Association Between Varicella-Zoster Virus Vaccination and Patient Experience in Elderly Japanese Outpatients: A Case-Control Study

Abstract

Background:

This study examined the association between patient experience (PX, events experienced by patients during primary care that are an indicator of patient-centered quality) of primary care and varicella-zoster virus (VZV) vaccine uptake in older adults.

Methods:

A case-control study of VZV vaccination was conducted at a community hospital in Ibaraki, Japan. Patients aged 65 years or older who had continuously been patients of the hospital between April 2018 and April 2021 were included in the study. The vaccinated group consisted of 166 VZV-vaccinated patients. The controls consisted of 29 age- and sex-matched patients who did not receive VZV vaccination. A self-administered questionnaire was distributed between August and September 2021. It included the Japanese version of the Primary Care Assessment Tool Short Form (JPCAT-SF) to evaluate PX and included questions about recommendations for VZV vaccination by a physician and the vaccination history of relatives. Multivariable and intermediate factor analyses were used to assess whether there was an association between VZV vaccination and PX.

Results:

Questionnaires were sent to 457 subjects. Responses from 228 (116 in the vaccination group and 112 in the non-vaccinated group) were included in the analysis. Multivariable analysis, which excluded physician recommendation for VZV vaccination as a variable because it was an intermediate factor in the analysis, showed an association between PX and VZV vaccination (odds ratio, 1.38; 95% confidence interval, 1.00-1.92; P = .049).

Conclusions:

PX was associated with past VZV vaccination. Physician recommendation for VZV vaccination was an intermediate factor between PX and VZV vaccination.

Keywords

case-control study patient experience varicella-zoster virus vaccine geriatrics primary care prevention

Introduction

Herpes zoster is a disease in which reactivation of the varicella-zoster virus (VZV) causes the appearance of painful blisters clustered in the innervated area. The pain might persist for years as postherpetic neuralgia.¹ The number of people affected by herpes zoster was increasing in Japan, which has an aging population.² Approximately 1 in 3 people are estimated to be affected by herpes zoster by age 80 years.² Herpes zoster in the elderly has been shown to significantly reduce activities of daily living and quality of life³; its impact is expected to become even greater in the future. In Japan, it became possible to administer live attenuated VZV vaccine in 2016, when it was approved for the purpose of preventing herpes zoster, approximately 10 years behind Europe and the United States. A prior study revealed that vaccination was effective in reducing the incidence of herpes zoster by 51.3%.⁴ We do not know the number of Japanese residents undergoing VZV vaccination, but it may be small. This can be primarily attributed to the voluntary nature of VZV vaccination and the absence of an official program. Consequently, the precise number of vaccinations administered in Japan has not been made publicly available.

In Japan, inactivated recombinant zoster vaccine was approved in 2020, but distribution had been slow, with little available as of 2022.

Several studies on factors related to influenza and pneumococcal vaccination in Japan and the United States have suggested associations between patient experience (PX, events experienced by patients during primary care) and subjective health perspectives.^5
-7 However, few studies have examined vaccination factors for the VZV vaccine. PX is a quality indicator of patient-centeredness. This indicator was increasingly used to assess the quality of primary care because patient-centeredness was a crucial concept, especially in the primary care setting.⁸ PX affects health outcomes through patient behaviors such as adherence to treatment and health care resource use.^9,10

If the association between PX and VZV vaccination can be clarified, the interventions needed to improve vaccination coverage can be identified, which might lead to higher VZV vaccine coverage. Therefore, this study evaluated whether PX is associated with VZV vaccination. Prior studies had shown that physician recommendation is strongly associated with receipt of other vaccinations,⁵ although it is more likely to be an intermediate factor in actual clinical practice. Therefore, this study evaluated whether PX is associated with VZV vaccination and whether physician recommendation could serve as an intermediate factor.

Methods

Study Design and Study Settings

This a case-control study of VZV vaccination involved a self-administered questionnaire. The study site was a 30-bed community hospital in X city, Ibaraki prefecture, Japan. At this hospital, the VZV vaccine has been widely promoted to outpatients since 2017 through digital signage, advertising posters, and pamphlets in the hospital. VZV vaccination is actively encouraged during outpatient consultations.

X city is a small city with a population of approximately 72 000 as of 2023. The percentage of the population aged ≥65 years was 33.0%. The city is making efforts to widely publicize the VZV vaccine to its citizens by including articles on the VZV vaccine in its city newsletter. As of 2023, VZV vaccination was self-paid in X city, as in most municipalities in Japan.

