Influenza vaccination coverage and factors affecting adherence to influenza vaccination among patients with diabetes in Taiwan

Abstract

The purpose of this study was to investigate influenza vaccination coverage and the factors influencing acceptance of influenza vaccination among patients with diabetes in Taiwan using the Health Belief Model (HBM). From January 1 to February 28, 2012, 700 patients with diabetes who visited National Cheng Kung University Hospital were invited to participate in the study. A total of 691 (99%) patients with diabetes were enrolled in the study. The mean age of the subjects was 64.7 years (SD = 10.7). The percentages of patients with diabetes who received seasonal influenza vaccination were 31%, 33%, and 35% in 2009–2010, 2010–2011, and 2011–2012, respectively. Multiple regression analyses revealed that patients with diabetes who were female, were older, had comorbidities, had a more positive perception of the benefits of the influenza vaccine and had lower perceived barriers to influenza vaccination were more likely to receive the influenza vaccine in 2011–2012 (adjusted R² = 0.47; Chi-square = 276.50; P < 0.001). Patients with diabetes perceived the risk of swine influenza to be similar to that of seasonal influenza. Consequently, in the absence of an increase in the perceived risk of influenza, a low level of actual vaccination against seasonal influenza is forecasted. Strategies to improve the uptake of influenza vaccination include interventions that highlight the risk posed by pandemic influenza while simultaneously offering tactics to ameliorate this risk.

Introduction

Influenza is a major cause of illness and death worldwide. The morbidity and mortality associated with influenza are high among the elderly, pregnant women, immunocompromised hosts, and those with chronic diseases such as diabetes mellitus (DM).^1-3 Diabetes is a serious and growing health problem worldwide. It is estimated that a total of 439 million adults will be affected by DM by 2030, with a 20% increase in developed countries and a 69% increase in developing countries from 2010 to 2030.⁴ Patients with diabetes are 2–4 times more likely to die from influenza and pneumonia compared with individuals without diabetes.^5-7 Influenza vaccination can reduce the risk of hospitalization and death in the elderly and in high-risk individuals.^8-10

The willingness of individuals to be vaccinated against a given infectious agent has been investigated extensively in the psycho-sociological literature on risk perceptions and illness, as this willingness is known to be a major issue in the success of vaccination programs.¹¹ Several models have been developed to account for the adoption of health protective behavior by individuals. These models include the Health Belief Model (HBM), the theory of reasoned action, the Triandis model, the multi-attribute utility (MAU) theory, and the subjective expected utility theory.^12-16 The HBM is one of the most widely used health behavior models.¹⁷

The original goal of the HBM was to focus the efforts of those individuals who sought to improve public health by understanding why people failed to adopt preventive health behaviors.¹² The model proposed that for an individual to take action to avoid a disease, the individual needs to believe that (1) he or she is susceptible to the disease (perceived susceptibility); (2) the disease could have at least a moderately severe impact on a certain component of his or her life (perceived severity); (3) certain behaviors could be beneficial to reduce his or her perceived susceptibility or severity in the event that the individual contracts the disease (perceived benefits); and (4) these behaviors are not impeded by factors such as tangible and psychological costs or pain caused by an advised action (perceived barriers). After 1985, to improve the model’s predictive ability, researchers expanded the HBM to include the concept of self-efficacy.¹⁸ Self-efficacy is a person’s belief in how capable he or she is of performing a certain behavior.

In Taiwan, free influenza vaccination was offered to the high-risk elderly population starting in 1998.¹⁹ Currently, the influenza vaccine is recommended for all individuals aged 65 y and older, and particularly those who are at increased risk due to an underlying chronic immunosuppressive disease, including patients with diabetes.²⁰ The average influenza vaccination rate for adults aged 65 y and older is approximately 40%,^19,21 far below the national goal of 68% set by the Centers for Disease Control in Taiwan (Taiwan CDC).²² To increase adherence to influenza vaccination among these high-risk individuals, we hypothesized that subjects who received influenza vaccination would perceive more benefits from vaccination and would have fewer barriers to receiving vaccination than would individuals who did not receive vaccination. However, few studies have explored the factors that affect the receipt of influenza vaccination among diabetes patients in Taiwan. Therefore, the aims of our study were to evaluate the uptake of preventive influenza vaccination in diabetes patients and to identify the reasons for these patients’ vaccination decisions.

