Acceptability of patient-centered digital vaccine safety monitoring: A Canadian Immunization Research Network study

ABSTRACT

This study examined the willingness of Canadians to use patient-centered digital reporting solutions for adverse events following immunization (AEFI) reporting. We identified the preferred medium for reporting, and any privacy and confidentiality concerns among prospective users. A geographically diverse panel of 2,036 Canadian adults 18 y of age and older was surveyed online in September 2024. Descriptive statistics and a multivariable regression model were used to identify factors associated with a willingness to report AEFI. Among respondents 85% (n = 1724) indicated a willingness to report AEFI, and most (n = 1137, 56%) preferred to report AEFI only when they occurred as opposed to answering survey on a regular basis for a short duration after vaccination (n = 458, 22%). The largest proportion of respondents (n = 911, 45%) indicated a preference to use an online fillable form through a secure government website to report AEFI. Living with a disability, age over 24 y and having a great deal of confidence in scientists were all significantly associated with a willingness to report, while having a great deal of trust in pharmaceutical companies was inversely associated. Our results emphasize the importance of considering convenience, privacy and confidentiality, and trust in public institutions when developing patient-centered digital reporting systems for AEFI. Future research should explore income and ethnic disparities in willingness to report AEFI and the effect of tailoring reporting systems to public concerns on willingness to report.

Plain Language Summary

This survey study examined whether Canadian adults are willing to report health events after vaccination. It also examined what type of events they wanted to report and what method they would like to use to report.

Introduction

Adverse events following immunization (AEFI) are any untoward medical occurrence that follows vaccination. AEFI may, but do not necessarily, have a causal relationship with the vaccine.¹ Ranging from mild symptoms, such as fatigue or numbness, to more severe events that may be life threatening, AEFI surveillance is crucial for improving vaccine safety and maintaining public confidence in vaccination.^2–4 While extremely rare,^5–10 serious AEFI do occur, and their timely and accurate reporting is critical to risk-benefit analyses and risk mitigation efforts during vaccine development and subsequent use.

While health-care systems around the world continue to recover from the COVID-19 pandemic, many are burdened by staffing shortages, burnout and limited financial resources.^11–15 The lack of sufficient capacity for AEFI reporting by health-care providers has been well documented,¹⁶ and coupled with decreasing public confidence in vaccination,¹⁷ has prompted demand for the development and refinement of patient-centered active reporting solutions for AEFI.¹⁸ These reporting solutions have been described as a compliment to existing AEFI surveillance and involve the self-reporting of AEFI or health status by individuals following vaccination.¹⁹ The development of patient-centered AEFI monitoring represents an important opportunity to broaden AEFI data collection and improve vaccine safety and public confidence in vaccination, while helping to alleviate the reporting burden that often falls on health-care providers.^16,18

In recent years, there has been growing interest in leveraging technology and digital solutions to improve uptake and accessibility of patient-centered AEFI surveillance.^20,21 The Technology Acceptance Model is a popular research model that is often used to understand technology adoption among patients,²² and it may help to analyze willingness to participate in patient-centered AEFI surveillance. First presented in 1989,²³ the model theorizes that perceived ease of use and perceived usefulness play a mediating role in the relationship between system uptake and other system characteristics. In short, the easier and the more useful a system is, the more likely patients are to use it.

Prominent examples of active, patient-centered systems include the Australian government-funded SmartVax, Vaxtracker and STARSS (AusVaxSafety),²⁴ the United States V-safe,²⁵ and the Canadian National Vaccine Safety network (CANVAS).^26,27 The most common mechanisms utilized by patient-centered digital solutions include the use of mobile apps, SMS messages or e-mails to prompt individuals to report the occurrence of any AEFI and/or to survey their general health status following vaccination.¹⁹ Many, but not all, systems were developed for seasonal influenza vaccine surveillance.¹⁹ Response rates ranging from 60–95% have been reported, and known factors influencing the response rate include the number of prompts, type of prompts, and frequency of follow-up.^28–33

To maximize uptake and response rates through patient centered digital reporting solutions, further research is required to better understand the factors that might impact an individual’s willingness to report AEFIs. For example, which reporting mediums are most preferred by users and any privacy and confidentiality concerns that users may have when reporting AEFIs. Furthermore, it has been well established that a lack of public confidence in various institutions is directly related to vaccine hesitancy,^34–36 and that confidence in one’s health-care provider plays a critical role in their willingness to share health information with them.³⁷ Further research is required to better understand how confidence in key institutions such as government, research scientists, universities, health-care professionals, and others, might impact one’s willingness to report AEFIs.

