The Effect of Satisfaction, Perceived Value, Image, and Perceived Sacrifice on Public Healthcare Service Institution’s Patient Loyalty

This research aims to investigate the simultaneous effect of satisfaction, perceived value, image, and perceived sacrifice on patient loyalty. This study is believed to be the first to develop and test patient loyalty model that includes satisfaction, perceived value, image, and perceived sacrifice. Quantitative research methodology was employed. Further, the research design of this study is cross sectional study. We performed survey to collect the empirical data. Convenience sampling technique was used to decide the resepondents of the survey. The respondents are 162 patients of two health care service institutions in Bogor and Bekasi, Indonesia. The validity of the questionnaire was tested by using factor analysis whi le the reliability was tested by using Cronbach alpha analysis. Multiple regressions analysis was performed to analyse the data. The findings showed that image has positive impact on patient loyalty. However, this research also found that satisfaction, perceived value, and perceived sacrifice do not have significant impact on patient loyalty. Thus, the management of public healthcare service institution should consider and manage the image of the institution proactively.


INTRODUCTION
Recently, patient loyalty has been becoming an important issue in healthcare service [1]. This can be understood since the competition among healthcare service institutions tends to be more competitive [1][2][3]. Furthermore, the rapid growth of internet technology makes the patient easily increases their healthcare service standard and/or spreads negative/positive word of mouth communication [1][2][3]. At that context, patient loyalty is identified to be a source of competitive advantage [1], [4][5].
Patient loyalty is important for both private healthcare service institution and public healthcare service institution [2][3]. Public healthcare service institutions will obtain some benefits if they focus on patient loyalty [6]. For example, public healthcare service institution with loyal patient can increase significantly the number of its patient [2], [3][4][5][6]. This can be used as the justification of the public budget they have used [7][8]. Furthermore, as a brand zealot, the loyal patient will assist the public healthcare service institution to overcome negative word of mouth communication from the other patients or parties [9]. Hence, research on public healthcare service institution"s patient loyalty becomes necessary to be carried out.
Some researchers have studied patient loyalty [e.g. 1, [10][11][12][13][14][15][16][17][18][19][20]. However, there are only few studies involving public healthcare service context. For example, Moliner studied patient loyalty using a model that includes satisfaction, trust, and perceived functional value [3]. The respondents of the study are 171 users of a private hospital and 170 users of a public hospital in Spain. He found that loyalty is influenced by satisfaction and trust while satisfaction and trust is affected by perceived functional value. Other researchers, Suki and Suki also investigated patient loyalty for both public and private healthcare service institution [21]. They surveyed 200 patients of the government hospital and clinics, as well as private clinics in The Federal Territory of Labuan, Malaysia. Their research results showed that (1) loyalty is influenced by commitment, (2) commitment is affected by trust, (3) trust is influenced by satisfaction and doctor reputation, and (4) satisfaction is affected by doctor reputation. Amin and Nasharuddin surveyed 216 patients of public and private hospitals in Malaysia [2]. They found that service quality affects satisfaction which subsequently affects behavioral intention. Patawayati et al. investigated 150 patients who have used healthcare services in the public Hospital of Southeast Sulawesi, Indonesia [6]. Their study showed that service quality affects satisfaction which then affects trust and commitment. In addition, Patawayati et al. also found that loyalty is influenced by trust and commitment. However, they found that satisfaction doesn"t affect loyalty even though satisfaction influences trust and commitment.
As previously explained, there are only few researches on patient loyalty involving public healthcare service context. Furthermore, we identified some research gaps on the topic. The previous studies didn"t include perceived sacrifice, which refers to "what is given up or sacrificed to obtain a [healthcare service]", as the antecedent of patient loyalty [22]. Meanwhile, in the context of public healthcare service, perceived sacrifice is important to be considered. This is because the public healthcare service institution is supported by public fund. The patient of public healthcare institution may be more aware of their sacrifice. The high awareness of the customers on their sacrifice may cause the direct effect of perceived sacrifice on patient loyalty [23]. Furthermore, in the other public services context, empirical studies have showed that perceived sacrifice influences loyalty [e.g. [24][25]. There is lack of previous studies that include image as the predictor of patient loyalty [26]. This is quite surprising since healthcare service is a credence service [1], [3], [26], [27]. On the other hand, in the context of credence service, image may influence a consumer"s behaviors, which include his/her post consumption behaviors [26][27][28]. Furthermore, it is also well documented that image influences customer loyalty positively [e.g. 26,[29][30][31][32]. Thus, it is important to include image in order to investigate patient loyalty.
To address the gap in the literature, this research aims to investigate patient loyalty using a model that integrates two variables that were not involved in the previous studies (image and perceived sacrifice) and two variables that were already used in the previous researches (satisfaction and perceived value). Satisfaction and perceived value are included in the model since both variables is widely known as the predictor of loyalty in healthcare service [3], [10], [1] and the other public services [e.g. [33][34].
More specifically, this research aims to answer four research questions. First, does image influence patient loyalty? Second, does perceived sacrifice influence patient loyalty? Third, does satisfaction influence patient loyalty? Fourth, does perceived value influence patient loyalty?
The reminder of this paper is organized as follows. The first section reviews the theoretical background and the proposed hypotheses. The second section presents the research methodology. The third section describes the results of the research. The fourth section presents the theoretical and managerial implications. The final section is the conclusion of this paper.

