Measuring Motivation for COVID-19 Vaccination: An Application of the Transtheoretical Model

Abstract

Purpose

In the United States (US), individuals vary widely in their readiness to get vaccinated for COVID-19. The present study developed measures based on the transtheoretical model (TTM) to better understand readiness, decisional balance (DCBL; pros and cons), self-efficacy (SE), as well as other motivators for change such as myths and barriers for COVID-19 vaccination.

Design

Cross-sectional measurement development.

Setting

Online survey.

Sample

528 US adults ages 18-75.

Measures

Demographics, stage of change (SOC), DCBL, SE, myths, and barriers.

Analysis

The sample was randomly split into halves for exploratory factor analysis using principal components analysis (EFA/PCA), followed by confirmatory factor analyses (CFA) to test measurement models. Correlation matrices were assessed and multivariate analyses examined relationships between constructs and sub-constructs.

Results

For DCBL, EFA/PCA revealed three correlated factors (one pros, two cons) (n₁ = 8, α = .97; n₂ = 5, α = .93; n₃ = 4, α = .84). For SE, two correlated factors were revealed (n₁ = 12, α = .96; n₂ = 3, α = .89). Single-factor solutions for Myths (n = 13, α = .94) and Barriers (n = 6, α = .82) were revealed. CFA confirmed models from EFAs/PCAs. Follow-up analyses of variance aligned with past theoretical predictions of the relationships between SOC, pros, cons, and SE, and the predicted relationships with myths and barriers.

Conclusion

This study produced reliable and valid measures of TTM constructs, myths, and barriers to understand motivation to receive COVID-19 vaccination that can be used in future research.

Keywords

COVID-19 transtheoretical model health behaviors vaccination decision making vaccine confidence

Purpose

Despite overwhelming evidence supporting clear benefits of vaccination for various diseases,¹ including the prevention of avoidable death and suffering,² many individuals fail or refuse to get vaccinated, while others may be unable to get vaccinated. Issues surrounding vaccination status persisted following the approval of the first COVID-19 vaccines, which were approved in December 2020 in the midst of the deadliest surge of the pandemic. Still, there is a wide range of readiness to engage in this health behavior change, even in the context of a global pandemic that has killed more than one million Americans.³ As of September 2022, close to two years from the start of wide COVID-19 vaccine availability, 23% of Americans remain unvaccinated, with the majority of this group (88%) reporting that they definitely would not get vaccinated for COVID-19.⁴ A robust body of literature suggests that interventions aimed at increasing vaccination willingness should focus on promoting benefits and debunking myths surrounding vaccinations.^5-8 Additionally, given the wide range of beliefs about vaccines, it is important that interventions are tailored to meet individual needs that might best support a decision to get vaccinated.^9-11 One model that is designed to guide health promotion at any stage of readiness for a behavior change is the transtheoretical model (TTM).

The TTM frames behavior change as a progression through five stages of change (SOC), which include Precontemplation (PC) (ie, not intending to change soon), Contemplation (C) (ie, change is being considered, but not definitely planned), Preparation (P) (ie, behavior change is about to occur), Action (A) (ie, behavior change is occurring), and Maintenance (M) (ie, behavior change has occurred).¹² Movement through the stages is initiated by changes in the TTM-related constructs of decisional balance (DCBL, pros and cons), which measures positive or negative consequences for self or others, and self-efficacy (SE), which assesses one’s confidence to engage in health behavior change under challenging conditions. While barriers and myths are not constructs typically studied under the TTM framework, they are worthy of empirical investigation in the context of vaccination readiness and vaccination promotion, in addition to those that are central to the TTM, given their relationship with increased vaccine hesitancy and decreased vaccine uptake.^13-15

Barriers, or obstacles that prevent an individual from successfully engaging in health behavior change, have also been investigated in the context of health behaviors, including vaccination readiness.^12,16-18 Barriers (eg, inability to schedule an appointment) to vaccination are distinct from cons or contexts related to self-efficacy as they may not be overcome or changed merely by a change in motivation or perspective when deciding whether to get vaccinated. Moreover, myths, or widely held but false beliefs, also have a pre-existing role in interventions for certain health behaviors, specifically those for increasing vaccination rates.⁶ Myths surrounding vaccination (eg, vaccines cause autism) are distinct from cons as they are not accurate negative consequences of being vaccinated, however, they may still cause hesitancy to engage in this health behavior.

Currently, measures exist for general attitudes toward COVID-19 vaccination,¹⁹ as well as for factors that may predict intention to receive a COVID-19 vaccination.^20-22 However, to date, no known research has specifically applied the TTM to studying readiness for COVID-19 vaccination. Such measures differ from those developed through application of the TTM as they are not grounded in a single, comprehensive model of intentional behavior change that has been shown to accurately predict relationships between one’s stage of change (ie, readiness) and core constructs involved in the decision-making process.^12,23-28 Moreover, applying the TTM to COVID-19 vaccination employs a deliberate approach rooted in a framework that has been shown to be effective in guiding the development of tailored interventions for a range of health behavior changes that can be delivered at both the individual and population levels.^23-25 For example, measures for DCBL and SE were reliable and valid for HPV vaccination among young women²⁴ and for pros and SE among young men.²³ Paiva and colleagues²⁵ extended this work and found that 91% of participants endorsed their intention to get vaccinated after a computer-tailored intervention based on the TTM was administered. This research suggests that the TTM constructs of SOC, DCBL, and SE can successfully be applied to vaccination as a health behavior and may be useful in informing future interventions aimed at increasing vaccination readiness.

