My goal in writing this editorial is not to be divisive or
controversial, but to bring the issue of mandatory influenza vaccination for
health care workers back to the level of a collegial discourse between people
of science and seekers of truth and justice.
In so doing, I hope to find some middle ground that will be acceptable
to the polarized camps on this issue. Neonates and vulnerable pediatric
patients are two of the groups who will benefit the most from a consensus.
However, to achieve that end, I must make some bold assertions and
observations, so please bear with me; reconciliation is possible between
William G. Buchta
As a board-certified practitioner in preventive medicine and public
health, one of the original authors of the guidance document on control of
influenza among health care workers (HCWs) by the American College of
Occupational and Environmental Medicine (ACOEM) in 2005 and, more recently, the
chair of the ACOEM Ethics Committee, I read with interest the article by
Ottenberg and colleagues regarding the ethical and legal rationale for a
mandate regarding influenza vaccination for HCWs. Before addressing ethics
specifically, I agree with the authors that any mandate must be predicated on a
rich and convincing body of science. They further assert that the scientific
justification for a mandate is overwhelming.
Frankly, I am incredulous. I honestly believe that the authors and I
have the same aim to protect HCWs and the patients they treat. However, the
evidence is lacking, and the logic is flawed regarding mandatory influenza
In the words of Dr. Tom Jefferson, lead author of the updated Cochrane
review of the efficacy of vaccination of HCWs, “the optimistic and
confident tone of some predictions of viral circulation and the impact of
inactivated vaccines, which are at odds with the evidence, is striking.”
In other words, the scientific discussion carried on in what is assumed to be
peer-reviewed literature has taken on the tone of a political campaign or
promotional speech. We need to bring the issue back to an evidence-based level
if we are ever to reach a common ground.
I am not what some would call an “anti-vaccinationist.” Quite
the contrary, I have promoted robust and successful influenza vaccination
programs in two health systems during the past 19 years as medical director of
occupational health. We have offered influenza vaccination to workers primarily
for their own sake and secondarily for the health of their families and
During the past 10 years, some of our colleagues in infection control
would have us believe that the primary way to protect vulnerable hospitalized
patients from the flu is to require HCWs to be vaccinated against what is
predicted to be the circulating strains of influenza. I propose that such an
assertion challenges cardinal rules of infection control on multiple levels.
First, not everything that acts like influenza and harms vulnerable
patients like influenza is, in fact, influenza. This may be obvious to most
readers, but Ottenberg and colleagues, such as so many authors before them,
have claimed that “each year in the United States, seasonal influenza
causes on average more than 200,000 hospitalizations and 36,000 deaths.”
And yet, the CDC, the sole source of this data, clearly differentiates that
annual average death rate from influenza is about 8,000 and that the figure of
36,000 deaths is related to “underlying pneumonia and circulatory
deaths.” Influenza-like illness (ILI) is the current term for the
condition caused by a host of organisms, including influenza, respiratory
syncytial virus, rhinovirus, coronavirus, etc.
A recent 3-year study in France revealed that influenza itself, the only
virus for which we have a vaccine, is the causative agent in only 23% of cases
of ILI, correlating with the numbers quoted previously. Thus, if we are truly
concerned about preventing the 36,000 deaths, we need to focus our attention on
control measures that not only include vaccination for influenza, but also hand
washing and wearing masks, gloves and gowns, which have proven to be very
effective controls in contrast to influenza vaccination. We have additional
controls available in the form of isolation, visitor control, ventilation,
housekeeping and sick-leave policies. Clearly, vaccination is only part of a
multipronged infection-control strategy, as outlined in the updated 2008 ACOEM
Second, mandating influenza vaccination creates a false sense of
security among those vaccinated, particularly during a year in which we later
learn that the vaccine was partially or totally ineffective. In such a
scenario, the vaccinated but coughing and unmasked HCW is a far greater risk to
patients than the unvaccinated HCW who practices other infection-control
measures and/or stays home when symptomatic. It would be more effective (and
ethical) to warn every vaccinee that the vaccine alone may not be effective in
preventing influenza, that it will not be effective against other forms of ILI,
and that other infection control measures will be necessary to prevent spread
Third, a mandate for influenza vaccination disguises the truth about the
current form of influenza vaccine, namely that it is arguably the worst
currently available vaccine in our armamentarium. A comparison of influenza
vaccine to vaccines such as those for hepatitis B and measles-mumps-rubella
(Merck) is truly an “apples and oranges” situation. Influenza vaccine
is the only one that needs to be administered annually, is the only one that
needs to be reformulated annually, is the only one that takes more than 6
months to manufacture in the face of an immediate need (2009 H1N1, case in
point), and has overall efficacy on par with the much-maligned bacillus
Calmette-Guérin vaccine for tuberculosis and no better than 70% to 90%
efficacy even when matched well with circulating viruses.
