Literature search
Our initial search yielded 231 publications. After screening titles and abstracts, 41 publications remained. These full-text articles were assessed and subsequently 23 studies were included. Additionally, 9 publications found through references of relevant literature were added. In total, 32 articles were reviewed (Fig. 1).
Of the selected studies, 14 were conducted in the USA, 5 in Italy, 3 in Australia, and 1 each in Canada, France, Germany, Israel, Japan, Korea, Qatar, Spain, Switzerland, and Turkey. Most (n = 28) studies compared VC before and after a specific vaccination campaign conducted at individual or clustered institutions. Other studies (n = 4) compared vaccination rates between an intervention group and a control group. The majority (n = 30) of the studies were performed in hospitals, while two studies only analyzed nursing homes. In the following sections, the outcome per key interventions is described. Further details such as number of subjects described in each study can be found in Table 1. One study is listed under two key interventions [11].
Table 1 Summary of interventions, study population, season (year) and vaccination coverage Key intervention: education and promotion
Among the selected studies, six built their campaign mainly upon educational and promotional aspects [12,13,14,15,16,17]. Overall, the key intervention education and promotion increased VC relatively by 65.9% (standard deviation (SD): ± 55.8%, range: 14.5–162.5%) (Table1).
In one randomized trial from Israel, the intervention group (n = 163) received a lecture session, recurring emails containing literature as well as reminders and an appointed key figure from each department personally talked to each participant of the intervention group. Compared to the initial VC of 27%, the final VC was 53% in the intervention group. The VC in the control group increased from 20 to 27% [12]. A cluster-randomized controlled trial conducted in French nursing homes included slideshows and posters regarding prejudices against and reasons for influenza vaccination. VC increased from 28 to 34% in the intervention group. VC decreased from 24 to 23% in the control group [13]. A 1-h training course for all participants concerning influenza vaccination guidelines, vaccine types and administration was used as an intervention in an Italian study. Subsequently, the initial VC of 16% increased to a final VC of 42% in the intervention group. In comparison, VC increased from 13 to 31% in the control group [14]. A different Italian hospital appealed on personal as well as patient safety. It comprised posters in frequented areas, distribution of factsheets and intranet presence. Most survey participants (66%) agreed that the information was useful. Following the implementation of the toolkit, vaccination coverage was 14% which corresponded to earlier VC of 10% [15]. During a Korean campaign, unvaccinated HCW were contacted via phone for a ten-minute educational presentation. As this had no effect, unvaccinated medical doctors then received one-on-one educational counseling with on-site vaccination (OSV). VC increased from 83 to 93% [16]. In a Spanish before-and-after-trial, the key intervention consisted of a “I’ve already been vaccinated” webpage showing humorous pictures of all heads of departments as well as a vaccinated pregnant woman promoting vaccination also during pregnancy. The authors concluded that the campaign encouraged the discourse on vaccination increasing VC from 23 to 37% [17].
Key intervention: incentives
Incentives were emphasized as a key intervention in one study. In six other studies, incentives were used as part of multifaceted campaigns [11, 17,18,19,20,21,22].
The above-mentioned study provided a 25 US Dollar gift card for every employee, if the overall VC reached 95%. This approach increased the VC from 87 to 92% [11]. The key intervention incentive increased VC relatively by 8.2% (Table1).
An employee-bonus program was implemented in two studies [18, 19]. A prize draw among vaccinated staff was part of two campaigns [17, 20]. In one multifaceted campaign, prizes where given to wards if the target VC was achieved [21]. One study took a different approach creating a disincentive for department leaders. Departments could lose budget allocations if vaccination rates were unsatisfactory. This increased vaccination rates from 87 to 92% [22].
The vaccine was offered free of charge in the respective prior season and during all included campaigns. Therefore, no aspect in this regard can be reported.
Key intervention: organizational strategies
Organizational aspects which facilitated access to the vaccine were implemented in eight studies [18, 21,22,23,24,25,26,27]; however, OSV was highlighted as a main intervention in only four campaigns [23, 25,26,27]. OSV as implemented key intervention increased VC overall by 113.6% (SD: ± 102.7%, range 30.8%–263.6%) (Table1). An Italian teaching hospital introduced OSV observing an increase in vaccination rates in medical residents from 10 to 18% [23]. In a different Italian study, the VC increased from 10 to 18% in the intervention group after offering OSV. Of note, out of the vaccinated HCW, 80% received vaccination on-site. In comparison, VC increased by 1.5% in the control group (without offered OSV). Initial and resulting overall VC was not provided by the authors for the control arm [26]. At another Italian hospital, a promotional campaign as well as OSV had already been in place in previous seasons with VC of 13%. Increased availability of the vaccine through extended OSV as well as longer timeslots at vaccination stations and at the occupational health department were added increasing VC to 17% [27]. After offering OSV, the VC increased from 11 to 40% in a Turkish children’s Hospital [25].
