Health information acquisition online and its influence on intention to visit a medical institution offline

Introduction

The number of individuals using the Internet to acquire health care information is constantly increasing, with more than 70,000 Websites providing health care information (Um 2010). According to Harris Interactive, an American market research company, the number of adults in the US searching for health care information using the Internet rose from 54 million in 1998 to 150 million in 2008. This represents 66% of the adult population. In addition, approximately 47% of those individuals who retrieved health care information through various Websites used the information when consulting physicians.

In the past, patients obtained health care information solely through contacting health care providers while visiting medical institutions. However, the Internet has not only allowed patients to acquire information through Websites; it has also helped patients to actively cope with their illnesses or medical conditions in U.S. (Robinson et al. 1998). The Internet is a medium of public relations, which supports content-oriented, rational decision-making by its users. Health care is a high-involvement product, and, to make prudent decisions, health care consumers aggressively collect information on the quality of various medical institutions (Wilkins 1999; Ninfa 2008). Medical institution Websites are important because i) the health care market is shifting from provider-oriented to consumer-oriented, ii) the institution's image and quality may significantly affect initial consumer trust, and iii) marketing activities through the Internet often lead to low-cost and high-utility effects (Eng et al. 1998; Gallagher 1999; Collier and Bienstock 2006).

Meanwhile, the Korean health care system views the purpose of Websites from a rather dissimilar perspective to the way Websites are viewed in other countries. Specifically, unlike the US that uses Webcam technology to provide medical treatment remotely, Korean health care institutions tend to avoid such services, due to the possibility of medical negligence caused by an ambiguous legislation system. Therefore, Korean health care-related Websites do not provide actual medical services online, but are exclusively used as a medium to provide information about treatment and to advertise the particular institution. Also, health care institutions focus on the opportunity of gaining consumer attention through the Internet. In Korea, there are strict rules against attracting patients for profit-making purposes and prohibition of advertisements via television and newspapers which do not apply to the Internet. More specifically, Website consumer attraction is an issue especially in health care institutions including dental, dermatological and cosmetic clinics, rather than in large hospitals. Therefore, identifying the key role of Websites as marketing means is imperative in understanding the Korean primary health care system.

The majority of previous studies focused on the quality of the Websites of medical institutions, customer satisfaction, and online behaviour including Website re-visits. However, research on Website service quality with respect to the effective acquisition of online health care information by consumers and its effect on the consumer intention to visit and purchase medical services has been insufficient. We defined the Website service quality as information necessary for delivering health care effectively. We also measured satisfaction from the health care information delivered online. Our research goal is to investigate the effect of consumer satisfaction from the health care information acquired online on consumer intention to visit medical institutions and purchase health care services. The structure of this study is as follows. The first section discusses the theoretical background and presents the research model. The second section presents the research methodology, and the third section shows the results of empirical analyses and hypothesis testing. The final section provides the conclusions of our study.

Theoretical background and research model

Overview

In the past, patients relied on health care providers, particularly physicians, to acquire health care-related information and knowledge (Goldsmith 2000). This limited patients' choice of health care information and adhering to physicians' preferences created an unequal relationship between the patient and physician (Shackley and Ryan 1994). However, the development of mass media and the advent of the Internet brought about an enormous change in the traditional patient-physician relationship. The Internet has not only facilitated health care information acquisition by patients (Brodie et al. 1999), but also made different patient-physician interactions and cost reduction possible (Robinson et al. 1998). As the number of patients and potential customers using the Internet to acquire health care information increases (Laing et al. 2005), it has opened up a new window of business opportunities for health care providers. This led to the need for Websites (Lin and Lu 2000). In order for the customer to acquire information, Palmer and Griffith (1998) argued that Websites need to be designed to allow quick and easy information retrieval, and Marwick (2000) emphasised that health care information available through the Website must be easy to access and interesting.

