health agencies [5]. In the UK, the Care Quality Commission, which regulates National Health Service hospitals, actively updates its website with comparative information on the quality of hospitals [6]. In South Korea, a national insurance review agency has been publicly releasing information regarding the rates of antibiotic use among health care organizations since 2006 [7]. In China, health care transparency is a work in progress. In 2009, the Chinese government further emphasized the importance of increasing regulation in the health care field including transparent regulation [8]. One of the regulations on prescribing practices was prescription comments – mainly referred to the hospitals to

Journal of Evaluation in Clinical Practice 21 (2015) 943–951 © 2015 John Wiley & Sons, Ltd.

943

How public reporting influence prescribing practices

count, analyse and evaluate the medicines that doctors have prescribed; and public reporting the comment results in some medical institution [9]. Although there was a package of regulations in health care field, the inappropriate use and overuse of medicines was still a serious concern in China because of the corruption in medical practices just as Hsiao [10] pointed out. Therefore, improving transparency in prescribing practices to reduce corruption and promote rational use of medicines is a most urgent necessity. Many local governments have attempted to develop such policies; however, few research results about efficacy and mechanism of transparency have been published thus far.

Transparency on prescription There are few studies focusing on public reporting of prescription quality indicators. Finkelstein et al. [11] conducted a prospective case-control trial to test the effect of an educational outreach intervention on antibiotic prescribing. In intervention group, feedback on previous prescribing rates was provided to doctors, and patients were mailed a brochure on antibiotic use, and supporting materials were displayed in waiting rooms. They found that the intervention reduced antibiotic use [11]. However, the antibiotic prescribing rates only provided to doctors not to patients. After that, Gerber [12] conducted a cluster-randomized trial evaluating the effect of clinician education coupled with audit and feedback of antibiotic prescribing to patients on antibiotic prescribing. They indicated that the intervention improved adherence to prescribing guidelines for antibiotics [12]. Nevertheless, they only audit and feedback one indicator – antibiotic prescribing rate. Our research team conducted a cluster-randomized controlled trial to evaluate the effect of public reporting on prescription since 2013. The intervention lasted for one year and there were three indicators of public reporting including the ranking of antibiotic prescription percentage, injection prescription percentage and average drug cost per prescription. Several research results have been published so far. Yang [13] suggested that public reporting resulted in a 9% point reduction in the use of oral antibiotics, while the use of injectable antibiotics remained unchanged. Wang [14] pointed out that public reporting led to a reduction of approximately 4% in the injection prescribing rate. Zhang [15] indicated that there was a significant reduction in the percentage of prescriptions requiring injections after publicly reporting prescription quality indicators.

Few related empirical research on the transparency mechanism Some studies assessed the effect of publicly releasing performance data on quality improvement [5,16–18]. A systematic review suggested that publicly releasing performance data stimulated quality improvement at the hospital level [16]. After that, other studies have been published. Jung [5] reported that public reporting exerted a significantly positive effect on the quality of Health Maintenance Organization markets. Chen [17] indicated that clinical outcomes can improve as a result of public reporting of hospital performance. Ikkersheim [18] provided empirical evidence that general practitioners’ (GPs) referral patterns were almost unaffected by the available quality information. Nevertheless, mere awareness whether public reporting works is far from the desired. 944

X. Du et al.

To improve the efficacy of transparency, clarifying the transparency mechanism, that is, how public reporting influences GPs’ prescribing practices is necessary. Although several well-known theories, such as the framework of quality improvements that proposed by Berwick, explained the response of providers to public reporting [19], only a few empirical studies on the transparency mechanism exist. Some studies showed that providers were generally antipathetic towards publicly released comparative data and lacked of evidence to support that public reporting affected providers’ behaviour [20,21]. However, Totten [22] indicated that when performance data were made public, providers would engage in activities to improve quality. McGivern [23] developed a model to explain how providers produced emotional reactions and, in turn, defensive reactivity to transparency. But the model development was just based on a limited number of semi-structured interviews, so the results may be somewhat subjective and not so sound [23]. Another research answered that GPs were not motivated to use comparative performance information because of doubts and uncertainty about the report’s content, validity and effect, but the results were also based on qualitative method [24]. Sherman et al. [25] evaluated the perceptions of surgeons at the hospital level and individual-level public reporting and noted that surgeons preferred to publicly report hospital-level data, but they did not answer why. Nevertheless, the abovementioned studies did not discuss how public reporting affected providers’ behaviour nor investigated the mechanism by empirical quantitative data.

