Exploring Patients' Consultation Behaviors in the Online Health Community: The Role of Disease Risk

Abstract

Introduction: An online health community (OHC) is a unique social network that provides a platform for sharing healthcare information, communicating about health and offering health services. While many researchers have examined OHCs, the impact of patient reviews on patient consultation behavior in the OHCs has rarely been discussed in the existing literature.

Methods: A computer program was developed to download the required information. This process was conducted monthly to capture the variation in the dependent variable.

Results: By analyzing data from 907 physicians from the “Good Doctor Online” website, we found that online and offline patient reviews significantly affected the number of patient telephone consultations. In addition, disease risk significantly moderated the relationship between reviews and telephone consultations. Patients with high-risk diseases tended to pay more attention to offline patient reviews. In contrast, patients with low-risk diseases paid more attention to online service reviews.

Conclusion: It is important to explore the underlying mechanisms by which online versus offline reviews influence patients to undertake certain behaviors. This study aimed to explore the impact of online and offline reviews on patients' consultation behaviors and examine the moderating role of disease risk. Our results have implications for existing health management and e-health literature.

Introduction

Based on Health Online 2013, 59% of American adults use the internet to research health information.¹ An increasing number of internet users recognize that the internet is useful for acquiring knowledge and information on diseases and treatments,² which has stimulated the development of online health communities (OHCs). OHCs have emerged as new health platforms that encourage people to share ideas and experiences with others in similar circumstances, exchange support,^3,4 and promote interactions between doctors and patients.⁵

In the context of healthcare, since patients lack specialized healthcare knowledge and it is relatively difficult to obtain relevant medical information,⁶ accessing online reviews from OHC and then accessing required information seem to be very important for patients. Marx⁷ suggested that health-related topics have characteristics that are unlike those of other research domains. Healthcare services have been regarded as credence goods, which are goods for which an expert knows how much the consumer needs better than the consumer.⁸ In addition, information asymmetry has become a significant issue that can interfere with the quality of a patient's medical treatment.⁸ OHC reviews can help patients acquire relevant information about specific diseases or treatments, thus decreasing information asymmetry. Research has examined the effect of various kinds of product reviews on consumer behavior. However, few studies have explored the role of online reviews in OHCs in influencing patients' online consultation behaviors.

Therefore, the purpose of this study is to investigate how online and offline reviews influence patients' consultations behaviors. In addition, patients' behaviors and decisions can be influenced by their characteristics such as fitness, health conditions, and disease risk. Therefore, in this article, we focus on the moderating effect of disease risk on the relationship between online and offline reviews and patients' consultation behaviors in the OHCs. We therefore propose two research questions:

Q1: How do online service reviews and offline service reviews affect patients' consultation behaviors?

Q2: Can disease risk exert a moderating effect on influencing the relationship between patient reviews and patients' consultation behaviors?

To address these research questions, we collected data on 907 physicians from the “Good Doctor Online” website (www.“Good Doctor Online.”com), which is one of the largest online health communities in China. We developed an empirical model to test the proposed hypotheses.

Theoretical Background and Hypotheses

Online Health Communities

As Web 2.0 technologies are used within the healthcare industry, patients have begun to participate more actively in health management.⁹ OHCs provide a platform where patients can research health information and obtain support without face-to-face communication with physicians.¹⁰ The advantages of using OHCs include cost reduction, greater privacy, avoiding embarrassment, and more efficient and effective access to health information and obtaining medical services.¹¹ In general, OHCs show promise for improving the interaction between physicians and patients.⁴ To some extent, OHCs facilitate the development of the healthcare industry by encouraging patients to communicate with physicians more proactively.¹² There are now many well-known OHCs, including “Good Doctor Online,” Guahao, and PatientLikeMe. These platforms provide different services, including healthcare forums for patients to communicate with their peers, online medical consultation services, and reservations for offline physician-patient medical services. Thus, these OHCs help patients research the healthcare information they require and have effective interactions with physicians.