The inactivated recombinant zoster vaccine was not used at the study site or other medical facilities in the area. Only the attenuated Oka VZV vaccine was used.

Participants

We first selected study participants for the vaccination group. Among all patients seen at the study site between April 2018 and April 2021, 297 who were vaccinated against VZV were chosen. VZV vaccination status was ascertained from electronic medical records. From the patients chosen, those who had already died (n = 21), were unable to respond to the questionnaire (n = 9), had not visited the hospital regularly for more than 6 months (n = 75), or aged less 65 years (n = 26) were excluded; 166 were selected as the vaccination group. Those who were unable to answer the questionnaire included patients who were listed as “unable to make decisions” or “unable to communicate” in the core symptoms of dementia column of the long-term care insurance attending physician opinion form¹¹ obtained from electronic medical records and municipal databases, as well as those who were unable to communicate. Information on death was obtained from electronic medical records.

The control (non-vaccinated) group was selected by hospital staff not involved in the study using a random number table. For the control group, we selected 2 persons of the same sex and age as each participant in the vaccination group to whom the questionnaire was sent. Matched persons who had already died, were unable to respond to the questionnaire, or unable to be vaccinated due to immunodeficiency (eg, use of immunosuppressive drugs) were excluded. A replacement was selected using a random number table. Information on immunodeficiency was obtained from electronic medical records.

If only 1 person was an appropriate match, only 1 person was selected. Since the non-vaccinated group was expected to have lower PX than the vaccination group, the questionnaire response rate was also expected to be lower. Therefore, the ratio of participants in the vaccinated group to the non-vaccinated group was set at 1:2. Two hundred ninety-one controls were selected to form the non-vaccinated group. A flowchart of the research participant selection process is shown in Figure 1.

Figure 1.

Flowchart of the research participant selection process.

Self-administered questionnaires were sent by mail to a total of 457 participants in the vaccination and non-vaccinated groups between August and September 2021. The responses were returned by mail.

Exclusion criteria consisted of the following: inability to answer the self-administered questionnaire, no consent for study participation, and herpes zoster within 5 years as self-reported in the questionnaire. Patients who had herpes zoster within 5 years were excluded because they had been shown to have high shingles antibody titers and vaccination was not recommended for them at the time of the study.¹²

Measurements

This study examined the association between PX and VZV vaccination. To measure PX, the Japanese version of the Primary Care Assessment Tool Short Form (JPCAT-SF) was used.

The Primary Care Assessment Tool (PCAT) is a widely used measure of PX developed by Drph et al¹³ at the Johns Hopkins Primary Care Policy Center. Based on the PCAT, the JPCAT was translated into Japanese and modified to fit the Japanese context.¹⁴ The JPCAT-SF was developed as a shortened version.¹⁵ Validation studies have confirmed that it has higher validity and reliability than the PCAT.¹³ The JPCAT-SF has been validated to have the same psychometric properties as the JPCAT.¹⁴ This 13-item tool comprises 6 multi-item subscales addressing the following primary care attributes: first contact, longitudinality, coordination, comprehensiveness (services available), comprehensiveness (services provided), and community orientation. The JPCAT-SF scoring system converts each response on a 5-point Likert scale (1 = strongly disagree, 2 = somewhat disagree, 3 = not sure, 4 = somewhat agree, and 5 = strongly agree) into an item score from 0 to 4. The average of the item scores is calculated for each subscale and multiplied by 25 to obtain a subscale score from 0 to 100. The total JPCAT-SF score is the average of the 6 subscale scores. It provides an overall assessment of primary care performance, with higher scores indicating better performance.

Items in the JPCAT-SF, educational background(5 choices: less than high school, high school, vocational school, junior college, and university or graduate school), annual household income(4 choices: <2.50, 2.50–4.99, 5.00–7.99, and 8.00 million JPY ), self-assessment of health status (4 items), and subjective well-being (abbreviated Japanese version of the 11-item¹⁶ revised Philadelphia Geriatric Center Morale Scale¹⁷) were included in the self-administered questionnaire. In addition, whether and how the subject was aware of the VZV vaccine, whether and when the subject had herpes zoster, whether or not friends or relatives had herpes zoster, whether or not friends or relatives had been vaccinated against VZV, and whether or not vaccination was recommended by a doctor, friends, or relatives were investigated with the self-administered questionnaire.

Sex, age, type of health insurance (latter-stage elderly healthcare insurance, national health insurance, Japan Health Insurance Association, or other), municipality of residence, pneumococcal and influenza vaccination status, and underlying diseases (cardiac, respiratory, renal, malignant tumor, or diabetes) were abstracted from electronic medical records as of April 2021. The classification of underlying diseases followed previous studies⁵ that investigated factors associated with influenza vaccination.

