Vaccination coverage for seasonal influenza among residents and health care workers in Norwegian nursing homes during the 2012/13 season, a cross-sectional study

Residents in nursing homes (NHs) and long-term care facilities (LTCF) are at higher risk of severe respiratory tract infections owing to old age, the number of underlying illnesses, and close living conditions [1‐3]. The main vaccine- preventable respiratory tract infection is caused by the influenza virus. The infection can be severe and even lethal, and viral infections often predispose sufferers to bacterial secondary infections and complications [4]. Further, although the lethality of influenza infection is low, the resulting reduced general condition following infection is of major importance in the elderly. Accordingly, the World Health Organization (WHO) defines NHs residents as a risk group for severe influenza and recommends vaccination coverage (VC) of at least 75%. The main goals are to reduce risk of severe disease and to prevent outbreaks. The WHO recommendation also includes influenza vaccination of health care workers (HCWs) in order to prevent introduction of the disease into healthcare institutions [5, 6]. The Norwegian guidelines for infectious disease control follow the WHO recommendations relating to seasonal influenza, the aim being to achieve a minimum VC of 75% in Norwegian NHs [7]. Seasonal influenza vaccination for residents in NHs is reimbursed by the state in Norway. This is in contrast to influenza vaccination of HCWs, for whom each employer decides whether it is to be given free of charge. The yearly influenza vaccination is normally conducted during October and November [7].

The Vaccine European New Integrated Collaboration Effort (VENICE) report, covering the 2011/2012 season, showed seasonal influenza VC for the general Norwegian population above 65 years and for all HCWs to be 36% and 12% respectively [8]. However, the VC among residents and HCWs in NHs is unknown. To guide development of vaccination programs for residents and HCWs in NHs, data on VC is crucial. We therefore conducted a cross-sectional study approaching all NHs in Norway in order to estimate VC for seasonal influenza vaccine among both residents and HCWs. In addition, we investigated whether there was an association between VC and response rate by county, yearly vaccination campaigns among residents and HCWs, free-of-charge vaccination of HCWs, NH size (number of residents) and geographic location.

How many of the 910 NHs actually received the invitation is unknown since distribution was by means of generic municipality e-mail addresses. However, 391 of the intended 910 invitees in Norway (43.0%) responded to our survey. We excluded 37 NHs on the basis of incomplete data; thus 354 (38.9%) NHs were included in the analysis. These 354 NHs were located in 244 (57%) of the 429 municipalities in Norway, with representatives from all 19 counties. The response rate by county ranged from 26.7% in Buskerud to 65.5% in Finnmark (Table 1).

The 354 NHs included in the study provided aggregated information on a total of 14,208 residents and 28,237 HCWs. The median number of long-term-care residents per NH was 33 (range 1-168). The median number of HCWs per NH was 60 (range 15-400). The high number of HCWs is explained by the fact that we included aggregated data about all NH HCWs in our survey, independent of NH department, since they usually rotate and hence work both in long- and short-term-care.

To the best of our knowledge, this is the first study to aim at providing data on VC for seasonal influenza vaccine among residents and HCWs in Norwegian NHs.

In around half of the NHs, VC for seasonal influenza vaccination among residents was below the 75% recommended by WHO and laid down in National guidelines. However, the VC of NH residents was markedly higher than that in the general population above 65 years of age [8]. This may reflect both the impact of the vaccination recommendations’ specifically focusing on NH residents and awareness of vaccination of residents as an important infection control measure. There are, however, still NHs with no residents vaccinated; thus a substantial effort is still required in order to reach the overall target of 75%.

