The Main Hypothesis
To establish and hypothesize the research framework for this study, we examined the relationships between variables of the MHAU and service quality in existing literature. Chang and Chang [28] used the term “technology-based service encounters” to describe the interaction between medical practitioners and patients during clinical process by e-commerce and Internet technology, such as online appointment system (OAS) and EHRs, revealing positive effect of technology-based service encounters service quality without effecting patient satisfaction. Perception of service quality has a positive impact on patient satisfaction via the adoption of technology-based service. While in a study of effect of OAS on patient satisfaction, Wang, Cheng and Huang [29] claimed that OAS technology-based service encounters, compared with the hospital image, had a stronger effect on patient satisfaction. Based on the above results, we hypothesize that:
H1a: MHAU has a positive effect on patient's perception of service quality.
H1b: MHAU has a positive effect on patient satisfaction.
Technology acceptance has a significant effect on relationship quality which is defined as the outcome of interactions between both parties [30-31]. Wu, Li and Li [32] applied the construction of “interaction quality” to describe how the service is delivered. It can be explained as the patient’s cognition of service quality in communication with doctors and nurses. Generally, it claimed that interactive quality has a positive effect on overall experience quality [33]. In addition, Petter and Fruhling [34] indicated that the system usage, as a variable to measure use of STATPack™(an information tool used to aid in the diagnosis of pathogens in hospitals), has positive effect on individual and organization. These two variables describe how effective and useful the individual and organization using STATPack™. The results suggest that the use of the technology or systems may result in the change of the interaction between patients and doctors, thereby affecting perceived service quality. Thus, in our study, we hypothesized that the way of the MHAU affecting the patient perceived service quality works by the mediating role of clinical process change perceived by patient. Consequently, based on the discussion above, we can draw a conclusion and put forward hypotheses as follows:
H1c: MHAU is positively push forward clinical process change perceived by patients.
H2: Clinical process change perceived by patients has positive effect on perceived service quality in hospitals.
In the investigation of relationship between service quality and patient satisfaction, some authors state that service quality directly effects on patient satisfaction [35-41], while others argue that service quality influences the patient satisfaction by mediating factors [42-45]. Faria and Mendes [46] confirmed that apart from the direct effect of service quality on patient satisfaction, the institutional reputation also fluctuate the relationship between service quality and patient satisfaction. Johnson and Russell [47] found the mediating role of healthcare provider and nurse/assistant within service quality and patient satisfaction. Thus, we hypothesize that:
H3: The perceived service quality is positively related to patient satisfaction.
Control Variable
To make the study more scientific, we identified four control variables: gender, age, education, and occupation. Usually, compared with the women, men is more rational and insensitive and more likely to satisfaction, they tend to simplify complex problems when they experience trouble. Compared with the older, the younger show skilled use in health information technologies, while the older tend to show fear for using these technologies, which can easily trigger dissatisfaction among older patients. The level of education is usually proportional with the ability to solve and identify problems, with the ability to proficient use in health information technologies. Patients' occupation is significantly associated with skilled use of health information technology. Based on this, the gender, age, education, and occupation of the respondents were included as control variables that might influence patient satisfaction.
Research Model
Based on the above hypotheses, the research model was established, as illustrated in Figure 1.
Research Design
Questionnaire Design
By reviewing a large amount of literature, we directly introduced, selected, and modified the relevant mature scale to design variables and items of this study. The total number of scales is four, with eleven dimensions, which are: MHAU, clinical process change perceived by patient (dimensions included: physician-patient interaction, and information accessibility), service quality (dimensions included: tangibility, reliability, responsiveness, assurance, empathy, and convenience) and patient satisfaction (dimensions included: treatment outcome and visit time). with measurement of control variables, some demographic variables may have an impact on patient performance. Each questions of the scale and its’ source reference table 1. Furthermore, the independent variable, MHAU, may be affected by the patient's gender, age, education level, and occupation, thus indirectly or directly affecting the patient satisfaction. In addition, the above control variable may also have effects on the patient-perceived clinical process change and service quality. Therefore, the gender, age, education, and occupation of the respondents were added to the measurement scale. The Likert 7-point recording method was adopted for each item, with 1 being “very disagree” and 7 being “very agree” [48].