Theoretical foundation and research hypotheses Weil and Kennedy [26] indicated that transparent policy involved a government authority requiring the collection and dissemination of information to alter the behaviour of the information disclosers as well as users. In 2005, Fung et al. [27] proposed the transparency action cycle (TAC) based on transparent policy. According to the TAC, effective public transparency systems triggered a virtuous chain of action and reaction. First, information disclosers provided information to the public. Second, information users reacted to new facts by changing their perceptions and behaviour. Third, information disclosers changed their perceptions and behaviour in response to users’ actions in order to improve their competitive advantage [27]. In our study, we mainly focused on the first two phases, that is, after public reporting of prescription quality indicators, GPs reacted to new facts by changing their prescribing practices. As regards the information users’ perception about public reporting that were mentioned in the TAC, previous studies showed that perceived value and perceived risk were important aspects of perception and often, these were studied in relevant research [28,29]. Behavioural intention was closely related to the actual behaviour, as that some researchers recommended studying behavioural intention instead of actual behaviour [30–32]. Therefore, perception in the present research refers to perceived value and perceived risk. Moreover, this study considers behavioural intention instead of actual behaviour. Using the discussions above as basis, we propose that information accessibility significantly influences perceived value and perceived risk, and perceived value and perceived risk significantly influences behavioural intention.

© 2015 John Wiley & Sons, Ltd.

X. Du et al.


antibiotics and injections, reducing outpatient prescription costs) under the context of public reporting of their prescription quality indicators. Based on the preceding theoretical foundation and definitions of the constructs of transparency mechanism, we test the eight hypotheses among GPs as follows: H1: The accessibility of the information of prescription quality indicators (information accessibility) significantly influences GP’s overall assessment of the worth of the public reporting (perceived value). H2: The accessibility of the information of prescription quality indicators (information accessibility) significantly influences GPs’ perceptions of the possible loss (perceived risk). H3: GP’s overall assessment of the worth of the public reporting (perceived value) significantly influences the GPs’ willingness to improve prescribing practices (behavioural intention). H4: GPs’ perceptions of the possible loss (perceived risk) significantly influence their willingness to improve prescribing practices (behavioural intention). H5: The accessibility of the information of prescription quality indicators (information accessibility) significantly influences GPs’ psychological reaction to public reporting (attitude). H6: GPs’ psychological reaction to public reporting (attitude) significantly influences GP’s overall assessment of the worth of the public reporting (perceived value). H7: GPs’ psychological reaction to public reporting (attitude) significantly influences GPs’ perceptions of the possible loss (perceived risk). H8: GPs’ psychological reaction to public reporting (attitude) significantly influences their willingness to improve prescribing practices (behavioural intention).

Knowledge, attitude and practice (KAP) model is a psychological cognitive model. In this model, knowledge means the ability to pursue and use information. Attitude indicates the result of making reactions through certain ways in some situations. Knowledge (information) is the foundation, and attitude is the motivation; a causal relationship exists between the two elements. The basic view denotes that the knowledge (information) accumulation will cause the change of attitude to a certain degree [33]. Therefore, we propose that the information accessibility significantly influences attitude. Perception is the process that interprets and organizes the sensation to produce a meaningful experience of the world. A study result indicated that perception was closely related to attitude [34]. Hence, perception is closely related to attitude whether it is perceived value or perceived risk. Based on this idea, we propose that attitude significantly influences perceived value and perceived risk. The effect of attitude on behaviour is one of the most popular topics among psychologists [35]. The theory of reasoned action (TRA) is one of the most influential theories in predicting human behavioural intention and has been widely applied to health issues [36]. TRA focuses on the formation process of attitude and mainly to analyse the effect of attitude on individual behavioural intention consciously. Based on the theory foundation, we propose that attitude significantly influences behavioural intention. According to the foregoing discussions, we integrated the TAC, KAP, TRA and other studies’ results to model the transparency mechanism in the health care environment. The constructs of transparency mechanism in this study was defined as follows: (1) information accessibility refers to whether or not the information of prescription quality indicators is accessible, acceptable and appropriate to information users; (2) perceived value refers to GP’s overall assessment of the worth of the public reporting, for example, sense of achievement, learn more knowledge about rational prescription, improvement of prescription quality; (3) attitude refers to the psychological reaction of GPs to the policy of public reporting of their prescription quality indicators, that is, whether GPs viewed public reporting as necessary; (4) perceived risk refers to the perceptions of the possible loss when public reporting GPs’ prescription quality indicators, for example, decreasing their income, damaging their reputation, and have some psychological loss; (5) behavioural intention refers to the willingness of GPs to improve their prescribing practices (prescribing less

Conceptual model All the eight hypotheses (H1–H8) together comprise a conceptual framework of the transparency mechanism, which is illustrated in Fig. 1 and it is also regarded as the structural component in the structural equation model (SEM).

Objective and significance The objective of this study was to investigate the transparency mechanism, that is, how the public reporting influence GPs’ pre-

Perceived Value

H1

H3 H6 H5

Information

Attitude

H8

Behavioural

Accessibility

Intention

H7 Figure 1 Conceptual framework of the relationship among information accessibility, perceived value, attitude, perceived risk and behavioural intention.


H2

H4 Perceived Risk

945


scribing practices. This is one of the first studies that focus on public reporting of prescription quality indicators. The present research significantly contributes to both theory development and practice. For theory development, the research develops a conceptual framework of transparency mechanism based on existing theories, it may enrich transparent regulation theory and provides basis for future studies. For practice, the study provides a reference for the investigation of transparency mechanism by quantitative research. In addition, well-known how public reporting influence GPs’ prescribing practices not only improve the efficacy of transparent regulation, but also promote the implementation of transparent regulation.