Existing literature has primarily focused on the adoption of online healthcare.^10,13 Some studies have examined the beneficial and motivational aspects of utilizing online healthcare.¹⁰ However, there is little information regarding patients' online consultation behaviors. In addition, few studies have investigated the moderating role of disease risk. To address this gap in knowledge, this study aims to explore the effects of online and offline reviews on patients' consultation behaviors and examine the role of disease risk during this process.

The Impact of Reviews on Patients' Consultation Behaviors

Existing studies have demonstrated that online reviews significantly influence people's decision-making processes.^14,15 A greater number of reviews implies greater participation in the review process, which reflects the popularity of a product or service.¹⁵ As OHCs have developed, an increasing number of internet users are obtaining health information and healthcare consultation from these sources. Many researchers have investigated the impact of online health information on patients' decisions.¹⁶ Fox and Raine¹⁷ found that most individuals believe that online health information has positive effects on the adoption of healthy behaviors. In addition, it has been shown that online health information can effectively influence individuals to make certain health decisions.¹⁸ Thus, OHCs have revolutionized the way patients access health information and health support outside hospitals.¹⁹ OHCs differ from traditional online services because healthcare services are credence goods, that is, goods for which physicians know their patients' needs better than the patients themselves. Therefore, the impact of patient reviews may be different under these circumstances than for other types of consumer reviews.

Patients who participate in OHCs often receive information about two types of service: online services and offline services. Online services refer to physician–patient telephone consultations. Patients can receive consultations on medical issues by calling doctors, and they pay for this service on the OHC platform. In contrast, offline services refer to face-to-face appointments between physicians and patients, often in hospitals. For these appointments, no fee is charged on the OHC. Online service reviews involve patients' opinions of their telephone medical consulting experiences. Therefore, online service reviews can directly influence the number of telephone appointments with physicians. In contrast, offline service reviews reflect patients' opinions on the face-to-face medical services provided by specific physicians. A patient's opinion on the attitude or expertise of a physician can induce other patients to select telephone consultations online. Given these factors, we propose the following:

Hypothesis 1(a): Online service reviews positively impact the number of telephone consultations with physicians.

Hypothesis 1(b): Offline service reviews positively impact the number of telephone consultations with physicians.

Because online service reviews of telephone-based services reflect more direct feedback than offline service reviews, we propose that patients will focus more on online service reviews than offline service reviews when they choose telephone-based medical services online. Hence, we propose:

Hypothesis 1(c): Online service reviews have a stronger effect on the number of telephone consultations with physicians than offline service reviews.

Moderating Effects of Disease Risk

The health information that an individual seeks is closely related to their individual characteristics.¹⁹ Individuals tend to collect information to reduce anxiety and uncertainty. People who perceive themselves as unhealthy may research health information more often.²⁰ Patients with different degrees of disease might behave differently. Baker et al.²¹ suggest that searching for health information online appears to be more prevalent among individuals in poor health. However, it is not clear whether the impact of reviews differs based on such patient characteristics. Therefore, we consider disease risk in our research model and examine its moderating effects on the relationship between online and offline reviews and the consultation behaviors of patients. We categorize disease risk into two types, high risk and low risk, to clarify the impact of reviews on patients' online medical service choices. Online service reviews represent direct feedback for online medical services. Patients with high-risk diseases might desire more direct information. Therefore, patients with high-risk diseases may be more likely to choose online medical services. However, patients with high-risk diseases may be less likely to be influenced by indirect information provided by offline service reviews. Based on the above factors, we hypothesize the following:

Hypothesis 2(a): High-risk disease positively moderates the relationship between online service reviews and the number of telephone consultations.

Hypothesis 2(b): High-risk disease negatively moderates the relationship between offline service reviews and the number of telephone consultations.

Our research model is shown in Figure 1. As the model shows, online service reviews (reviews about online services from physicians) and offline service reviews (reviews about offline services from physicians) determine patients' consumer behaviors. We also propose that disease risk moderates the effect of both online and offline service reviews on patients' consultation behaviors.

Fig. 1.

Research model.

