Is there a case for mandating influenza vaccination in healthcare workers?

Ten years ago, Virginia Mason Medical Center became the first major American healthcare employer to implement a mandatory vaccination scheme for influenza 1. Since then, other healthcare organisations, especially in the USA, have followed suit 2. The UK's General Medical Council has stated that doctors “should be immunised against common serious communicable diseases (unless otherwise contraindicated)” – though stopping short of advocating a mandate 3. The recent popularity of mandatory schemes partly reflects growing anxiety about the emergence of novel influenza strains, e.g. H1N1, and the widespread harm they can cause, and partly by the significantly higher vaccination rates achieved by mandatory schemes over voluntary ones 4. In this editorial we discuss the nature of influenza, the effectiveness and safety of influenza vaccines, the rationale for vaccinating healthcare professionals, the effectiveness of voluntary and mandatory vaccination schemes and, finally, the ethical arguments for (and against) the latter. We do not single out any individual group of staff but suggest that this issue is of major relevance to anaesthetists and intensivists through their exposure to, and treatment of, high-risk groups. Influenza is a common respiratory illness caused by influenza virus types A, B or, occasionally, C. Although often regarded as a minor, self-limiting illness, some groups (e.g. children < six months, immunosuppressed or pregnant patients, or those with underlying chest/cardiac conditions) are more prone to serious complications e.g. primary/secondary pneumonia, meningitis or encephalitis. In addition, many healthy adults suffer a moderately severe illness, with several days’ prostration and inability to work 5. The virus surface glycoproteins haemaglutinin (H) and neuraminidase (N) are the major antigens recognised by the immune system; subtypes of these provide the terminology for the major types of influenza A virus. Minor mutational changes of the H and N antigens occur continually, leading to the emergence of new antigenic strains (‘antigenic drift’) 6 and seasonal outbreaks, seen most winters. In contrast, the appearance of completely new antigenic strains (‘antigenic shift’), to which the human population has little immunity, is thought to be due to genetic reassortment of human and animal influenza viruses – most likely avian and/or porcine strains 6. When it occurs, it can lead to more severe and dramatic global pandemics. These occur every few decades, with four in the last century – 1918, 1957, 1968 and 2009. The severity of these pandemics varies; the 1918 pandemic is estimated to have killed 20–50 million people 5, 7 whereas the 2009 pandemic was much milder, with an estimated 285 000 deaths worldwide 8. Seasonal influenza can also be deadly, with tens of thousands of excess deaths in the UK most winters 6. Most of these occur among the elderly or debilitated, and spread of the infection is more likely in institutions such as nursing homes and hospitals. Vaccines against influenza have existed for about 50 years. Because new viral strains continually evolve, new vaccines must be developed each year, with annual vaccination necessary to provide protection. Vaccination each autumn, in preparation for the winter epidemic, is recommended for the elderly, the immunocompromised, those with serious chronic disease, pregnant women, and healthcare workers 6. The usual vaccines are inactivated (i.e. have no live virus and cannot cause influenza) and are given by injection 6. The vaccine is about 60-75% effective in reducing cases in those exposed to strains of virus included in the vaccine 9, 10. Influenza vaccine is safe: local side-effects (pain/swelling at the injection site) are seen in 15–20% of cases, but resolve in a few days. Systemic features, e.g. fever, malaise and myalgia, are reported in < 1% of cases and typically last < 1-2 days 6, 11. This may give rise to the misconception that the vaccine ‘causes influenza’. Serious side-effects are so rare as to be highly controversial. These include neurological complications such as convulsions or Guillain-Barré syndrome. The latter has been reported to occur in about one per million vaccinations, but some studies report no link 12. Vaccinating healthcare workers against influenza is recommended principally to reduce the spread of infection by staff to their patients or colleagues. The other rationale is to reduce the rate of staff absenteeism, which would inevitably threaten productivity and patient safety. There is growing evidence that vaccinating healthcare workers decreases nosocomial influenza in hospitals and reduces mortality in patients receiving long-term care. For example, Hayward et al. 13 found significantly fewer deaths in elderly residents of nursing homes with a staff vaccination programme in place (though