The purpose of the current study is to apply the TTM to readiness for COVID-19 vaccination among adults who maintain the autonomy to choose whether to get vaccinated for COVID-19 (ie, aged 18 and older). This study describes the development and validation of TTM-based measures of SOC, DCBL, and SE, as well as myths and barriers for COVID-19 vaccination. Hypotheses include: (1) Measure development for SOC, DCBL and SE scales regarding COVID-19 vaccination will demonstrate factor structures comparable to previous studies investigating the application of TTM to health behaviors changes with good model fit, (2) measure development for myths and barriers scales regarding COVID-19 vaccination will demonstrate factor structures with good model fit, (3) cons, SE, myths, and barriers will be independent, yet moderately correlated constructs, and (4) internally consistent TTM measures will be developed, demonstrating a pattern of results of pros (+1 SD from PC to A), cons (−.5 SD from PC to A), SE (+.8 SD from PC to A), myths, and barriers scales by stage. Both myths and barriers are expected to significantly decrease from PC to A. However, no specific metric of decline is expected as there is minimal previous research on these constructs by stage, thus limiting confidence in this prediction.

Methods

Design

A sequential measure development method was used to assess the reliability and validity of the developed scales, including stage of change, decisional balance, self-efficacy, myths, and barriers measures for COVID-19 vaccination.^23,26,27 To determine external validity, the patterns of the decisional balance and self-efficacy constructs, as well as myths and barriers across stage of change were compared to those established in previous research on vaccination of HPV, blood donation, living donor kidney transplant, and other patterns for a range of behaviors.^12,23-28

Sample

Survey Sample

The target population for this study included adults ages 18 and older with the autonomy to choose whether to get vaccinated for COVID-19. Participants were eligible regardless of previous vaccination status during the first burst of the survey. Additional survey bursts were done to ensure adequate distribution of the sample across SOC, therefore limiting eligible participants to those in specific stages (ie, Contemplation and Preparation) during these recruitment efforts. No participant was excluded based on race, ethnicity, gender, or sexual orientation. Participants were excluded if they were under the age of 18 and did not live in the United States.

Eligible participants who provided written informed consent were asked to complete a demographic questionnaire, a SOC questionnaire, and scale questionnaires for DCBL, SE, myths, and barriers regarding COVID-19 vaccination. All study procedures were approved by the university’s Institutional Review Board (IRB).

Recruitment

The study survey was developed and distributed using an online survey platform, Qualtrics (www.qualtrics.com). Recruitment and reimbursement were completed through the data collection platform, Prolific (www.prolific.co). Prolific participants have typically been recruited to Prolific via social media, flyer distributions on college campuses, and a now discontinued Prolific referral scheme that offered small monetary incentives to a referring party. All participants were compensated for their time through the payment structure recommended by Prolific. Eight bursts of the survey were conducted to ensure adequate distribution of the sample across SOC beginning January 26, 2022, and ceasing March 13, 2022. Prolific identified 64 986 eligible participants across survey bursts. Attention checks were included in the surveys and responses were rejected or approved per Prolific guidelines.

Measures

Item Development

A comprehensive literature review was conducted by the first author (A.S.) to inform item development for TTM-related constructs as applied to COVID-19 vaccination, as well as definitions for other pertinent constructs and item development related to COVID-19 vaccination readiness (myths, barriers). Initial item development and mapping of items onto constructs (ie, stage of change, decisional balance, self-efficacy, myths, and barriers) was conducted by two authors (A.S. and M.L.R.). Items were adapted from previously conducted studies identified during the literature review on the application of TTM to various health behaviors,^23,26,27 as well as informed by studies on attitudes toward vaccination,¹ readiness for vaccination,¹⁸ and beliefs regarding vaccination.^29-31 Items for each construct were then reviewed by the remaining authors for feedback. Upon completion of development and review, the decisional balance measure consisted of 32 items, the self-efficacy measure consisted of 28 items, the myths measure consisted of 18 items, and the barriers measure consisted of 12 items.

Stage of Change

Participants completed a questionnaire assessing their vaccination status. They were placed in one of four mutually exclusive categories for stage of change depending on the responses to this questionnaire. Four SOC categories were used, without the inclusion of Maintenance as a fifth stage, given that completion of the behavior (ie, being “fully vaccinated”) at the time of data collection was marked by receiving a single dose of the Janssen/Johnson & Johnson vaccine or two doses of the Pfizer BioNTech or Moderna vaccines. In this way, individuals who had received one vaccine dose and were scheduled for the second were not maintaining the behavior, but rather completing their time in the stage of “Action”. Maintenance was not investigated within the current study given the lack of knowledge about and availability of additional prevention efforts (ie, boosters) for COVID-19 at the time this study was conducted.^32,33 Definitions of each SOC categories can be viewed in Table 1. Additional questions were posed about positive COVID-19 results, type of vaccine received (if applicable), dose(s) received, including boosters, and reasons why dose(s), including boosters, were not received.

Table 1.

Definitions for Stage of Change.

Stage	COVID-19 vaccine series
Precontemplation (PC)	Unvaccinated, I am not planning to get vaccinated within the next 3 months.
Contemplation (C)	Unvaccinated, I am considering getting vaccinated within the next 3 months.
Preparation (PR)	Unvaccinated, but I am trying to schedule, or I am scheduled to get vaccinated.
Action (A)	Received 1 dose of the vaccine, scheduled for the second dose (Moderna or Pfizer BioNTech OR fully vaccinated (Moderna, Pfizer BioNTech, or Janssen/J&J)

Decisional Balance

A 32-item DCBL measure was developed to assess pros (advantages, eg, I would feel safer) and cons (disadvantages, eg, I might put my health at risk) of COVID-19 vaccination Participants were asked to rate how important each item was in their decision to get vaccinated for COVID-19 on a five-point scale, ranging from 1 (not at all important) to 5 (extremely important) after reading the phrase, “if I were to get vaccinated for COVID-10… [item]”.

Self-Efficacy

A 28-item SE measure was developed to assess an individual’s confidence to get vaccinated for COVID-19 in challenging situations (eg, I am depressed). Participants were asked to rate, “I am confident in my ability to get vaccinated for COVID-19 even if… [item]” on a five-point scale, ranging from 1 (not at all confident) to 5 (extremely confident).

