Assessing licensed nurses COVID-19 vaccine attitudes and intentions: a cross-sectional survey in the state of Hawaii

ABSTRACT

Nurses are the largest single occupation of health care providers and at greatest risk for exposure to and acquisition of Coronavirus Disease 2019 (COVID-19). In December 2020, nurses in Hawaii were recruited for an online survey that measured perceived risk/threat of COVID-19, vaccine attitudes, and perceived safety of COVID-19 vaccines, as well as level of intention: primary, secondary (i.e., delayed), or no intention to vaccinate. The final sample consisted of 423 nurses. Participants were primarily Asian (27.9%) and White (45.2%). The majority were 18–50 years (65.5%) and female (87.0%), held an RN license (91.7%), and identified as a staff nurse (57.7%) in the hospital setting (56.7%). Among participants, 52.3% indicated primary intention, 27.9% secondary intention, and 19.9% no intention to vaccinate. The strongest predictors of any level of intention were greater positive attitudes toward COVID-19 vaccination and lower concerns related to COVID-19 vaccine safety. Findings can guide interventions to support vaccine acceptance for those who initially decline vaccination.

KEYWORDS:

• Coronavirus disease
• COVID-19
• vaccine
• vaccination
• nurse
• Hawaii

Introduction

Nurses have been identified as the largest single occupation of health care providers who are at greatest risk for exposure to and acquisition of Coronavirus Disease 2019 (COVID-19).¹ In the spring of 2020, of the hospitalized health care providers diagnosed with COVID-19 in the United States (US), over one third were in nursing-related occupations, the mean age was 49 years, and the majority were female.² Nurses are likely at greatest risk due to extended patient contact in health care settings, plus the societal norm of women/nurse as caregivers within the community and family, leading to greater amassed exposure to COVID-19.

The BNT162B2 vaccine, developed by BioNtech and Pfizer, and the mRNA-1273 vaccine, developed by Moderna (termed COVID-19 vaccines), both received US Food and Drug Administration (FDA) Emergency Use Authorization (EUA) in December 2020.^3,4 Immediately after, the Advisory Committee on Immunization Practices (ACIP)/Center of Disease Control and Prevention (CDC) recommended that health care personnel and residents of long-term care facilities are offered the COVID-19 vaccination in the first phase of vaccine distribution.^5,6 Since that time, in 2021, an adenovirus vector vaccine developed by Janssen/Johnson & Johnson has received FDA EUA.⁷ Reports for all COVID-19 vaccines document initial safety and efficacy.^7–9

Despite reports of safety and efficacy for these vaccines and pre-dating the current pandemic, there have been growing concerns related to general vaccine hesitancy (described as a range of feelings/perspectives toward vaccines from deep skepticism about vaccine efficacy and safety to more mild concerns) in the US^10,11 and abroad. The World Health Organization (2019) recently deemed vaccine hesitancy as a major global threat.¹² National and international media coverage of ‘rushed’ COVID-19 vaccine development and the US federal government use of the term ‘warp speed’ for vaccine discovery efforts may have contributed to unwarranted skepticism and a belief that vaccine clinical trials have been cutting corners and sacrificing safety, resulting in the potential for particularly low confidence in the COVID-19 vaccines.¹³

Hesitancy has been evident among health care providers as well. Data collected during the vaccine development phase (prior to December 2020) indicated that one in three health care workers may be hesitant to receive a COVID-19 vaccine,^14,15 and persons working in health care may have lower intention to vaccinate.¹⁶ To date, the vast majority of reports describing health care provider vaccine intentions were conducted during vaccine development.^14–18 Recent data of post-vaccine approval/availability highlight ongoing hesitation, with three in ten health care providers still undecided or not planning on receiving a COVID-19 vaccine.¹⁹ Ongoing assessments of vaccine attitudes and intentions are needed to capture necessary information for vaccine communication interventions over time.

In the State of Hawaii, similar to other US states, nurses represent the largest proportion of the health care workforce.^20,21 As nurses are offered vaccination they may need ongoing educational and/or behavioral support in making informed vaccine decisions. Having high confidence in COVID-19 vaccine safety and efficacy, as well as positive attitudes toward vaccination among nurses will be key in 1) ensuring that the largest sector of our health care workforce remains healthy, 2) role modeling positive vaccine behaviors, and 3) communicating positive vaccine messages for the larger community.