Table 1. Source of the questions of each scale
Variable
|
Number
|
Items
|
Source
|
Mobile Health Apps Use (MHAU)
|
a
|
Make an appointment for doctor
|
[47]
|
b
|
Pay for medical related expenses
|
[39,43,45,52]
|
c
|
Check laboratory reports and medical records
|
[43,52]
|
d
|
Interact with the doctor you want to consult online at any time
|
e
|
Communicate with other patients
|
f
|
Ask medical staff questions via SMS
|
g
|
Learn about health education information and medical information pushed by hospitals
|
Clinical Process Change perceived by patient
|
Physician-patient Interaction
|
a
|
Now I can make an appointment to the doctor who I want to see every time
|
[45]
|
b
|
When I choose a doctor, I can get information about the doctor's background and experience online
|
[43,52]
|
c
|
When I choose a doctor, I can see other patients’ scores and assessments on the Internet
|
d
|
After seeing the doctors, I can rate the doctor online.
|
k
|
Now I can manage and treat my disease more effectively
|
l
|
Now I can keep in touch with other patients online
|
m
|
I can now ask the medical staff questions via SMS on the Internet
|
n
|
Every link in my medical treatment process is now more coherent
|
o
|
I am maintaining a continuous communication relationship with my doctor
|
cc
|
Now the queue has been reduced in every link of my medical treatment
|
[47]
|
dd
|
I took the initiative to participate in discussions with doctors about treatment options
|
[45-46]
|
ee
|
I actively seek other information related to my health
|
[43,52]
|
ff
|
I take the initiative to participate in learning preventive treatment information
|
gg
|
After I see a doctor, I will take the initiative to follow up and complete all the required treatments
|
hh
|
I took the initiative to help the doctor determine my health and problems
|
ii
|
The medical staff in the hospital now know the records of every link of my visit very well
|
jj
|
In every aspect of my medical treatment process, the data related to me can be checked in time
|
kk
|
In every aspect of my medical treatment, medical staff are now in harmony with each other
|
ll
|
It's easy for me and the medical staff to make a common agreement now
|
Information Accessibility
|
e
|
I can easily access and store my medical information now
|
f
|
I can now access and process my medical information anytime and anywhere
|
g
|
Now, when I am in the hospital, the medical staff in different departments are well coordinated with each other
|
h
|
Now even if I am looking for a different doctor in the hospital, the process of diagnosis and treatment is consistently standardized
|
i
|
I can now check the medical records of the past at each stage of the hospital
|
[35]
|
j
|
Now if the medical staff I'm looking for is not there, other medical staff can meet my needs
|
[40]
|
p
|
Now the doctor can give me the most suitable treatment according to my personal condition and complete historical information
|
[35]
|
q
|
Now my doctor's treatment plan is consistent with my changing needs and conditions
|
[46]
|
r
|
Now that I'm in the hospital, I know exactly who I'm looking for at every step
|
[46]
|
s
|
The doctor encouraged me to ask questions
|
[42,52]
|
t
|
The doctor will answer my questions adequately
|
u
|
The doctor actively encouraged me to participate in discussions with the doctor about treatment options
|
v
|
Doctors offer other information about my condition and treatment on their own initiative
|
[40]
|
w
|
Doctors actively provide information on preventive treatment
|
[42,52]
|
x
|
I can understand the explanation given by the doctor
|
[43,45,52]
|
y
|
The doctor was very considerate of me
|
[55]
|
z
|
The doctor made me feel at ease discussing my condition
|
aa
|
I feel that the doctor knows my medical history very well
|
bb
|
I feel that doctors know very well about my health care needs
|
Service quality
|
Tangibility
|
a
|
The hospital is clean
|
[50]
|
b
|
The hospital's medical equipment is very advanced
|
c
|
Doctors and nurses dress professionally and neatly
|
d
|
The signs of hospital facilities are very clear
|
[55]
|
e
|
The TV screen in the waiting area shows useful information for the patient
|
f
|
In many places in the hospital, you can see promotional materials that guide how to use the hospital WeChat application
|
Reliability
|
g
|
My doctor is very concerned about my personal situation
|
[50]
|
h
|
My doctor is based on my special condition
|
[56]
|
i
|
My doctor understands my specific needs
|
[55]
|
j
|