Methods Setting and participants This study was conducted in Q City, a typical city in central Hubei Province, which has 1 030 000 inhabitants and covers approximately 2004 km2. Ten primary care institutions, where the prescription quality indicators of GPs have been publicly reported, were included in the study. There are three prescription quality ranking indicators of GPs who came from the same department: (1) the ranking of antibiotic prescription percentage, (2) the ranking of injection prescription percentage and (3) the ranking of average drug cost per prescription. The ranking report was calculated based on the prescription data retrieved from the hospital information system and was updated monthly by our research team. The public reporting forms included a bulletin board with A3 colour-printed papers and A4 papers, which were directly distributed to GPs. GPs of all 10 primary care institutions who had the license to prescribe medicine were eligible for the study. We would survey all eligible GPs who were on duty during the duration of investigation that usually last four days. The research team conducted two questionnaire surveys with all eligible GPs. The first survey results that were collected in April 2014 were used to validate the reliability and construct validity of the instrument. The second set of survey data was gathered in October 2014 and was used to investigate the transparency mechanism by SEM.

X. Du et al.

attitude, the study measured the emotional reactions of GPs to the public reporting, that is, whether GPs viewed it as necessary. Perceived value has several dimensions: social, emotional, functional, epistemic and conditional values [40]. In view of the particularity of the health care field, we focused on measuring emotional value, functional value and epistemic value [29,41]. Perceived risk has been classified into financial, performance, physical, social, time and psychological risks [42,43]. We mainly paid close attention to financial risk, psychological risk and social risk, which were usually considered in previous studies. Zeithaml et al. [44] proposed that behavioural intention incorporated five dimensions: loyalty, switch, pay more, external responses and internal responses. The behavioural intention of GPs in this study was mainly measured using two dimensions: switch, for example, change prescribing practices; pay more, for example, pay more time and energy to improve prescribing practices.

Phase 2 – conceptual refinement and item modification All the items in the instrument were reviewed and checked for content validity by experts who were knowledgeable in instrument design and familiar with health care transparency regulations [45]. The content validity was established using content validity index (CVI), and the items with CVIs that have not reached 0.78 were removed [46]. Results showed that the CVI of three items did not reach 0.78 and were removed. Overall, 17 items were retained after phase 2. In addition, a pilot study was conducted involving 30 selected GPs smoothly; the respondents considered the items as easy to understand, and no ambiguous tautology or ambiguity was found.

Phase 3 – survey data collections The face-to-face survey involving all eligible GPs was conducted in April 2014. The survey was used to collect the large-scale data for instrument validation. All items were measured by a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Finally, 161 responses were received.

Phase 4 – data analysis and instrument validation

Instrument We used a systematic four-phase process to develop and validate the instrument of the transparency mechanism model based on Xia [37]. Phase 1 – conceptual development and initial item generation In this phase, the conceptual framework and an initial pool of instrument items were developed through literature reviews and focus group discussions. A total of 20 items were generated and served as the initial pool of items. High information accessibility means that the information is easy to access, has the right level of details, made of different types and convenient to use in a form [38,39]. This study designed the items mainly according to the definition. Based on the definition of 946

Exploratory factor analysis was conducted to examine the factor structure and to validate the reliability and construct validity of the instrument. The KMO (0.778) and Bartlett’s test (P < 0.05) indicated the suitability of the survey data to factor analysis. Reliability was indicated by Cronbach’s alpha and all Cronbach’s alpha for the five constructs were higher than 0.70 indicating adequate levels of reliabilities [47] (Table 1). Principle component method with varimax rotation was used in exploratory factor analysis. As shown in Table 1, five constructs with eigenvalues greater than one emerged from the analysis, which explained 70.46% of the variance. No cross-loaded items had loadings greater than 0.30. Overall, these results provided the initial empirical support to the convergent and discriminant validity of the instrument. We also calculated the average variance extracted (AVE) and composite reliability (CR) to test the instrument’s convergent and discriminant validities. All AVE and CR


X. Du et al.


Table 1 Results of factor structure and instrument validation

Constructs

Items

Standard factor loading

Information accessibility

1. 2. 3. 4. 5. 6.

0.77 0.93 0.82 0.84 0.79 0.72

Perceived value

Attitude

Perceived risk

Behavioural intention

The publicly reported prescription quality indicators are accessible. The publicly reported prescription quality indicators have appropriate details. The publicly reported prescription quality indicators are of different types. The prescription quality indicators have appropriate public reporting form. Public reporting of prescription quality indicators can give me sense of accomplishment. Public reporting of prescription quality indicators can motivate me to learn more knowledge about the prescribing practice. 7. Public reporting of prescription quality indicators can improve my prescribing practice. 8. It is necessary to publicly report drug reimbursement data. 9. It is necessary to publicly report drug prices data. 10. It is necessary to publicly report prescription quality indicators. 11. Public reporting of prescription quality indicators will reduce my income because my outpatient visit volume will decline. 12. Public reporting of prescription quality indicators will lead to a loss of status for me because my friends and relatives will think less highly of me. 13. Public reporting of prescription quality indicators will lead to my psychological loss because of the inconsistency of my self-image. 14. I am likely to prescribe fewer antibiotics. 15. I am likely to prescribe fewer flu antibiotics. 16. I am likely to prescribe fewer injections. 17. I am likely to reduce outpatient costs.

