"Can I Trust You, Doc?": User Perception of Online Health Information
"Can I Trust You, Doc?": User Perception of Online Health Information
"Can I Trust You, Doc?": User Perception of Online Health Information
Abstract—Klikdokter.com is one of websites that provides consists of editorial team which supports both general
information service about health. Besides, it also provides live practitioner and specialist doctors. Klikdoktercom also
chat service i.e., question and answer about health which is provides live chat service i.e., question and answer with doctors
directly claimed and answered by the specialist doctors. who are expertise in their field [3].
According to the survey conducted by IndoPacific Edelman and
Research Unit and Community Service of Public Health Faculty According to the facts above, it is interested to have further
of Indonesia University, the amount of 79% doctors believes in research about the users’ trust towards health information
the most accurate health information source on the internet. provided in klikdokter.com. In this case, the users do not know
Moreover, a study entitle Patients Use of the Internet for Medical whether that information comes directly from the doctors.
Information shows that 53,5% out of 512 patients use internet as Furthermore, they also do not know whether the information
a media to find out information about health. The trust of doctors they get is responsible and accountable. It means that the users
and internet users towards health information is indeed are expected to use the available information correctly.
influenced by several factors. Therefore, the aim of this research Moreover the users also should find an effort to understand and
is to analyze the factors that influence the trust of users towards to decide whether that information is correct or not. If the users
the health information uploaded on klikdokter,com using ten have limited knowledge about health, it is indeed not an easy
construct variables. The data is collected from klikdokter.com thing to do [4].
users with 250 respondents. The data is analyzed using structural
equation modeling analysis (SEM). The result shows that factors Another problem is related to the live chat service provided
that influence Trust are perceived information quality, source by klikdokter.com which is directly claimed by the doctors
expertise and argument quality. who are expertise in their field. In this case, the users can
virtually do the live chat without knowing whether it is the
Keywords - SEM;confirmatory research; online health information doctor who answers the questions from users.
According to the problems above, there is a connection
I. INTRODUCTION
between the user and doctor. If there is a mistake in diagnosing,
Many kinds of health information are available on the so the doctor’s answer would not be responsible and
internet. A research conducted by Pew Research Center accountable. Moreover, it would be getting worse if patient
California Health Care Foundation finds that 8 out of 10 health does not understand that the answer or medical suggestion from
questions begin with internet searching. To the amount of 72% the doctor is incorrect.
internet users search for health information related to serious
condition, general information, and specific health problem. In this case, a high level of internet using to search for
Another finding shows that 77% users begin with search engine health information is influenced by particular factors so that the
such as Google, Bing, or Yahoo. 13% said that they begin the users can trust the website. It makes the users not hesitant to
searching through website about health first. Only 2% who always use internet to search for health information.
begin the searching through Wikipedia while another 1% begin The research method used is adapted from some previous
the searching through social media such as Facebook [1]. studies, such as a study conducted by Koo and Wati [5] entitle
A high level of users’ interest to do the health information E-Healthcare Service: An Investigation of the Antecedents,
searching via internet is due to the number health website. Moderating Roles, and Consequences noted that the webs
There are more than 70.000 website that provide health quality and perceived credibility of a health site affects the
information [2]. One of them is klikdoktercom. Klikdokter.com users’ satisfaction towards the site so that it can increase the
is one of the health portals that promote Health, Information intention to use. It is supported by a study conducted by M. Y.
and Education about health based on Bahasa Indonesia which Yi et al. [6] entitle Untangling the antecedents of initial trust in
is designed not only for medical community but also society in Web-based health information: The roles of argument quality,
Indonesia. Practically, klikdoktercom has an editorial staff source expertise, and user perceptions of information quality
589 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
and risk which connects antecedents factors that affect users’ will affect the information credibility so that it will decrease
trust towards health information through online access. These Perceived Risk [6].
two studies show that there is a correlation between intention to
use and trust. Consequently, a study entitles Technology Perceived Risk
Acceptance Model for the Use of M-Health Services among Health information is directly connected with prosperity
health related users in UAE proves the correlation. That study and risk due to the acceptance and usage of the health
is conducted by Alloghani, M et al. [7] which states 4 factors information. However, before the acceptance and the usage,
that affects Intention to Use of health service users. They are users should ensure that they receive well-structured health
Perceived Usefulness, Perceived Ease of Use, Trust, and information so that users can minimalize the risk [6].
Security.
This research aims to examine whether the factors can User Satisfaction
influence the users’ trust towards the health website; Bhattacherjee, A [9] argues that user satisfaction is
klikdokter.com. psychology condition of the users on two times; before using
the site and after using the site.