My doctor is concerned about my unique needs
|
[50]
|
k
|
My doctor showed great sympathy for my condition
|
[56]
|
Assurance
|
l
|
My doctor has the ability to treat me well
|
[50]
|
m
|
When the doctor came to see me, I felt safe
|
n
|
The way and behavior of my doctor give me great confidence
|
o
|
My doctor has good medical knowledge
|
p
|
My doctor is trustworthy
|
[55]
|
q
|
My doctor is very experienced
|
[36]
|
Convenience
|
r
|
I can easily make an appointment with the doctor I want to see
|
[45]
|
s
|
I can easily make an appointment to the time I want to see a doctor
|
t
|
It's easy for me to find where I need to go in the hospital
|
u
|
I don't have to wait long in hospital
|
[47]
|
v
|
Every step of seeing a doctor in my hospital is very convenient and easy
|
[45]
|
w
|
The hospital staff will always help me whenever I need
|
[50]
|
Reliability
|
x
|
My doctor seldom makes mistakes
|
y
|
My doctor always explains the diagnosis and treatment to me very clearly
|
z
|
My medical record in the hospital is always accurate
|
[55]
|
aa
|
There are few inconsistencies in my medical records
|
[50]
|
bb
|
The service of medical staff is always reliable
|
Responsiveness
|
cc
|
My doctor can answer my question quickly
|
[54]
|
dd
|
I always get prompt answers when I contact the hospital
|
ee
|
Doctors and nurses are not too busy to answer my questions in time
|
ff
|
My doctor will keep updated of my condition changes
|
gg
|
My doctor will make quick adjustments to my condition
|
Patient Satisfaction
|
Treatment Outcome
|
a
|
I am satisfied with the medical services I received during my stay in this hospital
|
[42,44]
|
b
|
I'm very satisfied with the doctor's attitude
|
c
|
I am very satisfied with the quality of the doctor's treatment
|
d
|
My illness has been properly treated
|
e
|
After seeing the doctor, I have a better understanding of my condition
|
[43,52]
|
f
|
After talking with the doctor, I feel a lot better about my condition
|
g
|
The doctor's choice of treatment is the most appropriate for me
|
h
|
My condition will be completely improved
|
l
|
I'm satisfied with the doctor's consultation time
|
m
|
I am satisfied with the waiting time in the hospital
|
n
|
I'm satisfied with the total time spent on this visit
|
Visit Time
|
i
|
I'm very clear about how to recover when I get home
|
[46-47]
|
j
|
I am very clear about how to use the medicine
|
k
|
I know exactly when to see the doctor next time.
|
Control variable
|
gender, age, education level, and occupation
|
Data Collection
In order to improve the validity of the questionnaire as much as possible, the paper version of the questionnaire was adopted. The patients were required to fill out the questionnaires as long as the medical consultation was approaching to the end on the paperwork. From December 19, 2018 to January 22, 2019, 647 questionnaires were issued, which involved 618 valid questionnaires. The rate of sample validity was 95.5%. The reasons for the invalid questionnaires are: the patients were impatient to answer questions which lead to the filling time is significantly lower than the rational and reasonable filling time of 20 minutes; the patients were very concentrating at the beginning, but they were interrupted by some emergency issue; some critical questions were neglected.
Empirical Test and Result Analysis
Demographic Analysis
The demographic information of patients who use mobile health apps is summarized in Table 2. The proportion of patient gender was 39% (male) and 61% (female), respectively. The majority of patients’ age were between 31 and 40, accounting for 46.3% of the total patients interviewed, followed by those ages between 21-30 and 11-20, accounting for 27% and 10.2%, showing that the group of patients using mobile health technologies were younger generation, and basically matched with the age distribution of overall patients in the hospital. The ratio of the patients who have the bachelor degree and diploma was the largest, accounting for 41% and 21.4% of the total, indicating that the education level of the targeted patients was generally high. From the perspective of occupation, the highest proportion was non-government-related enterprises, such as foreign-funded enterprises, privately owned corporations and self-employed households, accounting for 59.5%. In summary, the distribution of the demographic information of the patient samples in this study is even and reasonable and can represent the overall patient population of the hospital.