no difference in diagnosed influenza cases), and Riphagen-Dalhuisen et al. 14 reported lower rates of nosocomial influenza and/or pneumonia in hospitals where staff vaccination was encouraged. Potter et al. reported that vaccination of healthcare workers was associated with reductions in total patient mortality from 17% to 10% 15. However, the evidence is not unequivocal. Indeed, a recent meta-analysis (published before the Riphagen-Dalhuisen study) failed to find compelling evidence that vaccinating healthcare workers protected patients, and concluded that more and better studies were needed 16. Whilst there is some evidence that higher vaccination rates may reduce the rate of absenteeism due to respiratory illness among healthcare workers 17, there is currently insufficient evidence to draw definitive conclusions in this regard. Over the past few decades, there have been many attempts to increase the number of healthcare workers vaccinated against influenza, ranging from simple educational programmes to positive incentive schemes offering rewards 18. Some schemes require staff who refuse to be vaccinated to sign statements that they understand they are imposing risks on their patients 19. Many hospitals also ensure that vaccination is available at convenient locations and times 1. Most recently, the State of New York has mandated all healthcare professionals who do not get vaccinated to wear masks 4. We are aware of anecdotal evidence that in some areas, colleagues unwilling to be vaccinated are being put under considerable pressure to do so by senior healthcare professionals and managers. This may not always amount to coercion and, since it is not officially sanctioned, it cannot be described as a formal policy. However, putting colleagues under serious duress is ethically problematic, and unless applied equally in all areas is unjust, since staff making the same choice will be treated differently in different areas. Traditionally, even where vaccination is encouraged, rates of uptake are discouragingly low 20. Such encouragement may increase vaccination rates by 5–45%, but overall rates often remain below 50% 21. Indeed, during the 2011–12 influenza season in the USA, only 67% of healthcare professionals were vaccinated against the disease in spite of widespread vaccination campaigns 22. In England, recent data suggest that uptake among healthcare workers is around 45%, though with considerable variation between institutions 23. Interestingly, this is significantly lower than the uptake in at-risk groups: just under 75% for the over-65s and ~50% for those under 65 with a specific clinical indication 24. It is not entirely clear why voluntary staff vaccination schemes fail to ensure widespread coverage. Some of the issues may be ‘systematic’, e.g. perhaps vaccines are offered at insufficiently convenient times or locations. Staff might refuse the vaccine because they erroneously believe it is ineffective or actually causes influenza 25. Others may believe, equally erroneously, that they ‘never get flu’ or that if they did, they would know about it before they became infectious and be able to avoid transmitting the disease to patients and colleagues by taking timely sick leave. Yet others may decline vaccination because they think that influenza is a mild illness or because they fear the vaccine's side-effects 26. Because of the poor efficacy of voluntary schemes, a number of hospitals around the world have introduced mandatory schemes to deal with the problem 2. Mandatory vaccination of healthcare workers has been discussed for several decades but interest in this policy has recently intensified because of the failings of voluntary schemes. The effectiveness of a mandate may be measured in two ways: first, by measuring the overall vaccination rate amongst healthcare professionals; and second, more importantly, by measuring the rate of influenza amongst staff and their patients – or at least some proxy measurement (e.g. rates of sick leave in staff and mortality amongst patients). The former outcome is easier to measure and there is increasing evidence that mandating vaccines significantly increases the number of healthcare workers vaccinated. For example, when Virginia Mason Medical Center implemented a mandatory influenza scheme, immunization rates rose from 56% to 96% within a year 27. Similarly, within a few months of making influenza vaccination a condition of employment in 2008, BJC HealthCare achieved a vaccination rate of > 98%, from ~50% over the previous decade 28. A report published in 2012 indicates that vaccination rate in US hospitals with an influenza mandate is higher than in hospitals with voluntary schemes (95% vs 68%) 22. One should also note that mandates requiring healthcare workers’ vaccination against other infections (varicella, hepatitis B, measles, mumps, rubella) have been very successful in achieving near-universal rates of