0.74 0.78 0.83 0.70 0.87

Cronbach’s alpha

AVE

CR

0.90

0.71

0.91

0.78

0.56

0.79

0.80

0.60

0.81

0.73

0.52

0.75

0.77

0.50

0.79

0.75 0.48 0.41 0.77 0.86 0.72

AVE, average variance extracted; CR, composite reliability.

were higher than 0.5 and 0.7, respectively, and indicated that the instrument has a perfect convergence and discriminant validity [48] (Table 1). The final instrument consisted of five constructs and 17 items: information accessibility (item 1 to item 4), perceived value (item 5 to item 7), attitude (item 8 to item 10), perceived risk (item 11 to item 13) and behavioural intention (item 14 to item 17). The survey also included seven demographic variables: gender, age, working experience, education levels, job title, monthly income and weekly working time (Table 1).

(No. IORG 0003571). Permit was also issued by the local health bureau and the township hospital’s Institutional Review Board. The personal data of each participant were kept anonymous and confidential and were used only for research purposes in compliance with the spirit of the Declaration of Helsinki, 2008.

Results Description of the study sample

The SPSS 12.0 (SPSS, Inc., Chicago, IL, USA) and LISREL 8.70 (Scientific Software International, Inc., Chicago, IL, USA) statistical software packages were used for data analysis and processing. Descriptive statistical analysis was used to determine the sample characteristics, and SEM was used to test the hypothesized model. In the SEM analysis, multiple items were summed together for each construct. Seven common model fitting indexes were used to assess the model’s overall goodness of fit: the ratio of χ2 to degrees of freedom (χ2/d.f.), root mean square error of approximation, goodness of fit index, adjusted goodness of fit index, comparative fit index, normalized fit index and incremental fit index [49].

A total of 160 questionnaires were distributed in October 2014 and all of them were collected, the response rate was 100%. Table 2 includes the demographic characteristics of the study sample in the SEM model. More male GPs (63.75%) participated than female GPs (36.25%), and 30- to 50-year-old GPs accounted for 70%. Most GPs (75%) have worked for more than 10 years; 15% of whom have worked for more than 30 years. In terms of educational level, a minority of the respondents (21.88%) have completed bachelor’s education or higher. In terms of job title, majority (93.12%) has a primary title or lower. The income of most GPs (78.75%) was less than 2500 RMB yuan per month (USD 403, the exchange rate was US $1 = RMB 6.21 in May 2015), but their working time was longer, and nearly 75% GPs worked for more than 40 hours a week.

Ethical considerations

Descriptive statistics of the construct scoring

This study was approved by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology

The results of the descriptive statistics (Table 3) show that the perceived risk of GPs has the lowest average scoring rate, account-

Statistical analysis


947


X. Du et al.

Table 2 Respondents’ demographic characteristics Number Gender Male Female Age (year) 60 Experience (year) 30 Educational level High school or less Some college or associates Bachelors or higher Title Not certified Certified doctor House doctor Doctor in charge of a case Assistant director doctor and above Monthly income (RMB yuan) 3000 Weekly working time (hour) 60

%

63.75 36.25

Index name

Evaluating standard

26 75 37 18 4

16.25 46.88 23.12 11.25 2.50

χ /d.f. RMSEA GFI AGFI CFI NFI IFI

0.80 >0.90 >0.90 >0.90

40 67 29 24

25.00 41.88 18.12 15.00

40 85 35

25.00 53.12 21.88

11 32 55 49 11

6.88 20.00 34.38 30.63 6.88

29 55 42 22 12

18.12 34.38 26.25 13.75 7.50

41 65 31 23

25.63 40.62 19.38 14.37

102 58

2

Table 3 Results of the descriptive statistics of each construct scoring

Construct

Number of items

Range of summed score

Mean score (SD)

Average scoring rate %

Information accessibility Perceived value Attitude Perceived risk Behavioural intention

4 3 3 3 4

0–20 0–15 0–15 0–15 0–20

14.71 (2.23) 11.58 (2.27) 12.36 (2.38) 6.46 (2.31) 16.19 (2.92)

73.55 77.17 82.38 43.08 80.97

Average scoring rate = mean score/summed score, denoting each construct’s score level. SD, standard deviation.

ing for 43.08% of the summed score of this construct. The lowest average scoring rate indicated that the risk awareness of GPs was weak and that their perceived risk mean score was at a low level. For other relevant factors, the average scoring rate of information accessibility, perceived value, behavioural intention and attitude 948

Table 4 Evaluating standards and the evaluating result of the fitness of SEM Actual fitting value

Result

2.11 0.08 0.85 0.80 0.95 0.91 0.95

Ideal Ideal Not bad fit Ideal Ideal Ideal Ideal

SEM, structural equation model; χ2/d.f., ratio of χ2 to degrees of freedom; RMSEA, root mean square error of approximation; GFI, goodness of fit index; AGFI, adjusted goodness of fit index; CFI, comparative fit index; NFI, normalized fit index; IFI, incremental fit index.

were higher, accounting for 73.55%, 77.17%, 80.97% and 82.38% of each summed score of the construct, respectively.