Myths

An 18-item myths measure was developed to assess an individual’s degree of agreement with certain widely held, but false beliefs (eg, vaccines can cause autism). Participants were asked to “Please rate how much you agree with these statements on a scale of 1 (not at all) to 5 (strongly) … [item]”.

Barriers

A 12-item barriers measure was developed to assess the extent to which certain obstacles (eg, inability to get childcare) was a barrier to getting vaccinated for COVID-19. Participants were asked to rate, “For me personally, how much of a barrier to getting vaccinated for COVID-19 is… [item]” on a five-point scale, ranging from 1 (not at all) to 5 (extremely).

Analysis

A sample of 528 adults ages 18 and older were recruited considering previous findings regarding sufficient sample size for measurement development studies (ie, 100-400 participants),^34,35 as well as to account for an estimated 15%–20% of missing data.³⁶ The sample was randomly split into halves for exploratory (n₁ = 270) and confirmatory factor analysis (n₂ = 258) (EFA/principal components analysis (PCA) and CFA) to test measurement models. In the EFA/PCA, items with poor (<.40) or complex (>.40 on 1+ component) loading were removed, in addition to any unclear or redundant items. Scale reliability was determined using the overall Cronbach α. In the CFA, criteria for model fit included Comparative Fit Index (CFI), Chi Squared (χ²), the Root Mean Square Error of Approximation (RMSEA), and the Standardized Root Mean Square Residual (SRMR), where CFI ≥ .95, SRMR ≤ .08, and RMSEA close to .06 are considered acceptable values indicating good fit.³⁷ All retained items can be found in Table 2.

Table 2.

All Retained Items From Confirmatory Factor Analysis.

Construct	Item
Decisional Balance	Health and Safety Pros
	I would feel safer
	I would be less likely to get infected
	I would be less likely to experience severe symptoms
	I would worry less about getting my loved ones sick
	I would worry less about getting sick
	I would feel more comfortable interacting with others or being near others
	My body would develop an immune response to the virus
	I would be less likely to spread the virus
	Health and Safety Cons
	I might experience negative side effects
	I might put my health at risk
	I would be worried about putting my health at risk
	I would feel uncomfortable because I do not trust the people promoting the vaccine
	I would be worried about potential side effects
	Social Cons
	I would not fit in with my friends
	Others would disapprove of my decision
	My family would disapprove of my decision
	My loved ones would disapprove of my decision
Self-Efficacy	General Self-Efficacy
	My family does not approve of my decision
	I am under a lot of stress
	It is inconvenient
	My friends do not approve
	I am depressed
	Injections or needles make me uncomfortable
	It is difficult to make an appointment
	The celebrities I look up to do not promote vaccination
	I am too busy
	My political leaders do not approve
	My religious leaders do not approve
	I am anxious
	Mistrust
	I do not trust the companies that developed the vaccine.
	People around me are experiencing serious side effects from the vaccine.
	I do not trust the government organizations promoting the vaccine.
Myths	I do not have to get vaccinated if I am young and healthy.
	Vaccine development is rushed and therefore not safe.
	Vaccines can increase the risk of autoimmune disease.
	Vaccines can change my DNA.
	It is better for me to become naturally immune to the virus than to become immune from the vaccine.
	Getting too many vaccines will overwhelm my immune system.
	Vaccines can make it more difficult to become pregnant.
	Vaccines contain harmful ingredients.
	My personal medical information will be exposed after getting vaccinated.
	Vaccines given in early pregnancy increases the risk of miscarriage.
	I do not have to get vaccinated if everyone around me gets vaccinated.
	Vaccines can cause autism.
	I can be tracked by the government after getting vaccinated.
Barriers	Inability to take time off from school.
	Lack of access to transportation to a vaccination site.
	Inability to take time off from work.
	Inability to get childcare.
	Inability to take time off from current responsibilities.
	Inability to schedule an appointment.

Next, a correlational matrix assessed correlations between constructs. Correlations between scales are presented in Table 3. Finally, multivariate analyses (MANOVA and ANOVA) examined relationships between SOC and each construct, and strongest predictors of SOC after conversion of scale scores to T-scores (M = 50, SD = 10) for uniformity of measurement.

Table 3.

Correlations Between Scales.

		Health and safety pros	Health and safety cons	Social cons	General self-efficacy	Mistrust (self-efficacy)	Myths	Barriers
Health and Safety Pros	Pearson Correlation	1
Health and Safety Pros	Sig. (2-tailed)
Health and Safety Cons	Pearson Correlation	−.508^**	1
Health and Safety Cons	Sig. (2-tailed)	.000
Social Cons	Pearson Correlation	−.117^**	.286^**	1
Social Cons	Sig. (2-tailed)	.007	.000
General Self-Efficacy	Pearson Correlation	.506^**	−.374^**	−.235^**	1
General Self-Efficacy	Sig. (2-tailed)	.000	.000	.000
Mistrust (Self-Efficacy)	Pearson Correlation	.330^**	−.319^**	.005	.456^**	1
Mistrust (Self-Efficacy)	Sig. (2-tailed)	.000	.000	.917	.000
Myths	Pearson Correlation	−.629^**	.712^**	.333^**	−.451^**	−.258^**	1
Myths	Sig. (2-tailed)	.000	.000	.000	.000	.000
Barriers	Pearson Correlation	.125^**	.027	.243^**	−.101^*	−.062	.052	1
Barriers	Sig. (2-tailed)	.004	.542	.000	.021	.156	.237

**Correlation is significant at the .01 level (2-tailed).

*Correlation is significant at the .05 level (2-tailed).

Results

The final survey sample used for analysis included 528 participants (N = 528). The sample was 75.4% White (n = 398) and 50.9% female (n = 269). The mean age was approximately 36 years old (SD = 12.8) and the age range was 18-75. Select demographic details can be found in Table 4.

Table 4.