Cox and Dale (2001) argued that customers may acquire information and purchase products through the Website, but the Website itself does not provide customers with tangible products. Furthermore, they identified e-commerce with intangibility. Alba et al. (1997) asserted that tangible exchange of products cannot take place online and face-to-face interaction cannot be achieved through a Website.

In the case of online shopping malls, value is created online because the transaction takes place on the Internet. For medical institution Websites, however, value is created when the patient actually visits the institution and purchases health care. Kim et al. (2006) reported that an environment supporting fast information browsing facilitates information comparison by customers, and this strongly affects consumer intention to visit and purchase services offline. Jeong et al. (2001) reported that customers satisfied with the information provided through Websites show positive purchase intention to purchase services offline. Cronin and Taylor (1992) investigated the causality between service quality and satisfaction, and confirmed that service quality precedes satisfaction. Studies by Taylor and Baker (1994) and Grönroos (1984) are representative of research on the relationship between service quality and customer satisfaction. They verified that customer satisfaction is affected by service quality.

Usability is one of the determinants of e-service quality (Dabholkar 1996). It influences customer satisfaction by improving the efficiency of service utilization (Xue and Harker 2002). Hoffman et al. (1999) confirmed that download speed and Internet user satisfaction are positively related to each other. Gummerus et al. (2004) reported that responsiveness is expected to affect customer trust, and security, as a signal of the integrity of service providers and is directly related to customer trust. Ambre et al. (1997) asserted that Website service quality affects information acquisition by the consumer. Many researchers including Bauer et al. (2006), Kaynama and Black (2000), and Liu and Arnett (2000) found that information acquisition by the customer is one of the most important variables among the Website service quality factors. Therefore, we expect information acquisition by the customer will affect customer intention to visit and purchase health care services.

Information acquisition

Angehrn (1997) argued that the fundamental goal of the Website is to display and provide information on the products and services offered by companies through new Internet-based channels. From a consumer perspective, therefore, the Website has allowed a new and effective means for collecting information and comparing between different products. Huizingh (2000) reported that consumers access Websites that may contain useful information or provide links to such information. According to Boyer et al. (1998), customers visit health care-related Websites to acquire health care information such as treatment procedures and methods, and medical consultation. Wyman et al. (1997) reported that the location of information and appropriate use of graphics within the content are important, and that distracting elements need to be eliminated from the Website. In this study, information acquisition is defined as the degree to which consumers perceive ease of acquisition and usability of the particular information.

Website service quality

According to Zeithaml, Parasuraman and Malhotra (2000), customers evaluate Websites based on the interactions with the seller online, but also on the service provided following such interactions. Parasuraman et al. (2005) reported that e-service quality encompasses efficient and effective purchase, fast delivery, and consumer recognition of high-quality services. In online auctions, Yen and Lu (2008) reported that the e-service quality is related to customer satisfaction. Most of the variables suggested in previous studies were developed for Websites such as Internet shopping malls where financial transactions occur. These variables including price information, online shopping mall policies, security of payment methods, indemnity, and shipping and handling were considered inappropriate for measuring the e-service quality of medical institution Websites. From the many variables of Website service quality in reference to information acquisition, we adopted four variables: usability, aesthetics, responsiveness, and security.

Usability

According to Webster and Williams (2005) and Tarafdar and Zhang (2005), usability of the Website directly affects the Website service quality because if the user cannot access the Website, the information provided becomes meaningless, regardless of the quality of the Website. Pearson et al. (2007) defined usability as the measurement of the quality of user experience pertinent to utilizing goods or systems including Websites, software, and mobile communications. Usability can be properly measured by the Website's navigation, customization and personalization, download speed, accessibility, and user convenience. Among these five measurements, customization and personalization facilitates personalized services to meet customer needs and wants by securing individual customer information (Franzak et al. 2001). However, customization and personalization are inappropriate in the health care industry because protecting user privacy and developing services with anonymity in mind are imperative (Gummrus et al. 2004). As demonstrated by Pearson et al. (2007), Website accessibility implies that information on the Website must be made available to every potential user, including individuals with visual or hearing impairment. We defined usability as systematic services experienced by the user while utilizing the Website, and adopted three which are ease of navigation, download speed, and convenience. The following hypothesis is established in terms of usability of the Website:

Aesthetics

Apart from pursuing usability, Websites need to provide visual pleasure and satisfaction. Wolfinbarger and Gilly (2002) and Zeithaml et al. (2000) defined this property as e-scape arguing that it directly influences the Website service quality. Aesthetics implies those attributes related with the appearance of Websites. It reflects how information is delivered to users through clarity and legibility of colour mix, font, animation, image, sound effects, and text (Abels et al. 1997, Gummerus et al. 2004). In this study, aesthetics of the Website is based on customer evaluation of the audio-visual measures used to deliver health care information, and thus we propose the following hypothesis:

Security

Security is the degree of customer trust in the safety and protection of personal information on the Website (Zeithaml et al. 2000). Minocha et al. (2003) defined security as confidentiality of information associated with online payment and customer privacy. Although security is often compared with reliability (Reichheld and Schefter 2000), reliability is not always based on security policies and information protection on the Website (Gummerus et al. 2004). Security, in this study, is defined as user trust in Website administrators or managers to protect user information on the Website, and the following hypothesis is proposed:

Responsiveness

Responsiveness is defined as the depth of the information that the customer requested, response time taken across the feedback inquiry process, and ability of the service provider to assist the customer when needed (Parasuraman et al. 1988). Yang and Jun (2002) and Zhu et al. (2002) emphasised the importance of the service quality perceived by the customer and customer satisfaction in terms of responsiveness of Internet-based services. Yang and Jun (2002) reported that prompt response by Website administrators or managers helps customers solve problems and make timely decisions. When compared to face-to-face transactions, consumer inquiries through e-mail on a Website often lead to delayed responses. Minocha et al. (2003) defined responsiveness as the length of time it takes for the service provider to respond to customer inquiries, requests, orders, and complaints, or to deliver the goods or services to the customer through the Internet. We defined responsiveness as the volition to react to customer needs pertinent to health care information quickly and accurately, and propose the following hypothesis:

Customers accessing a Website are satisfied when they obtain the desired information from the site. Anderson and Srinivasan (2003) asserted that satisfaction is a distinct assessment derived from consuming resources to purchase particular goods or services, and defined e-satisfaction as the satisfaction of the customer with previous purchase experiences through e-commerce entities. When compared to satisfied customers, dissatisfied customers tend to search for and switch to alternative Websites. Muylle et al. (2004) reported that Website satisfaction is a series of items concomitant of online contents, structure, delivery methods, service-related user satisfaction, and positive experience. Website satisfaction may be measured by usability shown in the decision-making process of the end user. It is defined as user behaviour toward the Website with an emphasis on the user's necessity of direct interaction. Flavin et al. (2006) defined satisfaction as the user's emotional state toward the Website based on the evaluation of every characteristic constituting the customer relationship. In this study, satisfaction is defined as the overall user assessment of the usability of the website. Based on previous studies in which the Website service quality was reported to affect Website satisfaction, we propose the following hypotheses:

Health care service expectation and intention to visit health care provider

Oliver (1980) reported that expectation, as a basis for comparative assessment, is generally developed before purchasing goods and that expectation affects an individual's perception of quality. Parasuraman et al. (1988) defined expectation as the consumer desires or wants directed toward service providers. Dick and Basu (1994) reported that expectation reflects current services delivered by the service provider and prospective services demanded, based on consumer needs. Grönroos (1982, 1988, 1990) reported in his study using a perceived service quality model that four elements (market communication, image, word-of-mouth, and customer needs) affect expectation. Zeithaml et al. (1996) selected word-of-mouth, personal needs, past experiences, and marketing communication as factors affecting expectation. Schneider and Bowen (1985) asserted that service quality is determined by the degree of customer expectation, and expectation is, in turn, formed by the experiences of individual customers. They reported that customers retain intrinsic standards for quality, and these perceptions induce relative assessments of service quality. Parasuraman et al. (1985) discovered that customer expectation is developed based on a combination of standards rather than on one unitary standard. They further classified service expectation into two dimensions of i) expectation for desired services and ii) expectation for adequate services through their conceptual SERVQUAL model.

