Policy Statement 2006-3: Preparing for Pandemic Influenza1
Policy Statement 2005-2: Developing a Comprehensive Public Health Approach to Influenza2
Policy Statement 2004-11: Threats to Public Health Science3
Policy Statement 2000-11: The Precautionary Principle and Children’s Health4
Policy Statement 2000-23: Need for Continued and Strengthened Support for Immunization Programs5
Policy Statement 96-06: The Precautionary Principle and Chemical Exposure Standards for the Workplace6
Policy Statement 89-06: Recommendations for Adult Immunization7
Purpose
The American Public Health Association (APHA) has repeatedly endorsed the precautionary principle as a cornerstone of preventive public health policy and practice, both in the United States and throughout the world.4–6,8,9 Recognizing immunization as a premier public health intervention of the 20th century, APHA has advocated for effective implementation of universal immunization2,5,7 and recommended requiring all health and laboratory workers, as well as students in these fields, to be immunized against all vaccine-preventable diseases.7 This position statement examines influenza prevention in the United States today in the context of this longstanding APHA position. Immunizing health workers against influenza has an impact on the workers themselves and their coworkers and families, on patients in the healthcare facilities and community-based settings where they work, on overall communitywide immunity, and on the health system’s capacity to provide safe care and its readiness to meet both routine and emergent service demands.
The Problem
Influenza and its complications account for the greatest number of vaccine-preventable deaths worldwide: one-quarter to one-half million deaths occur every year from approximately 3 to 5 million infections that cause severe disease and hospitalization10–12 The actual number of symptomatic cases is much higher, with the greatest burden falling on the least privileged in countries without access to the same level of preventive and medical care that is available in developed countries, such as the United States.13 The World Health Organization’s (WHO’s) Strategic Advisory Group on Immunization puts control of seasonal influenza into the same priority as cholera, typhoid, and yellow fever.14
In the United States, influenza annually affects approximately 15% of adults; 15 to 60 million cases lead to 250,000 or more hospitalizations and from 20,000 to more than 40,000 deaths.15–23 Influenza-related death toll estimates in some years have reached as high as 80,000. Together with pneumonia, it is the 8th leading cause of death in the United States and the 5th leading cause of death among those 65 years old and older24—with annual direct medical costs of $3 billion to $10.4 billion and $16.3 billion in indirect costs.25–30 Local epidemics are frequent.
As its population ages, the United States has been experiencing higher influenza-related mortality, including influenza pneumonia and cardiopulmonary disease. Although not more susceptible to infection, people older than 65 years are typically at highest risk for complications and death from the disease. From 1979 to 2000, influenza hospitalization rates for elderly patients were 17 times higher than the average rate, and more than 90% of the patients who died were elderly.31–33 Yet more than half of influenza-related hospitalizations are reported in people younger than 65 years.20,32 Estimated rates of influenza-associated hospitalization and death start to rise around 50 years of age and continue upward thereafter.24,32,33
Pregnant women experience more complications than others with influenza. Besides being less likely to become infected with influenza during infancy, the newborns of women who had influenza vaccine when pregnant weigh more and are healthier at birth than those whose mothers did not.34,35
Influenza can trigger the complications of chronic disorders. People with diabetes; cardiovascular disease; or chronic lung, renal, or liver conditions are at higher risk for influenza morbidity and complications.36–51 During periods of high influenza incidence, hospitalizations of adults with these high-risk medical conditions may increase 2- to 5-fold, depending on age group. People with cancer and other immunocompromising conditions are especially susceptible to severe complications.45,46,47 Influenza-related hospitalization rates in adults younger than 65 years with cancer are 5 to 10 times higher than for the general population and 3 to 5 times higher in people with cancer older than 65 years—higher than for other high-risk groups.45,48 With an estimated 7% to 10% death rate, cancer patients are 10 times more likely to die than others hospitalized with flu-related infections, and this mortality impact is particularly notable among those younger than 65 years.45 Residents in long-term care facilities have a greater risk for infection because they live in close quarters in closed settings and have contact with numerous caregivers. Because residents often have multiple underlying medical problems, long-term care facility outbreaks are associated with significant morbidity and mortality.20,52–55
Protecting Health Workers From Influenza
Health workers include all workers who, during the course of their work, have direct or indirect contact with the recipients of a preventive or restorative health service or related social or counseling services or with their caregivers, family members, or household members—regardless of the location where they perform their work. This contact may be a part of their normal work duties or may occur incidental to work activities—routinely or infrequently. They include full-time and part-time employees, contract or per diem workers, independent consultants, volunteers, trainees, and students. Those not directly involved in direct care may, nevertheless, be exposed to infected people; infectious materials; or contaminated supplies, equipment, or environmental surfaces (e.g., food and housekeeping service workers).
