Effect of multifaceted social norms on physicians’ use of clinical practice guidelines on antimicrobials: Evidence from secondary and tertiary general hospitals in central-western China

Abstract

Background
With the improper use of antimicrobials becoming a major public health concern globally, poor compliance of clinical practice guidelines(CPGs) on antimicrobials is still prominent, especially in less developed regions. Although social norms have received increasing attention as the determinants of physicians’ CPGs use, most studies set forth only a single level of social norms. Therefore, this study aims to investigate the impact of multifaceted social norms on physician’ use of CPGs on antimicrobials, and further reveal the temporal effects of social norms.

Methods
Based on integration of Theory of Planned Behavior and Theory of Normative Social Behavior, a questionnaire survey was conducted covering social norms at individual level (subjective norms), organizational level (organization criterion) and social level (social identity), as well as other potential factors (attitudes, behavioral intention, etc) for the use of CPGs on antimicrobials. Data were collected by multi-stage random sampling from 502 physicians in secondary and tertiary general hospitals in central-western China. Structural equation model (SEM) was used to link the three-level factors with physician's behavior. And with reflected by the moderating effects of professional titles in this study, the temporal effects of social norms were examined by multi-group SEM.

Results
Nearly 70% of the participants had a good practice of using CPGs on antimicrobials. Reliability and validity analysis shows that the questionnaire developed from the theoretical model is acceptable. Subjective norms, organization criterion and social identity were linked to higher behavioral intentions(β = 0.212, p < 0.01; β = 0.254, P < 0.01; β = 0.212, P < 0.01). The direct effect of behavior intentions on physicians' practice was 0.822, and the indirect effects of subjective norms, organizational criterion and social identity on practice were 0.308, 0.236 and 0.235. The effects of organization criterion and social identity on behavior were moderated by the professional title, and regarding effects would be weakened with the rise of professional title.

Conclusion
This study reveals the importance of multifaceted social norms in enhancing physicians’ use of CPGs on antimicrobials and the moderating effects of professional titles on the role of social norms at organizational level (organizational criterion) and social level (social identity).

Full text

Page 1/25
Effect of multifaceted social norms on physicians’
use of clinical practice guidelines on antimicrobials:
Evidence from secondary and tertiary general
hospitals in central-western China
Lingjie Wang
Fujian Medical University
Wenbin Liu (  wenbinliu126@126.com )
Fujian Medical University
Research Article
Keywords: Antimicrobials, Clinical practice guidelines, Social norms, Structural equation modeling, China
Posted Date: September 15th, 2023
DOI: https://doi.org/10.21203/rs.3.rs-3344140/v1
License:   This work is licensed under a Creative Commons Attribution 4.0 International License. 
Read Full License

Page 2/25
Abstract
Background
With the improper use of antimicrobials becoming a major public health concern globally, poor
compliance of clinical practice guidelines(CPGs) on antimicrobials is still prominent, especially in less
developed regions. Although social norms have received increasing attention as the determinants of
physicians’ CPGs use, most studies set forth only a single level of social norms. Therefore, this study
aims to investigate the impact of multifaceted social norms on physician’ use of CPGs on antimicrobials,
and further reveal the temporal effects of social norms.
Methods
Based on integration of Theory of Planned Behavior and Theory of Normative Social Behavior, a
questionnaire survey was conducted covering social norms at individual level (subjective norms),
organizational level (organization criterion) and social level (social identity), as well as other potential
factors (attitudes, behavioral intention, etc) for the use of CPGs on antimicrobials. Data were collected by
multi-stage random sampling from 502 physicians in secondary and tertiary general hospitals in central-
western China. Structural equation model (SEM) was used to link the three-level factors with physician's
behavior. And with reected by the moderating effects of professional titles in this study, the temporal
effects of social norms were examined by multi-group SEM.
Results
Nearly 70% of the participants had a good practice of using CPGs on antimicrobials. Reliability and
validity analysis shows that the questionnaire developed from the theoretical model is acceptable.
Subjective norms, organization criterion and social identity were linked to higher behavioral intentions(β =
0.212, p < 0.01; β = 0.254, P < 0.01; β = 0.212, P < 0.01). The direct effect of behavior intentions on
physicians' practice was 0.822, and the indirect effects of subjective norms, organizational criterion and
social identity on practice were 0.308, 0.236 and 0.235. The effects of organization criterion and social
identity on behavior were moderated by the professional title, and regarding effects would be weakened
with the rise of professional title.
Conclusion
This study reveals the importance of multifaceted social norms in enhancing physicians’ use of CPGs on
antimicrobials and the moderating effects of professional titles on the role of social norms at
organizational level (organizational criterion) and social level (social identity).

Page 3/25
Background
Antimicrobials play a vital role in the treatment, which can effectively reduce the incidence and mortality
of bacterial infections (1, 2). However, with the widespread unnecessary and excessive use of
antimicrobials, the challenge of antibiotic resistance (AMR) has become a major global public health
problem and caused a series of problems, including increased mortality, prolonged hospital stay, and
increased economic burden on patients, causing a variety of negative effects on individuals and society
(3–6).
Although the effectiveness of the Clinical Practice Guidelines (CPGs) on antimicrobials has been fully
demonstrated in standardizing clinical treatment and improving prescription behavior, and well expect in
preventing the further deterioration of AMR and promoting patient health, the expansion of the regarding
guidelines implementation was still stagnated with poor compliance. For example, China, one of the
largest consumers of antimicrobial drugs, has issued the Guiding Principles for the Clinical Application of
antimicrobials since 2015, and intends to reduce inappropriate antimicrobial prescription by auditing and
testing the prescription behavior of physicians (7). However, the per capita use of antimicrobials is still
much higher than the international level of 30% (8). Especially in the less developed regions of China,
namely the central and western part, the phenomena of antibiotic overuse is more common (9–11). To
reverse this serious situation, the underlying causes are worthy of further investigation.
Social norms are dened as informal and shared behavioral rules, such as pressure, beliefs or emotions
of individuals, culture or features of social groups, mandatory guidelines or regulations and so on. As
implicit or explicit behavioral standards widely perceived and recognized by members of a certain group
in a specic situation, social norms play an important role in behavioral decision-making, and even affect
behavior by affecting personal personality traits. These action paths had been widely conrmed by
classical theories or frameworks in the eld of social psychology, such as the theory of planned behavior
expansion (TPB) and the theory of normative social behavior (TNSB) (12–14). Furthermore, according to
the different sources, social norms could be divided into individual level (subjective norms),
organizational level (organization criterion) and social level (social identity), and the roles of social norms
at different levels are also different (15, 16).
Attributing to the high professionalism and normativity of physician group, the social norms’ role on
physicians' cognition and clinical practice have received increasing attention (17, 18). Some efforts also
have been made to determine the impact mechanism of social norms on antimicrobials prescription or
regarding CPGs compliance (19). Hallsworth’s study found that providing social norm feedback from a
high-prole messenger to high-prescription antimicrobials in general practice could effectively reduce the
proportion of antimicrobials in prescriptions (16). And under the guidance of TPB, Fatemeh et al.found
that social norm feedback interventions based on subjective norms improved antimicrobials prescribing
behavior that lasted at least three months after the intervention (20). Furthermore, by integrating TPB and
other related theoretical models, Deng et al. found that subjective norms had signicant direct and
indirect effects on intention to use CPGs on antimicrobials, and the impact of subjective norm on