A baseline assessment of vaccine hesitancy and its correlates is critical to shaping current and ongoing vaccine program planning for nurses. It is also critical to look beyond vaccine hesitancy to potential facilitators for, and/or barriers to, COVID-19 vaccination so that interventions can be tailored to address them. The purpose of this study, conducted at the same time as the first two COVID-19 vaccines were approved, was to assess the intentions of licensed nurses in the State of Hawaii to obtain a COVID-19 vaccine and identify factors that are associated with nurses’ intention to vaccinate.

Methods

Participants and recruitment

In December 2020, we conducted an electronic survey of nurses in the State of Hawaii. Data were collected during the same time frame as the FDA approval and ACIP/CDC recommendation for COVID-19 vaccines; however, vaccination had yet to be offered or distributed nationally. Data were collected and managed using a secure Research Electronic Data Capture (REDCap) web-based survey and database hosted at the primary investigator’s institution. Recruitment e-mails, using Dillman’s Tailored Design Method approach,^22,23 were sent to the Hawaii State Center for Nursing e-mail list (approximately 2000 recipients), plus recruitment flyers were distributed via the Center’s social media pages (Twitter and Facebook). These e-mails/flyers included a link to the study webpage for further study information, informed consent procedures, and assessment of eligibility. If a potential participant was eligible and provided consent, they were directed to continue the 10-minute study survey. Participation was voluntary and remuneration was not offered. Individuals were eligible to participate if they were 1) currently licensed as a nurse at any level in the State of Hawaii (e.g., Licensed Practical Nurse, Registered Nurse, Advanced Practice Nurse), 2) currently employed or actively seeking employment in the State of Hawaii, 3) able to read and understand English, and 4) aged 18 years or older. This study received approval from the University of Hawaii at Manoa Institutional Review Board as exempt.

Measures

In addition to demographic information, the study team collected data on participants’ intention to vaccinate for COVID-19, personal beliefs regarding risk/threat of COVID-19 infection, and COVID-19 vaccine attitudes/perceived safety. We adapted scales, reported as valid and reliable,¹⁶ for this survey. All scales utilized were 5-point Likert scales (1 = strongly disagree to 5 = strongly agree)

Demographics

Age, gender, race/ethnicity, title, setting, and place of employment were collected. Age was categorized as 18–50 years and 51 years or older. Age was categorized based on data indicating that the risk of dying from COVID-19 for those aged ≥ 50 years is at least 3 times higher as compared to those aged 40–49 years and 30 times higher as compared to those aged ≤ 30 years.²⁴ Based on its distribution, race was categorized as White, Asian, or other. Other included Native Hawaiian and other Pacific Islanders (NHOPI), American Indian or Alaska Native (AI/AN), Black, and mixed race. Title was categorized as staff nurse and other (e.g., advanced practice nurses, administrators, educators, and researchers). Setting was categorized as hospital-based or other community-based locations. Place of employment was determined by zip code and categorized to Oahu versus other islands. Oahu can be considered the most urban area in Hawaii.

Intention to vaccinate

Intention to vaccinate was assessed using the following two questions: “Will you get a COVID-19 vaccine as soon as it becomes available/offered to you?” and “Would you get a COVID-19 vaccine within a year after it became available/offered for you?” The second question was provided only to those participants who responded ‘no’ to the first question. Vaccine intention was categorized into three groups: 1) primary intention if a participant responded ‘yes’ to the first question; 2) secondary intention if a participant responded ‘yes’ to the second question, and 3) no intention if a participant responded ‘no’ to the second question.