Methodology

Research Context

In this study, the research context is the “Good Doctor Online” website. This site is one of the most influential physician–patient interaction platforms in China. Since 2016, the “Good Doctor Online” website has included more than 343,900 physicians from 3,310 standard hospitals. “Good Doctor Online” provides patients with several health services, including online consultations, online telephone appointments, and offline appointment booking. Online consultation services allow patients to receive consultations online for medical issues. Telephone consultation services allow patients to communicate with physicians over the telephone, and fees are paid through the “Good Doctor Online” platform. Offline appointment booking services let patients make an offline appointment with a specific doctor in certain hospitals using the online booking system of “Good Doctor Online.”

“Good Doctor Online” includes specific homepages for physicians. On these homepages, patients can access information about the doctors such as their affiliated hospital, titles, expertise, and telephone consultation fees. In addition, “Good Doctor Online” provides two different review forums for patients who have received online versus offline services. Online service reviews are also called service evaluations. In these reviews, patients who receive telephone consultation services express their feelings and attitudes regarding the services. In contrast, offline service reviews are called treatment experience sharing. Patients who received a face-to-face diagnosis with a doctor can share their personal treatment experiences with others.

Sample and Data Collection

We collected information on physicians and patients from the “Good Doctor Online” website (Figure 2). This process was conducted monthly to capture variations in the dependent variable. “Good Doctor Online” categorizes each physician based on their departments and expertise. We focused on several diseases in different departments and categorized them as high risk or low risk based on mortality rates. According to the China Health Statistics Yearbook (2013),²² cancer has a mortality rate of 166.33, so we labeled it a high-risk disease; in contrast, gynecological disease has a mortality rate of 0.09, so we labeled it a low-risk disease. The large difference between the mortality rates in these two diseases is significant enough for us to distinguish between them. Finally, we collected data on 907 physicians based on our research settings.

Fig. 2.

“Good Doctor Online” website home page.

Constructs and Measures

We use the number of telephone consultations (ΔLn Telephone Consultation Amount) to evaluate patients' online consultation behaviors. We made this decision because telephone consultation services are the only paid service that we can use to validate patients' consumer behaviors. Good Doctor Online provides two separate forums for evaluating online services and offline services. Accordingly, online service reviews (online service evaluation) and offline service reviews (face-to-face experience sharing) are generated from the same platform. We use the number of online service reviews (Online Service Reviews) and the number of offline service reviews (Offline Service Reviews) as our primary independent variables. To distinguish between disease risk, we used cancer (high risk) and gynecological disease (low risk), which have significantly different mortality rates, as moderating factors. For cancer, we chose lung cancer, breast cancer, lymphoma, gastric cancer, liver cancer, brain tumors, and osteoma. For gynecological diseases, we chose menstrual disorders, endocrine disorders, cervicitis, vaginitis, dysmenorrhea, pelvic inflammatory disease, and ovarian cysts. We used a dummy variable for disease risk (Disease). When a disease belongs to the high-risk group, Disease is 1; for the low-risk group, Disease is 0. Descriptions of the major constructs are given in Table 1.

Table 1.

Variables Description

VARIABLES	DESCRIPTION
Online service reviews	Patients write online service reviews after online telephone consultations with physicians. We use the number of online service reviews to quantify online service reviews.
Offline service reviews	Patients write offline service reviews after face-to-face appointments with physicians in hospitals. We use the number of offline service reviews to quantify offline service reviews.
Disease risk	Patients have different types of diseases. We divided diseases into high-risk and low-risk categories. We classified diseases with higher mortality rates as high risk, including lung cancer, breast cancer, lymphoma, gastric cancer, liver cancer, brain tumors, and osteoma. We classified diseases with lower mortality rate as low risk, including menstrual disorders, endocrine disorders, cervicitis, vaginitis, dysmenorrhea, pelvic inflammatory disease, and ovarian cysts.
Telephone consultations	The number of patients who have telephone consultations with physicians.
ΔLn Telephone Consultation Amount	The number of telephone consultations
Title	Physician titles
Fee	Telephone fees
Online_Consultation	Online consultation services
Offline_Appointment	Offline face-to-face appointments
Disease	Disease risk