In this study, customer expectation is defined as the degree of expectation retained by customers accessing medical institution Websites for prospective services to be provided. We propose the following hypotheses for the relationship between the customer satisfied with health care information provided by the Website and customer expectation for health care services to be provided.

Kasper and Lemmink (1989) regard purchase intention as a preceding variable to purchase behaviour, and assume that it leads to actual purchase. Customer loyalty is interpreted as a type of intentional purchase behaviour and is often used with behavioural measures. Because measuring loyalty with purchase intention alone can be very difficult, Zins (2001) recommended using preference toward service provider as a supplementary tool, while Yu and Dean (2001) and Zeithaml et al. (1996) suggested using behavioural measurement tools, such as positive word-of-mouth, along with purchase intention. Gummerus et al. (2004) saw this suggestion as particularly valid because loyalty arises even among consumers accessing the Website intermittently. In this study, purchase intention is defined as the intention to visit a medical institution in person and to purchase services provided by the institution. Many of the previous studies focused on the relationship between customers satisfied with the Website and their intention to re-visit the site. This study extends the scope by measuring the satisfaction of customers and their intention to visit health care providers, and purchase services. Therefore, we established the following hypotheses.

Based on these hypotheses, we constructed a research model as illustrated in Figure 1.

Research methods

Questionnaire development

We reviewed all pertinent previous studies pertaining to a variety of Websites, and adopted four variables of usability, aesthetics, security, and responsiveness constituting the Website service quality as shown in Table 1.

For information acquisition, we reviewed the survey items developed Huizingh (2000), Misic and Johnson (1999) and Cho and Kang (2006), and created six revised items to fit the purpose of our study. As a measure of Website satisfaction, we developed three items based on the measurement tools for e-satisfaction proposed by Sahadev and Purani (2008). For health care expectation, we adopted four items among the measurement items developed by Oliver (1993). With regard to intention to visit, we developed three items based on the measurement tools suggested by Poddar et al. (2009). Each of the survey items was measured on a five-point Likert scale (see Appendix for more detail).

Data collection and analysis

We surveyed individuals between the ages of 20 and 30 who had substantial experience in using medical institution Websites and who were primarily students or office workers. The respondents were to answer based on their past experience in searching Websites of primary health care institutions (dental, dermatological, cosmetic clinics etc.) online. In order to enhance reliability, professional pollsters supervised the survey. The survey was conducted from October 1 to October 14, 2009 in Seoul, South Korea. Out of the 240 survey questionnaires distributed, 217 responses excluding 23 incomplete ones were ultimately used for our research. We used SPSS 18.0 and AMOS 18.0 for statistical data analysis.

Empirical analysis and hypothesis testing

Measurement model analysis

Before assessing the validity of the measurement model, we first investigated the internal consistency of each construct by calculating the Chronbach α. Reliability, in general, is thought to be secured when α is between 0.7 and 0.9 (Hair et al. 2006). The Chronbach α values of usability, aesthetics, security, and responsiveness were 0.88, 0.86, 0.87, and 0.86, respectively, thus complying with the criterion suggested by Hair et al. (2006). We then conducted a confirmatory factor analysis for each of the four Website service quality factors in order to assess the validity of our measurement model by using AMOS 18.0. We aimed to select relevant items with increased accuracy by performing a validity assessment on individual constructs prior to investigating the validity of the overall measurement model (Singh and Rhoads 1991). To determine the goodness of fit of the overall model, we used Goodness-of-Fit Index (GFI), Adjusted Goodness-of-Fit Index (AGFI), , p-value, Root Mean Square Residual (RMSR), and Normed-Fit-Index (NFI) (Hair et al. 2006). As a result of this process, usability item 6 and security item 2, which decreased the goodness of fit of the model, were eliminated. As a result of the confirmatory factor analysis, the goodness of fit elements in our study mostly satisfied the fit criteria of GFI ≥ 0.9, AGFI ≥ 0.9, p-value ≥ 0.05, RMSR ≤ 0.05, and NFI ≥ 0.9 suggested by Hair et al.. This is shown in Table 2.