II. MODEL STRUCTURE AND HYPOTHESES
Perceived Usefulness
The methodology used in this research is confirmatory
According to Alloghani, M et al [7], perceived usefulness
quantitative that is used to examine the model and hypothesis
conducted by (Koo dan Wati [5]; M. Y. Yi et al. [6] and is defined as to what extent the m-Health is beneficial for the
users. If the patients and doctors feel that the online health
Alloghani, M et al. [7]). Data is analyzed using Structural
information service is useful and able to increase their
Equation Modelling (SEM) that is used to analyze independent
productivity, as a result they will have intention to use it.
variable and dependent variable which connect each other to
form a model. The SEM analysis is conducted in 2 steps;
Perceived Ease of Use
structural model test and measurement model test. Structural
When users feel that a particular system can ease the users
model shows the correlation between constructs which is
in doing their tasks, the users will have intention to always use
generally linear. Measurement model is used to measure the
it [7].
correlation between indicator and variable.
Model in the Fig. 1 is a model that is used in this research.
Security
It shows a correlation among each construct. There are 9
constructs investigated in this research and 34 manifest The ultimate goal of online health service is not only to
increase the quality life of patients but also to get security and
variables or indicators. The nine constructs and 34 manifest
privacy. The distribution of health service via online device
are presented in Table 1.
can cause some problems such as lost control, data security,
data encryption, information distribution and privacy [7].
A. Definition of each construct
Web Quality Intention to Use
Koo and Wati [5] note that five dimensions of Website Intention to use is defined as a readiness to utilize the site.
Quality; Information Quality, Information Presentation, Hassenzahl and Trautman [10] suggest that character of the
Website Attractiveness, Navigation, and Technical Support site influence the users’ interpretation.
have direct influence in increasing Perceived Credibility.
Trust
Argument Quality Alloghani, M et al.[7] finds a direct effect between trust
In the consultation and suggestion, there is a high level of and intention to use. Besides, another study also found that
consequence from the trust and incorrect implementation of trust is a social factor that influences a large effect towards the
information. Therefore users should search high quality health use of mobile commerce technology, one of them is Online
information in the internet. Users tend to measure component health information [11].
of argument which is relevant, accurate and useful [8].
Perceived Quality
The high level of health information quality represents a
more relevant, accurate, up-to-date and complete information.
When those aspects are completed, it can decrease the risk [6].
Source Expertise
Some studies has identified that Source Expertise is
necessary to build credibility from online health information.
Some researchers believed in that statement. In the
management of online health consultation, source expertise
590 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
591 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
592 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
B. SEM Analysis
Structural equation modelling (SEM) statistical analysis is
used to analyze the data that is collected through
questionnaire. SEM is a series of statistical method that allow
complex relationships between one or more independent and
dependent variables. Though there are many ways to describe
SEM, it can be remarked that SEM allows one to perform
some type of multivariate data.
The ten proposed hypothesis were tested using AMOS.
AMOS is a software program used for structural equation Figure 2. Overall Model Fit after being modified
modeling (SEM). In order to answer the hypothesis, there are
several steps that should be completed as explained on the Table 5. The Result of Measurement Model Fit Test
methodology.
1) Missing data and Outlier Test Indeks Limit Value Note References
Chi-
Missing data test in this research shows that there is no square
>0.05 986.561 Good Fit [23]
missing data. On the other hand, outlier test is used to find the CMIN/
<3.00 1.981 Good Fit [23]
gap of a data with measuring score of mahalanobis distance. DF
Outliers are found by first computing a Mahalanobis Distance GFI > 0.8 0.805 Good Fit [23]
NFI > 0.8 0.849 Good Fit [23]
and once that is done the Mahalanobis scores are screened in
CFI > 0.9 0.918 Good Fit [23]
the same manner that multivariate outliers are screened. To RMSEA < 0.08 0.063 Good Fit [23]
compute Mahalanobis distance in SPSS you must use
regression/linear under analyze. By measuring it with level of
mistake, the result shows 72.443. The data that has
mahalanobis distance more than 72.443, it is called outlier and
it must be omitted. Out of 250 questionnaires, there are 37
outliers. Consequently, the data that can be analyzed is 213
data.
593 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
594 https://sites.google.com/site/ijcsis/
ISSN 1947-5500
International Journal of Computer Science and Information Security (IJCSIS),
Vol. 14, No. 12, December 2016
595 https://sites.google.com/site/ijcsis/
ISSN 1947-5500