Promotion of flu vaccination among healthcare workers in an Italian academic hospital: An experience with tailored web tools

ABSTRACT

Background: Influenza causes significant mortality particularly among the elderly and high-risk groups. Healthcare workers (HCWs) are at risk of occupational exposure due to contact with patients. Aims of this study was to promote flu shot among HCWs through a multimedia campaign in a large North-Eastern Italian Hospital.

Methods: The 2013/2014 flu vaccination multimedia campaign addressed to HCWs was developed by maintaining pre-existing tools (letters in pay slip and poster displayed in wards) and creating 4 on-line spots (30") delivered trough the hospital intranet. Campaign effectiveness was assessed in terms of changes in knowledge, attitude and practice comparing data of pre (10 items) and post test (20 items) survey on a randomized sample of HCWs.

Results: Response rates were 92.6% (464/501) in pre-test and 83.2% (417/501) in post-test. 93.8% (391/417) of HCWs reported to awareness of the campaign to promote vaccination. Spots were seen by 59.6% (233/391) of HCWs. Some reasons for vaccine denial, “not believing in vaccine efficacy” (34.7% to 14.9%), “not considering flu as a serious problem” (from 24% to 12.6%), “thinking not to get sick” (28.7% to 18.2%) or “being against the vaccine” (32.7% to 21%), showed a statistically significant reduction after the exposure to the campaign. The “intention to get vaccinated in the next year” instead, raised effectively (13.1% to 36.6%). Vaccinated HCWs rate in 2013-2014 season was 7.6% (221/2910), and 5.6% (164/2910) in 2012-2013 (p<0.005).

Conclusions: The multimedia campaign succeeded with regard to KAP outcomes, but the vaccination rate is still far from the goal of 90%. Due to their impact, especially on younger age groups, web tools deserve to be better studied as effective approach to convey health information among HCWs.

Introduction

Influenza is an important public health problem that causes significant mortality particularly among the elderly and high-risk groups.¹ Healthcare workers (HCWs) are at risk of occupational exposure due to contact with patients who may carry the virus.² This fact implies that HCWs may act as vectors for the transmission to vulnerable patients (elderly, immune-compromised patients and infants), for whom the infection may be associated with considerable morbidity and mortality.³

Absenteism peaks up to 20% are reported during influenza outbreaks among HCWs; staff shortages lead to discontinuance in assistance to patients⁴ and heavy work overloads for those who remain on duty⁵ This results in extra costs for the hospital due to overtime work and longer stays of patients.⁶ Due to the risk of influenza transmission up to 1 day before developing any symptom, it is not sufficient for the HCWs to simply stay home from work when they begin showing symptoms in order to prevent transmission in hospital settings.⁷

The most efficient strategy to prevent a significant number of influenza infections is an annual vaccination of HCWs, which is associated with reduction of morbidity and mortality among patients as well.⁸ HCWs are required to take influenza vaccine to protect themselves and their patients.^9,10 Vaccination is highly effective with minimal adverse effects and is highly cost-effective.⁵

Despite recommendations by public-health authorities and studies demonstrating the dangers of flu and the benefits of vaccination,^11,12 the rates among HCWs worldwide show no improvements.^13-15 The acceptance of influenza vaccination remains low in most of western countries, with coverage rates ranging from 4 to 40%.¹⁴ Italy has among the worst results in terms of uptake: 12.2% in 2006/2007 season and 10.9% in the 2007/2008.^16,17

It is crucial to understand the reasons for accepting or declining flu shot.¹⁸ Literature identifies as reasons for acceptance: self-protection, awareness of the severity of influenza, its complications, and the efficacy of the vaccine¹⁹; protection of patients;^20-22 avoiding absenteeism at work and serving as a role model.²³ The reasons for declining include: concerns about side effects, allergic reactions or infection from the vaccine itself, doubts about the efficacy of the vaccine and beliefs not being at risk.²⁴

Strategies and interventions to improve vaccination uptake in healthcare settings are required.⁵ HCWs are generally invited to get vaccinated trough standard strategies or communication techniques, including written reminders in pay slips, posters or flyers, letters or e-mails.^25,26 No experiences of wide and integrated employ of multimedia tools, to promote campaigns and reach higher rates of compliance, are reported .