Results

Between January 2012 and February 2012, 700 diabetes patients met the inclusion criteria. Of these patients, 99% (691) completed a questionnaire and were enrolled in the study. Table 1 shows the baseline demographic characteristics and comorbidities of the study subjects. The mean age of the subjects was 64.7 y (SD = 10.7), and the male-to-female ratio was 1:1.08. Many participants (49%) had a primary school education, and most of the participants were married (94%). The percentages of participants who said that they had received influenza vaccination in 2009–2010, 2010–2011, and 2011–2012 were 31%, 33%, and 35%, respectively. Compared with the non-vaccinated participants, the diabetic patients with influenza vaccination were significantly more likely to be female and aged 65 y and older and to have coronary heart disease (CHD) or chronic obstructive pulmonary disease (COPD) as a comorbidity.

Table 1. Baseline demographic characteristics and comorbidities in diabetic patients with and without influenza vaccination in 2011–2012.

Variable	No. vaccinated (%)	No. unvaccinated (%)	P value
Sex			<0.001
Male	99 (30)	234 (70)
Female	142 (40)	216 (60)
Age (years)			<0.001
40–64	58 (16)	314 (84)
≥65	183 (57)	136 (43)
Education (years)
≤6	115 (33)	230 (67)	0.395
7–9	48 (35)	91 (65)	0.924
10–12	42 (36)	75 (64)	0.799
≥13	36 (40)	91 (60)	0.237
Marital status			0.059
Married	232 (36)	417 (64)
Other	9 (21)	33 (79)
Concomitant disease
CHD	110 (53)	97 (47)	<0.001
COPD	34 (52)	32 (48)	0.003
CKD	10 (48)	11 (52)	0.213

CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease.

Table 2 presents the distribution of the subscale totals and the mean item scores, the standard deviations and ranges, the number of scale items, and the Cronbach’s α coefficient for the HBM scales among the study participants. The mean scores for the items were obtained by dividing the subscale scores by the number of items in the scale. The Cronbach’s α coefficients were acceptable for all of the multi-item measures (0.76–0.96). The perceived benefit items had the highest Cronbach’s α coefficient, and the perceived barrier items had the lowest coefficient.

Table 2. Subscale scores for the Health Belief Model for the intention to receive influenza vaccination among diabetes patients in Taiwan.

Subscale	No. of items	Range of scores	M ± SD	Cronbach’s α	Item score M ± SD
SUS	4	4–20	9.58 ± 3.67	0.92	2.39 ± 0.91
SEV	5	5–25	17.09 ± 5.00	0.89	3.40 ± 1.00
BEN	5	5–25	17.97 ± 3.99	0.96	3.59 ± 0.80
BAR	8	8–40	19.92 ± 5.94	0.76	2.49 ± 0.74
Cues to action	5	5–25	14.26 ± 4.81	0.80	2.85 ± 0.96

SUS, perceived susceptibility; SEV, perceived severity; BEN, perceived benefits; BAR, perceived barriers; M, mean.

Table 3 shows a comparison of the subscale for the HBM for the individuals who received influenza vaccination and the individuals who did not receive influenza vaccination. Compared with those individuals who did not receive influenza vaccination, those who received vaccination had a significantly higher score for perceived benefits and a lower score for perceived barriers and cues to action.

Table 3. Comparison of subscale scores for the Health Belief Model for those individuals who received and those who did not receive influenza vaccination.

Subscale	Those who received	Those who did not receive	P value
	M ± SD	M ± SD
SUS	9.69 ± 3.85	9.53 ± 3.59	0.582
SEV	17.02 ± 4.92	17.01 ± 5.06	0.989
BEN	19.57 ± 4.18	17.29 ± 3.70	0.001
BAR	16.49 ± 6.27	21.44 ± 5.07	0.001
Cues to action	13.63 ± 4.92	14.56 ± 4.69	0.018

SUS, perceived susceptibility; SEV, perceived severity; BEN, perceived benefits; BAR, perceived barriers; M, mean.

Table 4 shows comparisons of the factors that influenced the HBM. Personal experience with influenza in the past year (yes/no) or influenza in the family in the past year (yes/no), influenza vaccination in the past year (yes/no) or vaccination in the family in the past year (yes/no) and health status (good/poor) were the factors influencing the HBM.

Table 4. Comparison of health beliefs according to explanatory variables.