Educational background, annual household income, type of health insurance, and self-assessment of health status were investigated because they are known to be associated with PX^18,19 and adult influenza and pneumonia vaccination.⁶

Statistical Methods

We compared the vaccination and non-vaccinated groups. Descriptive statistics of the measured variables were performed. To examine factors related to VZV vaccination, we also examined differences between the vaccination and non-vaccinated groups using the t-test for continuous variables and the χ² test for categorical variables in univariate analyses. Multivariable logistic regression with the forced entry method was then performed using sex, age, variables with a significant association in the univariate analysis (P < .10), and variables examined in the directed acyclic graph (DAG) based on background knowledge. In a DAG, an arrow connecting 2 variables indicates a causal relationship. Variables without a direct causal relationship are not connected. We illustrated this approach to explore variables associated with VZV vaccination status in a case-control study. The DAG (Figure 2) was created using the DAGitty program.²⁰

Figure 2.

DAG for exploring variables associated with VZV vaccination status.

Since there are no previous studies on factors contributing to VZV vaccination, we considered a small number of variables with high predictive ability based on background knowledge, referring to previous studies on other types of vaccination.^5,6 We selected VZV vaccination history of family members, VZV vaccination history of friends, family member recommendation for VZV vaccination, and friend recommendation for VZV vaccination.

Next, an intermediate factor analysis²¹ was conducted. Figure 3 illustrates the flow of the intermediate factor analysis. This analysis examined the intermediate factors that intervene between PX and VZV vaccination. In addition to having no significant associations between PX and VZV vaccination in the multivariable analysis that included intermediate factors, the analysis needed to meet the following criteria: (1) a multivariable analysis with the intermediate factor as the dependent variable that shows an association with PX, (2) a multivariable analysis with VZV vaccination as the dependent variable that shows an association with the intermediate factor, and (3) a multivariable analysis that excludes the intermediate factor as an independent variable that indicates an association between PX and VZV vaccination.

Figure 3.

Flow of the intermediate factor analysis.

Sample size was estimated using the rule of adherence to 10 events per variable in multivariable analysis.²² We assumed 10 independent variables to be entered into the multivariable analysis; a sample size of 200 (100 cases and 100 controls) was calculated to be sufficient to fit a logistic regression model.

Missing values were addressed by performing an available-case analysis. In each analysis, we included all study participants with data available for the variable of interest.²³

Statistical analyses were performed with SPSS version 28.0-J (SPSS-Japan, Tokyo, Japan).

Ethical Considerations

This study was approved by the ethics committee of the Faculty of Medicine, University of Tsukuba (approval no. 1652). All study participants were mailed a description of the research study and its ethical considerations, a consent form, and the questionnaire. The consent form was collected by having study participants return the signed form along with the questionnaire.

Results

Descriptive Statistics and Univariate Analysis

A self-administered questionnaire was sent to 457 patients; 50.1% (n = 228; 69.8% (n = 116) in the vaccination group and 38.8% (n = 113) in the non-vaccinated group) returned the questionnaire. A study participant in the non-vaccinated group was excluded because the self-administered questionnaire revealed a history of herpes zoster within 5 years. Thus, 116 members of the vaccination group and 112 members of the non-vaccinated group were included in the final analysis.

Characteristics of the study participants and results of the univariate analysis are summarized in Table 1.

Table 1.

Characteristics of the Study Participants and Results of the Univariate Analysis.

