Narrowing the policy gap: lessons from years 2 and 3 of the British Columbia influenza prevention policy

ABSTRACT

Influenza can be potentially fatal to vulnerable populations, particularly those in the hospital. Canada’s National Advisory Committee on Immunization recommends that health-care workers (HCW) be immunized against influenza partly to avoid infecting high-risk populations. However, influenza immunization rates among HCW remain suboptimal. In 2012, health authorities across British Columbia (B.C.) implemented a province-wide influenza prevention policy requiring HCW to either be immunized or wear a mask when in patient-care areas during the influenza season. This paper describes the second of two studies focused on what was learned from years 2 and 3 of the policy. A case study approach was used to examine this policy implementation event. Qualitative data were collected through key documents and key informant interviews with members of leadership teams responsible for policy implementation. Framework analysis and Prior’s approach were used to analyze data from interviews and documents, respectively. Policy implementation varied by geographic region and gaps persist in immunization tracking and discipline for noncompliance. Debate regarding the scientific evidence used to support the policy fuels resistance from particular groups. Despite these challenges, findings suggest that the policy has been habituated, largely due to consistent policy objectives. This study emphasizes the importance of ongoing inter-professional and cross-sectoral program evaluation. While adherence may be routine for many, implementation processes must continue to respond to contextual issues to narrow the gap in policy implementation and to continue to engage stakeholders to ensure compliance.

KEYWORDS:

• British Columbia
• health personnel
• influenza vaccines
• immunization
• vaccination
• public health
• policy

Introduction

Influenza is a serious vaccine-preventable disease that is estimated to result in over 12,000 hospitalizations, and around 3,500 deaths in Canada annually.¹ Many of these serious health events could be prevented by immunization, which is considered one of the most effective tools to prevent influenza and its complications. Because of the risk that influenza poses to health-care workers (HCW) and their patients, Canada’s National Advisory Committee on Immunization (NACI) recommends annual seasonal influenza vaccination for HCW in facilities and community settings as part of the standard of care for protecting patients.² Despite this recommendation, influenza vaccination coverage among Canadian HCW remains below the established target of 80%.^3,4 Voluntary programs with educational interventions, extensive publicity campaigns, accessible vaccination via mobile immunization carts and vaccination champions, prizes for HCW and for hospital leaders based on institutional influenza vaccination rates, and required declination signature forms have been demonstrated to improve vaccine coverage, but none of these interventions have been successful in achieving target coverage rates.^5-8

Responding to suboptimal coverage rates, some jurisdictions introduced condition-of-service policies, which require employees to comply with occupational health policy to protect the patient population. For example, Barnes-Jewish/Christian (BJC) HealthCare (St. Louis, Missouri) and Virginia Mason Hospital (Seattle, Washington) experienced large boosts in uptake with immunization rates around 98% after mandatory influenza control policies were implemented, with few employees seeking medical or religious exemptions.^5,6 While examples of successful condition-of-service influenza prevention policies exist, few areas in Canada have adopted these practices and there is a paucity of large-scale studies of related policies in the Canadian context.

Providing timely and effective healthcare is the shared responsibility of the Ministry of Health in British Columbia (B.C.) in partnership with the provincial health services authority, five regional health authorities, and First Nations health authority.⁹ In 2012, B.C. became the first province in Canada to implement a province-wide vaccinate-or-mask influenza prevention policy for all HCW working in patient-care areas. The B.C. Influenza Prevention Policy (hereafter referred to as the policy) is broadly applicable across clinical settings, professional categories, and terms of employment. The goals of the policy were to: (1) increase influenza immunization rates in HCW employed in health authorities in B.C., (2) prevent transmission of influenza from HCW to patients/residents/clients and to other HCW in health-care facilities in B.C., and (3) reduce influenza-related absenteeism in HCW employed by health authorities in B.C. Implementation of this policy generated controversy: a labor grievance against the mandatory nature of the policy was filed by the Health Sciences Association (HSA). The HSA is a union representing over 20,000 health professionals in a variety of jobs at over 250 institutions across the province of B.C.¹⁰ Despite the grievance, the policy was upheld by an arbitrator in 2013. Since that time, the policy has been enforced in all health authorities in B.C.

Many jurisdictions are interested in understanding the experiences of HCW in B.C., and the decision to implement similar policies may be informed by these results. We began examining these issues in our previous publication by delineating the key factors associated with policy implementation during its inaugural year based on the insights of eight leadership implementation teams.¹¹ We now present results from years 2 and 3 of the policy based on key informant interviews with health-care leaders at different organizational levels from each of the health authorities. The primary research question for this study was: How do you implement a province-wide condition of service HCW seasonal influenza immunization policy?