Explanatory variable	SUS	SEV	BEN	BAR
	M ± SD	M ± SD	M ± SD	M ± SD
Influenza in the past year Yes No	9.91 ± 3.92** 9.04 ± 3.31	17.07 ± 4.99 17.01 ± 4.73	19.82 ± 4.14* 16.88 ± 3.61	18.29 ± 5.21 18.28 ± 5.11
Influenza in the family in the past year Yes No	9.70 ± 3.93 9.67 ± 3.84	17.04 ± 4.91 17.01 ± 5.07	18.79 ± 4.13 18.31 ± 4.23	18.43 ± 5.33 18.27 ± 5.43
Vaccination in the past year Yes No	9.71 ± 3.95 9.57 ± 3.62	17.21 ± 4.63 17.01 ± 4.58	19.59 ± 4.19* 17.31 ± 3.70	16.53 ± 6.12* 21.48 ± 5.39
Vaccination in the family in the past year Yes No	9.35 ± 3.80* 9.72 ± 3.91	17.05 ± 4.93 17.01 ± 5.07	18.79 ± 4.30 18.43 ± 4.27	18.29 ± 5.28 18.79 ± 5.31
Estimation risk without vaccination Yes No	9.72 ± 3.93* 9.37 ± 3.84	18.05 ± 4.83* 17.01 ± 5.07	19.48 ± 4.17* 17.50 ± 3.56	18.33 ± 5.28 18.45 ± 5.29
Health status Good/fair Bad	9.74 ± 3.92* 9.35 ± 3.81	18.01 ± 4.93* 16.95 ± 4.57	18.46 ± 4.27 18.21 ± 4.14	18.11 ± 5.13 18.66 ± 5.49

SUS, perceived susceptibility; SEV, perceived severity; BEN, perceived benefits; BAR, perceived barriers. *P < 0.01.

Self-experience with influenza in the past year and poor health status significantly increased the perceived susceptibility to illness. The results indicate that experience with vaccination in the family and influenza in the family in the past year significantly decreased the perceived susceptibility to influenza. A higher perceived probability of contracting influenza without vaccination and a poor health status led to a significantly higher perceived severity of disease. The perceived benefits of vaccination were positively associated with self-experience with influenza vaccination in the past year, a higher perceived probability of infection without vaccination and a lower perceived probability of infection with vaccination. Those individuals who had experience with influenza vaccination in the past year also had significantly decreased the perceived barriers to receiving influenza vaccination.

Table 5 shows the multiple regression analyses of receiving influenza vaccination in 2011–2012 as a function of the demographic variables, comorbidities, and the HBM constructs. Within this data set, the subjects who were female and aged 65 y or older and had CHD or COPD as a comorbidity, higher perceived benefits from and lower perceived barriers to receiving influenza vaccination were more likely to accept influenza vaccination (adjusted R² = 0.47; Chi-square = 276.50; P < 0.001). Compared with the subjects who did not receive influenza vaccination, the subjects who did receive vaccination had higher perceived benefits from vaccination and had fewer barriers to receiving vaccination. Insufficient knowledge concerning the safety of the influenza vaccine (58%), consideration of the potential side effects of the vaccine (54%), the vaccine’s effectiveness (38%) and the health effects of the vaccine (30%) were the most common reasons for not receiving influenza vaccination among the diabetes patients (data not shown).

Table 5. Multiple regression analysis of the intention to receive influenza vaccination in 2011–2012 as a function of demographic variables and the Health Belief Model constructs (n = 691).

	OR	95% CI
Demographic factor
Gender
Female		referent
Male	0.58*	0.36–0.92
Age (years)
40–64		referent
≥65	7.27**	4.53–11.68
Education level (years)
≤6		referent
7–9	1.03	0.66–1.61
10–12	1.13	0.71–1.79
≥13	1.40	0.80–2.21
Marital status
Unmarried		referent
Married	2.02	0.90–4.66
Comorbidity
CHD	3.06**	2.15–4.25
COPD	2.15*	1.25–3.69
CKD	1.73	0.67–4.45
Health Belief Model
SUS	1.01	0.95–1.07
SEV	0.97	0.93–1.02
BEN	1.13**	1.07–1.20
BAR	0.86**	0.83–0.90
Cues to action	1.04	0.99–1.09
Chi-square for model	276.50**
Adjusted R2	0.47

CHD, coronary heart disease; COPD, chronic obstructive pulmonary disease; CKD, chronic kidney disease; SUS, perceived susceptibility; SEV, perceived severity; BEN, perceived benefits; BAR, perceived barriers. The dependent variable was influenza vaccination in 2011–2012 (yes/no). The available independent variables were gender, age, education level, marital status, CHD, COPD, CKD, and the Health Belief Model constructs. *P < 0.05; **P < 0.001.