Page 4/25
individual attitude was also conrmed (21). It is worth noting that although recently social norms have
received special attention as determinants of antimicrobials prescribing behaviors or regarding CPGs
compliance, most studies have only proposed a single level of social norms: either individual normative
beliefs, or acceptance pressure at the social normative level (22–26). It may lead to failure of fully
identifying the different roles of social norms at different levels. Besides, some scholars have pointed out
that unlike formal institutions and top-down regulations, social norms are usually strengthened or
weakened over time(27) However, there are few studies have focused on the strengthening or weakening
effects of other factors on social norms for prescribing behavior, such as the temporal effects.
Therefore, to bridge these knowledge gaps, this study aimed to construct a theoretical framework based
on TPB and TNSB theories to comprehensively investigate the effect of social norms at multiple levels for
physician’ use of CPGs on antimicrobials, as well as determine whether social norms would be
strengthened or weakened by the time of action. These ndings not only add evidence to the impact
mechanism of social norms on improving CPGs compliance with antimicrobials, but also provide a
reference for tailoring social normative strategies to promote the implementation of CPGs on
antimicrobials in future.
Methods
Study setting
This study was conduct in secondary and tertiary hospitals of central and western China. Since the
provision of vast medical services is heavily dependent on tertiary and secondary hospitals, they were the
major consumers of antimicrobials with widespread irrational use phenomenon. Especially in central and
western China, this situation is more prominent. For example, the use rate of prophylactic antimicrobials
for inpatients was 54.6% in tertiary hospitals and 60.7% in secondary hospitals, both higher than the
national standard of 30% (28).Thus, for reducing AMR and promoting patients’ health, it is necessary to
regulate the physicians’ use of antimicrobials preferentially in secondary and tertiary hospitals of
regarding regions.
Theoretical framework
Since individual behavior is usually constrained by relevant social norms at multiple levels, the theoretical
framework of this study was adapted from the integration of TPB and TNSB to comprehensively
investigate the effect of social norms at the individual level (subjective norms), organizational level
(organization criterion), social level (social identity) on physician’ use of CPGs on antimicrobials. And the
temporal effects of social norms were also taken into account. Figure1 illustrates the theoretical
framework.
{insert Fig.1 here}
Social norms at individual level

Page 5/25
Subjective norms are social norms at individual level, which refers to social pressures that individuals
perceived as coming from parents, spouses, workmates, etc(29). As postulated by TPB, subjective norms
indirectly inuence behavior though behavioral intentions, which is a commonly known prerequisite of
nal behavior (30). Similar role are also played by attitude and perceived risk, while the former is an
individual's positive or negative evaluation of particular behaviors, and the latter reects the person’s
belief that an action is under his or her control, such as perceived risk (31, 32). Additionally, many studies
also showed that except for direct effects, subjective norms could inuence the behavioral intentions of
the physicians though attitude (22–24).
Social norms at organizational level
Organizational criterion is understood as prescribing or prohibiting behavioral norms or cultural
understandings of group members (33). TNSB believes that organizational criterion is one of the most
important factors inuencing the behavior intention of organization members, which subsequently
inuence the nal practice (14). Meanwhile, the indirect effect of organizational criterion on behavior
intention with attitude as a key intermediate factor was also conrmed in previous studies (21, 24).
Hospital managers could set organizational criterion to restrain and limit physicians' intention of
nonadherence to CPGs on antimicrobials. Also, organizational criterion could be used to slowly change
physicians' attitudes toward regarding CPGs compliance, which would bring effect on their intention and
nally the real practice.
Social norms at social level
At the social level, social identity refers to the recognition and approval of an individual from a wider
group in society, including peers, celebrities, and other groups (34, 35). Since individuals often desire
more approval from the outside world, social identity plays an important role in work and daily life. As it
increases, individuals become more accepting of their own behavior and the social norms they adhere to.
Research has demonstrated that the inuence of social identity on intention or behavior is mostly
moderated through attitudes, and it also has been conrmed that social identity is an independent
predictor of intention (36).
Temporal effects of social norms
Unlike formal instructions and top-down regulations, social norms cannot achieve the desired effect in a
short period of time and are usually strengthened or weakened over time, which has been also reported in
previous studies exploring social norms’ effects on physicians’ behavior (37). Accordingly, it is expected
that in the beginning, the effect of social norms on the improvement of physicians' antimicrobials
prescribing behavior were persisted and slowly strengthened over time. As time goes on, physicians
become tired and numb to the same social norms, and the role of social norms could be weakened after
reaching tipping points.
Measurements