Subsequently, we conducted a confirmatory factor analysis using a covariance matrix on the measurement model, including all four research constructs. As a result, the goodness of fit of the model mostly satisfied the aforementioned criteria suggested by Hair et al. (2006), as shown in Table 3. In addition, factor loadings of the measurement items of the four research constructs were all statistically significant, with p < 0.01.

In order to investigate the convergent validity, we calculated construct reliability (CR) and average variance extracted (AVE). As a result, CR values of usability, aesthetics, security, and responsiveness were 0.91, 0.91, 0.90, and 0.90, respectively, with AVE values of 0.57, 0.56, 0.68, and 0.63 for the constructs in the same order. The measurement items of each construct satisfied the universal standards of CR ≥ 0.7 and AVE ≥ 0.5, and thus the convergent validity was secured for the measurement items of each construct (Hair et al. 2006).

In reference to the discriminant validity, Fornell and Larcker (1981) suggests it exists for two particular constructs when the AVE values of each of these constructs are greater than the coefficient of determination (r? of the constructs. Table 4 shows the correlation coefficients among the eight research constructs. We calculated r?among the four extraneous variables, where the largest r?of 0.35, between usability and aesthetics, was smaller than the smallest AVE value of 0.56 for aesthetics. Therefore, we determined that the discriminant validity was secured in our study.

Hypothesis testing

Research hypotheses were tested through a structural equation model (SEM) using AMOS 18.0 as shown in Figure 2. As a result, the goodness of fit was satisfied with ײ = 272.5, d.f = 173, GFI = 0.893, AGFI = 0.857, NFI = 0.904, IFI = 0.963, CFI = 0.962, and RMSR = 0.038, (Hair et al., 2006). We reviewed the modified exponential and found it possible to improve the goodness of fit, to a certain degree, by correlating among particular error items. However, this had a potential risk of hindering the discriminant validity and unidimensionality of corresponding items, and we decided not to alter the model (cf. Hair et al. 2006: 782).

In terms of the relationship between e-SERVQUAL and information acquisition (H1, H2, H3, and H4), three of the four e-SERVQUAL variables, excluding aesthetics, were shown to significantly affect information acquisition. Among the three constructs, usability (0.33) had the most influence on information acquisition, followed by responsiveness (0.31), and security (0.14). In the relationship between e-SERVQUAL and Website satisfaction (H5, H6, H7, and H8), only usability and security were found to have a statistically significant effects, with usability (0.29) having the biggest effect on Website satisfaction. Moreover, the relations among the mediators of information acquisition, Website satisfaction, and health care expectation and the dependent variable of intention to visit, H9 through H14, were all statistically significant.

According to the decomposition of total effects, the total effects of usability and security among e-SERVQUAL toward Website satisfaction were 0.430 and 0.196, respectively, with direct effects of 0.294 and 0.140 in the same order as shown in Table 5. This implies that information acquisition was an important mediator. In addition, the direct effect of information acquisition on intention to visit was 0.216, while the total effect, with the mediator effect of Website satisfaction and health care expectation taken into account, was 0.504. Therefore, we determined that these two variables are important in the relationship between information acquisition online through Website and intention to visit.

