The participants were 6,086 patients including 2,951 outpatients and 3,135 inpatients over the age of 18 who used regional hub public hospitals (regional medical center: n =34; Red Cross Hospital: n =5) across the country. Table 1 shows the participants’ demographic characteristics. For the outpatients, there were more male patients (n = 1,705; 57.8%) than female patients (n = 1,246; 42.2%). The most common age group was the 60s (n = 812; 27.5%), followed by over 70 years (n = 761; 25.8%). The most common education level was high school graduation (n = 1,094; 37.1%), followed by under middle school graduation (n = 968; 32.8%) and over college graduation (n = 889; 30.1%). The most common health status perceived by patients was “moderate” (n = 1,091; 37%), followed by “relatively good” (n = 1,024; 34.7%).
For the inpatients, there were more male patients (n = 1,686; 53.8%) than female patients (n = 1,449; 46.2%). The most common age group was those over 70 (n = 1,130; 36.1%) and was followed by those in their 60s (n = 822; 26.2%). The most common education level was under middle school graduation (n = 1,469; 46.9%), followed by high school graduation (n = 1,005; 32.1%) and over college graduation (n = 661; 21.0%). The most common health status perceived by patients was “moderate” (n = 950; 30.3%), followed by “very bad” (n = 265; 8.5%).
Table 1
General Characteristics of Participants
|
Outpatient
|
Inpatient
|
N
|
%
|
N
|
%
|
Gender
|
Male
|
1,705
|
57.8
|
1,686
|
53.8
|
Female
|
1,246
|
42.2
|
1,449
|
46.2
|
Age(yr)
|
18~29
|
204
|
6.9
|
144
|
4.6
|
30~39
|
236
|
8.0
|
153
|
4.9
|
40~49
|
316
|
10.7
|
271
|
8.6
|
50~59
|
622
|
21.1
|
615
|
19.6
|
60~69
|
812
|
27.5
|
822
|
26.2
|
≥70
|
761
|
25.8
|
1,130
|
36.1
|
Education
|
≤Middle school
|
968
|
32.8
|
1,469
|
46.9
|
High school
|
1,094
|
37.1
|
1,005
|
32.1
|
≥College
|
889
|
30.1
|
661
|
21.0
|
Health State
|
Very good
|
284
|
9.6
|
317
|
10.1
|
Good
|
1,024
|
34.7
|
895
|
28.5
|
Moderate
|
1,091
|
37.0
|
950
|
30.3
|
Poor
|
436
|
14.8
|
708
|
22.6
|
Very poor
|
116
|
3.9
|
265
|
8.5
|
Total
|
2,951
|
100
|
3,135
|
100
|
2. Goodness-of-Fit Test of Research Model
The association analysis among major variables showed that the association among all latent variables were significant with P<0.01. The multicollinearity problem was not found because the Variance Inflation Factor was under 10; it was under four for all the measured variables in the outpatient group, and under six for all the measured variables in the inpatient group.
The convergence validity and discriminant validity were tested to determine whether the target concept or attribute was measured. The average variance extraction (AVE) and the concept of construct reliability (CR) were used to test validities. The validities were supported because AVE values were 0.5 or higher and CR values were 0.7 or higher in both the outpatient and inpatient groups.
The results of testing the overall structural model used in this study were as follows: For the outpatient group, χ2=1303.176, TLI=0.948, CFI=0.961, and RMSEA=0.053. For the inpatient group, χ2=3581.292, TLI=0.939, CFI=0.949, and RMSEA=0.055. These results indicate that the indices were within the recommended level and thus suggesting that the model is suitable.
3. Structural Model Analysis
Fig. 1 illustrates the structural model analysis results for the outpatient group. It was found that physician's practice service had a positive (+) significant effect on patient satisfaction (β=0.377, p<.001) and intent to revisit (β=0.243, p<.001). Medical staff's kindness and consideration had a positive (+) significant effect on patient satisfaction (β=0.303, p<.001) but not on intent to revisit (β=0.036, p=.286). Physical environment of the hospital had a positive (+) significant effect on patient satisfaction (β=0.186, p<.001) and intent to revisit (β=0.049, p=.046). Patient satisfaction had a positive (+) significant effect on intent to revisit (β=0.652, p<.001).
That is, the better the physician's and nurse's practice service, medical staff's kindness and consideration, and physical environment of the hospital, the more likely they are to increase patient satisfaction. It was also found that the better physician's practice service and physical environment of the hospital were likely to increase patient satisfaction and intent to revisit.
Fig. 2 displays the structural model analysis results for the inpatient group. It was found that physician's practice service had a positive (+) significant effect on patient satisfaction (β=0.223, p<.001) and intent to revisit (β=0.191, p<.001). Nurses' practice service had a positive (+) significant effect on patient satisfaction (β=0.236, p<.001) but not on intent to revisit (β=0.04, p=.089). Medical staff's kindness and consideration had a positive (+) significant effect on patient satisfaction (β=0.29, p<.001) and intent to revisit (β=0.096, p=.014). Physical environment of the hospital had a positive (+) significant effect on patient satisfaction (β=0.182, p<.001), but not on intent to revisit (β=-0.013, p=.51). Patient satisfaction had a positive (+) significant effect on intent to revisit (β=0.498, p<.001).
Thus, the better physician's and nurse's practice service, staff's kindness and consideration, and physical environment of the hospital, the more likely they are to increase patient satisfaction. It was also found that the better physician's practice service and the staff's kindness and consideration were likely to increase patient satisfaction and intent to revisit.