Structural equation model and hypotheses testing As Table 4 illustrates, all the eight hypotheses in this study were tested by SEM, and the model fit indexes were aligned with the theories and to be an acceptable model statistically [50]. All hypothesized propositions were simultaneously tested using SEM. The results support H1 as information accessibility has significantly positive effects on perceived value (r = 0.63, P < 0.001). Information accessibility did not significantly affect perceived risk (r = −0.09, P > 0.1); hence, H2 is not supported. However, information accessibility significantly influenced attitude (r = 0.32, P < 0.001), thus supporting H5. We also found that attitude has a significantly positive effect on perceived value (r = 0.45, P < 0.001) but did not significantly influence perceived risk (r = 0.01, P > 0.1); thus, only H6 is supported. The results also indicated that significant relationships exist between behavioural intention and both perceived value and attitude (r = 0.28, P < 0.05; r = 0.36, P < 0.001); thus, H3 and H8 are supported. Behavioural intention was negatively influenced by perceived risk (r = −0.07, P > 0.1) but had no statistical significance; therefore, H4 is not supported. Figure 2 shows the standardized LISREL path coefficients.

Discussions The instrument in the present study, which was developed based on literature reviews, focus group discussions and expert opinions, was tested using factor analysis. The instrument complied with the fundamental issues of instrument design, including question wording, question order and other aspects [51]. The developed instrument exhibited satisfactory reliability and validity, and the fit index of SEM was also at an acceptable level. The abovementioned indicators underscored the credibility of the results. Structural relationship analysis indicated that perceived value significantly influenced behavioural intention and was influenced


X. Du et al.


Perceived Value

r =0.28*

r =0.63*** r =0.45*** r =0.32***

Information

Attitude

r =0.36***

Accessibility

Behavioural Intention

r =0.01 r =–0.09 Perceived Risk

Figure 2 Final transparency mechanism model based on information accessibility, perceived value, attitude, perceived risk and behavioural intention.

Note: * P < 0.05; ** P < 0.01; *** P < 0.001

by information accessibility at the same time. This finding is consistent with TAC conducted by Fung. When transparency systems provide highly relevant, accessible, and comprehensible information and the cost of acquiring and using new information is sufficiently low, the users of the information may consider the information as valuable and may embed it in their decision-making processes to influence their behaviour [52]. The information perception, which leads to behavioural change, is the key aspect in transparent regulation. Previous studies showed that perceived value has a significantly direct effect on behavioural intention [28,53]. Hence, the accessibility of prescription quality indicators and the perceived value of GPs should be emphasized in the improvement of prescribing practices. Furthermore, SEM results indicated that the perceived value is largely related to attitude. This finding is consistent with those of previous studies, which state that a positive and significant relationship between attitude and perception exists and that a person’s perception is affected by his attitude [54,55]. The results also showed that information accessibility significantly influenced attitude, and attitude significantly influenced behavioural intention. The results of the present study are consistent with previous research results. A study suggested that information about the biological cause of homosexuality can have both positive and negative effects on the attitudes towards homosexuals [56]. The literature found that knowledge (information) directly affects attitude and that knowledge (information) affecting practice through attitude is better than knowledge affecting practice directly [33]. Another study showed that the participants’ attitudes towards their favourite clothing websites has a direct, positive effect on their intentions to search for information in those websites as well as on their intentions to purchase clothing items [57]. In addition, a study indicated that attitude is important in predicting an individual’s behaviour and significantly influences behavioural intention [58]. As a result, increasing the accessibility of prescription quality indicators can create a positive attitude on transparency regulation, which would increase the intention of GPs to improve prescribing practices. In the present study, H2, H4 and H7 were not supported; this finding is inconsistent with previous studies. Crocker [59] indicated that the amount of information available is the element that affects perceived risk. A study showed that risk plays an essential


r =–0.07

role in consumer behaviour, and it makes a valuable contribution towards explaining information-searching behaviour and consumer purchase decision making [60]. Another study stated that perceived risk reduces the intention of consumers to buy goods over the Internet [61]. However, the present results showed that the perceived risk of GPs was at a very low level and had no significant relationship with behavioural intention. The reasons may include the following: (1) GPs did not perceive the risk despite public reporting of prescription quality indicators. Several studies indicated that the way people deal with risk depends largely on their level of perception and that risk education could increase risk perception [62,63]. In China, the lack of a comprehensive occupational risk education system and the low occupational risk education coverage for medical personnel hinder GPs from perceiving risk [64]. (2) Although GPs perceived the risk, they have no stress or motivation to improve their prescribing practices. On the one hand, the public reporting policy is not associated with any formal incentives in Q City. A research indicated that attaching incentives to public reports can maximize the effectiveness of public reporting [65]. The absence of rewards and penalties may be an important reason for the low level of perceived risk of GPs, and GPs have no desire to improve their prescribing practices. On the other hand, a previous study showed that more than 75% of patients who aware of public report information but did not used it to choose providers [66]. The less stress brought about by fewer choices may account for the low level of perceived risk and the lack of an effect on behavioural intention.