Aims of this study was to plan, implement and evaluate a campaign promoting flu shot among HCWs, in terms of changes in knowledge, attitude and practice (KAP).

Results

Pre-test results

Pre-test questionnaire response rate was 92.6% (464/501). There are not significant statistically differences between the 2 sample of pre-test and post-test respondents stratified by occupation, age and gender (Table 1).

Table 1.

Pre and post-test respondents stratified by occupation, age and gender.

		PRE-TEST (N. 464)		POST-TEST (N. 417)
		N.	%	N.	%
OCCUPATION	Physician	102	22.0	90	21.6
	Residents	29	6.3	30	7.2
	Nurses	224	48.3	199	47.7
	Nurse assistants	81	17.5	76	18.2
	Technical staff	24	5.2	18	4.3
	Other	4	0.9	4	1.0
AGE	18-35	119	25.6	133	31.9
	36-45	163	35.1	122	29.3
	46-55	152	32.8	131	31.4
	>56	30	6.5	31	7.4
GENDER	Male	126	27.2	112	26.9
	Female	338	72.8	305	73.1

Pre-test survey assessed the familiarity with media among HCWs. Ninety-five percent (441/464) of the interviewed employees were aware of the hospital website, 93.3% (433/464) of e-mail, 80.0% (371/464) of social networks (Facebook, Twitter, etc.), 77.4% (359/464) of streaming video (YouTube, ecc) and 73.9% (343/464) of Apps.

In pre-test 50.9% (236/464) of respondents wanted to get information about vaccine safety, 26.5% (123/464) about influenza mortality in hospitalised population, 25.2% (117/464) about the protection of relatives from flu and 13.4% (62/464) required in-depth information on other topics (e.g. vaccine effectiveness, side effects).

In pre-test 29.5% (137/464) of interviewed respondents reported to have been vaccinated at least once in the last 5 years.

Post-test results

Post-test questionnaire response rate was 83.2% (417/501).

Post-test showed the intended audience exposure to the flu vaccination campaign through the HCWs' awareness: 93.8% (391/417) of respondents reported to know that a campaign to promote influenza vaccination had taken place in hospital, 5.8% (24/417) claimed not to know if a campaign had been organized, and 0.5% (2/417) thought that hospital had not organized it. 92.3% (361/391) of HCWs recalled to have seen the posters and 24.3% (95/391) recalled the letter in the pay slip.

The campaign spots were seen by 59.6% (233/391) of HCWs and all of them remembered that they contained short messages by colleagues. 94.8% (221/233) of HCWs saw the spots in the hospital website, 19.3% (45/233) also on social networks, and 12% (28/233) on mobile devices.

The campaign also stimulated the research of information about influenza and vaccine among the employees[29.9% (117/391)]. Collected informations were:vaccine safety [70.9% (83/117)], protection of relatives [41.0% (48/117)], influenza mortality in hospitalised population [17.1% (20/117)] and other topics (e.g., vaccine effectiveness, side effects) [4.3% (5/117)].

Regarding the influence of the campaign on the decision of HCWs to get or not to get vaccinated in the current season, 47.8% (187/391) of them reported that the campaign had not had any influence on their decision, 28.4% (111/391) referred a little influence, 22.8% (89/391) a moderate influence and 1% (4/391) a total influence.

Campaign impact

Campaign impact was measured with KAP outcomes comparing pre and post-questionnaires.