Vaccination of HCWs has been shown to play an important role in controlling transmission of influenza to residents [5]. A review from the Norwegian Knowledge Center for the Health Services came to the conclusion that influenza vaccination in HCWs reduced the risk of acquiring influenza- like illness in residents by 50% [11]. On the other hand, a recent Cochrane review shows that, when using laboratory confirmed influenza as an endpoint [12], there is insufficient evidence to prove the effect of vaccinating HCWs in relation to the transmission of influenza to residents. The Cochrane review, however, did not take influenza like illness (ILI) and all-cause mortality of residents into consideration, The impact of vaccinating HCWs has been discussed in several articles. Potter [13] and Carman [14] took ILI into consideration in their studies and found a significant association between HCWs’ vaccination against influenza and reduced rates of mortality in residents of NHs, showing how important it is to vaccinate HCWs. HCW absence from work owing to influenza is another important issue and has been studied by several authors [13, 15‐17]. These studies demonstrate reductions in work-time lost owing to illness among vaccinated HCWs. Moreover, in all of these studies, unvaccinated HCWs with influenza continued to work while symptomatic and may have infected residents. WHO also recommends the vaccination of HCWs against seasonal influenza in two other respects: One, as an effective measure in decreasing the extent of seasonal influenza and two, as a measure towards maintaining an active workforce during influenza epidemics [5, 6]. As studies in other European countries have shown [18‐22], we expected a low VC. Our results are markedly lower than the results of the VENICE project on influenza VC among all HCWs (12%), which is based on a telephone survey of the general Norwegian population and the numbers of influenza vaccines sold [8]. Most HCWs, however, are employed by hospitals, which conduct more comprehensive infection control. It is thus not unlikely that VC among HCWs in hospitals is higher than in NHs. We therefore suppose that our findings are more likely to reflect the actual VC among HCWs in NHs.

In order to raise VC among HCWs, several hospitals in the USA have already made seasonal influenza vaccination mandatory for HCWs with direct patient contact [23, 24]. In the Netherlands too, mandatory vaccination programs are being discussed owing to low VC among HCWs working in NHs. The authors of a Dutch study concerning the ethical aspects of mandatory vaccination conclude that NHs have a moral responsibility to implement vaccination programs with a view to getting their HCWs to accept voluntary vaccination, but that this does not exclude the possibility of introducing mandatory vaccination if VC rates don’t rise [25].

The results of our study show that having to pay for the vaccine does not affect the VC of HCWs while having annual vaccination campaigns does lead to increases in VC. In the Netherlands, vaccination campaigns have also been found to increase VC while, interestingly, having to pay for the vaccine as a HCW in a hospital actually decreased VC [22].

The low VC among HCWs warrants studies to see why they are so reluctant to be vaccinated. Specifically, research into the reason for low VC in HCWs, as well as into behavioral changes that are positively associated with VC are needed. In the Netherlands, studies have revealed a number of important, significant attitudes held by HCWs in relation to influenza vaccination. These include the responsibility not to harm patients, the knowledge that the vaccine has high efficacy, and the belief that they (HCWs) are in a high risk group in relation to contracting influenza infection [22, 25, 26].

One limitation in our study is the low response rate. We initially intended to reach all NHs through the generic e-mail addresses of the municipalities, as we do not have access to direct telephone and e-mail lists. More than half of the NHs did not reply and any generalization from the results should therefore be treated with caution. While the e-mail modality is fast and convenient and allows for the respondents to collect the requested data, it may also result in lower response rates compared to telephone modality interviews as undertaken in Venice [8, 27]. We did not, however, find an association between the response rate per county and the VC. All the same, non-participating NHs could be vaccinating fewer residents, resulting in an overestimation of the VC and introducing bias towards “the best in the class”.

The extent to which NHs document vaccination is unknown and we are unable to state if, and how much, this influences our results.

From the results of our study, we are able to conclude that the median influenza VC among residents of the participating NHs is not much lower than recommended. On the other hand, influenza VC among HCWs in the participating NHs is exceedingly low. VC in HCWs was significantly higher when NHs had an annual vaccination campaign. We recommend immediate measures be implemented to increase VC in residents and, especially, in HCWs. These measures should encompass vaccination campaigns and studies to identify potential barriers to vaccination. In order to achieve more comprehensive and accurate data in future, a direct communication channel with NHs should be established.