In fact, an Australian study of influenza vaccine effectiveness from
2003 to 2007 revealed age- and year-adjusted vaccine effectiveness of only 41%,
with a 95% CI of 19% to 57%.
Regarding the logistics of mandatory influenza vaccination, such
programs require employees to receive an invasive treatment every year of
employment. On the other hand, pertussis, MMR, and varicella (Varivax, Merck)
immunization and/or immunity verification can reasonably be mandated as a
condition of employment at the time of hire and completed within 6 months or
less with no subsequent need for boosters for the duration of employment.
Supply of vaccine is rarely a problem with any of them, and 95% efficacy is
Hepatitis B vaccination has similar merits but is not required of HCWs.
The Occupational Safety & Health Administration requires employers to offer
it to HCWs and document declinations of those who refuse. Consider how
profoundly vaccines for smallpox, polio, diphtheria, MMR have affected the
incidence of the associated diseases, whereas after decades of use, the
influenza vaccine has had little, if any, effect on the epidemiology of
influenza. I will have more to say later about the subpar quality of the
As for the allegedly dismal compliance rates for voluntary HCW
vaccination programs that “are rarely able to achieve vaccination rates
higher than 70%,” such a judgment is hardly fair when so many vaccination
programs have been hampered by shortages and delays in delivery of vaccine
during the past 10 years.
Although mandate advocates decry the seemingly irrational decisions of
HCWs who defer vaccination, imagine how motivated some of them are when they
have been repeatedly told to wait or return to the clinic multiple times until
the vaccine arrives (2000, 2001, 2006, 2009); told that the vaccine is in short
supply and that they do not qualify for vaccine (2000, 2001, 2004); told that
the previous year’s vaccine (2007) was, at best, 44% effective and not
effective at all against one of the influenza strains; told that there are two
vaccines this year (2009) but that only one is currently available; told that
the new inhaled version of the vaccine is probably more effective than the
injectable version, only to learn later that it probably is less effective in
adults; or told for years that the vaccine does not cause influenza, only to
learn later that the inhaled version has a high likelihood of inducing a mild
form of influenza.
One also has to wonder whether the problem lies more with the quality of
the product we are offering than the reasoning of those to whom we offer it.
Despite these issues, CDC reports that overall compliance during the 2009-2010
season was 62%, and 68% among those working in hospitals. A RAND survey
estimated compliance with the seasonal vaccine in hospital workers to be 71.4%.
That is a significant improvement over the often-quoted 35% to 40% long-term
compliance. Furthermore, the mandate advocates remain silent on the possibility
of herd immunity with more than 70% compliance or whether there is significant
marginal gain in desired outcome beyond 70% compliance that could not be
achieved by a broader focus on multiple control measures, as outlined
What is the desired outcome from mandatory vaccination for HCWs? Aside
from the obvious benefits to HCWs themselves, the issue has been framed as
patient safety, particularly protection of the vulnerable patients in a
hospitalized environment for whom influenza superimposed on pre-existing
disease could have disastrous results. Every published account of how an
institution implemented its mandatory program presents nearly 100% compliance
as the highly successful outcome. Personally, I thought 100% compliance was the
definition of mandatory and could be assumed.
What about an improvement in patient safety, which is the metric we
should all be measuring? None of them mentioned it; the authors of the Barnes
Jewish experience claimed that a change in nosocomial influenza incidence would
be too difficult to measure. I beg to differ; it is readily observed in the
hospital setting and easy to measure. However, only two published studies have
documented such a metric. Salgado and colleagues observed nosocomial influenza
over the span of 13 years with a solid case definition, and they observed an
association with statistical significance between HCW vaccination rate and a
rather curious metric of nosocomial cases of influenza as a fraction of all
hospitalized influenza cases. They also did not distinguish between nosocomial
cases linked to HCWs and those linked to other patients or visitors.