Characteristic	Total (n = 228)		Vaccination group (n = 116)		Non-vaccinated group (n = 112)		P value
Sex^§							.304
Male	85	37.3%	47	40.5%	38	33.9%
Female	143	62.7%	69	59.5%	74	66.1%
Age*
(years) median [interquartile range]	78	[73.0-83.0]	78	[73.0-83.0]	78	[73.0-83.0]	.526
Type of health insurance^§							.483
Latter-stage elderly healthcare insurance	155	68.0%	81	69.9%	74	66.1%
National health insurance	64	28.1%	29	25.0%	35	31.3%
Japan Health Insurance Association	6	2.6%	4	3.4%	2	1.8%
Other	3	1.3%	2	1.7%	1	0.8%
Municipality of residence^§							.519
X city	221	96.9%	112	96.6%	109	97.3%
Other	7	3.1%	4	3.4%	3	2.7%
Underlying disease^§							.134
Cardiac	157	68.8%	70	60.3%	87	77.7%
Respiratory	12	5.2%	9	7.8%	3	2.7%
Renal	15	6.6%	7	6.0%	8	7.1%
Malignant tumor	7	3.1%	6	5.2%	1	0.9%
Diabetes	31	13.6%	13	11.2%	18	16.1%
None	53	23.2%	33	28.4%	20	17.9%
Educational background^§							.375
Less than high school	57	25.0%	31	26.7%	26	23.2%
High school	122	53.5%	60	51.7%	62	55.4%
Vocational school	19	8.3%	11	9.5%	8	7.1%
Junior college	10	4.4%	5	4.3%	5	4.5%
University or graduate school	17	7.5%	6	5.2%	11	9.8%
Missing	3	1.3%	3	2.6%	0	0%
Annual household income (JPY, millions)^§							.703
<2.50 (<19 000 US dollars)	109	47.8%	58	50.0%	51	45.5%
2.50–4.99	85	37.3%	40	34.5%	45	40.2%
5.00–7.99	18	7.9%	9	7.8%	9	8.0%
≥8.00	6	2.6%	2	1.7%	4	3.6%
Missing	10	4.4%	7	6%	3	2.7%
History of herpes zoster in the participant^§							.645
Yes	50	21.9%	24	20.7%	26	23.2%
No	178	78.1%	92	79.3%	86	76.8%
History of herpes zoster in family members^§							.766
Yes	100	43.9%	50	43.1%	50	44.6%
No	125	54.8%	65	56.0%	60	53.6%
Missing	3	1.3%	1	0.9%	2	1.8%
History of herpes zoster in friends^§							.576
Yes	154	67.5%	81	69.8%	73	65.2%
No	66	28.9%	32	27.6%	34	30.4%
Missing	8	3.6%	3	2.6%	5	4.4%
Pneumococcal vaccination status (participant)^§							<.001
Yes	156	68.4%	105	90.5%	51	45.5%
No	72	31.6%	11	9.5%	61	54.5%
Influenza vaccination status (participant)^§							<.001
Yes	141	61.8%	102	87.9%	39	34.8%
No	87	38.2%	14	12.1%	73	65.2%
VZV vaccine awareness^§							<.001
Yes	138	60.5%	91	78.4%	47	42.0%
No	89	39.0%	24	20.7%	65	58.0%
Missing	1	0.5%	1	0.9%	0	0%
Reason for VZV vaccine awareness (% among those who know about the VZV vaccine) ^§							<.001
Saw the hospital board	22	16.2%	15	16.5%	7	14.9%
Read the pamphlet in the hospital	6	4.3%	5	5.5%	1	2.1%
Information from physician	67	48.6%	59	64.8%	8	17.0%
Information from family members	19	13.8%	8	8.8%	11	23.4%
Information from friends	29	21.0%	12	13.2%	17	36.2%
Other	8	5.8%	3	3.3%	5	10.6%
Physician recommendation for VZV vaccination^§							<.001
Yes	82	36.0%	71	61.2%	11	9.8%
No	143	62.7%	43	37.1%	100	89.3%
Missing	3	1.3%	2	1.7%	1	0.9%
Family member recommendation for VZV vaccination^§							.059
Yes	18	7.9%	13	11.2%	5	4.5%
No	208	91.2%	102	87.9%	106	94.6%
Missing	2	0.9%	1	0.9%	1	0.9%
Friend recommendation for VZV vaccination^§							.693
Yes	9	3.9%	4	3.4%	5	4.5%
No	217	95.2%	111	95.7%	106	94.6%
Missing	2	0.9%	1	0.9%	1	0.9%
VZV vaccination history of family members^§							.083
Yes	36	15.8%	23	19.8%	13	11.6%
No	187	82.0%	90	77.6%	97	86.6%
Missing	5	2.2%	3	2.6%	2	1.8%
VZV vaccination history of friends^§							.024
Yes	37	16.2%	25	21.6%	12	10.7%
No	184	80.7%	87	75.0%	97	86.6%
Missing	7	3.1%	4	3.4%	3	2.7%
Self-assessment of health status^§							.693
Very good	11		6		5
Good	163		79		84
Bad	50		29		21
Very bad	4		2		2
Subjective well-being
Agitation*	2.43		2.49		2.37		.481
Attitude toward own aging*	2.21		2.21		2.21		.955
Loneliness and dissatisfaction*	1.61		1.63		1.58		.568
Total score*	7.01		7.12		6.9		.488
JPCAT-SF
First contact*	57.4		57.7		57.1		.87
Longitudinality*	69.6		73.2		66		.012
Coordination*	55.3		57.9		52.7		.222
Comprehensiveness (services available)*	64.4		66.4		62.4		.204
Comprehensiveness (services provided)*	31.4		31.7		31.2		.916
Community orientation*	66.1		69		63.3		.071
Total score*	57.4		59.3		55.5		.047
Missing	29	12.7%	19	16.4%	10	8.9%

Abbreviations: JPCAT-SF, JapaneseVersion of the Primary Care Assessment Tool Short Form; VZV, varicella-zoster virus.