This study did not find any significant effect of the experience of contracting influenza or the experience of influenza vaccination on the perceived severity of illness. The perceived probability of contracting influenza and the experience of contracting influenza also had effects on the perceived barriers to receiving influenza vaccination.

Discussion

Our results show that influenza vaccination rates for diabetes patients were low (31–35%),^9,10,19 far below the national goal of 68% set by the Taiwan CDC.²² Patients aged 65 y and older had a higher rate of vaccination compared with patients aged younger than 65 y. The patients aged 65 y and older were vaccinated because of their age, and not because they understand that having diabetes and being older puts them in a high-risk category. The results of our survey are similar to the findings of a CDC survey of Medicare beneficiaries about the reasons for not receiving influenza vaccination.²³

In 2003, the coverage rate of the influenza vaccine was 49% among adults with diabetes in the USA.²⁴ Despite the recommendations of the Taiwan CDC, influenza vaccination coverage still did not achieve the goal of 90% set by the WHO.^24, In Taiwan, free influenza vaccination for high-risk and elderly individuals aged 65 y and older was implemented in 1998. However, the present study showed that the influenza vaccination rate for patients with diabetes was approximately 31–35% between 2009 and 2012. Insufficient knowledge about the influenza vaccine has led to the low rate of influenza vaccination. In addition, many health workers are not convinced of the safety and effectiveness of influenza vaccination.^9,25

Our study also found that adults aged 65 y and older were more likely to receive influenza vaccination from 2011–2012 than younger adults, aged 40–64 y, were (57% vs. 16%). That the Taiwanese government did not provide free influenza vaccination for people younger than 65 y of age may be one of the reasons for lower influenza vaccination coverage among younger diabetic patients. Alternatively, Li²⁶ found that older adults were more likely to receive influenza vaccination if they had received vaccination in the previous year and had more access to influenza vaccination at district hospitals and clinics than at medical centers. Collectively, providing free influenza vaccination, organizing influenza vaccination campaigns, and providing integrated service are the strategies available to increase vaccination coverage among diabetic patients.

The sociodemographic characteristics of individuals may influence their attitudes and health-related behaviors. Age is an important factor affecting willingness to receive influenza vaccination. Older age positively affects the decision to receive the influenza vaccine.²⁷ In addition, it has been found that men are more likely to receive vaccination than women are.²⁸ Our study found that patients aged 65 y and older had higher rates of receiving influenza vaccination. In contrast, our study found that women had a higher rate of influenza vaccination than men did.

According to the HBM, the perceived susceptibility to influenza, the perceived benefits of vaccination and fewer barriers to receiving vaccination are the factors affecting the acceptance of influenza vaccination. Similar to previous studies,^25,28 we found that subjects who received influenza vaccination perceived more benefits from vaccination and had fewer barriers to receiving vaccination than did individuals who did not receive vaccination. Inaccessibility reduced the willingness to receive influenza vaccination in the group of diabetes patients.

Rosenstock¹² suggested that individuals are likely to change their beliefs to be consistent with their behavior once they have adopted a particular behavior. However, Janz and Becker²⁹ have argued that once individuals have adopted a preventive measure, they should perceive themselves as being less susceptible to a disease. This theory may foster a paradoxical relationship between susceptibility and the likelihood of adopting a behavior. In line with this suggestion, our retrospective study found that diabetes patients who reported that they had received influenza vaccination perceived themselves as less susceptible to contracting influenza.

Cues to action refer to motivating factors, such as social norms and health education campaigns, which may include encouragements and enforcements to improve preventive behaviors.³⁰ Thus, to increase the intention to receive influenza vaccination, diabetes patients should be given a high level of encouragement and enforcement.

Consistent with the findings of a previous study by Shahrabani and Benzion,³¹ our results suggest that previous vaccination with the influenza vaccine had a significant effect on perceived benefits and perceived barriers. In addition, those individuals who had a family who had received influenza vaccination perceived a lower probability of infection. In particular, individuals who had experience with the influenza vaccine perceived more benefits from the vaccine and lower barriers to access to influenza vaccination. The possible reason for this outcome was that a positive experience with the vaccine helped these individuals to perceive more benefits from the vaccine and also decreased their perceived barriers to vaccination.