Methods

This study employed a case study approach with theoretical propositions, a methodology that has been described in the literature and that we used in our previous publication on this topic.^11-13 The case study method is considered a comprehensive qualitative research strategy as it helps explain complex social phenomena.¹⁴ This approach was particularly helpful in the organizational context of the policy in B.C. because it allowed us to trace meaningful organizational and managerial processes. Attention was paid to how the key factors involved in policy implementation evolved over the 2 years since the policy was introduced. The policy in B.C. was selected as a unique organizational case because it was the only Canadian jurisdiction with province-wide implementation of such a policy at the time of data collection. This segment of the case study focuses on years 2 and 3 of program implementation and is bound by time (2014–2016), place (province of B.C. health authorities’ boundaries as outlined in provincial legislation), organizational definitions (health organizations as defined by the health authorities for the policy) and policy definitions of HCW and policy implementation teams.

Consistent with a case study approach, theoretical propositions were used to narrow the scope of the research question to structure the data collection and analysis processes, and to ensure the feasibility of the study.¹⁴ Propositions were related to the substantive research question and informed by existing theory and empirical research on the implementation of HCW influenza immunization programming. The propositions guided the focus of the interview questions and were integrated into analysis to help answer the substantive research question. Please see Table 1 for a summary of the study propositions.

Table 1. Study of theoretical propositions.

Proposition 1	It is crucial that leaders continue to engage their program team in the influenza prevention policy.^Citation8,Citation15 The role of opinion leaders and physicians cannot be understated.^Citation15
Proposition 2	Clearly outlining and communicating a continued implementation plan will support policy sustainability.^Citation15
Proposition 3	The leadership team must determine appropriate program components and relevant tools to support policy sustainability over time. For example, program components such as the following may be deliberately considered: role models, vaccine access, education and promotion (including communication strategies), enforcement of the mask policy, and tracking and reporting immunization rates.^Citation16
Proposition 4	Maintaining human and financial resources is crucial in facilitating policy sustainability over time.^Citation15
Proposition 5	Developing an ongoing evaluation process and knowledge dissemination planning is crucial.^Citation15

Data sources included semi-structured interviews with health authority employees in B.C., and relevant documents that were either publicly available or shared with the study team by participants. Criterion sampling with maximum variation was used to seek sample diversity across participants who were involved with the implementation of the policy over years 2 and 3. Participants included members of leadership implementation teams (health authority teams at each of the five regional health authorities and at one provincial health authority, the provincial Occupational Health and Safety leadership team, and the provincial Public Health and Ministry of Health leadership team), health-care managers and directors, labor relations and communications personnel, HCW, health authority representatives, occupational health and safety representatives, and union representatives. Other criteria included gender, age, profession, years in professional role, years in a health-care setting, and role in the implementation of the policy. An emailed letter was used to describe the project and recruit participants. Thirty-five key informant interviews were conducted between February and May 2015. Each lasted approximately 1 h and was audio-recorded and transcribed. The interview guide consisted of nine questions and was iteratively updated in consultation with the study team as the interviews progressed (see Supplemental Text 1 for the final interview guide). Policy and other related documents were purposefully sampled between November 2013 and April 2015 and met inclusion criteria if they were relevant to policy implementation and/or were mentioned and shared by participants.

Data analysis was guided by Spencer, Ritchie, and O’Connor’s (2003) method of Framework Analysis, which was developed in the context of applied policy research and is increasingly used in health research in combination with case study methods.^11,13,17-19 Framework analysis involves the process of conceptual scaffolding, comprised of five iterative stages: (a) familiarization, (b) identifying a thematic framework, (c) indexing, (d) charting, and (e) mapping and interpretation.^17,19

Research assistants (AN and HFM) with experience in qualitative research conducted the key informant interviews and developed the coding structure for analysis using NVivo11 software.²⁰ Codes from individual research assistants were amalgamated into a coding dictionary that was used by HFM and AD to simultaneously code all 35 interviews. Codes and themes were reviewed by the primary investigator and the research team. Document sources were collected throughout the duration of the study and logged systematically. The documents were used to trace patterns of social exchange and the social networks that lie behind them as per Lindsay Prior’s approach.²¹ Particular attention was paid to: (a) content, (b) how they were produced, and (c) how they functioned or were used. Each document was systematically analyzed using a framework to address how the document was used and the perspective reflected in the document during implementation.