The total JPCAT-SF score was calculated as the average of the 6 subscale scores from 0 to 100.

Student’s t-test.

Chi-squared test.

There were significant associations between the 2 groups in terms of pneumococcal vaccination, influenza vaccination, physician recommendation for VZV vaccination, VZV vaccination history of friends, JPCAT-SF longitudinality, and total JPCAT-SF score.

Multivariable Analysis

Logistic regression was conducted with VZV vaccination as the dependent variable and adjustment for age and sex. In addition, physician recommendation for VZV vaccination, family history of VZV vaccination, history of VZV vaccination of friends, total JPCAT-SF score, and family member recommendation for VZV vaccination, which showed an association (P < .10) in the univariate analysis, were used as independent variables. Friend recommendation for VZV vaccination, selected based on the DAG, was also added. Based on discussions among researchers and on the DAG, we performed adjustments for variables with P < .10 in the univariate analysis, which included pneumococcal vaccination status, influenza vaccination status, VZV vaccine awareness, and reason for VZV vaccine awareness.

The results of the multivariable logistic regression analysis are summarized in Table 2.

Table 2.

Multivariable Logistic Regression Analysis With VZV Vaccination as the Dependent Variable.

	OR	95% CI	P value
Sex	0.993	0.940-1.049	.799
Age	0.478	0.223-1.028	.059
Physician recommendation for VZV vaccination	18.033	7.260-44.790	<.001
Family member recommendation for VZV vaccination	3.504	0.536-22.904	.073
Friend recommendation for VZV vaccination	0.118	0.011-1.223	.073
VZV vaccination history of family members	0.544	0.159-1.858	.332
VZV vaccination history of friends	1.193	0.364-3.907	.770
Total JPCAT-SF score	1.184^a	0.810-1.734^a	.374

Abbreviations: CI, confidence interval; JPCAT-SF, Japanese Version of the Primary Care Assessment Tool Short Form; OR, odds ratio; SD, standard deviation; VZV, varicella–zoster virus.

The total JPCAT-SF score was calculated as the average of the 6 subscale scores from 0 to 100.

Per 1SD increase.

Multivariable analysis showed no significant associations between VZV vaccination status and total JPCAT-SF score but a significant association with physician recommendation for VZV vaccination.

Mediation Analysis

Physician recommendation for VZV vaccination, which had a significant association with VZV vaccination in the multivariable analysis, was analyzed as an intermediate factor.

(1) Results of the logistic regression analysis with physician recommendation for VZV vaccination as the dependent variable are shown in Table 3. This analysis corresponds to the analysis of part (1) in Figure 3. A significant association was found between the 2 groups in terms of total JPCAT-SF score and physician recommendation for VZV vaccination.

(2) Results of the logistic regression analysis with VZV vaccination as the dependent variable are shown in Table 4. This analysis corresponds to the analysis of part (2) in Figure 3. There was a significant association between physician recommendation for VZV vaccination and VZV vaccination.

(3) We analyzed the association between total JPCAT-SF score and VZV vaccination in the logistic regression analysis in which physician recommendation for VZV vaccination, which was considered an intermediate factor, was not included as an independent variable and VZV vaccination was the dependent variable (Table 5). This analysis corresponds to the analysis of part (3) in Figure 3. There was a significant association between total JPCAT-SF score and VZV vaccination.

Table 3.

Logistic Regression Analysis With Physician Recommendation for VZV Vaccination as the Dependent Variable.

Characteristic	OR	95% CI	P value
Sex	0.969	0.920-1.020	.228
Age	1.288	0.641-2.590	.478
Family member recommendation for VZV vaccination	5.059	1.146-22.338	.032
Friend recommendation for VZV vaccination	0.590	0.076-4.588	.614
VZV vaccination history of family members	2.278	0.838-6.190	.107
VZV vaccination history of friends	3.649	1.367-9.742	.010
Total JPCAT-SF score	1.491^a	1.040-3.087^a	.029

Abbreviations: CI, confidence interval; JPCAT-SF, Japanese Version of the Primary Care Assessment Tool Short Form; OR, odds ratio; SD, standard deviation; VZV, varicella-zoster virus.