Page 6/25
Based on previous research and reliable scales, this study developed a questionnaire containing four
parts, including utilization, social norms, other potential determinants and personal information
(Additional le 1). The measures used in the questionnaires follow the TPB and TNSB constructs, and
represent an adaptation to this specic context of those used in previous research work on social norms
or social inuence. More specically:
Physicians’ utilization of CPGs on antimicrobials
Part 1 covered 3 items to measure physicians’ utilization of CPGs on antimicrobials in the past year. This
part were measured using a 5-point Likert scale, where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 =
agree, and 5 = strongly agree.
Social norms
Part 2 covered social norms at three different levels, nemely subjective norm, organizational criterion and
social identity. Among them, subjective norm refers to the perceived social pressure to which a physician
is subject in relation to CPGs on antimicrobials. Organizational criterion reects the mandatory guidelines
of the physician's organization regarding CPGs on antimicrobials. Social identity represents the social
group's approval of the physician's use of CPGs. Each social norm was measured by three corresponding
items. Responses were asked to tick the number that best ts their real feelings on a ve-point Likert
scale labeled as follows: 1 (Strongly disagree), 2 (Disagree), 3 (Neutral), 4 (Agree), and 5 (Strongly agree).
Other potential determinants and personal information
Part 3 covered three potential determinants of physician behavior regarding the use of CPGs on
antimicrobials, including attitude, perceived risk and behavioral intention. Attitude reects the degree to
which a physician is in favor of the use of CPGs on antimicrobials. Perceived risk refers to how easy a
prescriber feels in making a rational decision on CPGs on antimicrobials. Behavioral intention represents
the degree to which a physician is willing to use CPGs on antimicrobials. In this part, respondents rated
on a ve-point Likert scale, ranging from “1 = strongly disagree” to “5 = strongly agree”.
Personal information
Part 4 was a personal information card with 6 items, including several basic characteristics of
participants as gender, age, education, professional title, department and years of practice. Since the dual
attributes of professional titles as different titles representing different social role and length of working
time in the organization, and the temporal effects of social norms cannot be intuitively measure and
analyze, this study choose the moderating effect of professional titles to reect the temporal effects of
social norms.
Sampling
Due to the differences in economic development levels among regions, a cross-sectional study was
implemented by applying a multistage sampling strategy. First, Hubei, Yunnan & Sichuan provinces were
randomly selected from central and western regions of China, respectively. Second, ve secondary and

Page 7/25
tertiary general hospitals were selected from each of these regions. Lastly, participants were chosen on
the basis of the department size. There were 16–20 physicians randomly sampled from major
departments of internal medicine and surgery, respectively. And 3–5 physicians were randomly selected
respectively for other minor departments, such as obstetrics and gynecology, ophthalmology, and
orthopedics. Thus, about 45–60 physicians were selected from each hospital, and at least 450
physicians would participate in the survey, which would fully meet the basic requirement that the sample
size should be set at least ve times survey question (38).
Data collection
With the support of sampled hospitals, each round for lling out the questionnaire was accompanied by
trained facilitators. The purpose of the study and the use of the data were explained in detail to the
participants through professionally trained investigators. Additionally, all responses were anonymous to
protect their privacy. Written informed consent was obtained from all participants for this study. Data
collection lasted from April 2018 to January 2019, and a total of 502 physicians from the mid-west were
included in this study.
Data analysis
Data analysis was performed using SPSS 21.0 and AMOS 25.0. To better estimate model utility,
structural equation modeling (SEM) was applied, allowing for the creation of latent variables and relaxing
assumptions about data distribution and error (39). The data processing steps were as follows. First,
descriptive statistics (mean, standard deviation, absolute and relative frequencies) were performed on the
demographic characteristics of the participants and the scores of each variable measured. Second,
Cronbach's α, factor loadings, and CR were employed to discriminate the reliability and validity of all
constructs. Finally, SEM was utilized to estimate the relationship and mechanism among potential
inuencing factors. The path coecients calculated by path analysis are equivalent to the standardized
regression coecients and direct effects. The mediating effects were calculated by bootstrapping and
were signicant if the value of the mediation effect does not include 0 within its 95% condence interval.
Multi-group SEM was used to examine the moderating effect of professional title. The analysis samples
were divided into two categories according to the level of professional titles. This study would compare
the mediating effect in different groups. A two-tailed p-value for the comparison between groups less
than 0.05 indicated the signicant differences between groups.
Results
Descriptive characteristics
Table1 demonstrates the characteristics of the 502 participants. Among them, 51.0% (n = 256) were
female and 58.0% (n = 291) were younger than 35 years old. Regarding educational level, many
participants reported having a master's degree (45.6%, n = 229), followed by bachelor (44.2%, n = 222).
The proportion of the participants with the professional titles of non-senior and senior was 77.9% and

Page 8/25
22.1%, respectively. As for years in practice, nearly 90% of the participants had less than 15 years of
practice experience. And physicians in the central and west regions accounted for 52.8%(n = 265) and
47.2%(n = 237), respectively.
Table 1
Demographic characteristics of participants
Variable Category Frequency Percentage (%)
Gender Male 246 49.0
 Female 256 51.0
Age < 35 years old 291 58.0
 35–44 years old 155 30.9
≥ 45 years old 56 11.1
Education Junior college or below 10 2.0
 Bachelor 222 44.2
 Master 229 45.6
 Doctor 41 8.2
Professional title Non-senior 391 77.9
 Senior 111 22.1
Years in practice < 5 years 159 31.7
 5–10 years 150 29.9
 11–15 years 136 27.1
 16–20 years 49 9.8
 > 20 years 8 1.6
Region Central 265 52.8
 West 237 47.2
{insert Table1 here}
Reliability and validity
As shown in Table2, Cronbach's α values for each dimension ranged from 0.799 to 0.893, all of which
were greater than the recommended threshold of 0.7, which indicated that the scale had good internal
consistency. Moreover, the convergent validity was assessed by factor loadings, average variance
extracted (AVE), and composite reliability (CR). Most of the items had factor loadings greater than 0.7,

Page 9/25
and the AVE and CR values for each item exceeded 0.5 and 0.7, implying a satisfactory convergent
validity.
According to the formell-larcker criterion(40), it is known that values of the diagonal are greater than the
element below the diagonal implying that the construct has good discriminant validity. The square root of
AVE for each construct in Table3 is greater than its correlation coecient with the other constructs, which
meant the discriminant validity of the measurement model was acceptable.
Table 2
Results of reliability and convergent validity analyses
Construct Item Factor loading Cronbach’s αAVE CR
Attitude ATT1 0.816 0.893 0.716 0.883
 ATT2 0.857   
 ATT3 0.864   
Behavioral intention BI1 0.780 0.882 0.599 0.817
 BI2 0.817   
 BI3 0.722   
Organization criterion OC1 0.685 0.799 0.590 0.811
 OC2 0.847   
 OC3 0.764   
Social Identity SI1 0.665 0.823 0.6321 0.836
 SI2 0.871   
 SI3 0.834   
Perceived risk PR1 0.699 0.807 0.592 0.812
 PR2 0.857   
 PR3 0.743   
Practice Pra1 0.814 0.874 0.699 0.874
 Pra2 0.841   
 Pra3 0.852   
Subjective norms SN1 0.855 0.893 0.735 0.893
 SN2 0.867   
 SN3 0.850   