RESEARCH METHOD 2.1. Research design
This research used quantitative research methodology. Further, the research design of this study is a cross-sectional study. A survey with questionnaire was performed to collect the empirical data. The research includes five main variables, namely patient loyalty, satisfaction, perceived value, image, and perceived sacrifice. We measured the variables using multiple items since these variables can be categorized as latent variable with high degree of abstraction [35]. The operational definition of the variables and the number of the indicators can be seen in Table 1. Hu et al. [37] Perceived Value The difference or discrepancy between the benefits patient obtain from healthcare service provider with the sacrifices he/she has to perform that is perceived by him/her [22, 38 -41]. 3 Choi et al. [10] Image Patient"s overall (global) impression on the healthcare service provider [42] 3 Andreassen and Lindestad [43] Perceived Sacrifice "what is given up or sacrificed to obtain a [healthcare service]" [22] 3 Cronin et al. [44] The main variables were measured by suing a seven-point Likert scale. We used positive sentences to express the all indicators. Related to this matter, it should be noted that the use of positive sentences in expressing the indicators of perceived sacrifice caused the relationship between perceived sacrifice and patient loyalty must be interpreted as a positive relationship if the result of the statistical testing shows that the coefficient of the relationship is negative and vice versa.

Sampling technique
The intended population of this study is the patients of healthcare service institutions in Bogor and Bekasi, Indonesia. Due to the unknown true characteristics of the population of the patients of healthcare service institutions and budget and operational limitation, we employed convenience sampling technique [45]. However, it should be noted that since this research was designed to test the relationships among variables, the use of convenience sampling technique in this research could still be tolerated [46].
The data collection was carried out through survey. The survey was performed in two healthcare service institutions in Bogor and Bekasi, Indonesia. The respondents of the survey are the patients of the healthcare service institution. The participation of the patients in the survey is completely voluntary.
The sample of this research is 162 patients. The sample size fulfills the sample size required by the statistical analysis tool employed in this study (multiple regressions analysis) [47]. Table 2 shows the respondents" demographic profile.   Table 3, it can be concluded that all variables fulfill the construct validity criteria. Say positive things about the healthcare services to others people 0.881 b. Recommend the healthcare service provider to someone who seeks your advice 0.891 c.
Reuse the healthcare service provider if the service is needed 0.838 d. Consider the healthcare service provider as the first choice to travel 0.838 Cronbach Alpha analysis was performed to examine the reliability of the questionare. Table 4 presents the Alpha coefficient of each variable. In the literature, the cut off value of the Alpha coefficient that is generally used is 0.6 [34], [45], [47], [48]. Based on Table 4, all variables have alpha value above 0.6

Main analysis
In order to achieve the main objective of this research, this research performed multiple regressions analysis. The method was applied because this research examines the relationships between two or more independent metric variables and one dependent metric variable. Meanwhile, literature has revealed that multiple regression analysis is appropriate to perform that task [47]. The regression model is shown in Figure 1. The validity and reliability tests and the multiple regression analysis were conducted with the support of SPSS 16. This research is a cross sectional study. Furthermore, it can also be categorized as field study. Different with pure experimental study, the nature of field study is more focuss on the external validity of the research than the internal validity [45]. Thus, the existence of confounding variable may happen in field study. In order to reduce the effect of counfounding variable, we only involved the antecedent variable of regression model that are supported by strong theoretical background. Table 5 shows the descriptive statistics of the constructs of the research. All constructs have high mean value. The construct with highest mean value is perceived sacrifice (5.70). Furthermore, image has the lower mean value (5.16).