Descriptive Characteristics of Sample (N = 528).

Variable	Category	n	%
Stage of Change	Precontemplation (PC)	227	43
	Contemplation (C)	66	12.5
	Preparation (PR)	36	6.8
	Action (A)	199	37.7
Race/Ethnicity	White	398	75.4
	More than one	41	7.8
	Black or African American	39	7.4
	Hispanic/Latino	23	4.4
	Asian	20	3.8
	Not listed	4	.8
	American Indian or Alaska Native	2	.4
	Native Hawaiian or Pacific Islander	1	.2
Gender	Female	269	50.9
	Male	248	47
	Gender Non-Conforming	8	1.5
	Not listed	2	.4
	Transgender Man	1	.2
Age	18-29 years old	186	35.2
	30-49 years old	247	46.8
	50-64 years old	80	15.2
	65+	15	2.8

Exploratory and Confirmatory Analyses

A split half approach was used in which a randomly generated half of the sample (n = 270) was used as the exploratory sample to investigate the number of items and components present. Estimates of the correlation between items and components, as well as estimates for the factor loadings of each item were provided. Complex items (ie, items loading more than .40 on more than one component) and items with poor loading (ie, items loading less than .40) were eliminated. Scale reliability was determined using the overall Cronbach α. Principal components analysis (PCA) with varimax rotation was used to examine item correlation matrices for decisional balance, self-efficacy, myths, and barriers. The number of components to retain was based on the minimum average partial procedure (MAP) and parallel analysis (PA).^23,28 Final item selection considered maximizing item clarity, lack of redundancy, and breadth of construct.

Decisional Balance

Through a series of seven iterative PCAs with varimax rotation, 32 DCBL items were reduced to 17 items with three final factors. Examination of the content revealed that the first factor (eight items) clearly represented health and safety pros, the second factor (five items) clearly represented health and safety cons, and the third factor (four items) clearly represented social cons of COVID-19 vaccination. Internal consistency was excellent for the health and safety pros scale (α = .97) and cons scale (α = .93) and good for the social cons scale (α = .84). The three factors accounted for 77.39% of the total variance (38.13% for health and safety pros, 23.04% for health and safety cons, 16.22% for social cons).

The CFA produced a three-component correlated model with good factor loadings and an adequate model fit, χ² (116) = 375.37, CFI = .94, RMSEA = .09, SRMR = .08 (α = .97, Health and Safety Pros; α = .92, Health and Safety Cons; α = .85, Social Cons). Correlations between Health and Safety Pros and Health and Safety Cons was r = −.51 and Health and Safety Pros and Social Cons was r = −.12. The correlation between Health and Safety Cons and Social Cons was r = .29.

Self-Efficacy

Through a series of five iterative PCAs with varimax rotation, 28 SE items were reduced to 15 items. Examination of the content revealed that the first factor (twelve items) clearly represented general self-efficacy and the second factor (three items) clearly represented mistrust about COVID-19 vaccination. Internal consistency was excellent for the general self-efficacy scale (α = .96) and mistrust scale (α = .89). The two factors accounted for 72.98% of the total variance (56.66% for general self-efficacy, 16.32% for mistrust).

The CFA produced a two-component correlated model with good factor loadings and an adequate model fit, χ² (89) = 368.89, CFI = .92, RMSEA = .11, SRMR = .05 (α = .96, General Self-Efficacy; α = .91, Mistrust). The correlation between General Self-Efficacy and Mistrust was r = .46.

Myths

Through a series of five iterative PCAs with varimax rotation, 18 myths items were reduced to 13 items. Here, researchers erred on being more inclusive in retaining items to ensure breadth of the construct was achieved. Examination of the content revealed that one factor (thirteen items) clearly represented myths about COVID-19 vaccination. All item loadings were above .6. Internal consistency was excellent for the scale (α = .94). The one factor accounted for 59.31% of the total variance.

The CFA produced a one-component model with good factor loadings and an adequate model fit, χ² (65) = 483.03, CFI = .82, RMSEA = .16, SRMR = .07 (α = .94).

Barriers

Through a series of five iterative PCAs with varimax rotation, 12 barrier items were reduced to 6 items. Similar to myths, researchers erred on being more inclusive in retaining items to ensure breadth of the construct was achieved. Items were then removed due to low loadings and repetition based on findings from a CFA prior to conducting the final PCA. Examination of the content revealed that one factor (six items) clearly represented barriers to COVID-19 vaccination. All item loadings were above .75. Internal consistency was good for the barriers scale (α = .82). The one factor accounted for 54.46% of the total variance.

The CFA produced a one-component model with good factor loadings and an adequate model fit, χ² (9) = 87.95, CFI = .88, RMSEA = .18, SRMR = .07 (α = .85).

External Validation

Decisional Balance by Stage

A series of MANOVAs revealed that individuals at different stages of readiness for COVID-19 vaccination differed significantly on their subjective importance of the pros and cons of COVID-19 vaccination (F (9, 1270) = 72.20, P < .001, η² = .29). Follow up ANOVAs indicated that those in different SOC differed significantly on the Health and Safety Pros of COVID-19 vaccination (F (3, 8493) = 163.50, P < .001, η² = .48), Health and Safety Cons of COVID-19 vaccination (F (3, 8157) = 151.41, P = 0 < .001, η² = .46), and Social Cons (F (3, 491) = 5.02, P = .002, η² = .03). In comparison to individuals in Contemplation, Preparation, and Action, post-hoc analyses showed that individuals in Precontemplation rated Health and Safety Pros of COVID-19 vaccination as significantly less important. Individuals in Precontemplation also rated Health and Safety Cons as significantly more important in comparison to individuals in Contemplation, Preparation, and Action. Additionally, those in Precontemplation rated Social Cons as significantly more important than those in Action. Overall, Health and Safety Pros increased by 1.52 SD from Precontemplation to Action, Health and Safety Cons decreased by 1.51 SD from Precontemplation to Action, and Social Cons decreased by .34 SD from Precontemplation to Action (Figure 1).