The following studies used special organizational strategies as part of their campaigns and are discussed under their respective subheading. An approach using peer-to-peer vaccination was taken by two hospitals [22, 24]. A flu kit including the vaccine, consent forms and stickers was handed out to appointed team leaders of individual departments [24]. In the second clinic nurse managers could receive vaccines from the Occupational Health Department to distribute among their personnel [22]. A “flu-stop-shop” in a main area was organized in one Australian study. During the campaign, HCW could receive vaccination at the “flu-stop-shop” at all times without appointment [21]. In another study, a “blitz” campaign was conducted during the first 2 weeks of October. Vaccination stations were set up at all entrances of the hospital. Consequently, about 70% of all employees were vaccinated in the first 2 weeks [18].
Key intervention: policies
Among the selected studies, 15 included policies as key interventions. Overall, policies increased VC relatively by 25.0% (SD: ± 31.3%, range 31.1%–97.9%) (Table1). One study conducted in the USA analyzed the effect of several different policies from 2008 to 2016. During seasons, in which policies included a signed declination option, the VC varied from 62 to 66%. Upon addition of educational aspects, VC increased to 86%. After a state-wide mandate in 2013, requiring unvaccinated staff to wear a mask, a maintained VC of 92–96% over the course of three seasons was reached [28].
In five studies, influenza vaccination was mandatory for HCW [29,30,31,32,33]. These publications were exclusively from the USA. Before implementation of the mandate, multifaceted vaccination campaigns had already been in place in all five studies with VC ranging from 54 to 80%. After influenza vaccination was made an employment requirement VC was 93–98%. In every study, “mandatory” implied that contracts with unvaccinated staff without exemptions were to be terminated. Overall, none to 0.14% of staff contracts were terminated due to the mandate. All five campaigns granted medical or religious exemptions. Egg allergy, history of Guillain–Barré syndrome and previously reported severe vaccine reaction were among the regarded exemptions. Exemptions due to medical reasons were acknowledged to 0.7–1.9% of staff and religious exemptions to 0.13–0.3%. One study declared that exemption requests reflected misinformation regarding the vaccine. These exemption requests included immunosuppression or pregnancy as reasons, although vaccination is recommended for both of these conditions [32]. Except for one hospital [29], the exempted unvaccinated staff had to wear a mask during influenza season.
A vaccinate-or-wear-a-mask approach was a key intervention in six publications [22, 24, 34,35,36,37]. A deadline for vaccination was set, after which unvaccinated staff had to wear a mask for the duration of the influenza season [34,35,36]. Vaccinated staff partially had markings on identification badges [22, 35, 36]. Supervisors were informed of their employees’ vaccination status and were held accountable in three campaigns [22, 24, 36]. One study implemented a 100 US Dollar fine for noncompliant staff [34]. Another study initially implemented contract termination as consequence of noncompliance, but was forced to retract due to litigation [35]. A sustained VC of 90–97% over 4 years, was achieved through a vaccinate-or-wear-a-mask policy in combination with a decentralized vaccine supply (complete vaccine kits for appointed team captains of different departments) in one study [24]. An Australian pilot study applied a vaccinate-or-wear-a-mask mandate in the nephrology department increasing VC from 47 to 93% (n = 208) [37]. Amid the six studies, three also included a declination form [22, 34, 36]. Overall, remarkable increases up to 97% in VC were observed after mask mandate [22, 24, 34,35,36,37].
Declination forms as a key intervention were used in three of the reviewed studies performed in Japan and the USA [11, 38, 39]. HCW refusing vaccination had to complete a declination form stating their reasons in all three studies. In a Japanese study noncompliant HCW, who neither received vaccine nor handed in declination forms, were interviewed by the hospital vice president. After implementing the mandatory declination form in this study, VC increased from 87 to 97% [38]. A pilot study conducted in a US Veterans Affairs facility included a signed statement acknowledging the personal risks and risks to others in their declination form. This study reported VC increasing from 54 to 77% [39]. Another study evaluated the impact of declination forms. Here, HCW refusing vaccination had to complete a 30-min educational module, receive one-on-one counseling and sign an attestation statement in presence of an occupational health or infection prevention staff. In cases of non-compliance, HCW were required to meet with their managers and a disciplinary letter was included in their employee file. This penalty-based approach increased VC from 92 to 96% [11]. Declination forms also played an important role in four multifaceted campaigns, which are discussed under the subheading “Combined interventions” [18, 19, 21, 40].