Risk/threat of COVID-19 infection

We measured personal beliefs regarding COVID-19 infection risk and threat using two scales: 1) COVID-19-related worry and 2) perceived severity of COVID-19. COVID-19-related worry was measured by two items: “I am scared about getting COVID-19” and “The possibility of getting infected in the future with COVID-19 concerns me.” These showed excellent reliability (Cronbach’s alpha = 0.91) in this study. These two items were averaged to measure overall COVID-19-related worry. Perceived severity of COVID-19 was measured using two items: “I am afraid that I may die if I contract COVID-19 (or if I contract COVID-19 again)” and “I am at greater risk of dying if I contract COVID-19 because of my general health.” They demonstrated good reliability in this study (Cronbach’s alpha = 0.84). We averaged these to measure overall perceived severity of COVID-19.

We also queried, in single-item questions, if the participant had ever been tested for COVID-19, and what those results were; if they had a health concern that would make infection with COVID-19 more severe; if they had a family member or friend who had been infected with COVID-19; if they provided care for a patient with an active COVID-19 infection; and if they believed COVID-19 was a major problem for their local community and/or primary workplace setting.

COVID-19 vaccine attitudes/perceived safety

Two scales were used to measure COVID-19 vaccine attitudes/perceived safety: 1) general COVID-19 vaccine attitude and 2) lack of perceived safety. Four items were used to measure general COVID-19 vaccine attitude. The four items were “The COVID-19 vaccine will be important for my health,” “Getting a COVID-19 vaccine would be a good way to protect me from COVID-19,” “Getting the COVID-19 vaccine will be important for the health of others in my local community,” and “Getting COVID-19 vaccine will be important for the health of others at my primary workplace setting.” They showed excellent reliability (Cronbach’s alpha = 0.97) and were averaged to measure general COVID vaccine attitude. Lack of perceived safety was measured by three items: “I am concerned that the COVID-19 vaccine has not been around long enough to be sure it is safe,” “I am concerned about serious side effects of the COVID-19 vaccine,” and “I am concerned that the COVID-19 vaccine might cause lasting health problems for me.” They showed excellent reliability (Cronbach’s alpha = 0.97) and were averaged to measure lack of perceived safety.

Analysis

Descriptive statistics were reported by frequencies and percentages for categorical variables and means and standard deviations for continuous variables. Bivariate associations between intention to vaccinate and each of the variables were evaluated using a univariate multinomial logistic regression on intention to vaccinate, treating secondary intention as the reference. Next, we conducted a two-step hierarchical multivariable multinomial regression analysis. Demographic covariates were included in the first step. In the second step, we added COVID-19-related behavioral variables to the model. We computed odds ratios (ORs) and 95% confidence intervals (CIs) to evaluate association with intention to vaccinate. For a sensitivity analysis, we conducted a multinomial regression by backward selection method using p ≥ 0.05 as the cutoff in dropping a variable. All analyses were implemented in SAS 9.4 (SAS Institute, Cary NC) and p ≤ 0.05 was considered statistically significant.

Results

Sample

A total of 606 participants opened the survey, 602 provided consent, and 550 met eligibility criteria. Of these, 509 nurses engaged in the survey to some degree, but 86 did not respond to the vaccine intention items. A total of 423 nurses completed the survey and responded to the vaccine intention items. There was no significant difference in characteristics between people who did or did not complete the survey in full.

The final analytic sample included 423 participants. The majority of participants were 18–50 years (65.5%), identified as female (87.0%), reported their highest level of license as an RN (91.7%), identified as a staff nurse (57.7%), worked in the hospital setting (56.7%), and worked on Oahu (72.3%). Race was identified as 45.2% White, 27.9% Asian, and 26.9% other. Other included 10.6% NHOPI and 16.3% mixed or others (e.g., < 2.5% Black and < 2.5% AI/AN). Only 6% of the sample indicated ethnicity as Hispanic. Among participants, 52.3% indicated primary (immediate), 27.9% secondary intention (within a year), and 19.9% no intention to receive the COVID-19 vaccine (see Table 1 for further details).