Figure 1.

Decisional balance by stage of change.

Self-Efficacy by Stage

A series of MANOVAs revealed that individuals at different stages of readiness for COVID-19 vaccination differed significantly in their confidence to get vaccinated (F (6, 1046) = 22.71, P < .001, η² = .12). Follow up ANOVAs indicated that those in different SOC differed significantly on the General Self-Efficacy of COVID-19 vaccination (F (3, 3121) = 37.74, P < .001, η² = .18) and Mistrust of COVID-19 vaccination (F (3, 2226) = 25.35, P < .001, η² = .13). Post-hoc analyses showed that individuals in PC for COVID-19 vaccination rated General Self-Efficacy significantly lower (ie, showing less general confidence) than those C, PR, and A. Additionally, those in PC rated Mistrust significantly lower (ie, showing less mistrust-related confidence) than those in PR and A. However, ratings of General Self-Efficacy and Mistrust did not significantly differ among those in C and PR. General Self-Efficacy increased by .94 SD from PC to A and Mistrust increased by .76 SD from PC to A (Figure 2).

Figure 2.

Self-efficacy by stage of change.

Myths by Stage

A series of MANOVAs revealed that individuals at different stages of readiness for COVID-19 vaccination differed significantly on their endorsement of myths (F (3, 7807) = 139.71, P < .001, η² = .44). Post-hoc analyses showed that those in PC endorsed Myths surrounding COVID-19 vaccination significantly higher than those in C, PR, and A. Those in C also endorsed Myths significantly more than those in A. Myths decreased by 1.45 SD from PC to A (Figure 3).

Figure 3.

Myths by stage of change.

Barriers by Stage

A series of MANOVAs revealed that individuals at different stages of readiness for COVID-19 vaccination differed significantly on their endorsement of barriers (F (3, 1246) = 13.33, P < .001, η² = .07). Post-hoc analyses showed that those in PC and C endorsed significantly less Barriers to COVID-19 vaccination than those in PR. Those in PR also endorsed Barriers significantly more than those in A. Barriers increased by 1.05 SD from PC to PR and decreased by .01 SD from PC to A (Figure 4).

Figure 4.

Barriers by stage of change.

Discussion

In the present study, measures for constructs specific to the transtheoretical model, including decisional balance and self-efficacy, in addition to myths and barriers, were developed through successful application of the transtheoretical model to COVID-19 vaccination. Results of this application were consistent with previous TTM research²⁷ and yielded measures for decisional balance, self-efficacy, as well as other key constructs (myths and barriers) that are particularly relevant for such a behavior that may be influenced in certain social and political contexts.

While it is common for DCBL measure development studies to yield two-factor solutions,³⁸ three-factor solutions have been found in several studies.^26,39,40 The current study’s findings is consistent with the latter, yielding two cons scales and one pros scale. Pros and cons items consisted of similar language and phrasing, which may account for the retained items’ high internal consistency and model fit values. Moreover, the retention of solely social-related cons items, vs two distinct scales for social pros and social cons is notable. These results imply that cultural context appears important for many considering vaccination, and the potential for social disapproval may be a powerful con for this behavior, given the social climate in the US surrounding vaccination. Thus, questions arise regarding whether social disapproval may act as a driving factor to engage in various health behaviors specifically among American adults.

External validity was demonstrated for all three DCBL scales developed in the current study, as the majority of patterns for changes in construct and sub-constructs across stage of change were confirmed. The magnitude and direction of change in both pros (+1.52 SD) and cons (−1.51 SD) is consistent with expected changes by stage for these constructs found in previous studies.³⁸ Similarly, the correlation found between Health and Safety Pros and Health and Safety Cons (r = −.51) is consistent with previous research related to the TTM, which indicates that as pros of a behavior change increase in importance, cons decrease in importance.³⁸ These findings suggest that health- and safety-related pros and cons significantly contribute to one’s decision to get vaccinated for COVID-19. Notably, while the .34 SD decrease in social cons across stage of change also closely adheres to that found in previous research,³⁸ the difference in the magnitude of the decrease in Health and Safety Cons vs Social Cons should be considered. Social Cons thus appear to be very challenging regardless of what stage one is in. This may be occurring because of the highly politicized nature of COVID-19 vaccination in the US.⁴¹ Moreover, the cross-over in which the pros begin to outweigh the cons in Preparation occurred as well, for both cons scales. Thus, these findings are consistent with TTM theory and past results of measurement development for a range of health behaviors.²⁸

Results showed two factors within SE: general self-efficacy and mistrust. Previous research supports the idea that mistrust in COVID-19 vaccination may be related to vaccine hesitancy.^42,43 However, such previous findings underscore how medical mistrust, specifically, is at the forefront of this sub-construct. Instead, the current study found that items related to medical mistrust, such as “I do not trust the medical field” did not load to the point of retention in the EFA. In a post-hoc follow-up analysis, ratings of confidence to get vaccinated for COVID-19 despite one’s mistrust in the medical field did not differ among participants in Precontemplation, Contemplation, and Preparation though these stages were found to significantly differ from one another based on the items retained. While other items, such as “people around me are experiencing serious side effects from the vaccine” may imply mistrust of the medical field, future research should investigate various ways of phrasing items related to mistrust and whether the current phrasing accurately represents the challenge that not having trust in the medical field poses in deciding whether to get vaccinated for COVID-19. Similarly, it is critical to consider whether items that are more directly related to medical mistrust may have been retained in a more diverse sample given the historical mistreatment by the medical community against people of color.^44-46 Additionally, in a post-hoc follow-up analysis, ratings of confidence to get vaccinated for COVID-19 despite one’s mistrust in the government organizations promoting the vaccine were significantly lower among individuals in Precontemplation and Contemplation compared to those in Preparation and Action. Interestingly, most participants in Precontemplation identified as republican (n = 87) or independent (n = 106), while most in Action identified as democratic (n = 123). These results are consistent with prior research showing that a lack of trust and confidence in government authorities increases the likelihood of vaccine hesitancy and refusal.⁴³