Community health workers (also called health advisors, health representatives, liaisons, promotores de salud, and related titles) work in diverse, informal, community-based settings (such as homes, schools, churches), often in unpaid capacities for health or social service agencies or other community organizations serving high-disparity populations. They provide the frontline links to health promotion and maintenance for many of the most vulnerable and needy populations. Their activities include, but are not limited to, outreach, informal counseling, social support, advocacy, and education of both clients and staff to facilitate access to culturally competent service delivery.
APHA advocates protecting US workers on the job as a top priority for the president and Congress. This agenda includes putting worker health and safety first, ensuring protection through tough enforcement of existing regulations, establishing new worker protections, and increasing worker participation in workplace safety and health programs.55 With more than 12 million workers, health care is the second fastest growing sector of the US economy.56 More than 13% of US workers have jobs in the healthcare sector, according to the US Department of Labor.57 Any workplace can be a setting for influenza transmission (not just healthcare facilities). Health workers can likewise be exposed to influenza anywhere in the community. Occupational exposure to infected patients, however, especially those with unrecognized infection, heightens risk for health workers. If infected at work, they can, in turn, unknowingly transmit infection to coworkers and carry infection home to family members.
Up to 25% of unvaccinated health workers may be infected each year.40,58,59 Health workers themselves (and their family members) frequently have medical conditions that raise their risk for influenza morbidity and mortality.60 In addition, the frequently noted aging of the healthcare workforce places an increasingly greater number and proportion of health workers in a higher-risk category.
Influenza infection is readily spread by respiratory droplets. It mainly spreads from person to person when a host coughs or sneezes, with greater contagion in semiclosed or crowded environments. Less efficient transmission also may occur through indirect contact, such as touching something already laden with virus, then touching the eyes or nose. Symptoms usually appear 1 to 4 days after infection, and an infected person is contagious during this asymptomatic period. Approximately 20% of cases remain subclinical.11,20 Thus, we cannot rely on signs of another person’s illness to alert us to use protective barriers, nor is an infected person necessarily even aware of having been exposed.
Annual vaccination is the most effective method for preventing influenza infection and its severe complications.21,22,61–69 Primary prevention by vaccination is therefore at the top of the influenza infection control hierarchy. Influenza vaccination of healthcare workers is the single most important measure for preventing occupation-acquired and nosocomial influenza from both known and unexpected sources. Other measures, such as hand hygiene and barrier precautions, are additional protective steps, not alternatives. Masks or respirators, whether worn by people with influenza-like illness (ILI) symptoms or those who are in proximity to them, are not as protective as preexposure immunization, especially given the high proportion of asymptomatic infectious people. Influenza occurs in healthcare workers even when there is high personal protective equipment (PPE) adherence.59 Improving influenza vaccination rates in health workers is thus essential for their safety and for infection control.
Social Justice Perspective—For Workers and Patients
Addressing the risk to patients is an especially salient social justice issue when poverty, poor health infrastructure, low health literacy, or lack of information influence their susceptibility status.70–72 In turn, unvaccinated people can experience a double jeopardy disparity when, in a healthcare setting, they are exposed to infected personnel—the healthcare system fails them twice.
Unvaccinated status may reflect a disparity in access to the primary care that should afford a person timely counsel from a healthcare provider to get vaccinated.73 Access to health care is a predictor of influenza vaccination, even among those at high risk for complications.43,71,74 Access limitations exist not only for people in medically underserved or low socioeconomic communities, but also for many others who, for whatever reason, lack a medical home or may not know that gratis vaccination is available or where to find it. A high proportion of health workers represent minority ethnic groups, including many immigrants, and health workers from different racial and ethnic groups have significantly different immunization vaccination rates.38,39,43,44,73–75
Chronic conditions that put people at higher risk for influenza-related morbidity and mortality, such as diabetes and asthma, are more prevalent in African Americans and Hispanics than in Whites. Yet influenza vaccination rates are lower in these populations, including among those with greater influenza risks, such as the elderly and people with diabetes or heart, lung, or renal disease.38,39,43,44,48–50,71,73–77 African Americans have a higher influenza hospitalization rate than other races/ethnicities.