Table 2. Summary of the demographic information of patients (N=618)
Variable
|
Category Description
|
Sample size
|
Ratio (%)
|
Gender
|
M
|
241
|
39
|
F
|
377
|
61
|
Age
|
10 and under
|
19
|
3.1
|
11~20
|
63
|
10.2
|
21~30
|
167
|
27
|
31~40
|
286
|
46.3
|
41~50
|
51
|
8.3
|
51~60
|
20
|
3.2
|
61 and above
|
12
|
1.9
|
Occupation
|
Government
|
11
|
1.8
|
Government-affiliated Institutions
|
49
|
8
|
State-owned Enterprises
|
44
|
7.1
|
Private-owned Company
|
149
|
24.2
|
Foreign Enterprises
|
54
|
8.8
|
Private Enterprises
|
64
|
10.4
|
Self-employed Households
|
99
|
16.1
|
Farmers
|
30
|
4.9
|
Students
|
62
|
10.1
|
Others: Retirement, Unemployment, Full-time mother
|
54
|
8.9
|
Education
|
Doctor
|
5
|
0.8
|
Master
|
35
|
5.7
|
Bachelor
|
253
|
41
|
Diploma
|
132
|
21.4
|
Vocational Technical School
|
37
|
6
|
High School
|
73
|
11.8
|
Junior Middle School
|
52
|
8.4
|
Primary School
|
28
|
4.5
|
Kinder garden
|
2
|
0.4
|
Measurement Reliability
We conducted the reliability test using Cronbach's α coefficient in this study. The α coefficient is usually expected to be greater than 0.7, indicating that the reliability is acceptable [48]. To determine the factors that affect the reliability, we may calculate the corrected item-total correlation (CITC) value. If the CITC is lower than 0.4, or the α coefficient increases after deletion, we could consider deleting the corresponding question. Reliability analysis requires a separate analysis for each variable. The α coefficients of the each variables or dimensions is greater than 0.7 (see table 3), indicated that the reliability of each variable is acceptable. The Cronbach's coefficient of the subscale is higher than 0.8 (see table 3), and the Cronbach's coefficient of the total scale is 0.846 (see table 3), indicating that it is relatively reliable.
Table 3. Reliability analysis of variables and dimensions
Variables
|
Dimensions
|
Cronbach’s α
|
Mobile health apps use (MHAU)
|
0.839
|
Clinical process change perceived by patient
α=0.822
|
Information Accessibility (IA)
|
0.848
|
Physician-patient Interaction (PI)
|
0.713
|
Patient satisfaction
α=0.845
|
Treatment Outcome (TO)
|
0.861
|
Visit Time (VT)
|
0.930
|
Service quality
α=0.861
|
Tangibility (Ta)
|
0.872
|
Empathy (Em)
|
0.886
|
Assurance (Ass)
|
0.857
|
Convenience (Co)
|
0.815
|
Reliability (Rel)
|
0.848
|
Responsiveness (Res)
|
0.796
|
Total scale
|
0.846
|
The results showed that the reliability of each variable and dimension was very high, and the CITC values of all the questions were greater than 0.4 and deleting any one of the questions did not improve the overall α coefficient. Therefore, those 11 variables (including all dimensions), as well as 41 question items do not need to be modified or deleted, and its results are reliable.
Measurement Validity
The validity test included content validity and construct validity in this study. The content validity of the scale adopts the mature scale content, and is supported by a lot of theories. Meanwhile, five hospital management experts were invited to examine the content of the scale to ensure the content validity. The items used in this study mainly refer to the scales used by previous researchers, as well as the experience and opinions of experts, after revision, the contents completely conform to the conceptual description of relevant variables.
SPSS 24 statistical software was used for the exploratory factor analysis to verify the validity of the questionnaire structure in this study. Before exploratory factor analysis, KMO (Kaiser-Meyer-Olkin) statistics and Bartlett sphere test were used to determine whether the data were applicable for the factor analysis. If the KMO value is greater than 0.6, which is the general standard, and the p value of Bartlett spherical test is 0.000, less than 0.01, it indicates that the statistics is applicable for the factor analysis. Results of the structural validity analysis of MHAU, clinical process change perceived by patient (physician-patient interaction and information accessibility), service quality (tangibility, reliability, responsiveness, assurance, empathy, and convenience) and patient satisfaction (treatment outcome and visit time) are shown in table 4 to table 7.