Conclusions This study has successfully shown the existence of causal relationships between information accessibility, perceived value, attitude and perceived risk of GPs in relation to their prescribing behavioural intention. The strong effect of the relationship among information accessibility, perceived value and attitude on the prescribing behavioural intention of GPs emphasizes the importance of making prescription quality indicators more transparent and of paying attention to the perceived values and attitudes of GPs. Perceived risk did not influence behavioural intention in the 949


present study. We suggest that policymakers need to strengthen the risk education of GPs and attach incentives to the public report policy. The study is useful in understanding the transparency function mechanism, that is, the response of GPs to the public reporting of prescription quality indicators. Based on this study, future research should investigate the organizational determinants of transparency mechanism. Consequently, organizations can then maximize improve the efficacy and effectiveness of transparent regulation. Another promising future research direction is to study the robustness of the transparency mechanism using several panel data. Our study had several limitations. First, the study was based on a limited number of respondents in a county of the Hubei Province. Further research is needed to validate our transparency mechanism model and to test the generalizability of our empirical findings in China, other countries, and other professions. Second, despite our effort at maintaining strict anonymity to encourage unvarnished responses, our use of a survey may have led to inaccurate responses because the respondents could have been socially impelled to provide desirable answers.

Acknowledgements This research is supported by the National Natural Science Foundation of China (Grant Number: 71373092). The authors would like to thank the National Natural Science Foundation of China for the funding of this research, the GPs participating in this study and the Health Bureau of Q City. They also thank Lianping Yang, Chunyan Yang, Xi Yin, Shiru Yang, Yuqi Xiong, Fangying Zhong, Chenxi Liu, Xiaofei Zheng, Dan Wang, Fei Cai, Jing Ye, Jie Wan and Chao Yan who helped collect data of this study.

References 1. Hale, T. N. (2008) Transparency, accountability, and global governance. Global Governance, 14 (1), 73–94. 2. Weil, D., Fung, A., Graham, M. & Fagotto, E. (2006) The effectiveness of regulatory disclosure policies. Journal of Policy Analysis and Management, 25 (1), 155–181. 3. Jae-Oh, L. (2010) Anti-Corruption and Transparency: South Korea’s Pathway to Preeminence in the Asia-Pacific. Washington: Johns Hopkins University. 4. Brown, D. L., Clarke, S. & Oakley, J. (2012) Cardiac surgeon report cards, referral for cardiac surgery, and the ethical responsibilities of cardiologists. Journal of the American College of Cardiology, 59 (25), 2378–2382. 5. Jung, K. (2010) The impact of information disclosure on quality of care in HMO markets. International Journal for Quality in Health Care, 22 (6), 461–468. 6. Gravelle, H. & Sivey, P. (2010) Imperfect information in a qualitycompetitive hospital market. Journal of Health Economics, 29 (4), 524–535. 7. Health Insurance Review & Assessment Service (2010) Comprehensive Quality Report of National Health Insurance 2010. Available at: http://www.hira.or.kr/eng/news/02/1212222_25619.html (last accessed 7 January 2014). 8. National Health and Family Planning Commission of the People’s Republic of China (2009) Opinions on deepening medical and health system reform. Available at: http://www.nhfpc.gov.cn/tigs/s7846/ 200904/902e526039424a2ca87e8d78f307e9f5.shtml (last accessed 17 May 2015).