Table 1. Bivariate association between demographics and intention to vaccinate

	Total	Intention			Bivariate Association
Demographics	N (%)	Primary (n = 221, 52.3%)	Secondary (n = 118, 27.9%)	No (n = 84, 19.9%)	Primary vs. Secondary	No vs. Secondary
Age
18–50 yrs.	277 (65.5%)	125 (56.6%)	88 (74.6%)	64 (76.2%)	Ref	Ref
51+ yrs.	146 (34.5%)	96 (43.4%)	30 (25.4%)	20 (23.8%)	2.25 [1.38, 3.69]**	0.92 [0.48, 1.76]
Race
White	191 (45.2%)	120 (54.3%)	41 (34.7%)	30 (35.7%)	Ref	Ref
Asian	118 (27.9%)	56 (25.3%)	42 (35.6%)	20 (23.8%)	0.46 [0.27, 0.78]**	0.65 [0.32, 1.33]
NHOPI/Mixed/Other	114 (27.0%)	45 (20.4%)	35 (29.7%)	34 (40.5%)	0.44 [0.25, 0.77]**	1.33 [0.68, 2.59]
Title
Other	179 (42.3%)	115 (52.0%)	39 (33.1%)	25 (29.8%)	Ref	Ref
Staff	244 (57.7%)	106 (48.0%)	79 (66.9%)	59 (70.2%)	0.46 [0.29, 0.73]***	1.17 [0.64, 2.13]
Setting
Other	183 (43.3%)	108 (48.9%)	39 (33.1%)	36 (42.9%)	Ref	Ref
Hospital	240 (56.7%)	113 (51.1%)	79 (66.9%)	48 (57.1%)	0.52 [0.32, 0.82]**	0.66 [0.37, 1.17]
Place of Employment
Other	117 (27.7%)	60 (27.1%)	23 (19.5%)	34 (40.5%)	Ref	Ref
Oahu	306 (72.3%)	161 (72.9%)	95 (80.5%)	50 (59.5%)	0.65 [0.38, 1.12]	0.36 [0.19, 0.67]**

OR = Odds Ratio. CI = Confidence Interval. Ref = Reference. Column percentage. Bivariate association was explored using multinomial logistic regression model with each demographic variable as a predictor and intention as the outcome variable. +P < 0.10; *P < 0.05; **P < 0.01; ***P < 0.001.

Bivariate analysis

In bivariate analyses, all the demographic characteristics, age, race, title, setting, and place of employment, were associated with intention to vaccinate (see Table 1). Additionally, among the COVID-19 related predictors, ever been tested for COVID-19, perceived severity, worry, thinking that COVID-19 infection is a major problem in local community and workplace, general COVID-19 vaccine attitude, and lack of perceived safety, were associated with intention to vaccinate (see Table 2). The mean score on the perceived severity of COVID-19 scale for those who indicated primary intention (M = 2.8, SD = 1.1) and secondary intention (M = 2.8, SD = 1.1) was greater than that who indicated no intention to vaccinate (M = 2.0, SD = 1.0). The mean scores on the COVID-19 related worry scale for those indicating primary intention (M = 3.9, SD = 0.9) and secondary intention (M = 3.8, SD = 1.0) were greater than those who indicated no intention to vaccinate (M = 2.9, SD = 1.4). Mean general COVID-19 vaccine attitudes (favorable) decreased from primary (M = 4.7, SD = 0.4) to secondary (M = 3.7, 0.8) to the no (M = 2.1, SD = 0.9) intention groups. Lastly, the lack of perceived safety increased from primary (M = 3.0, SD = 1.1) to secondary (M = 4.4, SD = 0.7) to the no (M = 4.7, SD = 0.6) intention groups.

Table 2. Bivariate Association between COVID-Related Predictors and Intention to Vaccinate