Agreement with sentences expressing beliefs about influenza and flu vaccination was considered as a knowledge indicator (K outcome). Statements obtaining a statistically significant difference in percentage distribution of all respondents of pre-test and post-test were those related to: “not believe in vaccine efficacy,” “not consider flu as a serious problem” and “think not to get sick” (Table 2). Stratified by age and occupation, the difference remained statistically significant for the statement "not believe in vaccine efficacy" among age groups 36-55 [p < 0.001], 46-55 [p < 0.001], >56 [p < 0.05], physicians [p < 0.01], residents [p < 0.05], nurses [p < 0.001] and nurse assistants [p < 0.001]; for the statement "not consider flu as a serious problem" among age groups 18-35 [p < 0.001], 45-55 [p < 0.05],nurses [p < 0.01], nurses assistants [p < 0.001] and technical staff [p < 0.001]; for the statement “think not to get sick” among age group 18-35 [p < 0.01], physicians [p < 0.05] and technical staff [p < 0.01].

Table 2.

Agreement/disagreement with sentences expressing beliefs about influenza and flu vaccination in post-test survey (tot = 391) compared with pre-test results (tot = 464).

	PRE TEST (N. 464)					POST TEST (N. 391)
	Agree		Disagree			Agree		Disagree
	Strongly	Mildly	Strongly	Mildly	Not expressed	Strongly	Mildly	Strongly	Mildly	P value
BELIEFS
Not believing in vaccine efficacy	7,1% (33)	27,6% (128)	34,7% (161)	28,0% (130)	2,6% (12)	3,6% (14)	11,3% (44)	36,3% (142)	48,8% (191)	<0.0001
Not considering flu as a serious problem	6,5% (30)	17,5% (81)	41,2% (191)	29,5% (137)	5,4% (25)	3,1% (12)	9,5% (37)	54,2% (212)	33,2% (130)	<0.0001
Thinking not to get sick	9,3% (43)	19,4% (90)	28,9% (134)	26,9% (171)	5,6% (26)	6,4% (25)	11,8% (46)	36,8% (144)	45,0% (176)	<0.001
Thinking not to be part of a high-risk group	13,1% (61)	14,2% (66)	56,0% (260)	11,0% (51)	5,6% (26)	12,8% (50)	14,3% (56)	58,8% (230)	14,1% (55)	n.s
ATTITUDES
Being against the vaccine	7,5% (35)	25,2% (117)	40,5% (188)	23,3% (108)	3,4% (16)	5,4% (21)	15,6% (61)	48,1% (188)	30,9% (121)	<0.0001
Not having time to get vaccinated	2,4% (11)	7,1% (33)	67,9% (315)	17,0% (79)	5,6% (26)	1,5% (6)	4,9% (19)	69,6% (272)	24,0% (94)	n.s
Not having contacts with patients	8,4% (39)	4,1% (19)	75,0% (348)	5,2% (24)	7,3% (34)	9,7% (38)	4,3% (17)	78,8% (308)	7,2% (28)	n.s
Being afraid of needles	1,3% (6)	3,4% (16)	81,3% (377)	8,4% (39)	5,6% (26)	0,8% (3)	3,3% (13)	88,2% (345)	7,7% (30)	n.s

The first indicator explored that concerned attitudes (A outcome) was the agreement with sentences expressing resistance factors in getting flu vaccination. The only statement obtaining a statistically significant difference in percentage distribution of all respondents of pre-test and post-test was this related to "being against the vaccines" (Table 2). Stratified by age and occupation, the difference remained statistically significant among age group 36-45 [p < 0.01], nurses [p < 0.05] and technical staff [p < 0.0001].

Another attitude indicator was the intention to get vaccinated (Table 3). There was a statistically significant difference in percentage distribution of all respondents of pre-test and post-test about intention to get vaccinated in the next year; stratified by age and occupation, the difference remained statistically significant among age groups 18-35 [p < 0.0001], 36-45 [p < 0.0001], 46-55 [p < 0.0001], physicians [p < 0.01], residents [p < 0.01], nurses [p < 0.0001], nurse assistants [p < 0.05] and technical staff [p < 0.0001].

Table 3.

Comparison between percentages of HCWs respondents with regard to self-reported vaccination practice in the current year and to intention to get vaccinated in the next year in post-test survey (tot = 417) compared with pre-test results (tot = 464).