950

X. Du et al.

9. The Central People’s Government of the People’s Republic of China (2010) Standardized regulation of hospital prescription comments (trial implementation). Available at: http://www.gov.cn/gzdt/2010-03/ 04/content_1547080.htm (last accessed 17 May 2015). 10. Hsiao, W., Li, M. Q. & Zhang, S. H. (2014) Universal health coverage: the case of China. Geneva, The United Nations Research Institute for Social Development. 11. Finkelstein, J. A., Davis, R. L., Dowell, S. F., et al. (2001) Reducing antibiotic use in children: a randomized trial in 12 practices. Pediatrics, 108 (1), 1–7. 12. Gerber, J. S., Prasad, P. A., Fiks, A. G., Localio, A. R., Grundmeier, R. W., Bell, L. M., Wasserman, R. C., Keren, R. & Zaoutis, T. E. (2013) Effect of an outpatient antimicrobial stewardship intervention on broad-spectrum antibiotic prescribing by primary care pediatricians: a randomized trial. Journal of the American Medical Association, 309 (22), 2345–2352. 13. Yang, L. P., Liu, C. J., Wang, L. J., Yin, X. & Zhang, X. P. (2014) Public reporting improves antibiotic prescribing for upper respiratory tract infections in primary care: a matched-pair cluster-randomized trial in China. Health Research Policy and Systems, 12 (61), 1–9. 14. Wang, X., Tang, Y. Q., Zhang, X. P., Yin, X., Du, X. & Zhang, X. P. (2014) Effect of publicly reporting performance data of medicine use on injection use: a quasi-experimental study. PLoS ONE, 9 (10), 1–7. 15. Zhang, X. P., Wang, L. J. & Zhang, X. P. (2014) Application of propensity scores to explore the effect of public reporting of medicine use information on rational drug use in China: a quasi-experimental design. BMC Health Services Research, 14 (1), 1–7. 16. Fung, C. H., Lim, Y. W., Mattke, S., Damberg, C. & Shekelle, P. G. (2008) Systematic review: the evidence that publishing patient care performance data improves quality of care. Annals of Internal Medicine, 148 (2), 111–123. 17. Chen, J. (2010) Public reporting improves healthcare. Available at: http://www.bhi.nsw.gov.au/__data/assets/pdf_file/0005/168368/ Public-Reporting-Improves-Health-Care.pdf (last accessed 7 January 2014). 18. Ikkersheim, D. & Koolman, X. (2013) The use of quality information by general practitioners: does it alter choices? A randomized clustered study. BMC Family Practice, 14 (95), 1–7. 19. Berwick, D. M., James, B. & Coye, M. J. (2003) Connections between quality measurement and improvement. Medical Care, 41 (1 Suppl. ), I30–I38. 20. Ketelaar, N. A., Faber, M. J., Flottorp, S., Rygh, L. H., Deane, K. H. & Eccles, M. P. (2011) Public release of performance data in changing the behaviour of healthcare consumers, professionals or organisations. Cochrane Database of Systematic Review, 11, 1–62. 21. Davies, H. T. O. (2001) Public release of performance data and quality improvement: internal responses to external data by US health care providers. Quality in Health Care, 10 (2), 104–110. 22. Totten, A. M., Wagner, J., Tiwari, A. O., Haire, C., Griffin, J. & Walker, M. (2012) Closing the Quality Gap: Revisiting the State of the Science, Vol. 5: Public Reporting as a Quality Improvement Strategy. Rockville: Agency for Healthcare Research and Quality Publication. 23. McGivern, G. & Fischer, M. D. (2012) Reactivity and reactions to regulatory transparency in medicine, psychotherapy and counselling. Social Science & Medicine, 74 (3), 289–296. 24. Ketelaar, N. A., Faber, M. J., Elwyn, G., Westert, G. P. & Braspenning, J. C. (2014) Comparative performance information plays no role in the referral behaviour of GPs. BMC Family Practice, 15, 146. 25. Sherman, K. L., Gordon, E. J., Mahvi, D. M., Chung, J., Bentrem, D. J., Holl, J. L. & Bilimoria, K. Y. (2013) Surgeons’ perceptions of public reporting of hospital and individual surgeon quality. Medical Care, 51 (12), 1069–1075.


X. Du et al.

26. Weil, D. & Kennedy, J. F. (2002) The benefits and costs of transparency: a model of disclosure based regulation. Boston, Boston University, 1–43. 27. Fung, A., Graham, M., Weil, D. & Fagotto, E. (2005) From food to finance: what makes disclosure policies effective? Taubman Centre Policy Briefs, 1–4. 28. Chen, H. S., Chen, C. Y., Chen, H. K. & Hsieh, T. (2012) A study of relationships among green consumption attitude, perceived risk, perceived value toward hydrogen-electric motorcycle purchase intention. AASRI Procedia, 2, 163–168. 29. Chen, W. (2013) Perceived value of a community supported agriculture (CSA) working share. The construct and its dimensions. Appetite, 62, 37–49. 30. Bluedorn, A. C. (1982) A unified model of turnover from organizations. Human Relations, 35 (2), 135–153. 31. Shore, L. M. & Martin, H. J. (1989) Job satisfaction and organizational commitment in relation to work performance and turnover intentions. Human Relations, 42 (7), 625–638. 32. Price, J. L. (2001) Reflections on the determinants of voluntary turnover. International Journal of Manpower, 22 (7), 600–624. 33. Chien-Yun, D., Wan-Fei, C., Yu-Hsi, Y. & Chia-Hung, Y. (2011) A study on modification of knowledge, attitude and practice on vocational high school electronics courses integrated with nanotechnology concept. International Journal of Thermal and Environmental Engineering, 4 (1), 73–79. 34. Pickens, J. (2005) Attitudes and perceptions. In Organizational behavior in health care (ed. N. Borkowski), pp. 43–76. Sudbury: Jones and Bartlett Publishers. 35. Ji, S. Q. (2013) Factors of audience information dissemination behavior: conceptual development and empirical study [in Chinese]. Journal of Intelligence, 32 (3), 30–36. 36. Crosby, R. A., Kegler, M. C. & DIClemente, R. J. (2002) Understanding and applying theory in health promotion practice and research. In Emerging Theories in Health Promotion Practice and Research: Strategies for Improving Public Health (eds. R. J. DiClemente, R. A. Crosby, M. C. Kegler), pp. 1–18. San Francisco: Jossey-Bass. 37. Xia, W. D. & Lee, G. (2005) Complexity of information systems development projects: conceptualization and measurement development. Journal of Management Information Systems, 22 (1), 45–83. 38. Teo, H. H., Chan, H. C., Wei, K. K. & Zhang, Z. J. (2003) Evaluating information accessibility and community adaptivity features for sustaining virtual learning communities. International Journal of HumanComputer Studies, 59 (5), 671–697. 39. Fidel, R. & Green, M. (2004) The many faces of accessibility: engineers’ perception of information sources. Information Processing & Management, 40 (3), 563–581. 40. Sweeney, J. C. & Soutar, G. N. (2001) Consumer perceived value: the development of a multiple item scale. Journal of Retailing, 77 (2), 203–220. 41. Sánchez, J., Callarisa, L., Rodríguez, R. M. & Moliner, M. A. (2006) Perceived value of the purchase of a tourism product. Tourism Management, 27 (3), 394–409. 42. Tsiakis, T. (2012) Consumers’ issues and concerns of perceived risk of information security in online framework. The marketing strategies. Procedia – Social and Behavioral Sciences, 62, 1265–1270. 43. Chang, E. & Tseng, Y. (2013) Research note: e-store image, perceived value and perceived risk. Journal of Business Research, 66 (7), 864– 870. 44. Zeithaml, V. A., Berry, L. L. & Parasuraman, A. (1996) The behavioral consequences of service quality. Journal of Marketing, 60 (2), 31–46. 45. Hardesty, D. M. & Bearden, W. O. (2004) The use of expert judges in scale development. Journal of Business Research, 57 (2), 98–107.