It has been found that considerations about the vaccine’s safety and fears of the side effects of vaccination are associated with a lower willingness to receive vaccination.²⁸ Our results indicate that the main reason for not receiving influenza vaccination among diabetes patients was that they did not have sufficient knowledge about the vaccine’s safety and its side effects. These findings are consistent with the findings in Hoeney and Moorre’s study,³² which may suggest that the group that refused the vaccine wanted more information about influenza and vaccination before being vaccinated because of the lack of safety and efficacy data on the new vaccine.

Regional immunization policies may play an important role in affecting influenza vaccination coverage.³³ Chang et al.²¹ found a positive relationship between preventive health examinations and vaccination rates. Another population-based study in Taiwan also showed that older people who regularly undergo health examination are more likely to receive influenza vaccinations than those individuals who do not.³⁴ The authors concluded that this phenomenon might be attributable to those individuals receiving the influenza vaccine having a better knowledge of the vaccine, leading to increased adoption of health protective behaviors and thus increasing influenza vaccination coverage. Furthermore, if general practitioners provide free influenza vaccination while providing routine physical examinations and preventive health education to patients, older diabetic patients may be more willing to receive influenza vaccination.³⁵

Our study provides a more accurate assessment of actual influenza vaccination practices than do surveys of the general population. Potential limitations of our study include the following: (1) the measure for receiving influenza vaccination relied on self-reports and may have overestimated action, despite efforts to reduce this effect; (2) the study sample was limited to a teaching hospital, so the results may not be generalizable to all patients with diabetes; (3) the nature of cross-sectional studies may pose difficulties in interpreting the results; (4) the questionnaire was administered in 2012 regarding events that had happened 3 y before, so recall bias could have affected the accuracy of the information; and (5) the validity of the influenza vaccination status should have been evaluated. However, recent studies found more than 90% sensitivity and 56% specificity for self-reported influenza vaccination.³⁶ Therefore, it is strongly recommended that a prospective study be conducted in the future.

In conclusion, our study indicates that diabetes patients who are female, are older, with comorbidity, perceive more benefits from the influenza vaccine and perceive fewer barriers to receiving influenza vaccination are the most likely to receive influenza vaccination. Increasing confidence about the vaccine and its safety, providing free influenza vaccination to diabetes patients younger than 65 y old and reducing the barriers to vaccination (e.g., by increasing the accessibility of vaccination) may be strategies to increase influenza vaccination coverage.

Methods

Settings

This study was conducted at National Cheng-Kung University Hospital, a tertiary hospital located in southern Taiwan. This hospital is one of the largest health care facilities in the country and has a catchment population of 2 million people. Approximately 4000–4500 diabetes patients in the area receive care at this facility each month. Certain patients are referred to the hospital by other health care providers, but the majority of patients present on their own, either for the diagnosis and treatment of their diabetes or because treatment obtained elsewhere has failed.

Study subjects

From January 1 to February 28, 2012, all diabetes patients who visited National Cheng-Kung University Hospital were invited to participate in the study, regardless of the reason for their visit. The inclusion criteria were diabetes diagnosed more than 3 y ago and being aged 40 y or older. The exclusion criteria were a diagnosis of diabetes less than 3 y before, type 1 diabetes and incomplete information on the questionnaire (e.g., age, vaccination history). Approval was obtained from the Institutional Review Board for human subjects at National Cheng-Kung University Hospital, Tainan, Taiwan. After signing an informed consent form, the patients were given a self-administered questionnaire addressing their relevant sociodemographic characteristics, the HBM constructs, and their intention to receive and action regarding influenza vaccination. We calculated an estimated necessary sample size of 460 patients. The calculation of sample size assumed that the study should have 80% power at an α level of 0.05 to detect influenza vaccination coverage of 40% among subjects who visited the study hospital and to ensure a maximal sample size and an allowable error of 10% (±5%).³⁷

Age was split into 2 groups (40–64 y and 65 y and older). Educational level was categorized as ≤6 y, 7–9 y, 10–12 y, or ≥13 y. Marital status was classified as married or other. Influenza in the family in the past year was used to assess the experience of influenza in the family, which was self-reported and categorized as yes or no. Health status was self-assessed and categorized as good/fair or poor.