	Total	Intention			Bivariate Association
Predictors	N (%)	Primary (n = 221, 52.3%)	Secondary (n = 118, 27.9%)	No (n = 84, 19.9%)	Primary vs. Secondary	No vs. Secondary
Tested for COVID-19
No	141 (33.3%)	68 (30.8%)	35 (29.7%)	38 (45.2%)	Ref	Ref
Yes	282 (66.7%)	153 (69.2%)	83 (70.3%)	46 (54.8%)	0.95 [0.58, 1.55]	0.51 [0.29, 0.92]*
The Result of COVID-19 Test ^a
Tested Positive	8 (2.8%)	3 (2.0%)	3 (3.6%)	2 (4.3%)
Tested Negative	272 (96.5%)	149 (97.4%)	79 (95.2%)	44 (95.7%)
Still Awaiting the Results	2 (0.7%)	1 (0.7%)	1 (1.2%)	0 (0%)
Health concern make infection more severe
No	304 (71.9%)	153 (69.2%)	84 (71.2%)	67 (79.8%)	Ref	Ref
Yes	119 (28.1%)	68 (30.8%)	34 (28.8%)	17 (20.2%)	1.10 [0.67, 1.79]	0.63 [0.32, 1.22]
Mean Perceived Severity (2 items)	2.6 ± 1.1	2.8 ± 1.1	2.8 ± 1.1	2.0 ± 1.0	1.02 [0.84, 1.24]	0.54 [0.41, 0.71]***
Mean Worry (2 items)	3.7 ± 1.1	3.9 ± 0.9	3.8 ± 1.0	2.9 ± 1.4	1.07 [0.85, 1.34]	0.49 [0.38, 0.63]***
Had a family member or close friend become infected with COVID-19
No	279 (66.0%)	146 (66.1%)	80 (67.8%)	53 (63.1%)	Ref	Ref
Yes	144 (34.0%)	75 (33.9%)	38 (32.2%)	31 (36.9%)	1.08 [0.67, 1.74]	1.23 [0.68, 2.22]
Provided care for a patient who has tested positive for COVID-19
No	210 (49.6%)	118 (53.4%)	52 (44.1%)	40 (47.6%)	Ref	Ref
Yes	213 (50.4%)	103 (46.6%)	66 (55.9%)	44 (52.4%)	0.69 [0.44, 1.08]	0.87 [0.49, 1.52]
Think COVID-19 infection is a major problem in your local community
No	129 (30.5%)	55 (24.9%)	33 (28.0%)	41 (48.8%)	Ref	Ref
Yes	294 (69.5%)	166 (75.1%)	85 (72.0%)	43 (51.2%)	1.17 [0.71, 1.94]	0.41 [0.23, 0.73]**
Think COVID-19 infection is a major problem in your primary work place setting?
No	211 (49.9%)	105 (47.5%)	53 (44.9%)	53 (63.1%)	Ref	Ref
Yes	212 (50.1%)	116 (52.5%)	65 (55.1%)	31 (36.9%)	0.90 [0.58, 1.41]	0.48 [0.27, 0.85]*
Mean General COVID-19 Vaccine Attitude (4 items)	3.9 ± 1.2	4.7 ± 0.4	3.7 ± 0.8	2.1 ± 0.9	10.93 [6.58, 18.17]***	0.11 [0.06, 0.20]***
Mean Lack of Perceived Safety (3 items)	3.7 ± 1.2	3.0 ± 1.1	4.4 ± 0.7	4.7 ± 0.6	0.21 [0.15, 0.30]***	2.19 [1.35, 3.54]**

OR = Odds Ratio. CI = Confidence Interval. Ref = Reference. ^a Among those who received COVID-19 test. Measures were averaged over items, indicating 1 = Strongly disagree to 5 = Strongly agree. Column percentage. Bivariate association was explored using a multinomial logistic model with each demographic variable as a predictor and intention as the outcome variable. +P < 0.10; *P < 0.05; **P < 0.01; ***P < 0.001.

Multivariate analysis

Hierarchical multinomial regressions were conducted (see Table 3). In Step 1, race, setting, and place of employment were significantly associated with vaccine intention. Compared to White, other racial groups were less likely to show primary intention as compared to secondary intention to vaccinate (Asian: OR = 0.51, 95% CI = 0.29–0.91; Other: OR = 0.48, 95% CI = 0.27–0.86). Nurses who work at a hospital were less likely to show primary intention as compared to secondary intention, compared to nurses who work at other settings (OR = 0.57, 95% CI = 0.35–0.94). However, no significant differences were found in demographics between nurses who identified secondary intention and nurses who identified no intention to vaccinate, except location (island) of employment. Compared to nurses working on other islands, nurses who work on Oahu were less likely to be in the no intention group than the secondary intention group (OR = 0.38, 95% CI = 0.20–0.73).