Qualitative methodological rigor was established using Lincoln and Guba’s trustworthiness criteria (1985) (i.e., credibility, transferability, and dependability).²² Credibility was enhanced by having two independent coders review and code the interview transcripts separately before collaborating and comparing results.²² We addressed transferability by providing “thick” qualitative descriptions of our findings to give readers enough information to determine whether or not our results were applicable to their contexts.²² Transcripts and coding memos were kept to ensure that our research process was clearly documented.²² An iterative process was used to triangulate interview and document data, which helped to support code and theme development. Consistent with a case study approach, a chain of evidence was systematically established during data analysis and interpretation, including consistent testing against propositions.¹⁴ There was a deliberate focus on divergent patterns, negative instances, alternate themes, and rival explanations.¹⁴

Results

The sample included 35 participants and 33 documents. The key informant interviews consisted of members of the eight leadership teams, some of whom were involved in Phase 1 of this study. Interviewees were health-care managers and directors (n = 19), labor relations and communications personnel (n = 4), HCW (n = 3), health authority representatives (n = 1), occupational health and safety representatives (n = 5), and union representatives (n = 3). Sample values are approximate as some interviewees fit under several of the categories identified. Members of this group were responsible for implementing the policy at the facility level, organizing influenza vaccination and prevention campaigns, and enforcing mask-wearing for unimmunized HCW. Important documents recommended by key informants were retrieved and analyzed, including policy and implementation documents (n = 1), internal memos (n = 5), government and authority frequently asked question pages (n = 4), online news articles (n = 16), and other communication materials (n = 7). Three themes were identified: (1) policy uniformity, (2) policy of choice with consequences, and (3) tension within the comfort zone.

Policy uniformity: does one size really fit all?

The policy was authored and brought into effect by the B.C. Leadership Council, comprised of leaders from each of the health authorities and the Ministry of Health (MoH). Initially, it took a one-size-fits-all approach to ensure consistency across regions in terms of the policy objectives and requirements of employees. However, the policy did not provide much detail with respect to how it should be implemented in local health authorities.

The one-size-fits-all approach of the policy was considered a strength inasmuch as it afforded flexibility to implement the policy differently in different contexts. The ambiguity in the implementation directive allowed health authorities and workplaces the opportunity to adapt implementation strategies to their unique contexts. For example, health authorities offered different types of immunization services and enabled health authority employees to report vaccinations received outside of the workplace.

… [a] thing that’s been helpful has been the broadness with which staff have been able to have access to the vaccine because we’re not restricted to being able to get vaccines within just the Health Authorities itself or within Health Authority clinics, but they can attend a community pharmacy, doctor’s office, or public health clinic and receive the vaccine for free with just identifying that they are healthcare worker. (Participant 9)

Urban institutions with more employees were able to use staff engagement and employee density to their benefit to save money and resources in supporting the policy. Others took advantage of surrounding resources. For example, one health authority outsourced their flu clinics to a nearby 24-h retail pharmacy near their hospital. This was both convenient for HCW and cost-saving for directors and managers.

A staff member could go at 1 o’clock in the morning to the pharmacy just around the corner and get a vaccination if that’s when they had the time. So for us it was a very convenient thing, and it saved us a bucket load of money. We went from over $100,000 to I think our flu campaign this past year was maybe $1,000. (Participant 17)

Lack of detail regarding how the one-size-fits-all approach of the policy should be operationalized also led to some confusion. For example, the policy required that employers provide masks for unvaccinated employees in patient-care areas which are defined in the policy documents.

A patient care area/location is defined as an area within a healthcare facility, including a contracted facility that is accessible to patients, residents or clients who are there to access care or services – including, for example, hallways or lobbies. It includes any other location where care is provided, such as home and community care locations (including a client’s home). It does not include locations such as administrative areas or private offices which are not generally accessed by patients, residents or clients. (Documents 1.2 & 3.2)

This definition was broad and led to confusion among managers, employees, and visitors about whether or not they were in a patient-care area and therefore needed to be wearing a mask.

… a lot was left up to interpretation and when that happens it can get quite confusing for sites because they’re all different sizes and they look a little different, and what do we really expect? Like some of our small sites it would be considered a patient care area, it doesn’t matter where you go in the building because you have patients. Whereas some of your big ones you have very clear administration areas that you don’t have patients in. (Participant 22)

Another challenge that the one-size-fits-all approach presented to managers was the difficulty of identifying unvaccinated employees who needed to mask. A provincial approach to address this implementation challenge was a self-reporting system used by all health authorities starting in the 2014/15 season.

All health-care workers (even those getting immunized at a Northern Health clinic) will need to self-report their flu shot. It takes less than a minute! Just get your shot and then visit flu.northernhealth.ca to report it. (Document 2.4)

A provincially coordinated self-reporting system was developed to enable employees to report their vaccination status through an online form that directly populated the health authorities’ occupational health databases. This shifted the onus onto HCW to ensure that their vaccination was reported and relieved occupational health teams of a substantial workload of record keeping.