vaccination 29. It is harder to prove that mandatory vaccination schemes actually reduce the rate of infection amongst staff and patients, though there is at least some evidence to suggest that vaccines reduce the rate of influenza, and since mandatory schemes increase vaccination rates, it is reasonable to conclude that mandatory schemes will lead to reduced rates of infection. The prima facia ethical argument in favour of mandatory vaccines for healthcare workers is relatively straightforward. Healthcare workers have a deontological duty, based on the principle of non-maleficence, not to harm their patients (so long as the steps to be taken are safe, feasible and reasonable), and influenza vaccines represent a way of reducing the probability of unintentionally doing so by transmitting a potentially serious infection. As well as a duty to protect patient safety, healthcare professionals also have (more limited) duties to remain healthy in order to provide care for their patients. Thus, taking the influenza vaccine forms part of the duty owed to a patient 30. Healthcare professionals also have duties to each other (though weaker than duties to patients), and it can be argued that these include protecting colleagues from infection by taking reasonable steps such as vaccination. Of course, most staff can protect themselves from their potentially infected colleagues by taking the vaccine. However, some will be unable to be vaccinated for medical reasons (e.g. allergy to the vaccine) and others may not have had the vaccine when they are exposed to the virus. Further, healthcare workers also have a duty to ensure that their colleagues are not aversely affected by having to provide cover (or work without adequate support) when they are unwell with an easily preventable illness. There are also strong utilitarian arguments in favour of mandatory vaccination. First, a mandate is likely to be cost-effective and thus constitute good use of public funds. This argument relies partially on the general claim that ‘prevention is cheaper than cure’, but also upon the more specific claim that reducing absenteeism will save money and improve ‘productivity’. Second, implementing a mandate will reduce the number of patients and staff who contract a potentially serious infection and this can only serve to reduce suffering, ill-health and potentially, death. Third, it may be argued that issuing a mandate will improve net utility by improving public health, through sending out a strong message to the public that influenza vaccines are safe and effective. Where mandates do not exist, large numbers of healthcare professionals may not get vaccinated, setting a bad example and stoking anti-vaccine fears amongst the general public. A utilitarian calculation, of course, requires consideration of the possible negative impact of mandatory vaccination, e.g. the potential psychological harm caused by ‘forcing’ staff to take a vaccine they are unwilling to have, the possibility of reduced staff morale and resignations, as well as side-effects 31. However, the evidence suggests that in hospitals with a mandatory vaccination policy, very few staff refuse to have the vaccine and even fewer leave as a result of that refusal. For example, at BJC Healthcare, once the mandate was in place the majority of the < 2% of staff not taking the vaccine received medical or religious exemptions, with only 0.03% having their contracts terminated for non-compliance 28. The impact on morale is harder to calculate, but there is evidence that mandatory schemes are widely supported by staff. For example, Poland et al. 32 reported that 59% of nurses stated they would support a mandatory influenza vaccine policy. Finally, it is worth noting that policies requiring healthcare professionals to be vaccinated (or demonstrate immunity) against other infectious diseases (e.g. hepatitis B) have not been widely challenged. Communitarian arguments can also support influenza vaccine mandates. Communitarianism emphasises the fact that patients and professionals are bound to each other in important ways by belonging to a normatively important community of peoples. This ‘belonging’ brings many rewards but also obligations and responsibilities – and these extend to protecting one another from reasonably avoidable harms. (Interestingly, this argument might lead to patients’ having obligations to vaccinate themselves in order to protect healthcare professionals, though we do not press for that here). The arguments in favour of mandating vaccines are strong, but powerful counter-arguments also exist. The most obvious is that forcing professionals to take vaccines as a condition of employment violates their autonomy and freedom to refuse medical treatment, without suffering serious consequences 31. The choice not to be vaccinated is sometimes presented as almost whimsical, but opposition to vaccines may be deeper than whimsy and so