Likely correlation between sources of information and acceptability of A/H1N1 swine-origin influenza virus vaccine in Marseille, France

Abstract

Background: In France, there was a reluctance to accept vaccination against the A/H1N1 pandemic influenza virus despite government recommendation and investment in the vaccine programme.

Methods and findings: We examined the willingness of different populations to accept A/H1N1 vaccination (i) in a French hospital among 3315 employees immunized either by in-house medical personnel or mobile teams of MDs and (ii) in a shelter housing 250 homeless persons. Google was used to assess the volume of enquiries concerning incidence of influenza. We analyzed the information on vaccination provided by Google, the website of the major French newspapers, and PubMed. Two trust Surveys were used to assess public opinion on the trustworthiness of people in different professions. Paramedics were significantly more reluctant to accept immunisation than qualified medical staff. Acceptance was significantly increased when recommended directly by MDs. Anecdotal cases of directly observed severe infections were followed by enhanced acceptance of paramedical staff. Scientific literature was significantly more in favour of vaccination than Google and French newspaper websites. In the case of the newspaper websites, information correlated with their recognised political reputations, although they would presumably claim independence from political bias. The Trust Surveys showed that politicians were highly dis-trusted in contrast with doctors and pharmacists who were considered much more trustworthy.

Conclusions: The low uptake of the vaccine could reflect failure to convey high quality medical information and advice relating to the benefits of being vaccinated. We believe that the media and internet contributed to this problem by raising concerns within the general population and that failure to involve GPs in the control programme may have been a mistake. GPs are highly regarded by the public and can provide face-to-face professional advice and information. The top-down strategy of vaccine programme management and information delivered by the Ministry of Health could have aggravated the problem, because the general population does not always trust politicians.

Figure 1. Time distribution of the 994 HCWs vaccinated in the LCPH.

Figure 2. Distribution of vaccination uptake for the 994 HCWs vaccinated in the LCPH.

OMU: Occupational medicine unit. MVF: Mobile vaccination facility.

Figure 3. Time distribution of the vaccination by HCWs categories in the LCPH.

Figure 3A indicates the vaccination coverage and figure 3B indicates the probability of being immunized (chi square for trend; the period from October 31^st to November 6^th was considered as baseline). Event 1: First patient admitted in ICU. Event 2: Two pregnant women admitted in ICU. p-value (chi square for trend test): 0.30, <0.001, <0.0001 and <0.01 for medical HCWs, midwife nurses, nurses and auxiliary nurses respectively.

Figure 4. Comparison of Google enquiries in France with data from our Virology laboratory.

Figure 4A compares Google enquiry data about RSV infections and figure 4B compares data about influenza.

Figure 5. Comparison of Google enquiries in France with data from our Virology laboratory about the novel A/H1N1 Influenza virus.

Google enquiries were compared with the number of positive samples (figure 5A) and the number of samples tested (figure 5B). The peaks observed with GIFS were noted P1, P2 and P3.

Figure 6. Proportion of Google enquiries in France due to Influenza vaccination.