Measurements

A structured questionnaire containing questions on sociodemographic characteristics (including age, marital status, education, and comorbidity) and questions related to the HBM constructs was used as the data collection tool. The HBM consisted of the following 4 constructs: perceived susceptibility, perceived severity, perceived benefits, and perceived barriers. In addition, other important constructs, including cues to action, were added to this study.

To develop the HBM construct items, we reviewed the related literature^38,39 and interviewed 30 diabetes patients who were aged 40 y and older to identify their beliefs concerning influenza vaccination. The derived items were measured on a Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). A pilot study was performed on 30 diabetes patients who met the inclusion criteria to determine the appropriateness of the questionnaire for the target population. The subjects who participated in the pilot study were asked to respond to questions about the relevance, simplicity and clarity of each item. Following this pilot study, no items were omitted from the final questionnaire, but several items were rephrased to enhance clarity and understanding. In addition, the questionnaire was reviewed by 5 health education specialists who were experienced in high-risk driving behavior research, and their considerations were applied to the questionnaire. To test reliability, the internal consistency of the questionnaire was assessed using the Cronbach’s α coefficient. An α ≥ 0.70 was considered to be satisfactory.

Behavior

Behavior related to influenza vaccination was determined as a self-reported construct. One item was used to measure behavior: “Did you receive seasonal influenza vaccination in 2009–2010, 2010–2011, and 2011–2012?” This item was measured as a categorical variable (yes vs. no). In this study, only information on seasonal influenza vaccination was collected and analyzed. We did not collect information on H1N1 influenza vaccine uptake.

Perceived susceptibility

Four items were used to measure the perceived susceptibility, including “The probability that you will contract an influenza infection is very high during the pandemic.” The score ranged from 4–20. The Cronbach’s α for the perceived susceptibility was 0.92.

Perceived severity

Five items were used to measure the perceived severity, including “How serious do you think the influenza pandemic is?” The score ranged from 5–25. The Cronbach’s α for the perceived severity was 0.89.

Perceived benefits

Five items were used to measure the perceived benefits, such as “receiving influenza vaccination decreases the risk of influenza infection in Taiwan.” The score range of this scale was from 5–25. The Cronbach’s α for the perceived benefits was 0.96.

Perceived barriers

Eight items were used to measure the perceived barriers, including “There are too many risks in receiving the influenza vaccine.” The score ranged from 8–40. The Cronbach’s α for the perceived barriers was 0.76.

Cues to action

Five items were used to measure the cues to action, including “How often do participants who received influenza vaccination remind you to receive it?” These items were measured on a Likert scale ranging from 1 (never) to 5 (always). The score ranged from 5–25. The Cronbach’s α for the cues to action was 0.80.

After the subjects received clear instructions and an explanation of the study’s purpose, the questionnaire was administered. The study subjects completed the questionnaire while waiting for their physician appointments. The subjects took approximately 20–30 min to complete the questionnaire and were assured that the data that they provided would remain confidential.

Data analysis

All reverse-scaled statements on the questionnaire were recorded in the same direction. Therefore, for the HBM constructs, a high score represented high perceived susceptibility, high perceived severity, high perceived benefits, high perceived barriers, and high cues to action. A multiple logistic regression model was generated to identify the factors affecting the uptake of the influenza vaccine in 2011–2012. The dependent factor was set as receiving the influenza vaccine in 2011–2012. The independent variables were gender, age, educational level, marital status, the presence of comorbidities (CHD, COPD, or CKD), and the HBM constructs. A two-tailed P value of <0.05 indicated statistical significance. Statistical analyses were performed using the SPSS software package, version 11.5 (SPSS Inc.).

Glossary

Abbreviations:

HBM

Health Belief Model

DM

diabetes mellitus

CHD

coronary heart disease

COPD

chronic obstructive pulmonary disease

CKD

chronic kidney disease

SUS

perceived susceptibility

SEV

perceived severity

BEN

perceived benefits

BAR

perceived barriers

M

mean

10.4161/hv.27816

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

This study was supported by 2 grants: CLFHR-10110, from the Chi-Mei Medical Center, and NCKUH-10103021, from National Cheng Kung University Hospital, Tainan, Taiwan. We thank the Centers for Disease Control, Taiwan, for providing the National Surveillance and vaccination coverage data for influenza infections.