The strongest and most frequently asserted ethical principle is that the healthcare provider’s primary duty is to protect and avoid harming those served, often articulated as “First, do no harm.” The healthcare consumer has the right to assume that health workers, and the organizations that employ them, will take all reasonable measures to avoid transmitting communicable pathogens for which safe and effective vaccines exist.78–81 Bioethicist Arthur Caplan maintains, “Getting a flu shot is the least those who claim to be bound by professional ethics ought to do.”82 American Nurses Association President Rebecca Patton enjoins her colleagues, “As nurses, we have an ethical obligation to protect ourselves, our patients, and our families from illness. Vaccination is one simple step we can take to do that.”83 Matthew Wynia, Director of the American Medical Association Institute for Bioethics, adds that “patients should be informed when they are seeing a healthcare worker who has refused vaccination.”81
The public reporting of staff vaccination rates at healthcare facilities as both a quality measure and a matter of transparency to inform communities, patients, and visitors has also been recommended21,40,78,81 as a component of retrospective patient safety “report cards” with real-time, facilitywide, and unit-specific posting during influenza seasons. Current Medicare-Medicaid requirements for reporting of nursing home residents’ vaccination rates could be expanded to include staff coverage too.
Improving Vaccination Coverage of Health Workers
Improving influenza vaccination rates in health workers provides benefits to workers, patients, and health service agencies. Preventing both community and workplace influenza transmission to health workers is essential both for maintaining a safe work environment in healthcare settings and for ensuring staffing capacity. ILI-related absenteeism can cause or exacerbate significant staffing shortages, which can be especially problematic during influenza’s peak periods.28,29,59,60,84–86 Staff immunization is highly cost-effective and can be cost saving.25,26 Additional costs for healthcare organizations implicit during and after a nosocomial influenza outbreak are also relevant considerations.26,28,29
Since 1984, the Advisory Committee on Immunization Practices (ACIP) of the US Centers for Disease Control and Prevention (CDC) has continually recommended universal annual influenza vaccination for health workers.87 At the millennium, 60% was targeted as the national 2010 health objective for healthcare personnel and all adults younger than 64 years, with a 90% goal for older adults.77 Other professional groups have consistently supported and endorsed these recommendations and proposed ways for institutions to improve their personnel vaccination rates. A 2004 National Foundation of Infectious Diseases initiative called on healthcare institutions to ensure influenza vaccination is available and offered to every health worker every year.65 Since 2007, the Joint Commission has required accredited hospitals and long-term care facilities to offer influenza vaccination to staff and independent licensed practitioners as a patient safety and infection control standard.88,89 Unions, too, urge strong enforcement of prevention steps to protect health workers, including influenza vaccination,56,90,91 encouraging members working in health services to get influenza vaccination and calling for healthcare employers to provide free vaccination to employees.91 Nevertheless, the vaccination rate among health workers has remained dismally low, typically less than 30% (often much lower) and infrequently reaching 50%, even in hospital units caring for high-risk patients.92 By the 2008–2009 season, few institutions reported rates as high as 50%, even those with aggressive programs to promote staff vaccination.92–96 In fact, as a group, health workers are among the most poorly covered.68 By mid-January 2010, after unprecedented intense, communitywide promotion efforts across the country, the highest level ever was reached, though it was still less than 70%.97
Barriers to workers’ accepting getting vaccinated against influenza can be financial, structural, and attitudinal1.47,63–65,84,91,93,94,98–116 Best practices from effective promotional campaigns by hospital employers have been identified.65,83,88,89,95,118 However, even with incentives and vigorously implemented, intensive promotional campaigns that use a panoply of these strategies, vaccination rates can remain significantly below 50%; very few hospitals report achieving rates of 80% or more.65,84,929,118,119 A ceiling effect—below desired levels for group protection (herd immunity)—has been inferred.119 Reports of substantial and lasting impact are rare and anecdotal. Aggressive programs of recognized best practices have even experienced declines in vaccination rates over 1 to 2 years.95
Unfortunately, randomized controlled trials61,89,120,121 are all but absent, and few studies have provided a theoretical framework to guide replication and build our understanding of what makes best practices work. Most surveys, including preintervention–postintervention studies, have been cross-sectional, relying on convenience samples and self-reported recall. Programs that have achieved substantial improvements (i.e., >10% increases in coverage, but typically with rates still <50–70%) have invariably implemented multipronged efforts, making it difficult to tease out the components that contributed more (or less) to outcomes. Moreover, similar programs have yielded different effects. Progress sometimes takes several years, but even multiyear intensive efforts also demonstrate limited success. Rarely is more than 70% coverage achieved without a mandate, even with programs that bundle multiple documented best practices.89,118