To inform the development and refinement of patient-centered digital reporting solutions for AEFI, it is critical that we understand the public’s general willingness to use these systems and the factors that influence their acceptability, in a given context. This study examined the willingness of the Canadian public to report AEFI and identified the preferred mediums for reporting and any privacy or confidentiality concerns. Our primary research questions were whether privacy or data sharing concerns and confidence in key institutions influenced willingness to report AEFI and whether this varied by age or gender.

Materials and methods

Setting

Canada conducts active and passive vaccine safety surveillance after vaccines are approved for use. Active surveillance systems attempt to definitively identify all AEFI in a given population through an organized process that often involves regular contact with health-care providers or institutions administering vaccines.³⁸ Meanwhile, passive surveillance relies on the spontaneous reporting of AEFI by health-care providers or patients themselves.³⁹ Canadian health-care providers are required to report certain health events after vaccination to the passive monitoring system in most jurisdictions.

Sample and procedures

An online panel of Canadian adults 18 y of age and older was surveyed in English September 4, 2024 – September 18, 2024.¹ The sample was constructed to be balanced by gender (men and women) with age groups reflecting the distribution of adults in the Canadian population based on the 2021 census. Geographic diversity was sought from the ten Canadian provinces, but not enough panel participants were from Canadian territories to include those regions. The University of British Columbia (#H22-00260) and Ottawa Hospital Research Institute (#20210374-01 H) provided research ethics approval for this study.

Members of the online consumer research panel were sent one e-mail invitation with a survey link, which led to an informed consent form and the survey. Reminders were not sent. New members were invited until the survey sample size was reached. Respondents earned points redeemable for retailer gift cards from the panel provider. A total of 18,318 invitations were sent to obtain 2,287 respondents who met gender, age and residence criteria and proceeded onto the survey and 2,036 completed the survey (84% of those who were eligible).

Measures

Our primary outcome measure was the willingness to report an AEFI (binary variable yes/no). The secondary outcome was the preferred method of reporting. Survey questions about AEFI were developed based on literature review,²² key informant interviews (n = 5) conducted to understand potential benefits, drawbacks, and barriers to participant-centered digital AEFI reporting specifically in the Canadian context, and the key constructs of perceived ease of use and perceived usefulness from the Technology Acceptance Model²² (see Supplement for survey instrument). Demographic predictor variables of interest (age, gender, province of residence, etc.) were decided a-priori based on our previous research experience with vaccine safety surveillance.⁴⁰ Institutional confidence questions were adapted from a previous study⁴¹ to include 6-items assessing the respondent’s level of confidence in the government (provincial and federal), research scientists, universities, health-care professionals and universities. A 3-point Likert scale (hardly any confidence/only some confidence/a great deal of confidence) was used to measure each item. The survey was tested for face validity with fewer than ten participants and questions were adjusted for further clarity before the survey was launched. Responses from the testing were excluded from the final sample.

Analysis

We first computed descriptive statistics for the study variables. Second, we identified whether demographic or trust variables were associated with a willingness to report AEFI in univariable and multivariable logistic regression. All variables with a significant level ≤0.10 in the univariable analyses were added to a multivariable regression model. Backward selection algorithm with the Akaike Information Criterion (AIC) was used to select the model with the best fit.⁴² Due to the potential inflation of type 1 error, two additional sensitivity analyses were conducted: 1) using forward selection algorithm and 2) using every initial covariate. Absence of collinearity was confirmed using variance inflation factor. Goodness of fit was tested using Hosmer–Lemeshow method. To assess the quality of the backward selection procedure results, R function “anova” was used to compare full and reduced models. R version (V.4.1.1; The R Project for Statistical Computing, Vienna, Austria) was used for all analyses.