Table 3. Stepwise multivariable multinomial logistic regression

	Intention, OR (95% CI)
	Step 1: Demographics		Step 2: Demographics + Predictors
Variable	Primary vs. Secondary	No vs. Secondary	Primary vs. Secondary	No vs. Secondary
Age
18–50 yrs.	Ref	Ref	Ref	Ref
51+ yrs.	1.63 [0.95, 2.79]+	0.95 [0.47, 1.93]	1.15 [0.50, 2.67]	0.82 [0.26, 2.59]
Race
White	Ref	Ref	Ref	Ref
Asian	0.51 [0.29, 0,91]*	0.71 [0.34, 1.51]	0.88 [0.36, 2.12]	1.66 [0.47, 5.82]
NHOPI/Mixed/Other	0.48 [0.27, 0.86]*	1.36 [0.68, 2.71]	0.89 [0.37, 2.15]	1.23 [0.41, 3.71]
Title
Other	Ref	Ref	Ref	Ref
Staff	0.62 [0.37, 1.04]+	1.51 [0.78, 2.93]	1.02 [0.49, 2.15]	2.93 [0.96, 8.94]+
Setting
Other	Ref	Ref	Ref	Ref
Hospital	0.57 [0.35, 0.94]*	0.70 [0.38, 1.30]	0.81 [0.39, 1.70]	0.51 [0.19, 1.38]
Place of Employment
Other	Ref	Ref	Ref	Ref
Oahu	0.85 [0.48, 1.52]	0.38 [0.20, 0.73]**	0.64 [0.26, 1.55]	0.46 [0.16, 1.32]
Tested for COVID-19
No			Ref	Ref
Yes			1.16 [0.54, 2.52]	0.46 [0.18, 1.21]
Health concern make infection more severe
No			Ref	Ref
Yes			1.07 [0.45, 2.57]	0.72 [0.22, 2.33]
Mean Perceived Severity (2 items)			1.09 [0.69, 1.71]	1.10 [0.59, 2.04]
Mean Worry (2 items)			1.07 [0.64, 1.79]	0.61 [0.36, 1.05]+
Had a family member or close friend become infected with COVID-19
No			Ref	Ref
Yes			1.60 [0.76, 3.35]	0.86 [0.35, 2.13]
Provided care for a patient who has tested positive for COVID-19
No			Ref	Ref
Yes			0.95 [0.45, 1.98]	0.81 [0.29, 2.21]
Think COVID-19 infection is a major problem in your local community
No			Ref	Ref
Yes			1.15 [0.48, 2.73]	0.81 [0.27, 2.44]
Think COVID-19 infection is a major problem in your primary work place setting?
No			Ref	Ref
Yes			1.52 [0.68, 3.42]	1.73 [0.59, 5.09]
Mean General COVID-19 Vaccine Attitude (4 items)			7.36 [3.88, 13.94]***	0.11 [0.06, 0.23]***
Mean Lack of Perceived Safety (3 items)			0.22 [0.14, 0.35]***	1.19 [0.59, 2.41]

OR = Odds Ratio. CI = Confidence Interval. Ref = Reference. +P < 0.10; *P < 0.05; **P < 0.01; ***P < 0.001. Mean perceive severity, worry, general COVID vaccine attitude, and lack of perceived safety were scored to indicate 1 = strongly disagree to 5 = strongly agree.

In Step 2 of the hierarchical multinomial model, with all variables included, only the mean general COVID-19 vaccine attitude and mean lack of perceived safety were significant. Compared to secondary intention, the odds for indicating primary intention would be 7.36 times more (95% CI = 3.88–13.94) but the odds for indicating no intention would be decreased by a factor of 0.11 (95% CI = 0.06–0.23) for one-unit increase in general vaccine attitude. The odds for indicating primary intention over secondary intention would be decreased by a factor of 0.22 (95% CI = 0.14–0.35) for one-unit increase in lack of perceived safety (see Table 3).

As a sensitivity analysis, we also conducted reduced model. All variables were removed using backward selection method with the criterion of p-value < 0.05. Only mean general COVID-19 vaccine attitude and mean lack of perceived safety were left in the final model, and their significances and magnitudes were similar to those in the step 2 model in Table 3 (results are not shown).