46. Zaichkowsky, J. L. (1985) Measuring the involvement construct. Journal of Consumer Research, 12 (3), 341–352. 47. Slavec, A. & Mateja, D. (2012) A perspective on scale development in entrepreneurship research. Economic and Business Review, 14 (1), 39–62. 48. Fornell, C. & Larcker, F. D. (1981) Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18 (1), 39–50. 49. Hsu, I. Y., Su, T., Kao, C., Shu, Y., Lin, P. & Tseng, J. (2012) Analysis of business safety performance by structural equation models. Safety Science, 50 (1), 1–11. 50. Wu, L. H., Zhong, Y. Q., Shan, L. J. & Qin, W. (2013) Public risk perception of food additives and food scares. The case in Suzhou, China. Appetite, 70, 90–98. 51. Murray, P. (1999) Fundamental issues in questionnaire design. Accident & Emergency Nursing, 7 (3), 148–153. 52. Fung, A., Weil, D., Graham, M. & Fagotto, E. (2004) The political economy of transparency: what makes disclosure policies effective? Ash Institute for Democratic Governance and Innovation, John F. Kennedy School of Government, Harvard University, 1–45. 53. Gao, T. T., Leichter, G. & Wei, Y. S. (2012) Countervailing effects of value and risk perceptions in manufacturers’ adoption of expensive, discontinuous innovations. Industrial Marketing Management, 41 (4), 659–668. 54. Vroom, V. H. (1960) The effects of attitudes on perception of organizational goals. Human Relations, 13 (3), 229–240. 55. Engel, M. T., Marchini, S., Pont, A. C., Machado, R. & Oliveira, L. R. D. (2014) Perceptions and attitudes of stakeholders towards the wildlife refuge of Ilha dos Lobos, a marine protected area in Brazil. Marine Policy, 45, 45–51. 56. Oldham, J. & Kasser, T. (1999) Attitude change in response to information that male homosexuality has a biological basis. Journal of Sex & Marital Therapy, 25 (2), 121–124. 57. Seock, Y. K. & Norton, M. (2007) Attitude toward internet web sites, online information search, and channel choices for purchasing. Journal of Fashion Marketing and Management, 11 (4), 571– 586. 58. Kim, Y. J., Chun, J. U. & Song, J. (2009) Investigating the role of attitude in technology acceptance from an attitude strength perspective. International Journal of Information Management, 29 (1), 67– 77. 59. Crocker, K. E. (1986) Influence of the amount and type of information on individuals’ perception of legal services. Journal of the Academy of Marketing Science, 14 (4), 18–27. 60. Corbitt, B. J., Thanasankit, T. & Yi, H. (2003) Trust and e-commerce: a study of consumer perceptions. Electronic Commerce Research and Applications, 2 (3), 203–215. 61. Barnes, S. J., Bauer, H. H., Neumann, M. M. & Huber, F. (2007) Segmenting cyberspace: a customer typology for the internet. European Journal of Marketing, 41 (1/2), 71–93. 62. Ryckman, L. (2005) Effects of education on risk perception a study of Hurricane Katrina cleanup workers. Providence, Rhode Island, 1–65. 63. Boice, J. D. (2013) Risk assessment, risk perception and risk management. International Academic Conference. Fukushima, Japan. 64. Qian, Y., Xu, Y. & Wang, J. E. (2013) Analysis of the status quo and necessity of medical occupational risks to medical students [in Chinese]. Chinese Hospital Management, 33 (11), 42–43. 65. Marshall, M. N., Shekelle, P. G., Davies, H. T. O. & Smith, P. C. (2003) Public reporting on quality in the United States and the United Kingdom. Health Affairs, 22 (3), 134–148. 66. Harris, K. & Buntin, M. (2008) Choosing a health care provider: the role of quality information. Research Synthesis Report, 14, 1:25.

951