Results

A total of 2,036 participants completed the survey. The characteristics of the survey population are shown in Table 1. When compared to data from the 2021 Canadian census,⁴³ Ontarians were overrepresented in the study sample, equaling 48% of the sample compared to 38% of the 2021 census. Albertans and British Columbians were also moderately overrepresented, equaling 14% and 17% of the sample, respectively, versus 11% and 14% of the census. Quebec was underrepresented, with only 5% of respondents from Quebec versus 23% in the 2021 census. Age group distribution was similar to that found in the 2021 census, which showed 11% 15–24 y of age; 13% 25–24; 13% 35–44 y of age; 12% 45–54 y of age; 14% 55–64 y of age; 11% 65–74 y of age and 8% 75 y of age or older. The survey was balanced by women and men.

Table 1.

Characteristics of the survey population.

Characteristic	N = 2,036	Percent
Age
18–24 y	210	10
25–34 y	331	16
35–44 y	332	16
45–54 y	332	16
55–64 y	349	17
65–74 y	299	15
≥75 y	183	9
Gender
Man	1,028	50
Woman	980	48
Non-binary, genderqueer, agender, or similar identity	22	1
Prefer not to answer	6	0
Province of residence
Alberta	282	14
British Columbia	347	17
Manitoba	87	4
New Brunswick	41	2
Newfoundland and Labrador	39	2
Nova Scotia	66	3
Ontario	979	48
Prince Edward Island	11	1
Quebec	111	5
Saskatchewan	73	4
Race/ethnicity (not mutually exclusive)
Black (African, Afro-Caribbean, African Canadian descent)	72	4
East Asian (Chinese, Korean, Japanese, Taiwanese descent, other East Asian descent)	133	7
Indigenous (First Nations/Metis/Inuk/Inuit descent)	88	4
Latino/Latina (Hispanic)	26	1
Middle Eastern (Arab, Persian, Afghan, Egyptian, Iranian, Lebanese, Turkish, Kurdish, other West Asian descent)	29	1
Southeast Asian (Filipino, Vietnamese, Cambodian, Thai, Indonesian, other Southeast Asian descent)	43	2
South Asian (Indian, Pakistani, Bangladeshi, Sri Lankan, Indo-Caribbean, other South Asian descent)	168	8
White (European descent)	1,478	73
Self-identify as a person with a disability	308	15
Residence
Medium to large city (≥30k)	1,551	76
Rural area (<30k)	485	24
Household income
Under $35,000	298	15
$35,000 to less than $75,000	507	25
$75,000 to less than $100,000	353	17
$100,000 to less than $200,000	566	28
$200,000 or more	120	6
Prefer not to answer	192	9
Willing to report adverse event following immunization	1,724	85

A total of 426 respondents (21%) reported experiencing an AEFI and 1,724 (85%) indicated a willingness to report an AEFI if they experienced one. A significant difference existed with age: 90% (n = 750) of participants 55 y of age and old were willing to report compared to 69% (n = 145) of respondents 18–24 y of age (p < .001). Most respondents (n = 1,137, 56%) preferred to report an AEFI only when it occurred, rather than complete a survey on a regular basis for a short duration after vaccination (n = 458, 22%). When asked about the need to report AEFI of varying degrees of severity, 50% of respondents (n = 1,018) indicated all adverse events, including mild ones, were important enough to be reported. Only 37 respondents (2%) indicated only life threatening AEFI or death should be reported.

Regarding their preferred method to report AEFI, the largest proportion of respondents (n = 911, 45%) indicated a preference to use an online fillable form through a secure government website, while 29% (n = 592) preferred to use a mobile app and 9% (n = 185) preferred to report over the phone (Figure 1). Participants 55 y of age and older preferred a secure government website over a mobile app, while only participants 25–34 y of age preferred a mobile app over a government form (Figure 1).

Figure 1.

Preferred AEFI reporting method.