Although the self-reporting system was generally successful, interviewees explained that not all vaccination data were housed in the same database. Non-salaried physicians and other health-care staff (e.g., dentists, midwives) who were not health authority employees self-reported to a separate database that had to be developed during policy implementation. This database for health-care staff outside of the health authority was generally characterized as challenging to use. Since the database for staff outside the authority was independent of the health authority employee self-reporting system, it was difficult for facility managers and leaders to monitor vaccination coverage consistently across occupational groups. This ultimately resulted in an increased workload for those managers and leaders whose job it was to monitor vaccine coverage. In addition to the separate databases, and increased workload for those tracking vaccines, some interviewees questioned the reliability of this system. Self-reported data regarding vaccine uptake were rarely audited and verification of immunization was seldom requested of employees.

We had a very cumbersome process where we had a lot of physicians that were actually showing up as unvaccinated that were vaccinated and had reported, including myself one year, and we had a lot of physicians that were showing up as unvaccinated that had vaccinated that hadn’t reported. (Participant 33)

Different interpretations of the policy also led to some cases of non-uniform disciplinary measures for non-compliant staff. Document analyses revealed differences among health authorities in disciplinary action for non-compliant employees and health-care staff. For example, a health authority employer terminated the contract of a nurses’ aide after they declined to receive the influenza vaccine or to wear a mask in patient-care areas (Document 4.2). However, there were anecdotal reports of HCW in another organization who did not comply with the policy and for which they were not disciplined.

A doctor at [a B.C. hospital] who defied the government’s policy on flu vaccines for healthcare workers says non-compliance is likely far higher than the province admits. [The doctor] said the government’s assertion that only one healthcare worker ultimately refused to get vaccinated or wear a mask during the flu season – an aide, who was fired – isn’t true. [The doctor] said [they] and about 10 nurses at [the hospital] also ignored the policy and, even though they told their bosses, they were never disciplined. ‘Are we a freakishly rebellious group in [this city], or is this actually occurring all around the province?’ [The doctor] said in an interview. ‘I don’t know the answer.’ (Document 4.14).

Rural and remote communities struggled with policy enforcement due to a lack of resources, minimal staffing, and the thin dispersion of staff across large geographic areas. Engagement with the policy was reportedly minimal among employees who worked alone or on a small team.

We have [several] health centers where the nurses work alone by themselves, and they’re remote around the province, and the managers are in Vancouver, and so they don’t get out there very often to check if the nurses are, who haven’t been vaccinated, wearing a mask. And similarly we have [several] nursing stations where there’s two to three nurses per station, but there’s still no manager and I know in the provincial policy that the senior nurse supervisor doesn’t have to be management, it’s supposed to be the enforcement, but we’re really poor that way. (Participant 5)

Over time and as challenges relating to the one-size-fits-all approach arose, administrative supports were put in place to mitigate barriers to implementation. For example, one support consisted of a provincial committee chaired by the Provincial Health Officer with membership from all health authorities to respond consistently to emergent issues.

… so the decisions would be made as the provincial group. There might be some separate working groups of the provincial group to discuss it and then bring it back to the provincial conference call and the decisions would be made there. … The guideline papers [which outlined] what to do with speech language pathologists … [and] how to approach psych patients, etcetera … would be communicated via the provincial group to the health authority reps and we within our health authority had on our website, we certainly had a location dedicated to healthcare worker flu program and so we would post them there. (Participant 20)

Health authority representatives reported to the committee that speech language and music therapists unvaccinated against influenza were unable to carry out their job duties while wearing a surgical mask. The committee responded by sending a general information update to all managers at all regional health authorities with a protocol for providing clear face shields or other accommodation to those employees.

The other big issue was communication around masks, our approach to speech language pathologists and music therapist, our approach to discipline. [During] the whole range of issues, we had a consistent provincial approach and that’s a huge strength to the policy. (Participant 20)

Policy of choice but with consequence

The policy provided the choice for HCW to either be immunized or wear a mask when in patient-care areas with consequences up to termination for noncompliance.

Any covered individual found in violation of this policy may be subject to remedial and/or disciplinary action up to and including termination of employment, cancellation of contract and/or revocation of privileges. (Documents 1.2 & 2.2).

In formal communications to employees, the policy was framed as a policy of choice between vaccination and wearing a mask in patient-care areas. Participants deconstructed this framing and questioned HCW autonomy to choose how to comply with the policy.

Participants described how HCW perceived the policy’s narrative of choice to be a falsehood, offering only the illusion of choice, since wearing a mask during work hours was considered such a burden or barrier to effective work with clients.

On the one hand we were saying to people you have a choice but for some of our members – if you’re a speech therapist or you’re dealing with vulnerable communities of people to wear a mask in your daily work with those folks – it doesn’t work. (Participant 35)

Participants shared that workers who otherwise supported influenza vaccination for HCW opposed the policy simply because it was an edict. Some HCW felt that the health authority was being overly prescriptive about a personal health decision.