Self-reported vaccination practice in current year	PRE % (N.)	POST % (N.)	OR (CI 95%) p value
Yes	8.6% (40)	13.7% (57)	1.84 (1.17-2.88) p<0.005
No	91.4% (464)	86.3 (360)
Intention to get vaccinated in the next season	PRE % (N.)	POST % (N.)	OR (CI 95%) p value
Yes	13.1% (61)	36.6% (143)	3.15 (2.21-4.51) p<0.0001
No	72.0% (334)	63.4% (248)
Don't know	14.9% (69)	6.2% (26)

Getting vaccination was the indicator of behavior change (P outcome). Differences of percentage of HCWs reporting that they were vaccinated against flu in pre-test and post-test questionnaires were statistically significant (Table 3); stratified by age and occupation, the increasing trend remained statistically significant among age group 18-35 [p < 0.05] and nurses [p < 0.05].

Interviewed HCWs that declared to have had flu in the 2013-14 season were 13.7% (57/417) and 94.7% (54/57) of them had not got vaccinated; 86% (49/57) of HCWs who had had flu stayed at home losing working days [mean = 4.9 days (SD = 2.7 days; range 1-15 days)]

Finally, from the data provided by the hospital Prevention Department, the percentage of HCWs vaccinated against flu in the 2013-2014 season was 7.6% (221/2910), and 5.6% (164/2910) in the 2012-2013 flu season (p < 0.005).

The increase in vaccination rate was statistically significant for physicians [2013: 18.1% (115/636) vs 2012: 13.2% (84/635); p < 0.05] and for residents [2013: 16.2% (51/314) vs 2012: 4.4% (14/317); p < 0.001], whereas for the others HCWs (e.g. nurses, physiotherapists, nursing assistants) it was not statistically significant.

Discussion

The study obtained that almost all HCWs (94%) were aware of the campaign to promote vaccination. This confirms the utility of multimedia to disseminate health information and to support campaigns aiming to produce positive changes.²⁷

Although in pre-test survey almost all HCWs declared to use website, only half of post-test respondents saw the videos, which are the new tool used in the 2013-14 campaign. Indeed, posters remained the tool that reached most of the interviewed HCWs. This confirms²⁵ that print material, as posters and leaflets, have a relevant role to support and enforce multimedia and media campaigns. Pre-campaign survey outcomes suggested that HCWs were not well informed about the vaccine or its effectiveness. Only one third of interviewed HCWs strongly believed in vaccination efficacy and was sure that without flu shot they could get sick, about half of respondents had no doubts that flu is a serious problem and was totally conscious to be part of an high risk group.

This finding is similar to other literature that found misconceptions about the effectiveness and safety of influenza vaccination.^24,28,29

The campaign reached a statistically significant outcome in changing beliefs about vaccine efficacy, about the consideration that flu is a serious problem and about the possibility of getting sick without flu shot. However, there were no changes in thinking not to be part of a high-risk group. The stratification by age and occupation revealed that the best outcome in knowledge was obtained with regard to belief in vaccine efficacy, especially among HCWs over 36 years and among physicians, residents, nurses and nurse assistants. This is an interesting result indicating a potential need for tailoring future campaigns among HCWs. Also the literature reports that hospital employees accept vaccination if they have more evidence of vaccine effectiveness and information about side effects.³⁰ Moreover, the awareness of being a potential source to transmit influenza to patients is associated with vaccine receipt.³¹

Elements of resistance to getting flu vaccination were modified by the campaign among all HCWs, in particular among the nurses, the technical staff and the 36-45 age group. The data of the survey showed a success of the campaign in changing attitudes among nurses, generally considered a critical group due to low vaccination rates.^16,32 At the same time the data stressed the difficulty of persuading physicians and young HCWs to accept the vaccine. The literature indicates that programs may need to be profession-sensitive as well as specific in their approach because attitudes concerning influenza vaccination differ largely between occupational categories.^16,32