Page 10/25
Table 3
Results of discriminant validity analysis
Construct ATT BI OC SI PR Pra SN
ATT 0.846      
BI 0.656 0.774     
OC 0.497 0.585 0.768    
SI 0.495 0.609 0.559 0.795   
PR -0.468 -0.493 -0.314 -0.373 0.769  
Pra 0.548 0.761 0.632 0.543 -0.442 0.836 
SN 0.675 0.610 0.489 0.509 -0.372 0.621 0.857
{insert Table2 here}
{insert Table3 here}
Measurement scores of participants
As illustrated in Table4, the skewness ranged from − 0.747 to -0.088 implying that the data t a normal
distribution. The mean of 3.60 for physicians' practice to use CPGs on antimicrobials was slightly below
the median, while the remaining variables were above the median. Physicians' social identity (M = 4.11,
SD = 0.61) and organization criterion (M = 4.12, SD = 0.66) showed a strong tendency to support their use
of CPGs on antimicrobials. In addition, they had higher mean scores on subjective norms (M = 4.17, SD =
0.64) and attitude (M = 4.33, SD = 0.57), but lower on perceived risk (M = 2.50, SD = 0.911).
Table 4
Measurement scores of the participants
Measurements Mean SD Skewness Median N(%) of scores > 3
Attitude 4.33 0.57 -0.337 4 97.8
Behavioral intention 4.21 0.55 -0.188 4 96.6
Organization criterion 4.12 0.66 -0.088 4 94.8
Social Identity 4.11 0.61 -0.491 4 93.6
Perceived risk 2.50 0.91 -0.417 2 23.7
Practice 3.50 0.66 -0.088 4 69.5
Subjective norms 4.17 0.64 -0.747 4 93.8
{insert Table4 here}

Page 11/25
Structural equation modelling
The overall structural model for physicians to use CPGs on antimicrobials exhibited the following
goodness-of-t indices:
χ2
/df = 2.347 (< 5), GFI = 0.895 (> 0.8), AGFI = 0.863 (> 0.8), RMSEA = 0.045 (<
0.06), and CFI = 0.943 (> 0.9), which denoted that the research model has t the data well.
Model with standardized path coecients are summarized in Fig.2. At the individual level, physicians'
subjective norms regarding the use of CPGs on antimicrobials were linked to lower perceived risk (β =
-0.265, P < 0.01), better attitude (β = 0.534, P < 0.01) and higher behavioral intentions (β = 0.212, P < 0.01).
In addition, attitude had signicant positive impact on behavioral intentions (β = 0.217, P < 0.01), while
greater perceived risk was linked to lower intentions to use CPGs on antimicrobials (β = -0.175, P < 0.01).
At the organizational level, organization criterion signicantly inuenced attitude (β = 0.153, P < 0.05) and
behavioral intention (β = 0.254, P < 0.01). At the social level, social identity were linked to lower perceived
risk (β = -0.248, P < 0.01) and higher behavioral intention (β = 0.212, P < 0.01), while the impact of social
identity on attitude toward CPGs on antimicrobials was not signicant (β = 0.143, P > 0.05). Finally,
behavioral intention were linked to better practice about the use of CPGs on antimicrobials (β = 0.822, P <
0.01), indicating that intention signicantly inuenced behavior.
{insert Fig.2 here}
Table5 shows the standardized path coecients observed between the theoretical variables derived from
the previous structural analysis. Physicians' behavioral intention of the physician group had the highest
total effect (0.822) on the practice to use CPGs on antimicrobials, followed closely by the effect of
subjective norms on attitude (0.534). And three dimensions of social norms also had some indirect
effects on practice, namely subjective norms (0.308), organization criterion (0.236), and social identity
(0.235).

Page 12/25
Table 5
Results of standardized direct, indirect, and total effects
Paths Direct effects (path
Coecients)
Indirect effects Total effects
Subjective norms→ Attitude 0.534** 0 0.534**
Organization criterion→ Attitude 0.153* 0 0.153*
Social Identity→ Attitude 0.143 0 0.143
Subjective norms→ Perceived risk -0.265** 0 -0.265**
Social Identity→ Perceived risk -0.248** 0 -0.248**
Attitude→ Behavioral intention 0.217** 0 0.217**
Perceived risk→ Behavioral intention -0.175** 0 -0.175**
Subjective norms→ Behavioral intention 0.212** 0.162** 0.374**
Organization criterion→ Behavioral intention 0.254** 0.033* 0.287**
Social Identity→ Behavioral intention 0.212** 0.075** 0.287**
Behavioral intention→ Practice 0.822** 0 0.822**
Subjective norms→ Practice 0 0.308** 0.308**
Organization criterion→ Practice 0 0.236** 0.236**
Social Identity→ Practice 0 0.235** 0.235**
*p < 0.05; **p < 0.01
{insert Table5 here}
As Table6 and Fig.3 show, in senior group, organization criterion (β = 0.327, P < 0.05) were linked to
better attitude and attitude (β = 0.664, P < 0.01) was associated with higher behavioral intentions, while
the direct impact of organization criterion (β = -0.059, p > 0.05) and social identity (β = 0.114, p > 0.05)
toward behavioral intentions were not signicant. In non-senior group, organization criterion (β = 0.312, P
< 0.01) and social identity (β = 0.282, P < 0.01) were linked to higher behavioral intentions. However, the
inuence of organization criterion on attitude (β = 0.033, p > 0.05) and attitude on behavioral intentions (β
= 0.054, p > 0.05) were not signicant. Additionally, in senior group, the effect of organization criterion
toward attitude(χ²=2.201, P < 0.05) and attitude toward behavioral intention(χ²=4.463, P < 0.05) were
signicantly higher than that in non-senior group, while in non-senior group, the effect of organization
criterion and social identity toward behavioral intention(χ² = -3.627, P < 0.05; χ² = -2.249, P < 0.05) were
signicantly higher than that in senior group.