Combined interventions
The following studies are campaigns which did not focus on one key intervention but rather implemented three or more interventions as multifaceted strategies (education/promotion, incentive, organization, and policies) [18,19,20,21, 40,41,42]. Overall, combined interventions increased VC relatively by 14.4% (SD: ± 28.2%; range: − 20 to 88.1%) (Table1).
For one campaign a task force led by the Infection Prevention Department incorporating Employee Health, Pharmacy, and Nursing departments among others was created. A new policy was implemented which required employees to fill out either a consent, declination or exemption form. This included attestation of vaccination elsewhere. Vaccinated employees were asked to wear a badge saying “I’m vaccinated because I care”. If the badge wasn’t worn, employees had to wear a mask, regardless of vaccination status. Noncompliance was considered in performance evaluations hindering possible promotions or raises. As a financial incentive, an employee bonus program was implemented. This multifaceted campaign increased VC from 57 to 72% (in the 3 years prior to the campaign) to 92–93% sustaining for four years [18].
Similarly, an Australian campaign consisted of multiple key interventions. For 6 months each year, a full-time influenza vaccination coordinator was employed. Appointed nurses conducted the vaccinations in aforementioned “flu-stop-shop”. An intranet page with educational and promotional input was created. Promotions were spread via intranet, stickers and posters across the hospital. The chief executive officer sent emails and held presentations promoting the campaign. If wards achieved target rates, they received prizes. A mandatory declination form was implemented. Managers had access to the vaccination status of their employees via a database and were expected to hold their employees accountable. During the 6 active years of this campaign, VC was 79% to 82% compared to a VC ranging from 42 to 48% before [21].
One hospital implemented a new multifaceted strategy on top of OSV, a mobile cart, educational input and recurring e-mails. They added educational group meetings and a mandatory declination form. Also, progress reports on VC were sent to managers and heads of departments informing them of unvaccinated staff, yet without consequences for noncompliance. This increased VC from 71 to 93% [19].
Another Australian study introduced a database to track vaccination status of all HCW, identification of unvaccinated staff on ID badges, a declination form and awards for VC margins reached in wards (coffee machines in case of more than 80% VC). Following this campaign, VC increased from 56 to 80% [40].
A German hospital initiated the “Be a flu fighter” campaign, thereby managing to increase their VC by 4.5-fold. Key interventions included promotion and education, mobile vaccination teams and prize drawings as incentives among the vaccinated staff. Through the implementation of the campaign, VC reached 72% in physicians and 50% in nurses. Baseline values were not reported [20].
One hospital in Switzerland reported their influenza vaccination campaign being unsuccessful. The campaign included: vaccination daily during lunchtime in the cafeteria for 2 weeks, individualized mobile vaccination appointments at wards or during meetings, a “health week”, incentives such as free lottery ticket or a free lunch, educational and promotional flyers and posters, influenza vaccine logo, intranet presence including “frequently asked questions”, involvement of the head nurse, personal letters to employees and recurring lectures. According to the authors, the multitude of interventions, however, did not significantly increase VC (increase from 20 to 27% over 5 years). Among nurses the VC even decreased due to fear of potential short- or long-term side effects and doubts of efficacy of the vaccine [41].
As part of a quality improvement study, several plan–do–study–act (PDSA) cycles over the course of four seasons were implemented in one US study. The campaign consisted of educational aspects such as the distribution of a fact sheet and personal discussions on vaccination with HCW. Second, vaccine availability was increased in general and specifically for night shift staff and staff in remote clinics. Also, communicational aspects were enforced by sending out monthly emails showing current influenza epidemiology with a reminder of the availability of vaccination. Because “fear of needles” was identified as a barrier during a PDSA cycle, nasal vaccination was provided reducing this obstacle. Overall, VC increased from 70% to over 90% [42].
Descriptive comparison of key interventions
As shown in Fig. 2, key interventions such as education or promotion (n = 6) and organization (n = 4) were used as interventions in campaigns with initially low VC (range 15–25%). Policies (n = 15) combined interventions (n = 7) and incentives (n = 1) were applied in studies with initially high VC (> 70%). In studies with low initial VC, the key intervention led to an increase of the VC ranging from 11 to 18% for organizational interventions and 25–40% for education/promotion. In studies with high initial VC, the key intervention led to an increase of the VC from 79 to 92% for policies and from 85 to 92% for incentives. No change was observed for combined interventions. In the overall group (n = 32, all studies), VC increased from 71 to 87%.