We also introduced several control variables in our research model. Physicians have different titles in their affiliated hospitals, including chief physician, associate chief physician, and attending physician. We used three dummy variables for physician titles (Title1, Title2, and Title3), which are controlled in our research model. We also controlled for telephone fees (Fee), online consultation services, (Online_Consultation), and offline face-to-face appointments (Offline_Appointment) because these factors could influence patients' online consumer behaviors. The control variables are as follows: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} Title1 \, \, \left\{ { \begin{matrix} { \; = 1 , \; \;{ \rm{when \;physician \;title \;is}} \;chief \;physician} \\ { = 0 , \; \;{ \rm{others}} \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;} \\ \end{matrix} } \right. \end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} Title2 \, \, \left\{ { \begin{matrix} { \; = 1 , \; \;{ \rm{when \;physician \;title \;is}} \;associate \;physician} \\ { = 0 , \; \;{ \rm{others}} \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;} \\ \end{matrix} } \right. \end{align*} \end{document} \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} Title3 \,\,\left\{ { \begin{matrix} { \; = 1 , \; \;{ \rm{when \;physician \;title \;is}} \;attending \;physician} \\ { = 0 , \; \;{ \rm{others}} \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;} \\ \end{matrix} } \right. \end{align*} \end{document}

Online_Consultation \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \left\{ { \begin{matrix} { \; = 1 , \; \;{ \rm{when \;physicians \;provide \;online \;consultation \;services}}} \\ { = 0 , \; \;{ \rm{others}} \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;} \\ \end{matrix} } \right. \end{align*} \end{document}

Offline_Appointment \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \left\{ { \begin{matrix} { \; = 1 , \; \;{ \rm{when \;physicians \;provide \;face - to - face \;appointment \;services}}} \\ { = 0 , \; \,{ \rm{others}} \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \; \;} \\ \end{matrix} } \right. \end{align*} \end{document}

We collected data during two different time periods. The difference between telephone consultations during the two time periods was regarded as the dependent variable. The remaining variables in our research model were measured using the data collected during the earlier time period (t − 1). Our empirical model is as follows: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} \begin{split} & \Delta { \rm{Ln}} \ Telephone \ Consultation{} \ Amoun{t_t} \\ & = { \rm{Ln}} \,Telephone{ } \ Consultation{ } \ Amoun{t_t} - \\\quad & \quad { \rm{Ln}} \,Telephone{ } \ Consultation{ } \ Amoun{t_{t - \it1}} \\ & = { \beta _{i\it1}}Ttitle{1_{t - \it1}} + { \beta _{i\it2}}Ttitle{2_{t - \it1}} + { \beta _{i\it3}}Ttitle{\it3_{t - 1}} \\\quad &\quad +{ \beta _{i\it4}}Fe{e_{t- \it1}} + { \beta _{i\it5}}Online \_consultatio{n_{t- \it1}} \\ & + { \beta _{i\it5}}Offline \_Appointmen{t_{t- \it1}} + { \beta _{i\it6}}Online{ } \ Service{ } \ Review{s_{t - \it1}} \\\quad &\quad + { \beta _{i\it7}}Offline \ Service \ Review{s_{t - \it1}} \\ & + { \beta _{i\it8}}Disease + { \beta _{i\it9}}Online \ Service \ Review{s_{t - \it1}} \times Disease \\\quad &\quad + { \beta _{i\it10}}Offline \ Service \ Review{s_{t- \it1}} \times Disease + \varepsilon\end{split} \\\quad \tag{1} \end{align*} \end{document}

Analysis and Results

Table 1 shows descriptive statistics and correlations. The results demonstrate that both online and offline service reviews have significant and positive correlations with the number of telephone appointments. We used the ordinary least-squares model to test our hypotheses; the results are shown in Table 2. The adjusted R ² values and the significance of the F-values reflect the explanatory ability of the model. In addition, the variance inflation factor values are all are below 1, which indicates that multicollinearity is not a concern in our research model.

Table 2.