Understanding audience and tailoring educational efforts, according to its needs, to remove barriers and change behaviors is critical for an effective health communication. This suggests that investments for each hospital or organization, in terms of development of official channels, capable of reaching the vast majority of the HCWs, to spread institutional and influential message, deserve more in-depth analysis for cost-effectiveness. An hospital branded channel could also be used for the questions frequently asked by employees about vaccines, and other preventive measures, in order to offer a certified database with the most searched information's, in many cases related to misconceptions. The greater tendency of the video-exposed HCWs to the research of further information about vaccine related topics is certainly promising and support the idea that granting easier access to information can support the uptake of good practices. Although the literature demonstrated²⁴ that in certain cases time is as an important reason behind HCWs' failure to be vaccinated, in this experimental study not having time to get vaccinated was not a motivation for non-adherence, in both tests. This suggests that the strategy to offer flu shot in all the wards may be effective. In our experience, the resistance factors described in literature¹⁸ of not being in contact with patients and being afraid of needles were not reported as important points to be modified in future campaigns.

The reported intention to get vaccinated in the next season increased significantly in post-test survey in all occupational groups and in HCWs with an age between 18 and 55 years.

Percentages of HCWs reporting to get vaccinated in the 2013-2014 (13.7%), compared with the previous season (8.6%) increased in a statistically significant way, in particular among nurses and the 18-35 age group, even if the total absolute number of self-reported vaccinated personnel remained very low.

Although the results provided by the Prevention Department show an increase in terms of adhesion to the flu vaccination , the coverage is still limited and not significant in terms of controlling the flu spread. However in a perspective of continuous investments to achieve an higher and permanent rate of compliant HCWs, this little significant change in the attitudes could be a first step in a longer process.

Limitations of the study

The comparison between pre and post test shows a good homogeneity in terms of categories and age groups represented. Although both questionnaires were addressed to the same subjects, the anonymity of the forms doesn't allow to presume that the 2 groups fit exactly. Nevertheless this submission method prevented any kind of influence on the interviewed HCWs. The intranet spots have been viewed only by 59.6% of the sample population. This could lead to an underestimation of the effect of the multimedia tools in terms of modification of attitudes and beliefs of the HCW's.

Methods

Objectives and steps of the study

The experimental study was carried out from June 2013 to September 2014. It consisted in planning, implementing and evaluating a multimedia campaign in a large North-Eastern Italian Academic Hospital.

The campaign goal was to increase HCWs adherence to flu vaccination.

Steps adopted to develop communication campaign were those suggested by the US Department of Health and Human Services³³: planning, communication objective, setting, audience, channels and message, development and implementation, evaluation of outcomes.

Planning

The communication-persuasion model was adopted as the theoretical framework.

Channels and tools to reach the intended audience (HCWs) were established. An evaluation plan to assess the impact of the campaign was predisposed.

Theoretical framework: Communication persuasion model

The communication–persuasion model, which describes the steps that an individual must be persuaded to assimilate a desired behavior.³⁴ The chosen model was suitable for rapid mass media advertising.^34,35 The model can be characterized as an input-output matrix that can be manipulated and measured to achieve a change. Interdependency between output and input factors influences the results of the process. Output factors describe how conscious and unconscious cognitive processes are influenced by communication.³⁶

Setting, audience and channel definition

The 2013/2014 flu vaccination campaign was addressed to all HCWs of the Academic Hospital (N = 2910).

Past flu vaccination campaigns in the Hospital used print material to communicate importance, availability and schedules. Nevertheless vaccination coverage among HCWs never exceeded 15%.

The 2013/2014 flu vaccination campaign maintained pre-existing tools (letters in pay slip and poster displayed in wards) but new tools were developed: on-line spots hosted on intranet. Active vaccine offer in each ward was scheduled.

Message plan

Following the evidence-based features of effective messages,³⁷ the ones developed for the campaign resulted: accurate (reviewed by experts), consistent (same message, color scheme and actors in spots and posters), clear (messages were short - just one key point- and simple- language was basic). Final recommendation was explicit (Have a flu shot!) and repeated. The message also was: relevant (the recommended action ensured health benefits), plausible (trustworthy figures, chosen to represent the audience, in terms of demographics and educational backgrounds) and appealing (all actors wore their uniforms and were chosen to stimulate the curiosity of colleagues).