Table 4. Construct validity analysis of mobile health apps use (MHAU) a
No.
|
Items
|
Factor
|
1
|
2a
|
Make an appointment for doctor
|
.638
|
2b
|
Pay for medical related expenses
|
.720
|
2c
|
Check laboratory reports and medical records
|
.796
|
2d
|
Interact with the doctor you want to consult online at any time
|
.822
|
2e
|
Communicate with other patients
|
.767
|
2g
|
Learn about health education information and medical information pushed by hospitals
|
.715
|
Eigenvalue
|
3.335
|
Cumulative Variance Interpretation Rate
|
55.584
|
KMO
|
.793
|
Bartlett’s Test of Sphericity
|
1658.439
|
Sig
|
0.000
|
Extraction Method: Principal Component Analysis
a:1 component extracted.
Table 5. Construct validity analysis of clinical process change perceived by patient a
No.
|
Items
|
Factor
|
1
|
2
|
3f
|
Now I can keep in touch with other patients online
|
.817
|
|
3g
|
I can now ask the medical staff questions via SMS on the Internet
|
.844
|
|
3t
|
I can easily access and store my medical information now
|
.807
|
|
3u
|
I can now access and process my medical information anytime and anywhere
|
.800
|
|
3a
|
Now I can make an appointment to the doctor who I want to see every time
|
|
.635
|
3l
|
I actively seek other information related to my health
|
|
.647
|
3dd
|
The doctor will answer my questions adequately
|
|
.817
|
3ff
|
Doctors offer other information about my condition and treatment on their own initiative
|
|
.783
|
Rotated Eigenvalue
|
2.760
|
2.250
|
Rotated Variance Interpretation Rate
|
34.501
|
34.501
|
Cumulative Variance Interpretation Rate
|
28.123
|
62.623
|
KMO
|
.731
|
Bartlett’s Test of Sphericity
|
2224.485
|
Sig
|
0.000
|
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalizion
a: Rotation converged in 3 iterations.
Table 6. Construct validity analysis of patient satisfaction a
No.
|
Items
|
Factor
|
1
|
2
|
4b
|
I'm very satisfied with the doctor's attitude
|
.837
|
|
4c
|
I am very satisfied with the quality of the doctor's treatment
|
.815
|
|
4d
|
My illness has been properly treated
|
.842
|
|
4f
|
After talking with the doctor, I feel a lot better about my condition
|
.807
|
|
4j
|
I am satisfied with the waiting time in the hospital
|
|
.950
|
4k
|
I'm satisfied with the total time spent on this visit
|
|
.950
|
Rotated Eigenvalue
|
2.785
|
1.920
|
Rotated Variance Interpretation Rate
|
46.409
|
32.008
|
Cumulative Variance Interpretation Rate
|
46.409
|
78.417
|
KMO
|
.739
|
Bartlett’s Test of Sphericity
|
2128.748
|
Sig
|
0.000
|
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalizion
a: Rotation converged in 3 iterations.
Table 7. Construct validity analysis of service quality a
No.
|
Items
|
Factor
|
1
|
2
|
3
|
4
|
5
|
6
|
5a2
|
The hospital is clean
|
.784
|
|
|
|
|
|
5b2
|
The hospital's medical equipment is very advanced
|
.814
|
|
|
|
|
|
5c2
|
Doctors and nurses dress professionally and neatly
|
.799
|
|
|
|
|
|
5d2
|
The signs of hospital facilities are very clear
|
.755
|
|
|
|
|
|
5g2
|
My doctor is very concerned about my personal situation
|
|
.755
|
|
|
|
|
5h2
|
My doctor is based on my special condition
|
|
.730
|
|
|
|
|
5i2
|
My doctor understands my specific needs
|
|
.723
|
|
|
|
|
5j2
|
My doctor is concerned about my unique needs
|
|
.765
|
|
|
|
|
5o2
|
My doctor has good medical knowledge
|
|
|
|
.665
|
|
|
5p2
|
My doctor is trustworthy
|
|
|
|
.730
|
|
|
5q2
|
My doctor is very experienced
|
|
|
|
.768
|
|
|
5r2
|
I can easily make an appointment with the doctor I want to see
|
|
|
|
|
.812
|
|
5s2
|
I can easily make an appointment to the time I want to see a doctor
|
|
|
|
|
.834
|
|
5u2
|
I don't have to wait long in hospital
|
|
|
|
|
.602
|
|
5z2
|
My medical record in the hospital is always accurate
|
|
|
|
|
|
.709
|
5aa2
|
There are few inconsistencies in my medical records
|
|
|
|
|
|
.754
|
5bb2
|
The service of medical staff is always reliable
|
|
|
|
|
|
.618
|
5dd2
|
I always get prompt answers when I contact the hospital
|
|
|
.600
|
|
|
|
5ee2
|
Doctors and nurses are not too busy to answer my questions in time
|
|
|
.787
|
|
|
|
5ff2
|
My doctor will keep updated of my condition changes
|
|
|
.773
|
|
|
|
5gg2
|
My doctor will make quick adjustments to my condition
|
|
|
.651
|
|
|
|
Rotated Eigenvalue
|
3.187
|
3.011
|
2.623
|
2.292
|
2.256
|
2.125
|
Rotated Variance Interpretation Rate
|
15.177
|
14.340
|
12.491
|
10.912
|
10.744
|
10.121
|
Cumulative Variance Interpretation Rate
|
15.177
|
29.517
|
42.008
|
52.920
|
63.665
|
73.785
|
KMO
|
.943
|
Bartlett’s Test of Sphericity
|
7726.062
|
Sig
|
0.000
|
Extraction Method: Principal Component Analysis
Rotation Method: Varimax with Kaiser Normalizion
a: Rotation converged in 6 iterations.