Regarding the myths surrounding COVID-19 vaccination, the anticipated decrease of myths was observed across SOC, which may suggest that myths about vaccines are highly impactful in one’s decision to get vaccinated, specifically among those who have chosen not to get vaccinated. This is consistent with previous research, which suggests that myths surrounding vaccines may be more likely to influence the decision to receive vaccination when an individual is already hesitant to get vaccinated due to heightened side effect concerns.⁶ However, given the cross-sectional nature of the data, directionality of these findings cannot be definitively concluded. Moreover, given the increase in barriers from Precontemplation to Preparation, future research should investigate the imminent impact of barriers on those who are trying to schedule or already scheduled to get vaccinated for COVID-19, as results show that barriers seem to be most obstructive to those in Preparation. This need is highlighted by the US Department of Health and Human Services as researchers note that mere access to the COVID-19 vaccination is a critical barrier.⁴⁷ Older adults who are homebound or unfamiliar with technological needs associated with scheduling, or those who lack a social network that may be helpful in traveling to a vaccination appointment may be unable to get vaccinated despite readiness to engage in the behavior.⁴⁷ It is also important to note the model fit observed for all scales, and specifically that for myths and barriers given their relatively high RMSEA values. Several researchers caution others in their interpretation of fit indices, including RMSEA, SRMR, CFI, and χ² as the meaning of “good” fit is not well understood, in that many cutoffs are arbitrary or may be valued differently depending on the size of the sample and degrees of freedom.^48-50

Limitations and Future Directions

Limitations of the present study include the recruitment of a convenience sample. These data were collected cross-sectionally; while the results mirrored those in previous measurement development studies based on the TTM, causal relationships between variables cannot be made. Additionally, most of the sample was White and between 30-49-years-old, thus, the investigated constructs may not fully generalize to other populations. Inadequate representation of older adults (ie, over 65 years), a population shown to be more susceptible to the effects of COVID-19,⁵¹ in particular, may have influenced item endorsement and final measurement development. Future research should aim to establish convergent and divergent validity of the resultant scales prior to their application in subsequent work. Additionally, researchers should aim to investigate generalizability of developed measures to populations other than the majority represented in the current study. Given previous changes with vaccine availability and ongoing discussion of policies surrounding vaccination, this was a time-sensitive study and item-development was largely informed by literature review and expertise with applications of the TTM. To improve generalizability, future research should consider utilizing other qualitative methods, such as cognitive testing or focus groups, to ensure key elements of the vaccination process are included in the construct measures.

Conclusions

The current study developed TTM-based measures of COVID-19 vaccination among adults ages 18-75, including COVID-19 vaccination-related SOC, DCBL, SE, myths, and barriers scales. Similar to measures created for other health behavior changes, these measures can both enhance understanding of behavior change and aid in the development of tailored interventions for different levels of readiness to support vaccination.^52-54 Thus, the measures developed in the current study may assist in the development of future interventions aimed at increasing vaccination rates for COVID-19 and in turn, aid in reducing death and disability caused by this virus.

So What? Implications for Health Promotion Practitioners and Researchers

What is Already Known on This Topic?

Over one million lives have been lost due to COVID-19 in the United States, yet vaccination rates remain relatively low. Limited research exists on the development of assessment methods and interventions efforts related to COVID-19 vaccination grounded in a single, comprehensive model of intentional behavior change.

What Does This Article Add?

This study provides evidence that the TTM may be applied to COVID-19 vaccination as it systematically developed valid and reliable measures of TTM-related constructs, among others, to assess readiness for COVID-19 vaccination.

What are the Implications for Health Promotion Practice or Research?

Utilization of the current measures of distinct constructs may adequately assess readiness to get vaccinated for COVID-19 as they allow for consideration of factors that contribute to one’s decision to be vaccinated. In turn, tailored interventions may be developed to effectively accelerate individuals through the stages of change and toward readiness to get vaccinated. Increased vaccination rates may then decrease potential suffering and death caused by COVID-19.

Ethical Statement

Ethical Approval

All study materials and procedures were approved by the Institutional Review Board at the University of Rhode Island (approval #00000599, reference # 1743575-1).

Consent to Participate

All participants provided written informed consent prior to enrollment in the study.

Footnotes

Author Contributions

All named authors made substantial contributions to this work, drafted/revised it critically, participated sufficiently and have approved of submission and agreed to be accountable for all aspects of this work.

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.

ORCID iDs

Allegra Sacco

Kathleen Monahan

References

Yaqub

Castle-Clarke

Sevdalis

Chataway

. Attitudes to vaccination: a critical review. Soc Sci Med. 2014;112:1-11. doi:10.1016/j.socscimed.2014.04.018.

Report shows 20-year US immunization program spares millions of children from diseases. Centers for Disease Control and Prevention (CDC). April 24, 2014. https://www.cdc.gov/media/releases/2014/p0424-immunization-program.html. Accessed November 20, 2020.

WHO coronavirus (COVID-19) dashboard. World Health Organization. Updated December 16, 2022. https://covid19.who.int/. Accessed December 16, 2022.

Sparks

Lopes

Hamel

Montero

Presiado

Brodie

. KFF COVID-19 vaccine monitor: September 2022. September 30. 2022. https://www.kff.org/coronavirus-covid-19/poll-finding/kff-covid-19-vaccine-monitor-september-2022/. https://www.kff.org/coronavirus-covid-19/poll-finding/kff-covid-19-vaccine-monitor-september-2022/. Accessed December 15, 2022.

Maltz

Sarid

. Attractive flu shot: a behavioral approach to increasing influenza vaccination uptake rates. Med Decis Making. 2020;40(6):774-784. doi:10.1177/0272989X20944190.