Methodological and statistical flaws aside, an examination of their raw
data is compelling: In no single year were there more than eight cases of
nosocomial influenza (four to five on average), and during two consecutive
seasons in the middle of the 13-year interval, there were none. Could it be
that nosocomial influenza is a rare event and that the vast majority of cases
are community-acquired? The second, latest and most robust study came out of
France and was reported in January, spanning three influenza seasons. Vanhems
and colleagues measured actual incidence rates of nosocomial influenza and
distinguished between those cases in contact with infected HCWs and/or other
infected patients (unfortunately not taking into account visitors as potential
The nosocomial cases are more likely linked to infected patients than to
infected HCWs, and if using the Salgado and colleagues definition for a
nosocomial influenza case, the incidence rate for HCW-related nosocomial
influenza for a 400-bed hospital would be less than 1 per year. An unpublished
verbal report from an occupational health nurse at Virginia Mason Hospital, the
first US institution to mandate vaccination in 2004, reported no cases of
nosocomial influenza the year before mandating and one case the year influenza
vaccine was mandatory.
So, if patient safety (HCW-related nosocomial influenza) is the
dependent variable, it is difficult to measure a change when it is a relatively
rare event. One has to wonder about the magnitude of the alleged threat from
unvaccinated HCWs, particularly when there are other control measures at our
disposal to mitigate the threat, and when other patients and, most likely,
visitors are an even greater threat, which is typically left unmitigated.
Regarding the dearth of published evidence for nosocomial influenza, several
colleagues and I will be submitting that analysis for publication in the near
In summary, where is the compelling scientific evidence for mandatory
influenza vaccination? Where is the evidence that mandatory vaccination
programs actually improve patient safety? Why do the advocates for mandatory
vaccination feel compelled to inflate the statistics about influenza mortality
if the documentation of validity is self-evident? If we want to talk about
ethics, is it not more ethical for our vaccinologists to influence their
benefactors in the vaccine manufacturing industry to generate a vaccine that is
safe, quickly reproduced in large quantities, and targeted to
“universal” influenza antigens such that the vaccine would be more
predictably efficacious from year to year?
Members of the International Federation of Pharmaceutical Manufacturers
& Associations (IFPMA) admit that such vaccines are a possibility and that
the technology exists, but they also make no mystery of their belief that
demand drives supply and that health care professionals are the key to driving
that demand. I would expect nothing else from businessmen. They are likely well
aware that a steady or increasing demand for the existing product ensures full
utilization of that product and, in the case of shortages, rising prices and
marginal profits. I sincerely hope that our colleagues who advocate for
mandatory vaccination, many of whom report potential financial conflicts with
the members of the IFPMA, are not similarly motivated.
Any medical professional who promotes policy that has commercial
ramifications and reports potential financial conflicts regarding the involved
commercial parties is obligated to not only report the conflicting
relationships but also to resolve them. I personally have no such conflicts to
Coming to a compromise
So, for a compromise position, I make the following proposal. In 2007, I
participated in a collegial debate at the International Congress of
Occupational Health (ICOH) conference on HCW health in Vancouver, British
Columbia. The last comment during the ensuing discussion was made by Dr.
Hans-Martin Hasselhorn, a colleague from Germany, who was then the chair of the
ICOH Committee on HCW Health. He said, “In my country, we would not be
having this discussion.”
In 2010, another German colleague suggested that mandatory vaccination
may be ethical in specific situations as an application of the precautionary
principle. The consequences of nosocomial ILI for those patients with severely
compromised immune systems are sufficiently serious to support such a policy,
and even one case in that setting would be unacceptable. Therefore, mandatory
vaccination for HCWs in certain limited units, such as neonatal/pediatric ICU
or bone marrow transplant facilities, is justified as long as other control
measures are strictly applied and patients and visitors are required to be
vaccinated. Such a policy would allow HCWs who decline vaccination to opt for
transfer to other assignments without losing employment. We can discuss the
specifics of such a policy in other venues, but I believe nearly all parties in
this controversy can live with this compromise.
The CDC has told us that the influenza vaccine is the primary method for
control of seasonal influenza, but neither the CDC nor The Joint Commission
have endorsed mandatory vaccination for HCWs. I challenge the assertion that
the current iteration of the influenza vaccine, which has changed little since
the 1950s, is the best tool at our disposal.
Imagine the outrage if our gloves, respirators or masks failed to
protect HCWs more than 30% of the time; were frequently in short supply or
delayed in delivery; were unavailable for 6 months or longer in the face of
imminent need; and were obsolete within a year such that all remaining stock
had to be discarded with no assurance that a suitable replacement would be
available. Such is the current status of influenza vaccine.