Hence, the strategy needed to consistently achieve the immunization rates of 90% and higher needed for herd immunity is to require vaccination as a condition of healthcare employment. Like APHA,7 some professional associations have proposed such requirements: the Infectious Diseases Society of America (IDSA), American College of Physicians (ACP), and the National Patient Safety Foundation.65,79,122,123 These positions recognize that education is not enough to change unfounded beliefs or misconceptions108,124,125 and that knowledge is not enough to ensure healthful behavior or consistent adherence to good infection control practice.126 The American College of Occupational and Environmental Medicine (ACOEM) took the position that “education and adherence to infection control practices should be mandatory” in 2006 but questioned whether evidence regarding the benefit of healthcare worker vaccination to patient safety was then currently adequate to override workers’ autonomy to refuse.127 Since then, more than 100 institutions across at least 30 states, Puerto Rico, and the District of Columbia—small and large, public and voluntary—have successfully implemented mandates.79 Reports from these institutions and multifacility systems indicate that mandates are a highly effective intervention, resulting in the highest reported rates for any intervention designed to improve coverage.79,118,124–130 As more employers establish and implement requirements, reports demonstrate their effectiveness with little, if any, negative impacts. A CDC-sponsored RAND Corporation study found that when healthcare employers required staff to be vaccinated against seasonal flu, the vaccination rates were twice as high as when employers recommended vaccination but did not require it.97
Facilities that instead allow “informed declination” after education report mixed results, ranging from improved rates to no effect.84,93,94,119,128,131–136 Employee response has included nonparticipation, perceptions of coercion, and opposition to being required to sign a form and specify a reason for not getting vaccinated. A review of 22 hospitals’ use of such refusal statements found that vaccination rates improved only modestly.135 ACOEM thus discouraged declination statements as a poor use of resources that can distract from education and the mission of improving vaccination coverage.127 Indeed, an IDSA survey of 99 facilities across the country found that requiring declinations was less relevant to vaccine uptake than other program elements (such as providing free vaccine, targeting education, and ensuring program resources).84,93 After California mandated signed declinations, significantly more staff at Southern California Kaiser Permanente’s 12 hospitals signed declinations than with the previous system that allowed those “not wishing vaccination to [voluntarily] state their reasons,” but more than 35% of the targeted health workers still remained unvaccinated.134 By contrast, when BJC Healthcare in Missouri implemented a requirement for staff at its 13 hospitals to be vaccinated, fewer sought medical or religious exemptions than in previous years and fewer than 2% remained unvaccinated.128
Declination must be regarded as a last resort, not as an alternative. It is essential that getting vaccinated is made easier and more convenient for workers than opting out and, likewise, the procedures for obtaining an exemption must be just as rigorous as those involved in getting the vaccine.81,94,135–138 Given that unvaccinated clusters within a work unit, facility, or other group setting may compromise herd immunity, allowing broad philosophical or personal belief exemptions can threaten the effectiveness of vaccination programs.139–141 Some ethicists maintain that the bioethical principle of justice precludes refusing vaccination for personal ideological reasons, including conscientious objections.142
Declining vaccination must not be a simple, perfunctory process of completing a checklist or signing a form acknowledging risk and the right to reconsider and get immunized later. Hence, some institutions require an individual’s personal attestation (not a check on a form) of religious belief about vaccination, whereas others require documentation from clergy; with both of these approaches, additional verification that vaccine history demonstrates consistent application of the belief is also required by some. Documentation of a medical contraindication should be required, with review by the director of employee health services, not simply received and filed by employee health services administrative or clerical staff. Temporary exemption should be given, subject to reevaluation, when indicated by the nature of the medical condition. Although medical information is confidential, hospital administrators and unit supervisors must be systematically updated on staff vaccination status throughout the season.