Description and Correlation

	MEAN	STD.	1	2	3	4	5	6	7	8	9
ΔLn Telephone Consultation Amount	0.577	0.826
1. Title1	0.44	0.497	0.198^**
2. Title2	0.40	0.489	−0.096^**	−0.717^**
3. Title3	0.53	0.499	0.176^**	0.314^**	−0.141^**
4. Fee	141.92	58.576	0.460^**	0.300^**	−0.095^**	0.145^**
5. Online_Consultation	0.92	0.274	−0.116^**	−0.020	0.002	0.036	−0.223^**
6. Offline_Appointment	0.24	0.430	0.337^**	0.154^**	−0.029	0.120^**	0.288^**	−0.009
7. Disease	0.40	0.490	−0.071^*	−0.035	0.037	−0.028	0.062	0.070^*	−0.005
8. Online Service Reviews	2.175	1.299	0.768^**	0.255^**	−0.123^**	0.193^**	0.448^**	−0.074^*	0.416^**	−0.102^**
9. Offline Service Reviews	1.296	1.430	0.595^**	0.371^**	−0.127^**	0.210^**	0.450^**	−0.082^*	0.380^**	−0.166^**	0.678^**

***

Indicates p < 0.001, ^**Indicates p < 0.01, ^*Indicates p < 0.05.

The results for model 2 show that online service reviews had a significant and positive impact on the number of telephone consultations (β = 0.366, p < 0.000). The results also show that offline service reviews significantly affected the number of telephone consultations. Therefore, Hypotheses 1(a) and (b) are both supported, indicating that the number of reviews directly influences patients' online consultation behaviors. To compare the impact of the two kinds of service reviews, we used the lincom function in STATA to test the impact of online versus offline service reviews. The results show that online service reviews have a greater impact than offline service reviews (see formula below). Thus, Hypothesis 1(c) is supported. \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\usepackage{upgreek}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} Sig \ { \beta _{online \ service \ reviews}} \; - \;{ \beta _{offline \ service \ reviews}} \; = \;0.000 \end{align*} \end{document}

The results for model 3 in Table 3 indicate that disease risk had a significant moderating effect. However, disease risk had negative moderating effects on the relationship between online service reviews and the number of telephone consultations (β = −0.067, p < 0.05), so Hypothesis 2(a) is not supported. Disease risk also had positive moderating effects on the relationship between offline service reviews and the number of telephone consultations (β = 0.079, p < 0.05), so Hypothesis 2(b) is not supported. The hypotheses' results are shown in Table 3.

Table 3.

Empirical Model Results

VARIABLE	MODEL 1	MODEL 2	MODEL 3
Title1	−0.022	0.170^*	−0.164^*
	(0.076)	(0.057)	(0.057)
Title2	−0.085	−0.100	−0.096
	(0.071)	(0.052)	(0.052)
Title3	0.156^**	0.057	0.062
	(0.050)	(0.037)	(0.037)
Fee	0.005^***	0.002^***	0.002^***
	(0.000)	(0.000)	(0.000)
Online_Consultation	−0.101	−0.104	−0.104
	(0.089)	(0.065)	(0.065)
Offline_Appointment	0.418^***	−0.008	−0.013
	(0.058)	(0.045)	0.045
Disease		0.011	−0.071
		(0.036)	(0.069)
Online Service Reviews		0.366^***	0.391^***
		(0.017)	(0.021)
Offline Service Reviews		0.082^***	0.051^*
		(0.020)	(0.024)
Online Service Reviews^ Disease*			−0.067^*
			(0.034)
Offline Service Reviews^* Disease			0.079^*
			(0.037)
Adjusted-R2	0.263	0.614	0.615
F change	86.558^***	272.328^***	2.523

***

Indicates p < 0.001, ^**Indicates p < 0.01, ^*Indicates p < 0.05.

Online service reviews are contributed only by patients who consulted with physicians online. Offline service reviews are posted by patients who had face-to-face consultations with doctors in hospitals. Therefore, offline medical reviews may be more accurate. Online service reviews offer patients more information and reduce their uncertainty, which in turn may help them judge whether a telephone consultation would be helpful for them. Prior research has shown that people in worse health have more intense health information needs.²¹ Therefore, for patients with high-risk diseases, offline service reviews have a stronger impact on their online consultation behaviors. Conversely, for these patients, online service reviews may have a limited effect on their choice to schedule a telephone consultation.