The total JPCAT-SF score was calculated as the average of the 6 subscale scores from 0 to 100.

Per 1SD increase.

Table 4.

Logistic Regression Analysis With VZV Vaccination as the Dependent Variable and Total JPCAT-SF Score Excluded as an Independent Variable.

Characteristic	OR	95% CI	P value
Sex	1.033	0.986-1.082	.169
Age	0.447	0.641-2.590	.021
Physician recommendation for VZV vaccination	20.46	8.479-49.371	<.001
Family member recommendation for VZV vaccination	2.825	0.498-16.016	.241
Friend recommendation for VZV vaccination	0.084	0.008-0.845	.035
VZV vaccination history of family members	0.458	0.147-1.429	.179
VZV vaccination history of friends	1.890	0.642-5.563	.248

Abbreviations: CI, confidence interval; JPCAT-SF, Japanese Version of the Primary Care Assessment Tool Short Form; OR, odds ratio; VZV, varicella-zoster virus.

Table 5.

Logistic Regression Analysis With VZV Vaccination as the Dependent Variable and Physician Recommendation for VZV Vaccination Excluded as an Independent Variable.

Characteristic	OR	95% CI	P value
Sex	0.978	0.932-1.025	.347
Age	0.666	0.350-1.268	.216
Family member recommendation for VZV vaccination	5.968	1.176-30.289	.031
Friend recommendation for VZV vaccination	0.150	0.019-1.178	.071
VZV vaccination history of family members	1.056	0.405-2.753	.911
VZV vaccination history of friends	2.338	0.887-6.163	.086
Total JPCAT-SF score	1.381^a	1.000-1.917^a	.049

Abbreviations: CI, confidence interval; JPCAT-SF, Japanese Version of the Primary Care Assessment Tool Short Form; OR, odds ratio; SD, standard deviation; VZV, varicella-zoster virus.

The total JPCAT-SF score was measured from the average of the 6 subscale scores from 0 to 100.

Per 1SD increase.

Discussion

This study showed an association between PX and VZV vaccination in a multivariable analysis that excluded physician recommendation for VZV vaccination as a variable. The association between PX and VZV vaccination disappeared when physician recommendation for VZV vaccination was added as a variable. Multivariable analysis with physician recommendation for VZV vaccination as the dependent variable showed an association between physician recommendation for VZV vaccination and PX. Furthermore, multivariable analysis with VZV vaccination as the dependent variable showed an association between physician recommendation for VZV vaccination and VZV vaccination. Therefore, physician recommendation for VZV vaccination was found to be an intermediate factor between PX and VZV vaccination. This study might be the first to show an association between VZV vaccination and PX measured after VZV vaccination.

We found that PX can affect VZV vaccination if the physician’s recommendation for VZV vaccination was not included as a variable. Previous studies of pneumococcal and influenza vaccination⁷ did not include physician recommendations as a variable; our findings show the same trend. PX can also be improved by the actions of a single physician, for example, by taking sufficient time to talk about the patient’s concerns. However, PX is improved to a greater extent through multidisciplinary and multi-departmental activities to enhance the quality of the entire healthcare organization than action at the individual level.^24,25 To improve vaccination coverage and disease prevention behavior, it is not enough for each physician to be patient-centered. It is more important for multiple professions to work together to improve PX.

The study revealed that physician recommendations strongly influence vaccination behavior. This is consistent with previous findings about influenza vaccination⁵ and a study showing that healthcare providers were the most trusted source of information about vaccination.²⁶ Although professional advice has been reported to be a significant factor in behavior change,²⁷ it is also possible that the halo effect²⁸ might influence the information provided by the physician. For the less well-known VZV vaccine, it is particularly likely that the information was first provided in a physician’s recommendation. Thus, it is possible that physician recommendation was a stronger factor for VZV vaccination than for other, more widely recognized vaccines.

The findings of this study, which showed that patients who rate the quality of their primary care highly are more likely to be vaccinated when their physicians recommend vaccines, are novel. A good doctor-patient relationship was reported to be associated with the success or failure of behavior change.²⁹ However, preventive interventions by primary care physicians were not always sufficient.³⁰ Evidence-based suggestions by individual physicians for vaccine-preventable diseases, along with daily activities to increase PX, might increase vaccination coverage, especially for vaccines that are not well recognized, such as the VZV vaccine.^31,32 Considering that the COVID-19 vaccine and other new vaccines and evidence about them will continue to be developed, it will become even more necessary to promote actions to improve the quality of primary care and the dissemination of correct vaccine knowledge in the future.