Page 13/25
Table 6
Parameters comparisons between multi-group structural models across professional title
Paths Non-senior
(path
Coecients)
Senior (path
Coecients)
Non-senior vs Senior
(χ²)
Subjective norms→Attitude 0.581** 0.376* -1.784
Organization criterion→Attitude 0.033 0.327* 2.201*
Social Identity→Attitude 0.218* 0.121 -1.219
Subjective norms→Perceived risk -0.274** -0.168 0.699
Social Identity→Perceived risk -0.243* -0.307 -0.059
Attitude→Behavioral intention 0.054 0.664** 4.463*
Perceived risk→Behavioral intention -0.143** -0.173 -0.003
Organization criterion→Behavioral
intention 0.312** -0.059 -3.627*
Social Identity→Behavioral intention 0.282** 0.114 -2.249*
Subjective norms→Behavioral intention 0.275** 0.098 -1.794
Behavioral intention→Practice 0.827** 0.870** 0.070
*p < 0.05; **p < 0.01
{insert Table6 here}
{insert Fig.3 here}
Discussion
In order to bridge the knowledge gap of the inuence of multifaceted social norms on physicians’ use of
CPGs on antimicrobials, this study investigated potential social norms and mediating factors at the
individual level, organizational level and social level by integrating TPB and TNSB. In addition, it applied
multi-group SEM to determine whether the effects of social norms on nal regarding CPGs use would be
strengthened or weakened by the time of action.
Main ndings
This study showed that social norms at the individual level (subjective norms), organizational level
(organization criterion) and social level (social identity) had important impacts on the intention and
behavior of physicians’ use of CPGs on antimicrobials. Additionally, in the results of multi-group SEM, the
professional title signicantly regulated the effect of organization criterion to attitude, attitude to

Page 14/25
behavioral intention, organization criterion to behavioral intention and social identity to behavioral
intention.
Comparison with other studies
Effects of individual-level social norm
In this study, subjective norms, attitude and perceived risk were identied as important factors that
directly inuenced physicians’ intention to use CPGs on antimicrobials, and indirectly inuenced the nal
use behavior, which was in line with some previous studies that established the TPB models of
antimicrobials prescribing behavior (41–43). In the eld of medical services, subjective norms and
attitude are often emphasized because they are related to the sense of security of a particular behavior
(44). To a certain extent, the sense of security is determined by the external pressure and risk felt by
physicians (45), so it is not hard to understand the important role of subjective norms in CPGs use
behavior. Moreover, except for direct effects, subjective norms could have indirect effects on behavioral
intention through attitudes and perceived risk, which in turn inuence actual practice about the use of
CPGs on antimicrobials. This is mainly due to the exemplary role of key people, whose opinions or
behaviors regarding CPGs on antimicrobials or other health technologies can form a perceived norm and
exert pressure on other physicians around them (46).
Effects of organizational-level social norm
In this study, organizational criterion refers to the mandatory guidelines of the physician's organization
regarding CPGs on antimicrobials, which are inscribed provisions for specic CPGs usage practices (47).
Consistent with previous studies (16, 21, 26), this study also conrmed the important role of
organizational criterion on inuencing the application attitudes, intentions and behaviors towards using
CPGs on antimicrobials. The reasons for this result may originate from the following two aspects. On the
one hand, organizational criterion, as the most universal and mandatory norms, have a self-evident effect
on the members of a certain organization (3, 48). They can continuously exert direct or indirect inuence
on physicians’ behavioral intention of adhering CPGs on antimicrobials until they gradually adjust their
behavior to align with the organization. On the other hand, organizational criterion will slowly shape the
organizational culture of the hospital regarding CPGs on antimicrobials over time (12). During this long
process, organizational criterion could inuence physicians' attitudes, views and intentions on the use of
CPGs on antimicrobials, which would also promote physicians to adjust their antimicrobials prescription
behavior for ensuring consistency with the organizationally recognized norms (49, 50).
Effects of social-level social norm
The important direct effect of social identity on the intention to use of CPGs on antimicrobials was
proved in this study, and so as its indirect effects on regarding CPGs use practice through intention and
indirect effects on intention through perceived risk. Similar results were demonstrated in Hallsworth's
study that the recognition and evaluation of physicians from a high-prole gure could reduce their
unnecessary prescriptions of antimicrobials (17). Additionally, it is also reported that the recognition and

Page 15/25
praise from patients and their families often reinforce the physician's acceptance of his own behavior and
the norms he adheres to (16, 47). Since social identity mainly refers to the recognition and support of
physician from social groups such as peers, industry expert, patients and their family members(51, 52), it
is not hard to understand that in order to get more recognition from these social groups, physicians would
complete their work and comply with CPGs with higher behavior intentions, thus forming a virtuous circle.
In contrast to expectations, social identity was not shown to have a signicant inuence on attitude
which is different from the ndings of Liu (3). The plausible reason may be that attitudes generally
change more slowly than behavioral intentions and behaviors. Especially at the beginning of
implementation regarding CPGs, the impact of social identity may be too limited to change the
physicians’ attitude, although it has reversed their intentions or behaviors. Future studies are
recommended to further investigate whether there are other factors or reasons that affect the role of
social identity.
The moderating role of professional title
The temporal effect, namely the moderating role of professional title played in the social norm impact
mechanism, had been also conrmed in this study. With the rise of professional title, the effects of
organizational criterion on attitude and attitude on behavioral intention were strengthened, while the
effects of organizational criterion and social identity on behavioral intention were weaken. These results
are inconsistent with the ndings of a randomized clinical trial that the effect of the social norm-related
intervention for physicians’ antimicrobials prescription was strengthen over time (37). The underlying
reason may be that the social norm-related intervention in the trial was a strong external stimulus for the
physicians, so the effect of “Hawthorne effect” led to the enhancement of the effect of social norms in a
short time. In contrast for this study, rather than strong external stimulus, some effects of professional
title promotion may play more important role in contributing to the seemingly contradictory ndings. On
the one hand, for the physicians with lower professional titles, they are often eager to get more
recognition and praise from their peers, leaders, and other social groups. Thus, under the context with
CPGs compliance as organizational criterion and social identity, these physicians may deem CPGs as
decrees that must be obeyed, although they don’t have a positive attitude towards certain CPGs. So, it is
demonstrated that the impact pathway directly from social norm to behavioral intention dominated in the
group with lower professional titles. On the other hand, for the physicians with higher professional titles,
they tend to be affected by organizational criterion for a longer time with their attitudes and behavior
intention on the use of CPGs on antimicrobials successively changed (53). It is noteworthy that the
change of intention for senior group tends to follow a path of “norm-attitude-intention”, rather than
directly norm to intention. The plausible reasons may be that, with higher positions in hospitals and
higher social status, the physicians with higher professional titles swift their roles from passive obedients
to active implementers of norms, which made them capable of implementing or adjusting the
organizational code according to their understanding, or even of asserting their own judgment or behavior
in the face of corresponding rules. Under this situation, the organizational criterion to which they have