4. Analysis of the Mediating Effect
This study used covariance structural analysis to confirm the overall influence among factors as a causal effect. The total, direct, and indirect effects were measured to determine the effect of the independent variable “quality of healthcare” on the dependent variable “intent to revisit,” which revealed the importance of the mediating variable of “patient satisfaction.” The total effect was expressed as the sum of all direct and indirect effects of the independent variable on the dependent variable. The direct effect reflected the direct relationship between the independent and dependent variables; the indirect effect represented the effect of the independent variable on the dependent variable through a mediating variable by regression analysis.
Table 2 shows the analysis results of the total, direct, and indirect effects in the outpatient group. The values of the standardized direct effect on “patient satisfaction” were 0.377 for “physician’s practice service,” 0.303 for “medical staff's kindness and consideration,” and 0.186 for “physical environment of hospital,” which were statistically significant. The values of the standardized direct effect on “intent to revisit” were 0.243 for “physician’s practice service,” 0.036 for “medical staff's kindness and consideration,” 0.049 for “physical environment of hospital,” and 0.522 for “patient satisfaction” where the values for “physician's practice service,” “physical environment of hospital,” and “patient satisfaction” were statistically significant. The values of the standardized indirect effect of the independent variable on the dependent variable through the mediating variable “patient satisfaction” were 0.197 for “physician's practice service,” 0.158 for “medical staff's kindness and consideration,” and 0.097 for “physical environment of hospital,” which were statistically significant. “Patient satisfaction” was found to mediate the relationship between quality of healthcare perceived by patient (“physician's practice service,” “medical staff's kindness and consideration,” and “physical environment of hospital”) and intent to revisit.
Table 2
Effect Analysis (Outpatient)
Path
|
Direct effect
|
Indirect effect
|
Total effect
|
Doctor's medical services
|
Patient satisfaction
|
0.377**
|
-
|
0.377**
|
Intent to revisit
|
0.243**
|
0.197**
|
0.440**
|
Staff kindness and consideration
|
Patient satisfaction
|
0.303**
|
-
|
0.303**
|
Intent to revisit
|
0.036
|
0.158**
|
0.195**
|
Physical environment
of hospital
|
Patient satisfaction
|
0.186**
|
-
|
0.186**
|
Intent to revisit
|
0.049*
|
0.097**
|
0.146**
|
Patient satisfaction
|
Intent to revisit
|
0.522**
|
-
|
0.522**
|
Bootstrap standardized direct, indirect, total effect
*p<.05, **p<.01, ***p<.001
|
The direct effect on “intent to revisit” was highest for “patient satisfaction,” followed by “physician's practice service,” “physical environment of hospital,” and “medical staff's kindness and consideration.” The indirect effect on “patient satisfaction” through “intent to revisit” was highest for “physician’s practice service,” followed by “medical staff's kindness and consideration” and “physical environment of hospital.” The total effect was highest for “physician’s practice service,” followed by “medical staff's kindness and consideration,” and “physical environment of hospital.” These results indicated that “physician’s practice service” is the most significant factor influencing “patient satisfaction” and “intent to revisit.”
Table 3 shows the analysis results of total, direct, and indirect effects in the inpatient group. The values of the standardized direct effect on “patient satisfaction” were 0.223 for “physician’s practice service,” 0.236 for “nurse’s practice service,” 0.290 for “medical staff's kindness and consideration,” and 0.182 for “physical environment of hospital,” which were statistically significant. The values of the standardized direct effect on “intent to revisit” were 0.191 for “physician’s practice service,” 0.040 for “nurse’s practice service,” 0.096 for “medical staff's kindness and consideration,” -0.013 for “physical environment of hospital,” and 0.498 for “patient satisfaction” where the values for “physician's practice service,” “medical staff's kindness and consideration,” and “patient satisfaction” were statistically significant. The values of the standardized indirect effect of the independent variable on the dependent variable through the mediating variable of “patient satisfaction” were 0.111 for “physician's practice service,” 0.117 for “nurse’s practice service,” 0.145 for “medical staff's kindness and consideration,” and 0.090 for “physical environment of hospital,” which were statistically significant. “Patient satisfaction” was found to mediate the relationship between quality of healthcare perceived by the patient (“physician's practice service,” “medical staff's kindness and consideration,” and “physical environment of hospital”) and intent to revisit.
The direct effect on “intent to revisit” was highest for “patient satisfaction” and was followed by “physician's practice service,” “medical staff's kindness and consideration,” “nurse's practice service,” and “physical environment of hospital.” The indirect effect on “patient satisfaction” through “intent to revisit” was highest for “medical staff's kindness and consideration,” followed by “nurse's practice service,” “physician’s practice service,” and “physical environment of hospital.” The total effect was highest for “physician’s practice service,” followed by “medical staff's kindness and consideration,” “nurse's practice service,” and “physical environment of hospital.”
Table 3
Effect Analysis (Inpatient)
Path
|
Direct effect
|
Indirect effect
|
Total effect
|
Doctor's medical services
|
Patient satisfaction
|
0.223**
|
-
|
0.223**
|
Intent to revisit
|
0.191**
|
0.111**
|
0.302**
|
Nurse's medical services
|
Patient satisfaction
|
0.236**
|
-
|
0.236**
|
Intent to revisit
|
0.040
|
0.117**
|
0.158**
|
Staff kindness and consideration
|
Patient satisfaction
|
0.290**
|
-
|
0.290**
|
Intent to revisit
|
0.096*
|
0.145**
|
0.240**
|
Physical environment
of hospital
|
Patient satisfaction
|
0.182**
|
-
|
0.182**
|
Intent to revisit
|
-0.013
|
0.090**
|
0.077*
|
Patient satisfaction
|
Intent to revisit
|
0.498**
|
-
|
0.498**
|
Bootstrap standardized direct, indirect, total effect
*p<.05, **p<.01, ***p<.001
|