Discussion

In this article, we primarily examine the impact of online and offline service reviews on patient consultation behavior in the OHCs. We propose that prior reviews from patients are important factors that influence patients' online consumer behaviors. Our findings provide valuable insight into the role of reviews. The empirical results show that most of these hypotheses are supported and yield interesting findings for the unsupported hypotheses (Table 4). We found that online and offline service reviews are positively impacted by the number of telephone consultations with physicians, which is consistent with Hypotheses 1(a) and (b). This statistical evidence suggests that physicians with more online and offline service reviews from their patients will attract more patients in the future. In addition, Hypothesis 1(c) postulates that online reviews have greater influence than offline reviews on patients' consultations with physicians, which indicates that online reviews are direct feedback to those potential patients for online telephone consultations and have stronger effect on them than offline service reviews. Hypotheses 2(a) and (b) propose that patient characteristics (disease risk) will moderate the relationship between reviews and patients' consultations with physicians. It can be seen from Hypotheses 2(a) and (b) that disease risk has a moderating influence on the relationship between service reviews and patients' online consultation behaviors. Specifically, when patients have high-risk diseases, they care more about offline service reviews than online service reviews. In contrast, patients with low-risk diseases pay more attention to online service reviews than offline service reviews. Our results also suggest that patients are more affected by reviews contributed by peers facing similar disease. These results can be explained by empathic concern, which was greater among people who had the same or similar experiences.⁴ Specifically, when individuals seek information from the OHCs, they feel reliable and empathetic to find the evaluations and reviews shared by people who encountered or experience the same issues.

Table 4.

Summary of Results

HYPOTHESIS	RESULTS
Hypothesis 1(a): Online service reviews positively impact the number of telephone consultations with physicians.	Supported
Hypothesis 1(b): Offline service reviews positively impact the number of telephone consultations with physicians.	Supported
Hypothesis 1(c): Online service reviews have a greater effect on the number of telephone consultations with physicians than offline service reviews.	Supported
Hypothesis 2(a): Disease risk positively moderates the relationship between online service reviews and the number of telephone consultations with physicians.	Not supported
Hypothesis 2(b): Disease risk negatively moderates the relationship between offline service reviews and the number of telephone consultations with physicians.	Not supported

Conclusion

This study makes three main contributions. First, we verify that medical service reviews positively impact patients' online consultation behaviors. Second, our study contributes to existing theories of reviews and patients' online consultation behaviors by testing the moderating effects of disease risk. Third, we contribute to the body of research examining reviews by assessing two kinds of reviews, online and offline reviews. Although previous studies have indicated the importance of reviews and confirmed that reviews strongly influence people's decision-making processes,²³ few studies have attempted to distinguish between different kinds of reviews.

This study has significant practical implications. For physicians, our findings suggest that the online consultation behavior of patients can be influenced by prior reviews. Thus, physicians can attract more patients by encouraging their patients to write reviews. In addition, our results suggest that physicians may find different types of reviews more helpful depending on the disease risks of the patients they treat. Physicians should emphasize offline service reviews if they frequently diagnose high-risk diseases. In contrast, online service reviews may be more important for physicians who primarily treat low-risk diseases.

This article also has several limitations and suggests multiple avenues for future research. First, because we used cross-sectional analysis, future studies are suggested to use longitudinal data to investigate the relationships we examined in this study. Second, we only used the number of online and offline service reviews to examine the impact of reviews on patients' online consultation behaviors. Thus, future research can examine textual information and features, especially regarding the different characteristics of online and offline service reviews. Third, we excluded patient demographics as factors for consideration in this study. However, patients' consultation behaviors can also be influenced by individuals with different gender, age, and income. To explore the importance of online and offline reviews, future studies could consider patient demographics and compare the different effects of individuals with different characteristics.

Footnotes

Disclosure Statement

This article has not been published or is not under consideration of publication elsewhere. We have no conflicts of interest to declare.