If we can make the case that this is the right thing to do to help people yourself included because you don’t want to get the flu and we make it really easy for them to get the flu vaccine then they should, and you don’t need to have this very heavy-hand kind of approach because in some ways you almost get push back on that just because you’re saying do it or else, right? Which is not a very respectful way of trying to encourage people to do what somebody thinks is the right thing to do. (Participant 21)

The B.C. Nurses’ Union (BCNU) publicly opposed the policy and issued a position statement stating, “healthcare employers should not take punitive or coercive action to force behaviour that is contrary to the individual’s decision [to be vaccinated against influenza]” (Document 6.1). Most participants in managerial positions identified that, in later years of the policy, disciplinary action became a more substantive reality. Documents analyzed for this study indicated that there was an employee who refused to comply with the policy and whose contract was terminated as a result.

A healthcare worker [in a community] is the first person fired under a new B.C. policy that forces health workers to get a flu shot or wear a protective mask around patients … [the individual] had held a permanent part-time position … but said he refused instructions from his bosses at the [Health Authority] to either get vaccinated for influenza or wear a mask at work. (Documents 4.1 & 4.2)

This perceived escalation of consequence over time further cleaved the opinions of HCW of the policy.

Tension within the comfort zone

Scientific standoff

By year three, although habituated to the policy, implementation leaders, union representatives, and employees continued to debate whether there was sufficient evidence to support it. While referencing many of the same academic works, policy supporters and critics held firm and opposing interpretations of the body of evidence.

The position of the MoH was that evidence from four randomized controlled trials showed that increased influenza vaccination among HCW reduced the mortality of residents in long-term care facilities and the findings could be applied to acute care settings. As the health system’s governing body, the MoH was a powerful knowledge broker, and their interpretation of the evidence was held up in labor arbitration in 2013.

We look at the body of evidence: there are multiple studies in multiple settings, using multiple methods that have all shown benefit from immunizing HCW … and none of which show harm. (Document 9.2)

Other knowledge brokers in the province continued to challenge the MoH’s interpretation of the evidence, including unions, physicians and other clinical staff.

There’s not great evidence that this is an appropriate policy. What is out there is being pooh-poohed by our public health but without just saying that this is a biased study, well really, Cochrane is biased? And then what is put forward as being the source of truth is from an equally or you could argue more biased source. (Participant 21)

Physicians and nurses were consistently identified by participants as groups that continued to resist the policy due to doubt about the scientific evidence supporting it. In particular, these groups were skeptical of conclusions about the merits of the policy given that the vaccine effectiveness during 2014–15 (policy year 3) was estimated to be unusually low.

Media coverage of the vaccine “mismatch” emerged before the influenza season began, significantly impacting HCW views about the policy. The “mismatch” also created an influx of questions regarding the efficacy of masking for only those HCW who did not receive the vaccine, rather than all workers.

So I don’t think there’s any evidence for it and the whole thing is totally and utterly illogical. There’s no sense in it at all. You know if we’ve got a vaccine that’s only 10% effective then why are only people that are not getting vaccinated have [sic] to wear a mask? Because what you’re saying is it’s okay to be 90% unvaccinated and still walk around without a mask. (Participant 21).

Participants affiliated with the nurses’ union supported clinicians’ skepticism regarding the influenza vaccine and masking efficacy. They cited insufficient scientific evidence as a reason to discontinue the policy, despite the fact that the MoH collection of evidence had been held up in arbitration.

BCNU believes that … nurses and other healthcare workers should have the right to decide whether to be vaccinated against influenza, based on their understanding of the current evidence and in discussion with their own family physician or other care provider. (Document 6.1)

Public discourse around the scientific merit of the policy continued well after arbitration, within both the healthcare community and the general public. Power brokers presented and published their own summaries of the evidence, which came to contradictory conclusions. In 2012, the British Columbia Center for Disease Control issued a report detailing the evidence used to create and justify the policy. This document was made available to the public to allay skepticism; however, detractors of the available scientific evidence used this same resource to promote skepticism. One dissenting physician gave interviews to news media and a detailed presentation rebutting the evidence used by the province to support the policy and criticizing its clinical merit.

By leaving evidence behind and instituting a policy based on ideology, illogic and intimidation, Public Health officials have undermined their credibility and set the stage for spread of mandatory vaccination (flu shots to children, other vaccines for HCWs). (Document 9.1)

The mandatory nature of the policy was particularly frustrating to participants who took issue with the evidence used by the MoH to support the policy.