Discussion

Our data indicate that nurses practicing in the State of Hawaii have positive COVID-19 vaccine attitudes and high levels of intention to be vaccinated. Cumulatively, 80% of study participants indicated intention to be vaccinated, with over 50% planning on being vaccinated immediately (primary intention), and nearly 30% planning on being vaccinated within the next 12 months (secondary intention), while approximately 20% had no intention to be vaccinated at the time of the survey. The strongest predictors of vaccination among nurses were greater positive attitudes toward COVID-19 vaccination and lower concerns related to COVID-19 vaccine safety.

Published research on vaccine acceptability among health care providers, largely conducted in the spring of 2020, noted similar findings national and internationally. Survey results nationally, among medical students in Michigan and nursing home staff in Indiana, indicate that approximately 70–80% of health care providers would vaccinate at some point,^25,26 with increasing rates of intention over time (e.g., among staff in Indiana 45% indicated immediate intention and an additional 44% indicated secondary intention).²⁶ Internationally, rates of intention to vaccinate range from 63% to 78%.^17,18,27,28 Recently, in March 2021, one report highlighted that most health care providers (who were doctors or nurses) reported receipt of the COVID-19 vaccine (68%).¹⁹ Similar to our study, perceived vaccine safety continues to be the strongest predictors of vaccine intention, which seems to be a common factor despite type of health care provider or nationality.^{17,19,25,26,28} Of note, among previous studies that included nurses, nurses reported lower rates of vaccine intention as compared to other health care providers.^16–18,27

In the general public, nationally and internationally, rates of vaccine intention range from approximately 60–85%.^{15,16,29–31} Similarly, barriers to vaccination are associated with safety concerns²⁹ and negative vaccine attitudes,³¹ whereas facilitators are associated with health care provider recommendation.^16,32,33 Disparities associated with race are common, where those who identify as Black report lower rates of vaccine intention.^{19,29,31–33} In our study (in the first step of the multinomial regression), nurses who were White were more likely to report primary vaccine intention. Race clearly needs to be explored further in order to develop effective and culturally informed vaccine interventions and communication strategies.

Understanding factors associated with primary and secondary intention to vaccinate can assist in vaccine distribution planning. Health care providers are a priority population for COVID-19 vaccination efforts; however, due to limited vaccine supply and delays in distribution, it is critical to assure availability of vaccine for those indicating primary intention and then subsequently secondary intention. As supplies are replenished, even after the first several waves of vaccine distribution, there are ongoing opportunities for continued vaccine safety and efficacy education. Increasing vaccination rates over time, with the goal of herd immunity, will take both sustained educational efforts and continued opportunities to vaccinate individuals who were at first hesitant or whom initially had no intention.

It is imperative to develop campaigns to promote COVID-19 vaccine confidence and acceptance among nurses. Our study indicates that interventions should target ways to promote positive vaccine attitudes and provide ongoing education about vaccine safety. Increasing rates of vaccination will reduce COVID-19 related morbidity and mortality among nurses. Additionally, the public has repeatedly noted nurses to be the most honest and ethical profession in the US, holding the profession in high esteem;³⁴ therefore, nurses can lead vaccine promotion efforts for the general population as effective vaccine recommenders. It has been repeatedly demonstrated that a health provider’s recommendation strongly influences decision-making around vaccines,^35–38 including vaccinating for COVID-19.^16,32 The next step is to develop educational programs specifically to train nurses to provide strong vaccine recommendations as well as increase nurses' capability to dispel vaccine misinformation.^39,40