Nyhan

Reifler

. Does correcting myths about the flu vaccine work? An experimental evaluation of the effects of corrective information. Vaccine. 2015;33(3):459-464. doi:10.1016/j.vaccine.2014.11.017.

Lawes-Wickwar

Ghio

Tang

, et al. A rapid systematic review of public responses to health messages encouraging vaccination against infectious diseases in a pandemic or epidemic. Vaccines (Basel). 2021;9(2):72. doi:10.3390/vaccines9020072.

Whitehead

French

Caldwell

Letley

Mounier-Jack

. A systematic review of communication interventions for countering vaccine misinformation. Vaccine. 2023;41(5):1018-1034. doi:10.1016/j.vaccine.2022.12.059.

Zimmerman

Nowalk

Raymund

, et al. Tailored interventions to increase influenza vaccination in neighborhood health centers serving the disadvantaged. Am J Public Health. 2003;93(10):1699-1705. doi:10.2105/ajph.93.10.1699.

10.

Campbell

Quintiliani

. Tailored interventions in public health. Am Behav Sci. 2006;49(6):775-793. doi:10.1177/0002764205283807.

11.

Dempsey

Kwan

Wagner

, et al. A values-tailored web-based intervention for new mothers to increase infant vaccine uptake: development and qualitative study. J Med Internet Res. 2020;22(3):e15800. doi:10.2196/15800.

12.

Aveyard

Massey

Parsons

Manaseki

Griffin

. The effect of transtheoretical model based interventions on smoking cessation. Soc Sci Med. 2009;68(3):397-403. doi:10.1016/j.socscimed.2008.10.036.

13.

Kuehn

LaMori

DeMartino

, et al. Assessing barriers to access and equity for COVID-19 vaccination in the US. BMC Publ Health. 2022;22(1):2263. doi:10.1186/s12889-022-14636-1.

14.

Neely

Eldredge

Ersing

Remington

. Vaccine hesitancy and exposure to misinformation: a survey analysis. J Gen Intern Med. 2022;37(1):179-187. doi:10.1007/s11606-021-07171-z.

15.

Ullah

Khan

Tahir

Ahmed

Harapan

. Myths and conspiracy theories on vaccines and COVID-19: potential effect on global vaccine refusals. Vacunas. 2021;22(2):93-97. doi:10.1016/j.vacun.2021.01.001.

16.

Rhodes

Hergenrather

. Exploring hepatitis B vaccination acceptance among young men who have sex with men: facilitators and barriers. Prev Med. 2002;35(2):128-134. doi:10.1006/pmed.2002.1047.

17.

Schmid

Rauber

Betsch

Lidolt

Denker

. Barriers of influenza vaccination intention and behavior - a systematic review of influenza vaccine hesitancy, 2005 - 2016. PLoS One. 2017;12(1):e0170550. doi:10.1371/journal.pone.0170550.

18.

Suhadev

Nyamathi

Swaminathan

, et al. A pilot study on willingness to participate in future preventive HIV vaccine trials. Indian J Med Res. 2006;124(6):631-640.

19.

Alam

Melhim

LKB

Ahmad

Jemmali

. Public atitude towards COVID-19 vaccination: validation of COVID-vaccination attitude scale (C-VAS). J Multidiscip Healthc. 2022;15:941-954. doi:10.2147/JMDH.S353594.

20.

Chu

Liu

. Integrating health behavior theories to predict American's intention to receive a COVID-19 vaccine. Patient Educ Couns. 2021;104(8):1878-1886. doi:10.1016/j.pec.2021.02.031.

21.

Lachance-Grzela

Charbonneau

Jbilou

Dubé

Richard

. Factors related to the intention to get vaccinated against COVID-19 in the Province of New Brunswick, Canada. J Community Health. 2022;47(4):674-679. doi:10.1007/s10900-022-01093-5.

22.

Shiloh

Peleg

Nudelman

. Vaccination against COVID-19: a longitudinal trans-theoretical study to determine factors that predict intentions and behavior. Ann Behav Med. 2022;56(4):357-367. doi:10.1093/abm/kaab101.

23.

Fernandez

Amoyal

Paiva

Prochaska

. Motivation for HPV vaccination among young adult men: validation of TTM decisional balance and self-efficacy constructs. Am J Health Promot. 2016;30(3):163-171. doi:10.4278/ajhp.131108-QUAN-570.

24.

Lipschitz

Fernandez

Larson

, et al. Validation of decisional balance and self-efficacy measures for HPV vaccination in college women. Am J Health Promot. 2013;27(5):299-307. doi:10.4278/ajhp.110606-QUAN-240.

25.

Paiva

Lipschitz

Fernandez

Redding

Prochaska

. Evaluation of the acceptability and feasibility of a computer-tailored intervention to increase human papillomavirus vaccination among young adult women. J Am Coll Health. 2014;62(1):32-38. doi:10.1080/07448481.2013.843534.

26.

Burditt

Robbins

Paiva

Velicer

Koblin

Kessler

. Motivation for blood donation among African Americans: developing measures for stage of change, decisional balance, and self-efficacy constructs. J Behav Med. 2009;32(5):429-442. doi:10.1007/s10865-009-9214-7.

27.

Prochaska

Velicer

Rossi

, et al. Stages of change and decisional balance for 12 problem behaviors. Health Psychol. 1994;13(1):39-46. doi:10.1037/0278-6133.13.1.39.

28.

Waterman

Robbins

Paiva

, et al. Measuring kidney patients’ motivation to pursue living donor kidney transplant: development of stage of change, decisional balance and self-efficacy measures. J Health Psychol. 2015;20(2):210-221. doi:10.1177/1359105313501707.

29.

Geoghegan

O'Callaghan

Offit

. Vaccine safety: myths and misinformation. Front Microbiol. 2020;11:372. doi:10.3389/fmicb.2020.00372.

30.