Habituation

The policy became habituated, meaning the repeated exposure to the policy over the years resulted in a decrease in attention and reactivity to the policy in both HCW and upper level management. Despite ongoing dissonance regarding the evidence behind this policy, the sentiment of resistance to it began to wane during the second and third years of policy implementation as the health authority continued to enforce compliance.

I think it was a little more relaxed and it’s more accepted. People, are just now accepting that this [is] here and it’s not going away. I think the first year everybody thought it would be a flop and that all the Health Authorities would just say okay never mind, we’re not going to do this anymore. So I think staff has started to accept this …, it’s either a mask or a vaccine. (Participant 13)

Executive and senior leadership continued to issue a consistent message that informed HCW of the expectations of compliance while emphasizing the safety and effectiveness of the vaccine. Furthermore, policy implementation became more streamlined in successive years, as the division of labor in the enforcement of the policy increased. Employees were able to report their own vaccination status, frontline managers could work with individuals who were non-compliant, and human resources could enact disciplinary action for continued noncompliance.

… all employees get one warning if they’re not seen wearing a mask, they’re advised that they should be, there’s a conversation happening to ask about why they may not be wearing a mask, do they understand the policy but then they’re told that this is it and next time we find you not complying on the policy we would be making arrangements to meet with HR to talk about progressive discipline so I think that quick escalation and being very clear that folks are only getting one warning was helpful and it was helpful to the managers. (Participant 20)

Our findings indicate that HCW gradually accepted the choice of vaccination or masking and tensions were reduced as they became habituated to the policy over time.

While the condition-of-service nature of the policy significantly contributed to the habituation of the policy, habituation was also reinforced by the consistent nature of key aspects of the policy and its implementation over several years. Objectives, expectations, and means by which to remain in compliance with the policy year after year were well communicated to staff. Most participants identified frontline nursing and immunization staff as an essential reason for the success of the policy because these individuals distributed vaccines, masks, and information regarding influenza prevention. As the policy became habituated, there was less need for formal communication surrounding it.

[The last flu season] was kind of a non-event. I think in previous years it’s been a bit more formal, a bit more memo driven. This year we provided a brief communication to managers just to let them know that it was no longer in effect, just from the adherence standpoint, they didn’t need to worry about it at a certain day, and then we did some stuff informally for our staff, just putting information in the bulletins and really focusing on thanking people for their contribution throughout the flu season […] it’s become a lot more streamlined. We developed a pretty decent communication plan at the outset of the policy, so we’ve more or less been able to implement that each year, so it’s a bit more routine. (Participant 2)

Access to policy expectations and knowledgeable peers contributed to the successful implementation of the policy at most facilities.

It’s so visible. Having the clinics in facilities, posters, and signage everywhere, that’s where we get a lot of the questions answered. People will come to clinic for their flu shot or for information, and the nurses are really well versed in policy, and in the immunization process itself. (Participant 7)

Discussion

The focus of this study was on how to implement a province-wide condition-of-service HCW influenza prevention policy. Despite challenges surrounding immunization tracking, discipline for noncompliance, and ongoing debate about the scientific evidence used to support the policy, the policy has been habituated, in large part due to consistent policy objectives. The results offer insight into the importance of ongoing inter-professional and cross-sectoral program evaluation, prompt response to contextual issues, and the continued engagement of stakeholders.

Provincial-level mandate

In B.C., influenza immunization coverage among the staff of acute care facilities ranged from 35% to 46% between the 2007–08 and 2011–12 influenza seasons. Coverage increased after the implementation of the policy, ranging from 74% to 76% for seasons 2012–2013 through 2015–2016.²³ Similarly, coverage among the staff of residential care facilities in B.C. increased from 49% to 67% for the 2007–08 to 2011–12 influenza seasons to 75–79% for seasons 2012–13 through 2015–16.²⁴ While these rates fluctuate on a year-to-year basis, influenza immunization is much closer to achieving target coverage levels than before the policy was mandated.

Since the implementation of the policy in British Columbia, similar policies have been introduced in one of the two health authorities (Horizon Health Network) in New Brunswick, and in selected facilities across Ontario.^25,26 The success of these types of policies in preventing influenza is difficult to measure. In the absence of a measure of influenza prevention, success has been described as high vaccine coverage and staff support for the policy.²⁷ The vaccinate-or-mask policy implemented in Horizon Health Network in New Brunswick has been very successful in achieving anecdotally reported compliance by HCW of 100%.²⁸ A cross-sectional study of immunization among health-care professionals revealed a 16% increase in HCW influenza immunization after the introduction of vaccinate-or-mask influenza prevention policies in British Columbia and the Horizon Health Network.⁴ Areas without such policies saw a decrease in influenza immunization of 1% over the course of the 2012–2013 season.⁴ There is otherwise very little evaluative data about HCW immunization rates in these provinces.