the violation of autonomy cannot be dismissed so briskly. Moreover, since vaccination requires placing foreign biological material in a person's body, mandatory vaccination schemes inherently run the risk of undermining people's bodily integrity (although this already occurs e.g. in the context of mandatory hepatitis B vaccination). There are also side-effects to consider since the principle of non-maleficence extends to staff. Although serious side-effects are rare, this may be of little consolation to the individual who suffers these following mandatory vaccination. We do not deny the vital importance of autonomy, bodily integrity and non-maleficence where healthcare professionals are concerned. However, these arguments appear weak in the absence of evidence to suggest that failure to be vaccinated is an active, informed choice. It could be argued that the autonomy-rights of healthcare professionals to refuse vaccination is outweighed by the rights of patients not to be harmed by those caring for them – a situation that is accepted for mandatory hepatitis vaccination. It is also worth bearing in mind that the right to bodily integrity is not technically violated since individuals can always refuse a mandatory vaccination. However, the possibility of termination of a contract or refusal to pay for influenza absenteeism would appear to be both punitive and coercive and such policies undeniably restrict autonomy even if they do not undermine bodily integrity. Further, a degree of coercion and violation of autonomy already exists in the current ‘voluntary’ vaccination programme, and we have observed both individuals who have felt unduly pressured by colleagues and managers to accept vaccination, and those who have contracted influenza and felt blamed when asked if they had been vaccinated. A mandatory vaccination system that addresses concerns and provides adequate information, with the possibility of reassigning those who do not wish to be vaccinated away from high-risk areas, is less intrusive of personal autonomy than a voluntary system that may not be as voluntary as it claims. As for the principle of non-maleficence, we accept that vaccines have side-effects and that mandatory vaccination may cause some physical and psychological harm to those who do not want it. However, we also believe that healthcare workers already accept certain risks and sacrifice certain choices in order to care for their patients. The only other major counter-argument to the idea of a mandate is one derived from the principle of justice. Aristotle argued long ago that it was a cornerstone of justice that equals should be treated equally and unequals unequally. In this context, it might be argued that a mandate is unfair if it only applies to healthcare professionals; thus teachers, care home workers, police officers, transport workers and many others who also pose a potential risk to others should be forced to be vaccinated too. It should be noted that this does not constitute an argument against a mandate, but rather against restricting the mandate to healthcare professionals only. We also accept that mandates for teachers and care home workers in particular might well make much sense. However, in comparison with many other workers, healthcare professionals are caring for individuals who are usually acutely unwell and infection in these individuals is likely to be more serious, so it is reasonable to argue that a mandate requiring healthcare professionals to be vaccinated is not unjust. Mandating vaccination should only be implemented if there is sufficient reason to think that this will increase vaccination rates and decrease the prevalence of influenza amongst staff and patients. We have argued that there is good reason to think that the former claim is true, but only limited evidence to suggest that the latter is true. The ethical arguments presented appear to favour consideration of a mandatory scheme although we accept this is partly due to a lack of data suggesting staff would be strongly opposed to this. We would add that the existence of unofficial ‘coercion’, whereby those who refuse to be vaccinated are put under undue pressure by colleagues within the currently operating ‘voluntary’ schemes, also supports a change in the current system. We suggest that a pilot study should be carried out in the UK as a matter of some urgency, in order to determine the impact that a mandatory policy would be likely to have in this country. The study could look at the rates of vaccination and infection, and the views and opinions of the staff and patients affected by the policy. Such a study would provide valuable information on the practical feasibility and normative reasonability of introducing such a policy in the UK and it could help policy makers decide whether a mandatory scheme could (or should) be rolled out across the NHS. No external funding and no competing interests declared.