MacIntyre

Leask

. Immunization myths and realities: responding to arguments against immunization. J Paediatr Child Health. 2003;39:487-491. doi:10.1016/j.vaccine.2007.11.001.

31.

Pluviano

Watt

Della Sala

. Misinformation lingers in memory: failure of three pro-vaccination strategies. PLoS One. 2017;12(7):e0181640. doi:10.1371/journal.pone.0181640.

32.

Rubin

. Covid-19 vaccine makers plan for annual boosters, but it’s not clear they’ll be needed. JAMA. 2021;326(22):2247. doi:10.1001/jama.2021.21291.

33.

Burckhardt

Dennehy

Poon

LLM

Saif

Enquist

. Are COVID-19 vaccine boosters needed? The science behind boosters. J Virol. 2022;96(3):e0197321. doi:10.1128/JVI.01973-21.

34.

Guadagnoli

Velicer

. Relation of sample size to the stability of component patterns. Psychol Bull. 1988;103(2):265-275. doi:10.1037/0033-2909.103.2.265.

35.

Redding

Maddock

Rossi

. Measurement of theoretical constructs for health behavior. Calif Jour Health Promot. 2006;4(1):83-101. doi:10.32398/cjhp.v4i1.736.

36.

Enders

. Using the expectation maximization algorithm to estimate coefficient alpha for scales with item-level missing data. Psychol Methods. 2003;8(3):322-337. doi:10.1037/1082-989x.8.3.322.

37.

Bentler

. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Modeling. 1999;6(1):1-55. doi:10.1080/10705519909540118.

38.

Prochaska

Velicer

. The transtheoretical model of health behavior change. Am J Health Promot. 1997;12(1):38-48. doi:10.4278/0890-1171-12.1.38.

39.

Hoeppner

Redding

Rossi

Pallonen

Prochaska

Velicer

. Factor structure of decisional balance and temptations scales for smoking: cross-validation in urban female African-American adolescents. Int J Behav Med. 2012;19(2):217-227. doi:10.1007/s12529-011-9145-x.

40.

Kruzan

Whitlock

Hasking

. Development and initial validation of scales to assess decisional balance (NSSI-DB), processes of change (NSSI-POC), and self-efficacy (NSSI-SE) in a population of young adults engaging in nonsuicidal self-injury. Psychol Assess. 2020;32(7):635-648. doi:10.1037/pas0000821.

41.

Bolsen

Palm

. Politicization and COVID-19 vaccine resistance in the U.S. Prog Mol Biol Transl Sci. 2022;188(1):81-100. doi:10.1016/bs.pmbts.2021.10.002.

42.

Bogart

Ojikutu

Tyagi

, et al. COVID-19 related medical mistrust, health impacts, and potential vaccine hesitancy among Black Americans living with HIV. J Acquir Immune Defic Syndr. 2021;86(2):200-207. doi:10.1097/QAI.0000000000002570.

43.

Trent

Seale

Chughtai

Salmon

MacIntyre

. Trust in government, intention to vaccinate and COVID-19 vaccine hesitancy: a comparative survey of five large cities in the United States, United Kingdom, and Australia. Vaccine. 2022;40(17):2498-2505. doi:10.1016/j.vaccine.2021.06.048.

44.

Nong

Raj

Creary

Kardia

SLR

Platt

. Patient-reported experiences of discrimination in the US health care system. JAMA Netw Open. 2020;3(12):e2029650. doi:10.1001/jamanetworkopen.2020.29650.

45.

Rivenbark

Ichou

. Discrimination in healthcare as a barrier to care: experiences of socially disadvantaged populations in France from a nationally representative survey. BMC Publ Health. 2020;20(1):31. doi:10.1186/s12889-019-8124-z.

46.

Williams

Rucker

. Understanding and addressing racial disparities in health care. Health Care Financ Rev. 2000;21(4):75-90.

47.

Nye

Blanco

. Characteristics of Homebound Older Adults: Potential Barriers to Accessing the COVID-19 Vaccine. Office of the Assistant Secretary for Planning and Evaluation. U.S. Department of Health and Human Services; 2021.

48.

Kenny

Kaniskan

McCoach

. The performance of RMSEA in models with small degrees of freedom. Sociol Methods Res. 2014;44(3):486-507. doi:10.1177/0049124114543236.

49.

Lai

Green

. The problem with having two watches: assessment of fit when RMSEA and CFI disagree. Multivar Behav Res. 2016;51(2–3):220-239. doi:10.1080/00273171.2015.1134306.

50.

Moshagen

. The model size effect in SEM: inflated goodness-of-fit statistics are due to the size of the covariance matrix. Struct Equ Modeling. 2012;19(1):86-98. doi:10.1080/10705511.2012.634724.

51.

Older adults risks and vaccine information. Centers for Disease Control and Prevention (CDC). August 4, 2021. https://www.cdc.gov/aging/covid19/covid19-older-adults.html#:∼:text=Increased_Risk_of_Severe_Illness_from_COVID%2D19,-Older_adults_are&text=The_risk_increases_for_people,having_certain_underlying_medical_conditions. Published August 4, 2021. Accessed December 15, 2022.

52.

Chen

Shen

Sun

. Effectiveness of a transtheoretical model-based intervention to improve blood pressure control of hypertensive patients in China: a clustered randomized controlled trial. Front Public Health. 2022;9:760421. doi:10.3389/fpubh.2021.760421.

53.

Fried

Paiva

Redding

, et al. Effect of the stamp (sharing and talking about my preferences) intervention on completing multiple advance care planning activities in ambulatory care. Ann Intern Med. 2021;174(11):1519-1527. doi:10.7326/m21-1007.

54.

Romain

Bortolon

Gourlan

, et al. Matched or nonmatched interventions based on the transtheoretical model to promote physical activity. A meta-analysis of randomized controlled trials. J Sport Health Sci. 2018;7(1):50-57. doi:10.1016/j.jshs.2016.10.007.