One limitation of this study is that it was conducted at a single, small facility, and at a hospital with only a few physicians, which might have reflected the strong influence of specific physicians. As expected, the response rate for the questionnaires in the non-vaccinated group was low. Therefore, there is a possibility of bias in the results. Although there were 29 missing values for JPCAT-SF, we excluded them as part of a missing available-case analysis because JPCAT-SF was the most important factor in this study. This exclusion might have caused selection bias.

In addition, due to the study design, this study measured PX after VZV vaccination. PX, which covers a range of time, scores events experienced in primary care in the period prior to the time of measurement, so it is likely that events prior to VZV vaccination were assessed. However, it also reflects events experienced in primary care after VZV vaccination. Since there was a period of up to 3 years between VZV vaccination and the self-administered questionnaire survey, PX might have changed during that period. However, since there were no significant changes in physicians or hospital systems during this period, the impact of this change would be negligible. This study alone is insufficient to verify whether PX contributes to VZV vaccination. Further research is needed.

Furthermore, since the study used a self-administered questionnaire for elderly study participants, it is possible that the responses might differ from actual results due to recall bias or other factors. Elderly patients with cognitive decline who could not respond to the self-administered questionnaire were excluded from the study, but it is possible that patients with mild dementia were included. Since there was a period of up to 3 years between VZV vaccination and the self-administered questionnaire survey, it is possible that the respondents might have forgotten details such as whether or not their physicians had recommended the vaccine. Vaccination histories for influenza and pneumococcal vaccines were extracted from electronic medical records, but the date might have been incomplete because it is possible that vaccines given at other hospitals were not recorded.

As mentioned above, the survey was conducted in an environment where the VZV vaccine was widely publicized. The survey results might reflect a high level of awareness of the VZV vaccine compared with the general clinical setting in Japan today. However, since awareness of the VZV vaccine has been gradually increasing in Japan as a whole since the survey, the overall situation in Japan is expected to become closer to the setting of this study.

Conclusion

An association was found between PX measured after VZV vaccination and its vaccination. Physician recommendation for VZV vaccination is an intermediate factor between PX and VZV vaccination. A recommendation from a physician for VZV vaccination based on better PX is very useful for promoting VZV vaccination among the elderly.

Footnotes

Acknowledgements

The authors thank the community hospital and their staff which participated in our study. The authors would like to thank zenis corp () for the English language review.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Approval

This study was approved by the Ethics Committee of the Faculty of Medicine, University of Tsukuba (approval no. 1652). Study participants provided written informed consent.

ORCID iDs

Takashi Inaba

Ryohei Goto

References

Dworkin

Johnson

Breuer

, et al. Recommendations for the management of herpes zoster. Clin Infect Dis. 2007;44 Suppl 1:1. doi:10.1086/510206

Toyama

Shiraki

. Epidemiology of herpes zoster and its relationship to varicella in Japan: a 10-year survey of 48,388 herpes zoster cases in Miyazaki prefecture. J Med Virol. 2009; 81(12):2053-2058. doi:10.1002/jmv.21599

Johnson

Bouhassira

Kassianos

Leplège

Schmader

Weinke

. The impact of herpes zoster and post-herpetic neuralgia on quality-of-life. BMC Med. 2010;8(1):37. doi:10.1186/1741-7015-8-37

Oxman

Levin

Johnson

, et al. A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults. N Engl J Med. 2005;352(22):2271-2284. doi:10.1056/NEJMoa051016

Kajikawa

Kataoka

Goto

, et al. Factors associated with influenza vaccination in Japanese elderly outpatients. Infect Dis Health. 2019;24(4):212-221. doi:10.1016/j.idh.2019.07.002

Kamal

Madhavan

Amonkar

. Determinants of adult influenza and pneumonia immunization rates. J Am Pharm Assoc. 2003;43(3):403-411. doi:10.1331/154434503321831120

Kaneko

Aoki

Goto

Ozone

Haruta

. Better patient experience is associated with better vaccine uptake in older adults: multicentered cross-sectional study. J Gen Intern Med. 2020;35(12):3485-3491. doi:10.1007/s11606-020-06187-1

Wolf

Niederhauser

Marshburn

LaVela

. Defining patient experience. Patient Exp J. 2014;1(1):7-19. doi:10.35680/2372-0247.1004

Little

Everitt

Williamson

, et al. Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations. BMJ. 2001; 323(7318): 908-911. doi:10.1136/bmj.323.7318.908

10.