Multiple regression analysis result
The results of the multiple regressions analysis are shown in Table 6. The F statistic value is 16.396 with p-value (p= 0.000) lower than 0.05. This means that the regression model can be used to predict patients" loyalty. The first finding of this research showed that the unstandardized B coefficient of satisfaction is positive (B = 0.076) and the p value is higher than 0.05 (p value = 0.265). This means satisfaction doesn"t affect patient loyalty significantly. This finding doesn"t support the finding of the previous researches carried out by Moliner [3] and Amin and Nasharuddin [2].
The second finding of this research showed that perceived value doesn"t affect patient loyalty significantly (B = -0.095, p value = 0.295). This finding doesn"t support the finding of the previous researches carried out by Choi et al. [10].
The third finding of this research showed that image affects patient loyalty positively and significantly (B = 0.254, p value = 0.001). This finding supports the finding of the previous researches which tested the impact of image on loyalty [e.g. [29][30][31].
The fourth finding of this research showed that perceived sacrifice doesn"t affect patient loyalty significantly (B = 0.097, p value = 0.177). This finding doesn"t support the previous researches on perceived sacrificeloyalty, which are carried out in the other contexts than healthcare service [e.g. 24-25].

Theoretical implication
Recently, patient loyalty has been identified to be an important factor for public healthcare service institutions [6]. Thus, the managements of public healthcare service institutions need to understand the factors that influence patient loyalty. Some researchers have investigated patient loyalty of public healthcare service institution. However, there is no study that investigates patient loyalty using a model that involves satisfaction, perceived value, image, and perceived sacrifice. Given this, the paper fulfills the gap in the The results of this research showed that patient loyalty is only influenced by image while satisfaction, perceived value, and perceived sacrifice do not affect patient loyalty. The positive impact of image on loyalty supports the previous researches [e.g. 26,[29][30][31]49]. Meanwhile, the non-significant impact of satisfaction, perceived value, and perceived sacrifice on patient loyalty may due to the characteristics of healthcare services. Healthcare services can be categorized as credence service so that the service performance is difficult to be evaluated by the patients [27]. Meanwhile, the three constructs relate to the patient"s evaluation of the service [22], [39]. Based on information economics signaling theory, under credence service consumption, a customer tends to use credible signal provided by the producer [25][26]. On other hand, image is a type of a signal that may increase customer expected utility value [25][26].
Another explanation may come from consumer behavior literature. The consumer behavior literature has revealed that consumer tends to use heuristic way, such as relies only on service provider"s image, in their purchasing decision regarding credence services consumption [28]. This may make the patient doesn"t evaluate the overall performance of the healthcare service as well as the value of the services and the sacrifice they have to perform during their consumption.
Furthermore, the non-significance effect of satisfaction, perceived value, and perceived sacrifice may also be explained from the perspective of customer needs. The patient may only emphasize on the outcome dimension of the healthcare service, which is healthier after they obtain the service. This may make the patient doesn"t seriously consider the emotional atmosphere of the healthcare service, that is the root of satisfaction, and the sacrifice they have to perform as well as the value of the service they obtain.

Managerial implication
The results of this research give managerial implications for public healthcare managers in establishing patient loyalty. The findings show that public healthcare service institution should manage its image effectively. Furthermore, the managements of public healthcare service institutions should emphasize on marketing and operational strategy that can improve the image of the institution.

CONCLUSION
This research has tried to test the influence of satisfaction, perceived value, image, and perceived sacrifice on patient loyalty. This is important because there is limited literature on the topic. Based on the data analysis, this research found that image influences patient loyalty positively. However, this research also found that satisfaction, perceived value, and perceived sacrifice do not influence patient loyalty significantly.
Even though this research has generated interesting findings, we addressed some limitations. First, we employed convenience sampling technique. In addition, we performed the data collection only in two healthcare institutions in Bogor and Bekasi, Indonesia. Thus, the research results may not be generalized into other contexts. Second, this research only included four variables (satisfaction, perceived value, image, and perceived sacrifice) as the predictors of patient loyalty. The R 2 is 29.5%. Hence, we predict there are still other antecedent variables of patient loyalties. Based on the limitations, we suggest including others variables in the research model and involving more healthcare service institution to test the model in the future research.