Discussion

In this study, we aimed to discover the Website service quality factors affecting information acquisition online by customers, and to explore the influence of such information acquisition on customer intention to visit and purchase the services provided. While the majority of previous studies focused mainly on the relationship between the Website service quality factors and online customer behaviour, such as satisfaction and Website re-visit, we extended the existing spectrum by i) focusing on the needs of customers visiting Websites in terms of information acquisition, ii) confirming that the Website service quality factors affect Website satisfaction as in previous studies, iii) identifying the Website service quality factors facilitating information acquisition by consumers from a managerial perspective, and iv) studying the impact of information acquisition on intention to visit and purchase services.

The findings of this study provide the following implications. First, four service quality factors of usability, aesthetics, security, and responsiveness were extracted as a result of an exploratory factor analysis. From a total-effects perspective, usability, security, and responsiveness were important Website service quality factors for medical institutions, with aesthetics having relatively low importance. Discrepancies exist between the services provided by medical institutions and the services provided by online shopping malls in terms of service delivery. Health care, such as medical procedures and treatments, is directly provided by the health provider to the patient. In the general the retail service industry Websites, such as eBay and Auction, for which value is created online, aesthetic properties including font, colour combination, and image significantly affect the purchase intentions of customers; thus presentation of information is relatively important. Medical institution Websites, however, provide information mainly through text; thus aesthetic quality may be of less importance.

Second, responsiveness, one of the Website service quality factors, affected Website satisfaction, mediated by information acquisition. Health care information often includes technical terms that consumers may not understand. When consumers direct inquiries to Website administrators or managers regarding medical information, if responses are not made quickly, consumers will be less inclined to visit the Website because of their inability to acquire requested information. Therefore, the role of information acquisition, as a mediator, is extremely important in the relationship between responsiveness and Website satisfaction.

Third, as a result of examining the effect of the Website service quality factors on information acquisition and Website satisfaction, the service quality factors (except aesthetics) extracted through factor analysis affected Website satisfaction and intention to visit both directly and indirectly, with information acquisition as a mediator. In addition, the relationship between information acquisition by the consumer and intention to visit was statistically significant. Websites help consumers acquire information and enable service providers to secure potential customers. Health care providers operate Websites not only to provide customers with information, but also to help achieve effective medical institution marketing. This was confirmed through the decomposition of total effects (as shown in Table 5) because information acquisition affected intention to visit both directly and indirectly. This implies that health care providers can increase potential demand and create economic value through the customers who acquire information from their Websites.

Conclusion

This study indicates that the most important quality for improving information acquisition is usability of the Website. In other words, making the Website easy to use by improving its speed, organization, and browsing capability will help customers acquire information, and this, in turn, will lead to improved Website satisfaction and health care expectation. Furthermore, customers will be able to acquire information through the Website with ease when responsiveness and online security are assured. In conclusion, in order to improve information acquisition and reinforce intention to visit, special attention must be paid to usability, responsiveness, and security, which are Website service quality factors.

Limitations of this study are that the survey was conducted in Seoul, South Korea, a metropolis with high population density and medical institution concentration. Thus the findings of this study may not be applicable to all medical institutions. Only one age group was surveyed in the study, which may limit generalizability of our results to the overall population of Internet users. Lastly, we did not subdivide customer purpose of visiting medical institution Websites. Therefore, further research to resolve such matters is strongly encouraged. The future researcher should develop our proposed model by including the level of medical service quality provided by medical institutions.

About the authors

Dr. Kim, Sang-Man is a professor of school of management and also serves as the department chair of Medical MBA at Kyung Hee University, Seoul, South Korea. He received his PhD in management from the University of Nebraska-Lincoln, USA. His research has spanned a number of areas, including healthcare information management and information behaviour. He can be contacted at smkim@khu.ac.kr

Oh, Jae-Young is a Ph.D. student of Graduate School of Business Administration at Kyung Hee University, Seoul, South Korea. He received his Masters in Business Administration at Kyung Hee University, Seoul, South Korea. He can be contacted at jyoh@khu.ac.kr