Close to half of respondents expressed no privacy or confidentiality related concerns for AEFI reporting, regardless of the reporting method (Figure 2). All privacy/confidentiality concerns, except targeted advertisements, were reported most often for mobile app reporting (Figure 2). Stratification by age revealed that the youngest respondents were more likely to have privacy and/or confidentiality concerns for website (p-value < .001) or telephone (p-value < .001) reporting compared to older age groups.

Figure 2.

Concerns about privacy/confidentiality of personal health information.

When asked to indicate their level of confidence in various institutions (Figure 3), over one third of the sample indicated “hardly any” confidence in the federal (36%) and provincial (32%) governments and pharmaceutical companies (24%). Health-care professionals and research scientists received the highest levels of trust, with 63% and 52% reporting “a great deal of confidence,” respectively.

Figure 3.

Confidence in Canadian Institutions/Professions.

We also examined the factors that were associated with a willingness to report AEFI. Living with a disability, age over 24 y and having a great deal of trust in scientists were all significantly associated with a willingness to report (Table 2), while having a great deal of trust in pharmaceutical companies was inversely associated.

Table 2.

Factors associated with a willingness to report adverse events following immunization in multivariable regression.

	Univariate Odds Ratio	95% Confidence Interval		Adjusted Odds Ratio	95% Confidence Interval
Age Group
18–24 y	REF	REF	REF	REF	REF	REF
25–34 y	1.83	1.23	2.73	1.65	1.09	2.5
35–44 y	2.65	1.74	4.05	2.29	1.47	3.56
45–54 y	2.36	1.56	3.57	1.98	1.29	3.04
55–64 y	3.67	2.35	5.73	2.86	1.8	4.61
65–74 y	4.71	2.86	7.74	3.63	2.17	6.2
≥75 y	4.38	2.45	7.81	3.43	1.91	6.45
White (vs nonwhite race)	2.03	1.58	2.6	1.31	0.99	1.73
Disability (vs no disability)	1.49	1.02	2.17	1.47	1.01	2.2
Women	1.04	0.81	1.32
Urban (vs rural)	0.89	0.67	1.19
Income more than $75,000	1.31	1.01	1.69
Trust in Research Scientists
Hardly any confidence	REF	REF	REF	REF	REF	REF
Only some confidence	1.83	1.19	2.82	1.59	0.99	2.5
A great deal of confidence	2.79	1.82	4.27	2.54	1.57	4.06
Don’t know/Prefer not to answer	0.9	0.53	1.53	0.99	0.53	1.88
Trust in Pharmaceutical Companies
Hardly any confidence				REF	REF	REF
Only some confidence	1.36	1	1.85	1.14	0.81	1.58
A great deal of confidence	0.83	0.59	1.17	0.62	0.42	0.91
Don’t know/Prefer not to answer	0.47	0.29	0.76	0.77	0.41	1.45
Trust in Federal Government
Hardly any confidence	REF	REF	REF
Only some confidence	1.28	0.97	1.68
A great deal of confidence	1.25	0.86	1.81
Don’t know/Prefer not to answer	0.39	0.24	0.63
Trust in Provincial Government
Hardly any confidence	REF	REF	REF
Only some confidence	0.97	0.73	1.29
A great deal of confidence	0.88	0.61	1.26
Don’t know/Prefer not to answer	0.35	0.21	0.57
Trust in Health Care Professionals
Hardly any confidence	REF	REF	REF
Only some confidence	1.74	1.07	2.83
A great deal of confidence	2.12	1.34	3.38
Don’t know/Prefer not to answer	0.75	0.38	1.47
Trust in Universities	REF	REF	REF
Hardly any confidence	1.42	0.97	2.08
Only some confidence	1.75	1.19	2.57
A great deal of confidence	0.74	0.44	1.22
Don’t know/Prefer not to answer	1.42	0.97	2.08
Province
Ontario	REF	REF	REF
Alberta	1.07	0.74	1.54
British Columbia	1.4	0.97	2.01
Manitoba	0.85	0.48	1.51
New Brunswick	1.78	0.63	5.06
Newfoundland and Labrador	0.64	0.3	1.38
Nova Scotia	1.22	0.59	2.51
Prince Edward Island	1.92	0.24	15.13
Quebec	1.23	0.7	2.18
Saskatchewan	0.74	0.41	1.35