Doyle

Lennox

Bell

. A systematic review of evidence on the links between patient experience and clinical safety and effectiveness. BMJ Open. 2013;3(1):e001570. doi:10.1136/bmjopen-2012-001570

11.

Ministry of Health, Labour and Welfare. Long-Term Care, Health and Welfare Services for the Elderly. 2017. Accessed May 21, 2023. https://www.mhlw.go.jp/english/policy/care-welfare/care-welfare-elderly/

12.

Imafuku

Herpes zoster and subunit vaccine. Article in Japanese. Virus. 2021;71(1):45-54.

13.

Drph

Starfield

. Validating the adult primary care assessment tool. J Fam Pract. 2001;50(2):161.

14.

Aoki

Inoue

Nakayama

. Development and validation of the Japanese version of Primary Care Assessment Tool. Fam Pract. 2016;33(1):112-117. doi:10.1093/fampra/cmv087

15.

Aoki

Fukuhara

Yamamoto

. Development and validation of a concise scale for assessing patient experience of primary care for adults in Japan. Fam Pract. 2020;37(1): 137-142. doi:10.1093/fampra/cmz038

16.

Takahashi

, Ozawa T, eds. Evaluation of mental function, motivation, morale, and quality of life. In: Guide to the Assessment of Life Functioning of the Elderly. Ishiyaku Publishers, cInc; 1999;51-58.

17.

Lawton

. The Philadelphia Geriatric Center Morale Scale: a revision. J Gerontol. 1975;30(1):85-89. doi:10.1093/geronj/30.1.85

18.

Campbell

Ramsay

Green

. Age, gender, socioeconomic, and ethnic differences in patients’ assessments of primary health care. Qual Health Care. 2001;10(2):90-95. doi:10.1136/qhc.10.2.90

19.

Xiao

Barber

. The effect of perceived health status on patient satisfaction. Value Health. 2008;11(4):719-725. doi:10.1111/j.1524-4733.2007.00294.x

20.

Textor

van der Zander

Gilthorpe

Liskiewicz

Ellison

. Robust causal inference using directed acyclic graphs: the R package ‘dagitty’. Int J Epidemiol. 2016;45(6):1887-1894. doi:10.1093/ije/dyw341

21.

Baron

Kenny

. The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51(6):1173-1182. doi:10.1037/0022-3514.51.6.1173

22.

Peduzzi

Concato

Kemper

Holford

Feinstein

. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373-1379. doi:10.1016/s0895-4356(96)00236-3

23.

Penny

Atkinson

. Approaches for dealing with missing data in health care studies. J Clin Nurs. 2012;21(19-20):2722-2729. doi:10.1111/j.1365-2702.2011.03854.x

24.

AHRQ. The CAHPS Ambulatory Care Improvement Guide: Practical Strategies for Improving Patient Experience. 2016. Accessed April 7, 2023. https://www.ahrq.gov/cahps/quality-improvement/improvement-guide/improvement-guide.html

25.

Friedberg

SteelFisher

Karp

Schneider

. Physician groups’ use of data from patient experience surveys. J Gen Intern Med. 2011;26(5):498-504. doi:10.1007/s11606-010-1597-1

26.

Wada

Smith

. Mistrust surrounding vaccination recommendations by the Japanese government: results from a national survey of working-age individuals. BMC Public Health. 2015;15:426. doi:10.1186/s12889-015-1772-8

27.

Elder

Ayala

Harris

. Theories and intervention approaches to health-behavior change in primary care. Am J Prev Med. 1999;17(4):275-284. doi:10.1016/s0749-3797(99)00094-x

28.

Nisbett

Wilson

. The halo effect: evidence for unconscious alteration of judgments. J Pers Soc Psychol. 1977;35(4):250-256. doi:10.1037/0022-3514.35.4.250

29.

Longnecker

. Doctor-patient communication: a review. Ochsner J. 2010;10(1):38-43.

30.

Yarnall

KSH

Pollak

Østbye

Krause

Michener

. Primary care: is there enough time for prevention? Am J Public Health. 2003;93(4):635-641. doi:10.2105/ajph.93.4.635

31.

GSK. Survey of Japanese people’s awareness of shingles and vaccination. Article in Japanese.GSK. 2022. Accessed May 21, 2023. https://jp.gsk.com/ja-jp/news/press-releases/20220228_taijouhoushin-yobou/

32.

International Federation on Ageing. Shingles Awareness Week 2023 (Global) | International Federation on Ageing. 2023. Accessed May 21, 2023. https://ifa.ngo/news/shingles-awareness-week-2023/