Page 16/25
been subject may be changed or adjusted accordingly, which may also weaken the inuence of
organizational criterion (54, 55).
Policy implications
Based on understanding the impact of multifaced social norms on physicians’ use of CPGs on
antimicrobials, several intervention strategies can be emphasized to further improve the actual use of
regarding CPGs. Firstly, give full play to the demonstration effect of core members. Since the lead, senior
physicians usually have an exemplary role for other physicians, it is strongly recommended to mobilize
them to hold experience sharing sessions on their use of CPGs on antimicrobials. Through the active
advocacy of these inuential people to strengthen the role of subjective norms, physicians could be
motivated to actively comply with regarding CPGs. Secondly, create a good hospital organizational
culture to bring into full play of the role of organizational criterion. For hospital managers, they could
promote the use of CPGs on antimicrobials through explicit means such as formulating and issuing
formal regulations. Additionally, hospital managers were advised to pay attention to the practice of
organizational criterion among the medical staff within their institution, such as holding regular
education and training to improve relevant knowledge of regarding CPGs use for young physicians,
establishing monitoring and supervision mechanism to timely address emerging problems during the
process of organizational criterion and norms implementation. Thirdly, some concrete measures are also
recommended to let the physicians get more social identity, such as establishing a feedback mechanism
to regularly feedback good external evaluations to physicians, promoting the establishment of a good
public image and encourage physicians to comply with CPGs in order to maintain a good image.
Strengths and limitations
In addition to the policy implications, this study is also strengthened by some distinguished features.
Firstly, to the best of our knowledge, few studies have investigated the impact of social norms at the
multiple levels on physicians’ use of CPGs on antimicrobials, especially in developing countries. Based on
the integration of TPB and TNSB, this study enables us to systematically consider the inuence of social
norms on physicians' use of CPGs on antimicrobials from three levels, namely individual, organizational
and social levels. Secondly, this study also investigated the moderating effects of professional titles on
the role of social norms, which greatly expanded the understanding of the temporal effects of social
norms on physicians’ prescription behavior. Thirdly, the sample areas were from the central and western
provinces of China, which are relatively backward in social and economic development. It will benet
providing reference to reveal the effect of social norms on physicians’ use of CPGs on antimicrobials in
other less developed regions of the world.
This study also has some limitations. First of all, all data are obtained through self-reporting, and we
cannot rule out the social expectation bias that some physicians may be reluctant to voice negative
evaluations of themselves and hospitals. Secondly, cross-sectional studies have limitations in the
interpretation of causality. Future studies may involve samples at different time points to form panel

Page 17/25
data, which will be more conducive to determine the impact and changes brought about by social norms.
Thirdly, due to time and funding constraints, this study included a limited number of variables to
investigate the impact mechanism of multifaced social norms on physicians’ use of CPGs on
antimicrobials, and the inuence of other potential variables will be explored in future research.
Conclusion
By integrating TPB and TNSB, this study investigated the impact of multifaced social norms on
physician’ use of CPGs on antimicrobials, and determined the temporal effects of social norms. The SEM
method is used to verify the proposed conceptual research framework. The results of this study reveal the
signicant effect of multifaceted social norms, namely subjective norms at individual level,
organizational criterion at organizational level, and social identity at social level, on the use of CPGs on
antimicrobials. And the organizational criterion and social identity had a more signicant impact on
physicians' behavior at their early stage of practice, which demonstrated its different effect on physicians
with different professional titles. These ndings will not only help tailor pertinent social intervention
strategies to expand the use of CPGs on antimicrobials, especially for the less developed regions, but also
provide clues for future research about impact mechanism of multifaced social norms on physicians’
adoption behavior of certain health services or products.
Abbreviations
AMR
Antimicrobials resistance
CPGs
Clinical practice guidelines
TPB
The theory of planned behavior expansion
TNSB
The theory of normative social behavior
SEM
Structural equation modeling
Declarations
Ethics approval and consent to participate
Ethics approval was obtained from the medical ethics committee, Fujian Medical University, China.
Written informed consent was obtained from all study participants.
Consent for publication
Not applicable.

Page 18/25
Availability of data and materials
The datasets generated during and/or analyzed during the current study are available from the
corresponding author on reasonable request.
Competing interests
The authors declare no competing interests.
Funding
This research was supported by National Natural Science Foundation of China (Grant Number:
72274035) and the Soft Science Project of Fujian Provincial Department of Science and Technology
(Grant Number: 2017R0044). No funders had a role in study design, data collection, data analysis, or in
writing the manuscript.
Authors’ contributions
WL and LW contributed to the conception and design of the study. LW conducted the data reduction,
analyses and wrote the manuscript. WL guided the whole process and reviewed the manuscript. All
authors read and approved the manuscript before submission.
Acknowledgments
We are thankful to all coordinators and physicians for their participation in this study.
Authors’ information
Aliations
Department of Health Management, School of Health Management, Fujian Medical University, Fuzhou,
Fujian, China
Lingjie Wang, Wenbin Liu
Corresponding author
Correspondence to Wenbin Liu.
References
1. Chernov VM, Chernova OA, Mouzykantov AA, Lopukhov LL, Aminov RI. Omics of antimicrobials and
antimicrobial resistance. Expert Opin Drug Discov. 2019;14(5):455-468.
2. Bakhit M, Del MC, Gibson E, Hoffmann T. Exploring patients’ understanding of antibiotic resistance
and how this may inuence attitudes towards antibiotic use for acute respiratory infections: a