In our recent publication which describes how to develop and implement a pan-provincial policy, implementation gaps in processes and content were identified between the province-wide policy and local implementation, leading to personal impacts for health-care leaders.¹¹ Confusion surrounding the operational definition of a patient-care area is one such implementation gap that has persisted into years 2 and 3 of the policy. Another implementation gap, non-uniformity in the identification of vaccination status, has since been addressed with the introduction of a self-reporting system for health-care staff. While an improvement from the original solution, which was to have staff display stickers identifying their vaccination status,¹¹ concerns arose about the reliability of the self-reporting system and its exclusion of non-salaried physicians. A seemingly more systematic approach was taken at BJC HealthCare with a centralized occupational database to track employee vaccinations and immune status using badge scanners, consent forms with carbon copies, a real-time database of vaccinated employees, and pre-printed labels with barcodes.⁶ The approach at University Hospital in Ohio was more human resource-driven and consisted of an administrative employee tasked with documenting which employees were vaccinated, an infection control practitioner to evaluate exemptions, and managers whose job it was to keep track of vaccinated and unvaccinated employees.²⁹ Regardless of the method, having a comprehensive and consistent documentation process to track and report immunization rates is instrumental to the success of an influenza prevention policy (Proposition 3).²⁹

Non-uniform disciplinary measures for unvaccinated employees were identified as an emergent policy implementation gap in years 2 and 3. Inconsistent disciplinary action created a contradictory standard wherein one non-compliant health-care provider’s contract was terminated, while others were seemingly able to ignore the policy without any ramifications. This finding is consistent with Nunn and colleagues’ (2018) quantitative investigation of the policy, which found that there is an ongoing need for human resources to be able to enforce the policy and education about how to enforce compliance in B.C. (Proposition 3).²⁷

In an effort to bridge ongoing and emergent implementation gaps, a Health Ministry-supported committee was formed to address some of the challenges relating to the provincial one-size-fits-all approach. The committee was instrumental in the evolution of the policy to include a provincially coordinated self-reporting system, the streamlining of the division of labor in the enforcement of the policy, human resources to enact disciplinary action for noncompliance, and specialized face masks for those employees whose ability to perform their job was hindered by the donning of a surgical mask. Communication and guidance from the oversight committee provided health authorities with a structured avenue to resolve policy gaps and was attributed to the success of the policy as it evolved. Authors of a study of mandatory influenza vaccination at a hospital in Ohio traced a similar policy evolution, noting that they had to invent ways to accommodate the mandate as they went along with implementation because the decision to mandate was made swiftly and without much advance notice.²⁹

Points of contention

A vaccinate-or-mask policy allowed for two different ways to be considered compliant (vaccinate or mask). This provided enough flexibility to soften the appearance of the mandatory nature of the policy, but established boundaries that, if crossed, had potential for consequences. In this way, the policy may be seen as a “choice with consequences,” meaning that an HCW can either decide to be compliant or can choose to neither get vaccinated nor wear a mask and risk having to leave work either temporarily or permanently.²⁶

Among the skeptics of the policy for years 2 and 3 were clinicians (especially physicians) and unions, which is consistent with our recent publication outlining results from the first year of implementation.¹¹ There were objections to the framing of the B.C. policy as one of the choice from these power brokers (i.e., physicians, unions) because they felt they were “choosing” to be vaccinated under pressure to avoid the physical burden and stigma associated with mask-wearing, and therefore that this was not a true policy of choice. In the matter of the vaccine-or-mask policy at St. Michael’s Hospital in September 2018, an arbitrator found that although the policy was not coercive, it was unreasonable, unsustainable, and inconsistent with and/or contrary to Ontario’s collective agreement.³⁰ Specifically, the arbitrator found there to be no persuasive evidence that masks were a reasonable alternative to getting vaccinated.³⁰ The arbitrator thus ruled against one such policy at St. Michael’s; a decision that was binding to several other hospitals in Ontario with similar policies in place.³⁰ In Saskatchewan, a mandatory vaccinate-or-mask policy was finalized and introduced in September 2014.³¹ The Medical Health Officer’s Council of Saskatchewan has since repealed this policy while health authorities review one of the rulings out of Ontario.^32,33 Unlike union dissent that took place in B.C. and elsewhere,^5,26,29,34 New Brunswick experienced union support that was identified as critical in the success of a vaccine-or-mask policy. A regional manager of employee Health and Wellness identified the supportive memo from the nurses’ union issued during implementation as the “biggest success of all”.²⁸