Thus, influenza vaccine is a valuable tool, but no more than an adjunct
to other established infection-control measures of all ILI and, as such, does
not warrant mandatory application for all HCWs. Such a consideration violates
the core principles of ethics, including autonomy, equity, beneficence and
Elaboration of these principles could be the basis for further
discussion. However, the simple truth is that when vaccine manufacturers offer
us a product that is safe, readily available and persistently efficacious, we
will not need a mandate; everyone will get it.
William G. Buchta, MD, MS, MPH, is the medical director for the
occupational health service in the division of preventive, occupational and
aerospace medicine at the Mayo Clinic in Rochester, Minn.
Disclosure: Dr. Buchta reports no relevant financial disclosures.
- Babcock HM. Clin Inf Dis. 2010;50:459-464.
- CDC. MMWR. 2005;54(08):196-199.
- CDC. MMWR. 2005;54(08):196-199.
- CDC. MMWR. 2010;59(12):357-362.
- Doshi P. Am J Public Health. 2008;98:939-945.
- Heyme N. Vaccine. 2008;26(Suppl 4):D23-D26.
- Jefferson T. BMJ. 2006;333:912-915.
- Jefferson T. BMJ. 2009;339:b3675
- Kelly H. PLoS One. 2009;4:e5079.
- McLennan S. Vaccine. 2010;28:8061-8064.
- Ottenberg AL. Am J Public Health. 2011;101:212-216.
- Poland GA. N Eng J Med. 2011;364:97-99.
- Salgado CD. Infect Control Hosp Epidemiol.
- Swift MD. Seasonal influenza prevention in health care workers.
ACOEM. Nov. 17, 2008. Available at:
- Thompson WW. JAMA. 2003;289:179-186.
- Vanhems P. Arch Intern Med. 2011;171:151-157.
Thomas R. Talbot
While I do not agree with many Dr. Buchta's assertions and comments, I
do believe he has a right to express his opinion in a professional and cogent
manner. My general thoughts on his editorial are summarized below:
- " The statement about infection control professionals and protection
of hospitalized patients through influenza immunization misrepresents the
position put forth by these groups. As clearly stated by the Society of
Healthcare Epidemiology of America's (SHEA) 2010 position paper on the topic,
mandatory vaccination must be present as part of a comprehensive infection
control program designed to prevent transmission of all respiratory viruses:
"Preventing the spread of influenza involves initiation of a comprehensive
infection control program designed to identify and to isolate infectious
persons while using work practice controls to reduce the risk of influenza
transmission. Rapid identification and isolation of patients suspected to have
infection, adherence to hand hygiene and respiratory etiquette, source control
by the masking of persons with influenza?like illness (ILI), patient cohorting,
use of personal protective equipment, restriction of ill HCP from working in
the facility and of ill visitors from visiting, and antiviral prophylaxis and
treatment (when indicated) all play essential roles in the reduction of
transmission of any healthcare?associated respiratory infection, including
influenza. Unlike efforts to prevent transmission of other respiratory viruses,
however, vaccination of both patients and their contacts is the cornerstone of
efforts to prevent influenza transmission" (Talbot TR. Infect Control Hosp
- " While appropriately noting that not all ILI equals influenza
infection, Dr. Buchta does not note the data that show the relatively low
frequency that actual lab-confirmed influenza infections presents as ILI
(Babcock HM. Infect Control Hosp Epidemiol. 2006;27(3):266-70; Talbot HK.
Infect Control Hosp Epidemiol. 2010;31(7):683-8.) In addition, the issue of
asymptomatic shedding of viruses by infected health care personnel (HCP) is not
addressed, and this can be an important mode of transmission in the healthcare
- " As Dr. Buchta mentions, vaccination is only part of a multipronged
infection-control strategy, which is outlined in the updated 2008 American
College of Occupational and Environmental Medicine (ACOEM) guidance document he
mentions, as well as in position statements of SHEA and the Association for
Professionals in Infection Control and Epidemiology. These two infection
control groups advocate for a multipronged approach to prevent
healthcare-associated influenza and, as clearly evident in their position
statements on the topic, they do not solely "push the vaccine."
- " While often espoused by those against vaccination, there are no
data to show that vaccinated HCP are less likely to adhere to basic infection
control practices. If anything, one could reasonably argue that those who are
vaccinated are more conscientious about other safe practices like hand hygiene.