Page 19/25
qualitative study in Australian general practice. BMJ Open. 2019;9(3):e026735.
3. Liu C, Liu C, Wang D, Zhang X. Intrinsic and external determinants of antibiotic prescribing: a multi-
level path analysis of primary care prescriptions in Hubei China. Antimicrob Resist Infect Control.
2019;8:132.
4. Ghosh D, Veeraraghavan B, Elangovan R, Vivekanandan P. Antibiotic resistance and epigenetics:
more to it than meets the eye. Antimicrob Agents Chemother. 2020;64(2):e02225-e2319.
5. Smith CR, Pogany L, Foley S, Wu J, Timmerman K, Gale-Rowe M, Demers A. Canadian physicians’
knowledge and counseling practices related to antibiotic use and antimicrobial resistance: two-cycle
national survey. Can Fam Physician. 2017;63:e525-e535.
. Ferri M, Ranucci E, Romagnoli P, Giaccone V. Antimicrobial resistance: a global emerging threat to
public health systems. Crit Rev Food Sci Nutr. 2017;57(13):2857-2876.
7. Sun Q, Dyar OJ, Zhao L, Tomson G, Nilsson LE, Grape M, Song Y, Yan L, Lundborg CS. Overuse of
antibiotics for the common cold–attitudes and behaviors among doctors in rural areas of Shandong
Province. China BMC Pharmacol Toxicol. 2015;16:6.
. National Action Plan for Antimicrobial Resistance (2016-2020). In. Edited by commission NHaFP.
Beijing; 2016.
9. Dyar OJ, Yang D, Yin J, Sun Q, Stålsby Lundborg C. Variations in antibiotic prescribing among village
doctors in a rural region of Shandong province, China: a cross-sectional analysis of prescriptions.
BMJ Open. 2020 Jun 1;10(6):e036703.
10. Yan K, Xue M, Ye D, et al. Antibiotic prescribing practices in secondary and tertiary hospitals in
Shaanxi province, western China, 2013-2015. PLoS One. 2018;13(12):e0207229.
11. Xu S, Yuan S, Kabba JA, et al. Analysis of Antibiotic Use Patterns and Trends Based on Procurement
Data of Healthcare Institutions in Shaanxi Province, Western China, 2015-2018. Int J Environ Res
Public Health. 2020;17(20):7536.
12. Legros S, Cislaghi B. Mapping the Social-Norms Literature: An Overview of Reviews. Perspect
Psychol Sci. 2020;15(1):62-80.
13. Ajzen, I. The Theory of Planned Behavior. Organizational Behavior and Human Decision
Processes.1991, 50(2): 179-211.
14. Rimal, Rajiv Nand Kevin Real. "How behaviors are inuenced by perceived norms: A test of the theory
of normative social behavior." Communication research,2005:389-414.
15. Guerin RJ, Toland MD. An application of a modied theory of planned behavior model to investigate
adolescents' job safety knowledge, norms, attitude and intention to enact workplace safety and
health skills. J Safety Res. 2020,72:189-198.
1. Deng Q, Liu W. The Effect of Social Norms on Physicians' Intentions to Use Liver Cancer Screening: A
Cross-Sectional Study Using Extended Theory of Planned Behavior. Risk Manag Healthc Policy.
2022,15:179-191.

Page 20/25
17. Hallsworth M, Chadborn T, Sallis A, et al. Provision of social norm feedback to high prescribers of
antibiotics in general practice: a pragmatic national randomised controlled trial. Lancet.
2016,387(10029):1743-52.
1. Tang MY, Rhodes S, Powell R, McGowan L, Howarth E, Brown B, Cotterill S. How effective are social
norms interventions in changing the clinical behaviours of healthcare workers? A systematic review
and meta-analysis. Implement Sci. 2021 Jan 7;16(1):8.
19. Rusic D, Bozic J, Bukic J, et al. Evaluation of accordance of antibiotics package size with
recommended treatment duration of guidelines for sore throat and urinary tract infections.
Antimicrob Resist Infect Control. 2019;8:30.
20. Soleymani F, Rashidian A, Hosseini M, Dinarvand R, Kebriaeezade A, Abdollahi M. Effectiveness of
audit and feedback in addressing over prescribing of antibiotics and injectable medicines in a
middle-income country: an RCT. Daru. 2019 Jun;27(1):101-109.
21. Deng Q, Zeng Z, Zheng Y, Lu J, Liu W. Predictors of physicians' intentions to use clinical practice
guidelines on antimicrobial in tertiary general hospitals of China: a structural equation modeling
approach. Antimicrob Resist Infect Control. 2021 Jun 30;10(1):97.
22. Liu C, Liu C, Wang D, Deng Z, Tang Y, Zhang X. Determinants of antibiotic prescribing behaviors of
primary care physicians in Hubei of China: a structural equation model based on the theory of
planned behavior. Antimicrob Resist Infect Control. 2019 Jan 30;8:23.
23. Cortoos PJ, Schreurs BH, Peetermans WE, De Witte K, Laekeman G. Divergent intentions to use
antibiotic guidelines: a theory of planned behavior survey. Med Decis Making. 2012;32(1):145-153.
24. Huang J, Liu W, Huang Y. Inuencing mechanism of the use behavior of clinical practice guidelines
on antimicrobials: evidence from the integration of theory of reasoned action and organizational
readiness for change. BMC Med Inform Decis Mak. 2022;22(1):279.
25. Aghlmandi S, Halbeisen FS, Saccilotto R, et al. Effect of Antibiotic Prescription Audit and Feedback
on Antibiotic Prescribing in Primary Care: A Randomized Clinical Trial. JAMA Intern Med.
2023;183(3):213-220.
2. Schwartz KL, Ivers N, Langford BJ, et al. Effect of Antibiotic-Prescribing Feedback to High-Volume
Primary Care Physicians on Number of Antibiotic Prescriptions: A Randomized Clinical Trial. JAMA
Intern Med. 2021;181(9):1165-1173.
27. Andrighetto G, Vriens E. A research agenda for the study of social norm change. Philos Trans A Math
Phys Eng Sci. 2022 Jul 11;380(2227):20200411.
2. Zhao H, Wei L, Li H, Zhang M, Cao B, Bian J, et al. Appropriateness of antibiotic prescriptions in
ambulatory care in China: a nationwide descriptive database study. Lancet Infect Dis.
2021;21(6):847-57.
29. Wu H, Cao Q, Mao JM, Hu HL. The effect of information overload and perceived risk on tourists'
intention to travel in the post-COVID-19 pandemic. Front Psychol. 2022;13:1000541.
30. Lin L, Sun R, Yao T, Zhou X, Harbarth S. Factors inuencing inappropriate use of antibiotics in
outpatient and community settings in China: a mixed-methods systematic review. BMJ Glob Health.