Patient Safety Concerns in Healthcare Settings
The patient safety issue has been highlighted by the Joint Commission, the Society of Healthcare Epidemiology of America, IDSA, ACP, the Association of Professionals in Infection Control and Epidemiology, ACOEM, the National Patient Safety Foundation, the National Foundation for Infectious Diseases, the Immunization Action Coalition, the Society of Healthcare Epidemiologists of America, and the Hospital Infection Control Practices Advisory Committee of the CDC.78,79,88–90,118,122,127,143–145 The consensus among these national agencies and organizations is that influenza vaccination of health workers is crucial. Unvaccinated workers can introduce infection or propagate an outbreak in any facility or congregate community setting. Barrier precautions must be considered for unvaccinated workers (regardless of the reason for not being immunized) when they are within a specified proximity of susceptible patients.78,79,85,123
Hospitalized patients who develop nosocomial influenza have a high mortality rate.15,17,52,68,94,146 Patients can be at risk of infection when exposed to infected health workers—both those who have no symptoms and are unwitting vectors and those who work while feeling ill, even with ILI symptoms during flu season. The latter presenteeism is a well-documented problem.85,101,147 Ensuring that symptomatic staff remain off work until recovered is important—indeed essential—but because the silent incubation period allows them to infect others, it is even more important to prevent their infection. Indeed, unvaccinated healthcare workers have been implicated as sources of influenza infections in deadly outbreaks among adults and children in both acute and long-term care settings.44,52,61,146,148
Unfortunately, neither nosocomial influenza nor staff vaccination status has been routinely tracked at hospitals, but available data nevertheless demonstrate a link between staff vaccination and nosocomial infection.117,149–152 A national survey of 50 university-affiliated hospitals found 62% monitored healthcare-associated influenza, documenting a range of 0 to 5 cases per 10,000 inpatient days.119 A tertiary medical center that tracked hospital-acquired influenza for more than a decade found a strong association with the vaccination rate of healthcare workers: The nosocomial infection rate was totally eliminated when the staff vaccination rate rose 63% above its baseline rate.131 Staff vaccination rate in the emergency department setting has been associated with lower absenteeism; with higher vaccination coverage, staff take fewer sick leave days per person, and fewer staff take leave with ILI.86
Immunizing staff even adds complementary protection to the most vulnerable and those with weaker immune responses to vaccination.21,31,42,43,50,70–76,131,153 Epidemiology shows that staff immunization is necessary to control outbreaks in nursing homes, even when there are high immunization rates of residents.31,70–76,131,153,142 A RAND Corporation study of 301 nursing homes found that, regardless of facility size, only the immunization of both staff and residents reduced the rate of ILI cluster outbreaks.131 Even when 60% of patients have been vaccinated, vaccinating staff enhances mortality reduction.130 These findings are especially relevant when the season’s vaccine is not well matched to the most common virus strain; that is, even higher vaccine uptake is then needed to achieve group protection (herd immunity).
Influenza Vaccination as a Condition of Healthcare Employment Is Essential to the Health Sector’s Capacity for and Commitment to Worker, Patient, and Community Safety
A Joint Commission infection control standard requires accredited hospitals, long-term care facilities, and home health providers to evaluate staff vaccination coverage each year and take steps to increase it.88,89 Some debate, nevertheless, persists about how great a benefit will be gained from universal vaccination of workers in healthcare facilities and whether the benefit will outweigh the cost and effort involved. However, the suggestion is not that available evidence favors nonmandatory approaches. Rather, authors point to the relatively limited data documenting the influence of health workers’ vaccination status on morbidity and mortality in nursing homes, the lack of surveillance for nosocomial influenza, and the limited examination of the relative impact of different approaches to improving vaccination coverage.31,127,132,144–152,154 To be sure, the almost nonexistent baseline data and the lack of monitoring at and across most facilities must be recognized. Healthcare employers do not routinely assess, document, or track vaccination status (of staff or clients), nor is surveillance of nosocomial influenza or worker ILI routine. Moreover, the impact of improved coverage can be difficult to measure. For example, with high rates of presenteeism, lower absenteeism may not always follow improved worker vaccination rates (although improved absenteeism has indeed been reported28,29,86).
Nor can inferences about workers’ “refusing” vaccination be made from extant data about vaccination status. Action must therefore be guided by the precautionary principle,4,6 especially given the current ACIP/CDC recommendation for universal adult vaccination against influenza.87 Not all public health policy can have an extensive evidence base.154,155 While we seek quantifiable evidence on which to base decisions, uncertainty is endemic with respect to the impact of population-focused prevention interventions on reducing health inequalities.156 Indeed, evaluation of vaccine effectiveness (as opposed to efficacy) is inevitably retrospective157 and must likewise be so for immunization programs and policies. Individual-level vaccination data are important, but ecological designs are needed to evaluate real-world effectiveness of population-level interventions’ impact on herd immunity, transmission, and illness.
Proponents of mandates maintain that much better vaccination coverage can be achieved for the same expenditure of effort and resources used in aggressive voluntary programs that allow workers to remain unvaccinated.63,83,85,129,146,158,159 Available data support this view. Making protective measures enforceable is a mechanism to ensure compliance that does not exist with voluntary guidelines.160 Strict adherence to infection control is routinely expected as both a standard of care and a condition of employment. Compliance is considered preventive behavior, and noncompliance is a risk behavior. Poor compliance with vaccination standards or other basic infection control recommendations (e.g., hand hygiene and sharps and barrier precautions) may have its roots in failure to learn basic, essential practices while a student and then confusion about when and how to use them persists, especially when not unequivocally maintained as a workplace expectation and not consistently and universally modeled by more senior coworkers, even those recognized for their clinical ex