Findings from this study provide rich insights into predictors for both primary and secondary intention to vaccine among nurses in the State of Hawaii at the start of the US national COVID-19 vaccine distribution effort, December 2020. However, findings should be viewed in terms of study limitations. Our findings are potentially impacted by lower case rates of COVID-19 in Hawaii. While the total of COVID-19 cases for Hawaii reached approximately 24,000 in January 2021, there were a total of 24.5 million cases nationwide. Hawaii case rate as of January 2021 was 1,729 per 100,000, while some hard-hit states such as North Dakota had 12,610 per 100,000.⁴¹ Previous studies have noted that greater perceived threat of COVID-19, which could be correlated with higher case rates, is associated with vaccine intention among the general population.^15,16,30,32 However, perceived threat was not a significant predictor of intention among nurses in Hawaii. Generalization of our findings is also limited by the unique culture and racial composition among nurses in Hawaii in general, as well as in our sample. The approximate racial/ethnic identification of nurses in Hawaii is primarily Asian (48%) and White (28%), with only 1% identifying as Black. Additionally, 20% identify as 2 or more races. Nearly 8% identify as Hispanic/Latinx, and approximately 11% identify as any part Native Hawaiian Ancestry.²⁰ Our study had a greater proportion of participants who identified as White, but similarly, the majority were identified as White or Asian (see Table 4). Finally, our survey collected self-reported data and was cross-sectional, collecting data at one-time point and only at the beginning of vaccine rollout. Future studies should examine uptake of the COVID-19 vaccine over time among all health care workers and include greater proportion of male providers as well as female providers.

Table 4. Nursing characteristics: study sample and state of Hawaii data

Nurse Characteristics	State of Hawaii Data^a %	Study Sample %
Gender Female Male Other/prefer not to answer	86.16 13.22 0.6	85.88 12.77 1.65
Race White Asian Other	28.13 47.80 24.38	44.91 28.13 26.95
Hispanic/Latinx No Yes Prefer not to answer	92.25 7.75 0	90.31 6.38 3.1
License LPN/LVN RN APRN	6.94 87.22 5.84	1.65 82.27 16.08
Work Title Staff Nurse Other	66.9 33.1	53.43 46.57
Work Setting Hospital Community	48.03 51.88	56.26 43.74

Cells do not add to 100% due to rounding.

^aSource: Hawaii State Center for Nursing. Hawaii’s nursing workforce supply, 2019²⁰.

Conclusion

To our knowledge, this study is the first state-based sample of nurses examining factors associated with COVID-19 vaccine intention. The study was novel as it captured time-frame levels of vaccine intention (primary, secondary, or no intention to vaccinate). Findings can guide ongoing vaccine delivery and interventions to support vaccine acceptance among nurses. Educational programs should focus on 1) improving vaccine attitudes and reducing concerns related to vaccine safety and 2) developing nurses’ capacity to be leaders in delivering effective vaccine recommendations to the communities they serve.

Abbreviations

COVID-19	=	Coronavirus Disease 2019
US	=	United States
CDC	=	Centers for Disease Control and Prevention
ACIP	=	Advisory Committee on Immunization Practices
ORs	=	Odds ratios
CI	=	Confidence intervals
M	=	Mean
SD	=	Standard deviation
RN	=	Registered Nurse
FDA	=	Food and Drug Administration
NHOPI	=	Native Hawaiian/Pacific Islanders
AI/AN	=	American Indian or Alaska Native

Disclosure of potential conflicts of interest

Greg Zimet has received consulting fees from Sanofi Pasteur for work on the Adolescent Immunization Project, from Merck for work on HPV vaccination, and from Moderna for work on SARS-CoV-2 vaccination (all outside of the current work), and through Indiana University, has also received investigator-initiated grant funding from Merck related to HPV and adolescent vaccination. Holly Fontenot has received investigator-initiated grant funding from Merck related to adolescent vaccination. All other authors declare that they have no competing interests (financial or non-financial).

IRB

This study was approved by the University of Hawaii at Manoa Institutional Review Board (IRB) on November 20, 2020 (#2020-00862), as exempt. The authority for the exemption applicable to this study is documented in the Code of Federal Regulations at 45 CFR 46.104(d)2. This study’s methods/procedures adhered to the ethical principles articulated in both the Declaration of Helsinki and the Belmont Report. Additionally, informed consent was obtained from all subjects, all subjects were legal adults (aged 18 years or older).

Acknowledgments

The authors thank study participants and the Hawaii State Center for Nursing for their kind cooperation.

Additional information

Funding

University of Hawaii at Manoa, School of Nursing and Dental Hygiene. Eunjung Lim was partially supported by the National Institute on Minority Health and Health Disparities, grant number [U54MD007601] (Ola HAWAII).