Results of the factor analysis demonstrated that the co-relationship between the four variables and each factor basically met the professional requirements. The factor loading coefficient is all higher than 0.6, indicating that the structural validity of the questionnaire is attainable , and the data is valid.
Correlation analysis
Correlation analysis was conducted on the four variables and their dimensions, and all relevant results were summarized in Table 8.
Table 8. Correlation analysis of dimensions in research model
Dimensions
|
MHAU
|
IA
|
PI
|
Ta
|
Em
|
Ass
|
Co
|
Rel
|
Res
|
TO
|
VT
|
MHAU
|
1
|
|
|
|
|
|
|
|
|
|
|
IA
|
.813**
|
1
|
|
|
|
|
|
|
|
|
|
PI
|
.327**
|
.376**
|
1
|
|
|
|
|
|
|
|
|
Ta
|
.288**
|
.284**
|
.368**
|
1
|
|
|
|
|
|
|
|
Em
|
.264**
|
.259**
|
.481**
|
.523**
|
1
|
|
|
|
|
|
|
Ass
|
.205**
|
.220**
|
.434**
|
.600**
|
.605**
|
1
|
|
|
|
|
|
Co
|
.230**
|
.199**
|
.497**
|
.399**
|
.575**
|
.487**
|
1
|
|
|
|
|
Rel
|
.281**
|
.288**
|
.510**
|
.539**
|
.651**
|
.665**
|
.573**
|
1
|
|
|
|
Res
|
.185**
|
.194**
|
.424**
|
.404**
|
.589**
|
.514**
|
.532**
|
.602**
|
1
|
|
|
TO
|
.324**
|
.364**
|
.582**
|
.541**
|
.573**
|
.579**
|
.438**
|
.588**
|
.465**
|
1
|
|
VT
|
.401**
|
.398**
|
.331**
|
.326**
|
.378**
|
.237**
|
.396**
|
.360**
|
.317**
|
.358**
|
1
|
**. Correlation is significant at the 0.01 level (2-tailed)
Note: MHAU=mobile healthy applications use; IA=information accessibility; PI=physician-patient interaction; Ta=tangibility; Em=empathy; Ass=assurance; Co=convenience; Rel=reliability; Res=responsiveness; TO=treatment outcome; VT=visit time
The correlation analysis showed that MHAU had the strongest correlation with information accessibility at the dimension level (correlation coefficient=0.813, p<0.01). Except for the moderate correlation with visit time (correlation coefficient=0.401, p<0.01), the other factors were weak or no correlation. Apart from the strong correlation with MHAU, information accessibility was invisible or irrelevant to all other dimensions. There was a low correlation between physician-patient interaction and tangibility (correlation coefficient=0.368, p<0.01), and there was a moderate correlation between physician-patient interaction and other five dimensions of service quality. There was weak correlations between the six dimensions of service quality and visit time. The most relevant element was convenience (correlation coefficient=0.396, p<0.01), while the six dimensions of service quality and treatment outcome were all moderately correlated.