References

Liu

, Guo

, Wu

. The impact of individual and organizational reputation on physicians'appointments online. Int J Electr Commerce, 2016; 20:551–577.

Ziebland

, Chapple

, Dumelow

, Evans

, Prinjha

, Rozmovits

. How the internet affects patients' experience of cancer: A qualitative study. BMJ, 2004; 328:564.

Kumar

, Zhao

. Making sense of a healthcare forum—Smart keyword and user navigation graphs. In: International Conference on Information Systems. Milan, Italy, December 16–19. 2013.

Nambisan

Information seeking and social support in online health communities: Impact on patients' perceived empathy. J Am Med Inform Assoc, 2011; 18:298–304.

Van De Belt

, Engelen

, Berben

, Schoonhoven

. Definition of Health 2.0 and Medicine 2.0: A systematic review. J Med Internet Res, 2010; 12:e18.

Carlsson

ME.

Cancer patients seeking information from sources outside the health care system. Support Care Cancer, 2000; 8:453–457.

Marx

GT.

Hey Buddy can you spare a DNa? New surveillance technologies and the growth of mandatory volunteerism in collecting personal information. Ann Ist Super Sanita, 2007; 43:12.

Dulleck

, Kerschbamer

. On doctors, mechanics and computer specialists—The economics of credence goods. J Econ Lit, 2006; 44:15–42.

Swan

Emerging patient-driven health care models: An examination of health social networks, consumer personalized medicine and quantified self-tracking. Int J Environ Res Pub Health, 2009; 6:492–525.

10.

Cao

, Liu

, Zhu

, Hu

, Chen

. Online selection of a physician by patients: Empirical study from elaboration likelihood perspective. Comput Human Behav, 2017; 73:403–412.

11.

Cline

, Haynes

. Consumer health information seeking on the Internet: The state of the art. Health Educ Res, 2001; 16:671–692.

12.

Welch Cline

, Penner

, Harper

, Foster

, Ruckdeschel

, Albrecht

. The roles of patients' internet use for cancer information and socioeconomic status in oncologist-patient communication. J Oncol Pract, 2007; 3:167–171.

13.

Guo

, Zhang

, Sun

. The privacy–personalization paradox in mHealth services acceptance of different age groups. Electr Commerc Res Appl, 2016; 16:55–65.

14.

Zhu

, Zhang

. Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics. J Market, 2010; 74:133–148.

15.

Zhang

, Zhang

, Wang

, Law

, Li

. Factors influencing the effectiveness of online group buying in the restaurant industry. Int J Hosp Manag, 2013; 35:237–245.

16.

Wang

, Xiao

, Rao

. An exploration of risk information search via a search engine: Queries and clicks in healthcare and information security. Dec Supp Syst, 2012; 52:395–405.

17.

Fox

, Raine

. How internet users decide what information to trust when they or their loved ones are sick. Pew Internet Am Life Proj 2002, 2011. Available at http://www.pewinternet.org/2002/05/22/vital-decisions-a-pew-internet-health-report/ (last accessed June 14, 2018 ).

18.

Lorig

, Laurent

, Deyo

, Marnell

, Minor

, Ritter

. Can a Back Pain E-mail Discussion Group improve health status and lower health care costs?: A randomized study. Arch Intern Med, 2002; 162:792–796.0.

19.

Lambert

, Loiselle

. Health information—Seeking behavior. Qual Health Res, 2007; 17:1006–1019.

20.

Xiao

, Sharman

, Rao

, Upadhyaya

. Factors influencing online health information search: An empirical analysis of a national cancer-related survey. Dec Supp Syst, 2014; 57:417–427.

21.

Baker

, Wagner

, Singer

, Bundorf

. Use of the Internet and e-mail for health care information: Results from a national survey. JAMA, 2003; 289:2400–2406.

22.

China Health Statistics Yearbook. 2013. Available at www.nhfpc.gov.cn/htmlfiles/zwgkzt/ptjnj/year2013/index2013.html (last accessed January 15, 2018 ).

23.

Cui

, Lui

H-K

, Guo

. The effect of online consumer reviews on new product sales. Int J Electr Commerce, 2012; 17:39–58.