Tools

The previously planned message was conveyed by 4 on-line spots (30"), whose protagonists were HCWs. A group of volunteers was recruited among all categories: physicians, nurses, nursing assistants. The CEO of the Hospital appeared in all the spots. Each spot had the same structure. It started with a rational appeal, containing scientific and statistic informations and an irrefutable fact for the audience (e.g., "Protecting patients is my duty!").

The spots were promoted by: a browser pop-up and posters portraying the protagonists in the pieces of a puzzle with the message "Each piece is important!" and the instructions to receive the vaccination (days, time, place).

The on-line spots were hosted on the Hospital intranet.

The posters were displayed in all the wards.

Evaluation of campaign impact

Campaign effectiveness was assessed comparing data of pre and post test survey on intended audiences. Outcomes in Knowledge, Attitude and Practice (KAP)³⁸ were explored to define the impact.

Pre- and post-campaign questionnaires were developed and they are available on request. The pre-campaign test ("You and Flu") included 10 items focused on KAP and was tested in a group of HCWs (N = 10). The post-campaign test ("You after Flu") included 20 items including the campaign evaluation.

The sample size was calculated using the formula of estimating a single population portion, taking 15% proportion of 3% margin of error and 95% confidence level. The sample (N = 501) was obtained from the hospital HCWs population (N = 2910) by a stratified randomization (physician, residents, nurses, nurse assistants, technical staff, others).

The questionnaires were distributed to the same sample using drop-off methods³⁹: a staff of consultants and head nurses were involved to deliver and collect them, encouraging the completion.

Data were entered in an Excel spreadsheet and analyzed using the statistical software SPSS, v.20. Changes in the distribution of respondents' percentages (stratified by HCWs' age and occupation) about the same items in pre-test and post-test survey were analyzed with the Chi-square test. Statistical significance was defined as p ≤ 0.05.

Pre-test data were compared with post-test data to evaluate changes in KAP³⁸ among HCWs after the campaign.

Data about vaccinated HCWs during the 2012-2013 and 2013-2014 flu campaign were provided by the hospital health prevention team.

Conclusions

The multimedia campaign succeeded with regard to all KAP outcomes, but the vaccination coverage is still far from the goal standard of 90% of HCWs.

Mass media, particularly those based on web techniques, should be better studied as effective tools to convey health information. Key messages need to be designed carefully according to literature and knowledge of local beliefs and attitudes.

Younger age groups confirm their greater sensitivity to these communicative strategies. Investments in web and mobile technologies, as tools to spread institutional and influential messages, deserve more cost-effectiveness analysis.

Complementary policies, involving interpersonal techniques, should also be considered to maximize mass media communication effects.

Disclosure of potential conflicts of interest

This study was conducted without external funding. The campaign was developed as part of a PhD research. Activities were conducted on a smaller scale and using services available in the hospital including human resources and tools (printing materials, computer, etc).

None of the authors have declared any kind of conflict of interest.

Acknowledgments

We thank all volunteer actors of the campaign spots: Prof. Matteo Balestrieri, Dr. Valentina Capodicasa, Dr. Roberto Cocconi, Fabio Colle, Dr. Luigi Conte, Dr. Pierpaolo Dal Secco, Federica Del Fabro, Dr. Mauro Delendi, Laura Di Meo, Dr. Roberta Di Vora, Dr. Fabiola Feletto, Dr. Lili Bednarova, Dr. Sandra Bednarova, Dr. Francesco Londero, Daniela Manfredo, Andrea Marco, Dr. Maria Teresa Grillo, Dr. Massimo Mondani, Piero Pasini, Mellinda Picotti, Dr. Federica Pigani, Massimiliano Pividore, Prof. Andrea Risaliti, Carla Toffoli, Dr. Francesco Tuniz. We thank Dr Piero Pascolo and his equipe for the production of the spots.

Thanks to Prof. Umberto Gelatti and Prof. Peter Schulz for their very useful comments.