Regression analysis and validation of model hypothesis
Multiple linear regression analysis was used to verify the relationship between independent variables and dependent variables in this study. For the purpose of preventing the interference caused by the sample demographic information, the study included the patient's age, gender, occupation and education level those were used as control variables in the model for analysis. The analysis of the relationship between model variables is conducted at two levels. The first level is among four variables, and the second level is between the dimensions of variables. Therefore, we established three groups of independent variables and dependent variables during the regression analysis. They are:
Group 1: Independent variables: MHAU; dependent variables: clinical process change perceived by patient (physician-patient interaction and information accessibility).
Group 2: Independent variables: MHAU, clinical process change perceived by patient (physician-patient interaction and information accessibility); dependent variables: service quality (tangibility, empathy, assurance, convenience, reliability and responsiveness).
Group 3: Independent variables: MHAU, service quality (tangibility, empathy, assurance, convenience, reliability and responsiveness); dependent variables: patient satisfaction (treatment outcome and visit time).
The results of the three groups at variables and dimensions level are shown in Tables 9.
Table 9. Regression analysis of clinical process change perceived by patient and its dimensions as dependent variable
|
Dependent variable
|
Independent variable
|
Unstandardized
|
Standardized Coefficients
|
t
|
Sig.
|
|
|
|
B
|
Std. Error
|
Beta
|
R²
|
Adjusted R²
|
F
|
CPC perceived by patient and its dimensions as dependent variable
|
CPC
|
MHAU
|
.456
|
.026
|
.619
|
17.549
|
.000
|
.404
|
.384
|
20.008**
|
IA
|
MHAU
|
.717
|
.037
|
.660
|
19.436
|
.000
|
.447
|
.428
|
23.810**
|
PI
|
MHAU
|
.196
|
.029
|
.284
|
6.693
|
.000
|
.138
|
.109
|
4.719**
|
SQ and its dimensions as dependent variable
|
SQ
|
MHAU
|
-.011
|
.025
|
-.021
|
-.434
|
.665
|
.272
|
.246
|
10.392**
|
CPC
|
.316
|
.033
|
.471
|
9.533
|
.000
|
SQ
|
MHAU
|
.044
|
.024
|
.089
|
1.809
|
.071
|
.349
|
.325
|
14.110**
|
IA
|
.032
|
.023
|
.069
|
1.391
|
.165
|
PI
|
.342
|
.029
|
.477
|
11.961
|
.000
|
Ta
|
MHAU
|
.093
|
.035
|
.149
|
2.688
|
.007
|
.169
|
.137
|
5.330**
|
IA
|
.047
|
.032
|
.081
|
1.445
|
.149
|
PI
|
.212
|
.041
|
.236
|
5.219
|
.000
|
Em
|
MHAU
|
.029
|
.037
|
.040
|
.766
|
.444
|
.254
|
.225
|
8.925**
|
IA
|
.065
|
.035
|
.099
|
1.853
|
.064
|
PI
|
.399
|
.044
|
.388
|
9.080
|
.000
|
As
|
MHAU
|
.019
|
.033
|
.031
|
.581
|
.562
|
.219
|
.190
|
7.380**
|
IA
|
.035
|
.031
|
.062
|
1.128
|
.260
|
PI
|
.331
|
.039
|
.369
|
8.437
|
.000
|
Co
|
MHAU
|
.068
|
.042
|
.083
|
1.599
|
.111
|
.267
|
.239
|
9.554**
|
IA
|
-.031
|
.040
|
-.041
|
-.775
|
.439
|
PI
|
.520
|
.050
|
.441
|
10.415
|
.000
|
Rel
|
MHAU
|
.048
|
.034
|
.073
|
1.407
|
.160
|
.282
|
.255
|
10.308**
|
IA
|
.065
|
.032
|
.105
|
2.008
|
.045
|
PI
|
.409
|
.041
|
.423
|
10.077
|
.000
|
Res
|
MHAU
|
.008
|
.023
|
.018
|
.328
|
.743
|
.207
|
.177
|
6.856**
|
IA
|
.003
|
.022
|
.007
|
.127
|
.899
|
PI
|
.240
|
.027
|
.389
|
8.819
|
.000
|
|
|
|
PS and its dimensions as dependent variable
|
PS
|
MHAU
|
.162
|
.022
|
.246
|
7.479
|
.000
|
.521
|
.504
|
30.201**
|
SQ
|
.784
|
.044
|
.589
|
17.651
|
.000
|
PS
|
MHAU
|
.156
|
.022
|
.236
|
7.062
|
.000
|
.525
|
.503
|
23.780**
|
Ta
|
.185
|
.042
|
.175
|
4.379
|
.000
|
Em
|
.192
|
.043
|
.207
|
4.477
|
.000
|
Ass
|
.068
|
.049
|
.064
|
1.382
|
.168
|
Co
|
.120
|
.033
|
.148
|
3.635
|
.000
|
Rel
|
.133
|
.048
|
.135