Discussion

Our findings demonstrate the Canadian public was willing to use patient-centered digital reporting solutions to report AEFI. In spite of low levels of confidence reported in both federal and provincial governments, an online fillable form through a secure government website posed the fewest privacy and confidentiality concerns for prospective users and was the preferred medium for reporting. With most respondents expressing a preference to report AEFI through a website, rather than downloading a mobile app, and most preferring to report AEFI only if they occur, our findings highlighted convenience as a key factor when it comes to Canadians’ willingness to use patient-centered digital reporting systems to report AEFI. However, respondents’ lack of willingness to participate in general health surveys following vaccination suggests Canadians may benefit from health promotion initiatives raising awareness of vaccine safety surveillance methods in use and emphasizing why post-vaccination health reporting is important even if respondents do not experience anything unusual. Overall, our results coincide with previous research that shows convenience is a key motivator for patients to participate in health information exchange,⁴⁴ as well as for health-care delivery more broadly.⁴⁵

“Privacy and confidentiality” was not a significant factor in participants’ willingness to report AEFI, although there was a preference for a secure website over a health related mobile app. Privacy, confidentiality, and security concerns when using health-related mobile apps – ranging from privacy from family and friends to collection of excessive data from users’ devices – have been well documented,⁴⁶ and these concerns were demonstrated by our participants. We found Canadians had the most privacy concerns about AEFI reporting via a mobile app. This coincides with previous research demonstrating a fear of personalized advertisements as the most significant barrier to using health-related mobile apps.⁴⁷

Previous research has shown younger adults are more likely than older adults to withhold health information due to privacy and security concerns,⁴⁸ coinciding with our results showing that adults 18–24 y of age expressed greater privacy and confidentiality-related concerns with website reporting methods compared to other age groups. Young adults represent the next generation of parents and potential AEFI reporters, and their heightened concerns relative to other age groups suggest that addressing privacy and confidentiality-related barriers to AEFI reporting should be of the utmost importance as the development and refinement of patient-centered digital reporting systems continues.

This study makes a novel contribution to the AEFI literature by displaying a statistically significant association between age and willingness to report AEFI. Adults 25 y of age and older are more likely to be willing to report AEFI than those aged 18–24 y of age. While novel, this finding is unsurprising when considering previous research showing that older adults are often more willing to share health information.⁴⁹ Further, in part due to maturing through a time of rapid technological and informational change, privacy and security concerns often cause younger adults to be more reserved and cautious when it comes to health information sharing.⁴⁸

Our findings are also consistent with the Technology Acceptance Model,²³ which theorizes use of technology is influenced by perceived ease of use and by perceived usefulness. Both of these influence an individual’s attitude toward using a technology impacting their behavioral intention to use, ultimately resulting in use or nonuse of the technology. Through a survey approach, we are able to elucidate insights into perceived ease of use, perceived usefulness and behavioral intention. As the direct usefulness of reporting an AEFI is primarily for government rather than the individual, perceived ease of use is critical. Thus, respondents’ preference of a web portal over other modalities suggests this would be important in the design of such systems. Although not measured in our study, younger individuals may perceive less usefulness of reporting AEFI. Thus, any barriers to perceived ease of use (lack of convenience) may make them hesitant to utilize the technology. Accordingly, solutions need to be optimally designed to facilitate user experience and ease to maximize uptake across all age groups. Raising awareness of the importance of AEFI reporting for keeping our vaccinations safe may also raise perceived usefulness to individuals and therefore willingness to report.

Our findings highlight the need to tailor systems to address the privacy and security concerns of younger adults. Who is involved in reporting, how reporting is framed (e.g., research study vs government surveillance), and how data are collected (e.g., convenience, privacy) will be critical considerations when attempting to increase AEFI reporting. Our findings regarding low levels of trust in government despite willingness to report to a government website may demonstrate public desire for government-supported but arm’s-length evaluation of vaccine safety. This has emerged as an issue in multiple jurisdictions and will be important to consider when designing these programs. Higher levels of trust in scientists, health-care professionals compared to government and pharmaceutical companies suggest scientists and health-care professionals may be best positioned to lead and administer messaging regarding patient-centered AEFI reporting. For example, system administrators could make clear that even if a system is federally or provincially funded, it is carried out in partnership with institutions that remain very highly trusted, such as local health-care professionals (e.g., community physician), research groups or universities. Future research should study the implementation of any new systems and explore how tailoring patient-centered reporting systems to public concerns influences the willingness to report AEFI, especially among younger adults.