There are many instances of requirements for safety in health care that have
never been shown to cause the unintended consequences that Dr. Buchta
- " The issue with assessing vaccine effectiveness is a challenge.
Often, the studies examining vaccine effectiveness either use non-laboratory
confirmed outcomes (or poorly sensitive laboratory-confirmed outcomes, like
rapid antigen testing) or involve populations with known decreased
immunogenicity to the vaccine (eg, elderly and pediatric patients, who are the
targets by the prospective surveillance program used to assess vaccine
effectiveness). Arguably, effectiveness will be higher in healthy adults.
Critics also fail to note the dramatic reduction in influenza and
influenza-related outcomes in HCP and healthy adults noted in several
randomized controlled trials.
- " Dr. Buchta also mentioned the logistics of mandatory influenza
vaccination and requirement of such programs that require "employees to receive
an invasive treatment every year of employment." While implementation of a
mandatory program does involves substantial logistics and planning, including
clear support from facility leadership, in order to succeed, Dr. Buchta fails
to note that annual tuberculin skin testing, also an invasive intervention, is
also required annually for health care workers. In addition, he mentions that
"pertussis, measles-mumps-rubella (MMR) and varicella . . . immunization and/or
immunity verification can reasonably be mandated as a condition of employment,
can be completed within 6 months with no subsequent need for boosters, vaccine
supply issues are rarely a problem, and 95% efficacy is almost guaranteed."
While ignoring the likelihood of a requirement for booster dosing with the
pertussis vaccine, these statements underlie a critical flaw in his arguments.
How do the arguments against mandatory influenza vaccination mentioned by Dr.
Buchta suddenly no longer apply to vaccines that are easier and more convenient
to provide? I would argue that, ethically, such logistic issues should not be a
deciding factor as to whether a vaccine is required.
- " Regarding what Dr. Buchta calls the "subpar quality of influenza
vaccine because of its limited effected on the epidemiology of influenza," this
may be due to failure to vaccinate an adequate percentage of the population, as
well as the unique nature of the pathogen. The antigenic drift and shift of
influenza, not seen in the other pathogens named, also drives this and should
- " The issue of vaccine supply is a complicated one. However, even in
years of initial delays, millions of doses of vaccine are wasted.
- " As for the RAND survey data that is quoted, I believe this figure
is not accurate. A RAND report for 2009-2010 season noted seasonal vaccination
coverage of 49% at mid-season, while CDC rates were 61.9% for seasonal vaccine
last year. Also many feel that the pandemic led to a transient increase in
rates last year and that this year's rates at various facilities have decreased
back to pre-H1N1 levels (unpublished).
- " The issue of whether a threshold level of HCP influenza vaccination
coverage exists has been addressed by those advocating mandatory programs.
Specifically, this is addressed in the 2010 SHEA position paper: "Importantly,
modeling studies have estimated that in both acute care and long?term care
settings, there is no HCP vaccination rate above which additional HCP
vaccination coverage will not lead to further protection of patients. In these
studies, vaccination of 100% of HCP in the acute care model resulted in a 43%
reduction in the risk of influenza among hospitalized patients and a 60% risk
reduction among nursing home patients." (Talbot TR. Infect Control Hosp
- " Dr. Buchta's points on the need to assess the burden of health
care-associated influenza are important, yet I would argue with assumptions
that it is "easy to measure." In most institutions, a patient with onset of
respiratory symptoms well into admission is not tested by clinicians for
respiratory viruses, but for health care-acquired bacterial infections. In
addition, disease resulting from transmission may not manifest until
post-discharge and may not be reported back to the group tracking the outcomes
in the hospital.
- " And, finally, while Dr. Buchta wrote that neither the CDC nor The
Joint Commission have endorsed mandatory vaccination for HCWs. However, those
organizations have just published a draft standard requiring achievement of 90%
coverage of HCP by 2020.
In closing, I thank Dr. Buchta for his passionate discussion, as a free
exchange of ideas and evidence on important safety topics like HCP influenza
vaccination is essential. The mounting evidence regarding the problem of health
care-associated influenza, the role that a robust influenza infection control
program can have on improving patient and HCP safety, and the impact of
mandatory HCP influenza vaccination programs is clear. Most importantly, with
this issue it is key to put ourselves in the place of our patients: would you
rather have an unvaccinated or vaccinated HCP care for you, your loved ones, or
you friends in the hospital?
Thomas R. Talbot, MD, MPH
of Medicine and Preventive Medicine, Vanderbilt University School of