|
2.745
|
.006
|
Res
|
.042
|
.067
|
.027
|
.621
|
.535
|
TO
|
MHAU
|
.075
|
.022
|
.119
|
3.472
|
.001
|
.499
|
.475
|
21.395**
|
Ta
|
.175
|
.041
|
.173
|
4.216
|
.000
|
Em
|
.174
|
.042
|
.196
|
4.127
|
.000
|
Ass
|
.233
|
.048
|
.230
|
4.828
|
.000
|
Co
|
.012
|
.032
|
.015
|
.365
|
.715
|
Rel
|
.143
|
.048
|
.152
|
3.004
|
.003
|
Res
|
.033
|
.066
|
.022
|
.496
|
.620
|
VT
|
MHAU
|
.317
|
.047
|
.274
|
6.719
|
.000
|
.297
|
.264
|
9.087**
|
Ta
|
.205
|
.090
|
.110
|
2.271
|
.024
|
Em
|
.228
|
.092
|
.140
|
2.491
|
.013
|
Ass
|
-.262
|
.105
|
-.140
|
-2.495
|
.013
|
Co
|
.336
|
.071
|
.236
|
4.764
|
.000
|
Rel
|
.113
|
.104
|
.065
|
1.091
|
.276
|
Res
|
.060
|
.144
|
.022
|
.416
|
.678
|
*P<0.05,**P<0.01
Note: MHAU=mobile healthy applications use; IA=information accessibility; PI=physician-patient interaction; Ta=tangibility; Em=empathy; Ass=assurance; Co=convenience; Rel=reliability; Res=responsiveness; TO=treatment outcome; VT=visit time; SQ=service quality; PS=patient satisfaction; CPC=clinical process change.
In the regression analysis of clinical process change perceived by patient and its dimensions as dependent variable, MHAU had a positive effect on clinical process change perceived by patient (regression coefficient=.456, p<0.01) while the other variables had no effect. Therefore, H1c should be accepted. In the regression analysis of the two dimensions of clinical process change perceived by patient, it was found that MHAU had a significant positive effect on both information accessibility (regression coefficient=.717, p<0.01) and physician-patient interaction (regression coefficient=.196, p< 0.01).
In the regression analysis of service quality and its dimensions as dependent variable, it was found that the impact of MHAU on service quality is insignificant (P=.665>0.05), nevertheless, it had a positive effect on tangibility exclusively (regression coefficient=.093, p=.007<0.01). While physician-patient interaction had an extensive positive effect on the six dimensions of service quality, and information accessibility had a positive effect only on the reliability of service quality (regression coefficient=0.065, P=.045<0.05). Thus, the null H1a is rejected unverifiable and alternative H2 is accepted.
In the regression analysis of patient satisfaction and its dimensions as dependent variable, the regression coefficients of MHAU and service quality as independent variables were .162 and .784, respectively. The p-value of 0.01 confidence level was significant, indicating that the above two variables could impact patient satisfaction positively. Therefore, the H1b and H3 will be accepted.
In the study of the influence of various dimensions of service quality on patient satisfaction, we found that tangibility (regression coefficient=.185, P=.000<0.01), empathy (regression coefficient=.192, p=.000<0.01), convenience (regression coefficient=.120, p=.000<0.01), reliability (regression coefficient=.133, p=.006<0.01), have positive effect on patient satisfaction, while the influence of assurance (regression coefficient=.068, p=.168>0.01) and responsiveness (regression coefficient=.042, p=.535>0.01) is null.
In the study of the influence of MHAU and the dimensions of service quality on the treatment outcome and visit time, it is found that MHAU, tangibility and empathy had significant positive effect on the treatment results and the treatment time; assurance and reliability only had a positive effect on the treatment outcome, but have no significant effect on visit time; convenience only had a significant effect on visit time, while there was no effect on treatment outcome; responsiveness had no effect on both treatment outcome and visit time.