Interestingly, a minority of respondents with high levels of trust in pharmaceutical companies responded that they were less willing to report AEFI. While we did not ask why this would be the case, we hypothesize these individuals with more favorable views toward pharmaceutical companies may have a better understanding of vaccine pharmacovigilance and the temporal relationship of AEFI reporting and therefore be less concerned about reporting it. The inverse would also be true that those will less trust would be more willing to report. Future research, ideally with qualitative methods to explore the meaning and intent behind such responses, could explore this.

Canadians who identified as white were slightly more likely to be willing to report AEFI than those who did not. Similar differences can be observed in other aspects of the Canadian health-care system, and it has been well established that nonwhite Canadians often face significant barriers and inequities in accessing health-care services when compared to white Canadians.⁵⁰ Consistent with the longstanding harmful impacts of colonialism on the relationship between Indigenous Canadians, the Canadian health-care system, and research institutions, Indigenous Canadians are also less likely to trust the health-care system and health-care professionals compared to non-Indigenous Canadians.^51–53 It is likely our findings are a product of these inequities in some way as perceived usefulness of more data into a system that perpetuated harm is likely to be low.

While our survey sample included participation from all ten provinces, we did not sample the territories and the sample was not geographically representative. This was an electronic panel survey so the survey method excluded individuals without access to a computer or mobile phone. Finally, the survey was only in English, so Canadians who speak other languages would not have participated and our results do not capture their viewpoints. Notably, the lack of a French survey instrument is a limitation that may account for the low participation of individuals from Quebec, where attitudes toward AEFI reporting may differ from the rest of Canada.

Overall, Canadians are willing to participate in AEFI reporting, and most believe that all AEFI, even minor ones, should be reported. Additionally, willingness to report increases with age and most prefer to report through a secure government website. The continued development of patient-centered digital reporting systems for AEFI will need to pay special attention to convenience, privacy and confidentiality concerns, as well as trust in public institutions. Future research should seek to further explore income and ethnic disparities in willingness to report AEFI, in addition to how tailoring reporting systems to public concerns might influence willingness to report AEFI through patient-centered digital reporting solutions.

Biographies

Ellis-Legault is a recent MPH graduate of Simon Fraser University. He has previously held policy and administrative roles with the Canadian Public Health Association and Ottawa Valley Family Health Team.

Kumanan Wilson is CEO and Chief Scientific Officer of Bruyere Health Research Institute and Professor of Medicine at the University of Ottawa where he has been studying vaccination policy and technology.

Devon Greyson is an Assistant Professor in Population and Public Health and an information scientist who specializes in public health information, communication, and surveillance.

Hennady P. Shulha is a statistician at the Vaccine Evaluation Center.

Adhiba Nilormi is an MPH graduate of the University of Guelph. She is currently a research coordinator at the Bruyere Health Research Institute.

Julie A. Bettinger is a Professor in Pediatrics and an infectious disease epidemiologist who specialized in vaccine preventable disease surveillance, vaccine safety and vaccine confidence.

Funding Statement

The work was supported by the Canadian Institutes of Health Research [FRN#151944]; Public Health Agency of Canada.

Acknowledgments

BEL: Writing – original draft

KW: Conceptualization, funding acquisition, writing – review and editing

DG: Methodology, writing – review and editing

HPS: Formal analysis, writing – review and editing

AN: Project administration, writing – review and editing

JAB: Conceptualization, funding acquisition, supervision, validation, writing – original draft

Disclosure statement

KW is the co-founder of CANImmunize Inc. and has served as a member of the independent data safety advisory board for Medicago and Moderna. All other authors report no competing interests.

Supplementary Information

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2571277