Page 21/25
2020;5(11):e003599.
31. Chu H, Liu S. Integrating health behavior theories to predict American's intention to receive a COVID-
19 vaccine. Patient Educ Couns. 2021;104(8):1878-86.Mtenga SM, Exavery A, Kakoko D, Geubbels E.
Social cognitive determinants of HIV voluntary counselling and testing uptake among married
individuals in Dar es Salaam Tanzania: Theory of Planned Behaviour (TPB). BMC Public Health.
2015;15:213.
32. Zhang X, Wang L. Factors Contributing to Citizens' Participation in COVID-19 Prevention and Control
in China: An Integrated Model Based on Theory of Planned Behavior, Norm Activation Model, and
Political Opportunity Structure Theory. Int J Environ Res Public Health. 2022;19(23).
33. Room R. Normative Perspectives on Alcohol Use and Problems. Journal of Drug Issues.
1975;5(4):358-68.
34. Rimal RN, Real K. How Behaviors are Inuenced by Perceived Norms:A Test of the Theory of
Normative Social Behavior. 2005;32(3):389-414.
35. Lapinski MK, Anderson J, Shugart A, Todd E. Social inuence in child care centers: a test of the
theory of normative social behavior. Health Commun. 2014;29(3):219-32.
3. Henseler J, Ringle CM, Sinkovics RR. The use of partial least squares path modeling in international
marketing. In: Sinkovics RR, Ghauri PN, editors. New Challenges to International Marketing. Advances
in International Marketing. 20: Emerald Group Publishing Limited; 2009. p. 277-319.
37. Persell SD, Goldstein NJ, Knight TK, Hay JW, Doctor JN. Effect of Behavioral Interventions on
Inappropriate Antibiotic Prescribing Among Primary Care Practices: A Randomized Clinical Trial.
JAMA. 2016 Feb 9;315(6):562-70.
3. Gorusch RL. Factor analysis. Hillsdale: Lawrence Erlbaum Associates; 1983.p. 332
39. Wook M, Hasbullah NA, Zainudin NM, Jabar ZZA, Ramli S, Razali NAM, et al. Exploring big data traits
and data quality dimensions for big data analytics application using partial least squares structural
equation modelling. Journal of Big Data. 2021;8(1):49.
40. Azjen I. Understanding attitudes and predicting social behavior. Englewood Cliffs; 1980.
41. Heijden H. Factors inuencing the usage of websites: the case of a generic portal in The Netherlands.
Inform Manage. 2003;40:541-549.
42. Ahmed RR, Vveinhardt J, Streimikiene D, Awais M. Mediating and marketing factors inuence the
prescription behavior of physicians: an empirical investigation. Amteatru Econ. 2016;18:153–167.
43. Rantanen T, Lehto P, Vuorinen P, Coco K. The adoption of care robots in home care-a survey on the
attitudes of nnish home care personnel. J Clin Nurs. 2018;27(9–10):1846–1859.
44. Saigi-Rubió F, Jiménez-Zarco A, Torrent-Sellens J. Determinants of the intention to use telemedicine:
evidence from primary care physicians. Int J Technol Assess. 2016;32(1-2):29-36.
45. Wilcock M, Harding G. What do pharmacists think of MURs and do they change prescribed
medication? Pharm J. 2008;281:163-167.

Page 22/25
4. von Knorring M, de Rijk A, Alexanderson K. Managers' perceptions of the manager role in relation to
physicians: a qualitative interview study of the top managers in Swedish healthcare. BMC Health
Serv Res. 2010;10:271.
47. Kummer TF, Recker J, Bick M. Technology-induced anxiety: manifestations, cultural inuences, and
its effect on the adoption of sensor-based technology in German and Australian hospitals. Inform
Manage. 2017;54:73-89.
4. Deng Q, Liu W. Utilization of clinical practice guideline on antimicrobial in China: an exploratory
survey on multilevel determinants. BMC Health Serv Res. 2020 Apr 6;20(1):282.
49. Qureshi QA, Shah B, Ullah N, et al. The impact of top management support and e-health policies on
the success of e-health practices in developing countries. Scholarly J Med. 2013;3(3):27-30.
50. Rimal RN, Yilma H. Descriptive, Injunctive, and Collective Norms: An Expansion of the Theory of
Normative Social Behavior (TNSB). Health Commun. 2022;37(13):1573-1580.
51. Rimal RN. Modeling the relationship between descriptive norms and behaviors: a test and extension
of the theory of normative social behavior (TNSB). Health Commun. 2008;23(2):103-116.
52. Liu FC, Cheng KL, Chao M, et al. Team innovation climate and knowledge sharing among healthcare
managers: mediating effects of altruistic intentions. Chang Gung Med J. 2012;35(5):408-419.
53. Andrighetto G, Vriens E. A research agenda for the study of social norm change. Philos Trans A Math
Phys Eng Sci. 2022 Jul 11;380(2227):20200411.
54. Andreoni J, Nikiforakis N, Siegenthaler S. Predicting social tipping and norm change in controlled
experiments. Proc Natl Acad Sci U S A. 2021 Apr 20;118(16):e2014893118.
55. Kish Bar-On K, Lamm E. The interplay of social identity and norm psychology in the evolution of
human groups. Philos Trans R Soc Lond B Biol Sci. 2023 Mar 13;378(1872):20210412.
Figures

Page 23/25
Figure 1
The theoretical framework
Figure 2

Page 24/25
Determinants of physicians’ intentions to use CPGs on antimicrobials
Figure 3
Comparisons between multi-group structural models across professional title
Supplementary Files

Page 25/25
This is a list of supplementary les associated with this preprint. Click to download.
Questionnaire.doc