Further objections to the B.C. policy by clinicians and unions were related to the scientific evidence or perceived lack thereof, to justify the policy. These objections were noted in the implementation stages of the policy,¹¹ but have only become amplified in the years since with the BCNU publishing an article as recently as December 2018 entitled “B.C. Health Authorities Continue Non-Science-Based Vaccine-or-Mask Flu Policies”.³⁵ The scientific standoff between the MoH and unions, physicians, and other clinical staff is unique because they tended to use the same academic materials to either support or dispute vaccine effectiveness. There have been no new, large-scale studies that suggest otherwise. As such, there is limited empirical evidence of the effectiveness of this policy, which does not encourage a resolution in the scientific standoff between policy supporters and detractors. This conflicting interpretation of empirical evidence further reinforces the need for ongoing program evaluation and knowledge dissemination regarding policy effectiveness, as originally posited in study proposition 5 (see Table 1).

Habituation

Despite ongoing tensions surrounding the framing of the policy and the evidence for and against it, resistance waned with repeated exposure to the policy over years 2 and 3 and influenza vaccination became a routine part of a provincial and institutional “culture of safety”.⁵ Similarly, at Virginia Mason Hospital, the mandatory influenza vaccination program which was initially resisted by health-care providers, was increasingly seen as just another fitness-for-duty requirement alongside others like tuberculosis skin testing.⁵ Data from our study suggest that having sufficient human and financial capacity to be able to distribute responsibilities across a diverse team of influenza champions, peer immunizers, employees, and managers plays a significant role in the habituation of this policy. This finding supports study propositions 1, 3, and 4. Communicating consistent messaging about the objectives, expectations, and ways to be compliant with the policy were also identified as key factors in the habituation and success of the policy, as outlined in proposition 2.

Strengths and limitations

One of the major strengths of this study was the use of theoretical propositions, which allowed us to narrow the scope of our research question and focus the data collection and analysis. Another strength was that participation was voluntary and therefore we trust that, despite the contentious topic, the perceptions expressed in this study are candid and reflective of interviewees’ experiences.

This study was conducted in a single province in Canada. The reader is cautioned to consider the case parameters in judging the level of transferability to other jurisdictions. This research is also limited by the relatively small number of interviews with each occupational group since our sample was designed for maximum diversity rather than to reach saturation of themes within occupational groups.

Future directions

Other jurisdictions seeking to implement a similar policy ought to consult with their proposed program team, senior leadership, unions, and physicians about its suitability and rationale in the context of their practice.¹¹ The consultation process should also involve careful examination of the settings and contexts where the policy has been successfully implemented (i.e., British Columbia, New Brunswick), and those in which it has been repealed (i.e., St. Michael’s Hospital, Sault Ste. Marie Hospital). Should policy makers decide to develop and implement one such policy, they might consider the system, organizational, and individual-level learnings from our first paper on this topic.¹¹ Based on the lessons from years 2 and 3 of the policy, it is critical that policy makers allow the program to evolve to better suit the unique needs of jurisdictions and local organizations. This evolution, spurred by continued challenges in policy implementation, has allowed for facility-level innovations to improve uptake of the influenza vaccine and has drawn attention to the challenges of policy implementation in a geographically diverse province.

While the policy improved immunization coverage among staff, rates are still below established targets. Policies that have been successful in achieving higher coverage rates than B.C. have mandatory approaches including systematic documentation of vaccination, formalized application and review of medical or religious exemptions, and termination of contract for employees neither vaccinated nor exempted for failure to meet their condition of employment.^5,6

The objectives of the policy have remained the same since its first year of implementation, which has helped habituate the policy despite its evolution and ongoing debate about the evidence to support the policy. It would be valuable to have evidence from large-scale studies that can investigate the effect of the policy on influenza morbidity and mortality rather than vaccine coverage alone.

The future direction and activities of the B.C. program will depend in part on an understanding of policy implementation experiences and lessons learned, including barriers and enabling factors, indicators of success and challenges, and best practices explored in this paper. Despite a growing understanding of the need for improved HCW influenza immunization coverage, some jurisdictions in Canada have been reluctant to implement a vaccine-or-mask policy for a variety of reasons including conflicting arbitration decisions about the merits of these policies. The policy implementation in B.C. is an example of how a controversial policy can be adopted successfully. The continued evolution of the policy provides an opportunity to continue to examine and understand how condition-of-service influenza vaccination policies are implemented and accepted (or not) in Canada.

Disclosure of potential conflicts of interest

The authors report no financial interest or benefit that has arisen from the direct applications of this research.

Ethics

The research protocol was approved by Research Ethics Boards at St. Francis Xavier University (#21885), the University of British Columbia (# H12-03649), and the University of Toronto (#28544).

Additional information

Funding

This study was funded by the Public Health Agency of Canada and the Canadian Institutes of Health Research- Influenza Research Network (PCIRN).