A Survey on Health and Chemical Exposures
Nurses' Health: About the Survey
Design and Dissemination of the Survey
In collaboration with Health Care Without Harm, the American Nurses Association, and the Environmental Health Education Center of University of Maryland's School of Nursing, Environmental Working Group (EWG) scientists designed an online survey to assess nurses' extent and frequency of exposures to 11 different common health care hazards, as well as the prevalence of a wide variety of health problems among nurses and their children. The survey collected detailed information on job and health history from participating nurses and their children. The survey also collected basic occupational, demographic, and lifestyle data, and probed nurses' attitudes concerning occupational health and their own workplace environments.
The American Nurses Association, Health Care Without Harm, the Environmental Health Education Center of University of Maryland's School of Nursing, and the American Public Health Association's Nurses Section disseminated information about the survey at meetings, conferences, and workshops, and in nursing journals, state nurses associations' newsletters, and specialty nursing organizations' publications. Links to the survey were posted on various nursing websites, including those of the collaborating nursing organizations. Any interested nurse was free to participate in the survey via the website. EWG collected survey responses between March 26, 2005 and March 20, 2006.
Analyzing Survey Data
EWG scientists received complete survey data from 1,552 nurses representing all 50 U.S. states. Eighty percent of respondents were Registered Nurses (RNs); 92 percent were currently working in nursing; and 94 percent were women. The number of years that respondents reported nursing was broadly distributed, with 44 percent of nurses practicing for between 20 and 35 years. Age of respondents was also broadly distributed, with 22 percent reporting an age between 50 and 55 years. Respondents were primarily white/caucasian and non-hispanic (88 percent). Thirty-two percent of respondents had two children, and 42 percent lived in suburban locations. States with a large number of respondents included Florida (80 nurses), Maryland (108 nurses), New York (114 nurses), and Ohio (98 nurses).
EWG scientists analyzed survey responses using basic statistical methods. Researchers assembled information on health effects associated with each hazardous hospital exposure from publicly available government databases, including the Hazardous Substances Data Bank (HSDB) and the Registry of Toxic Effects of Chemical Substances (RTECS), as well as peer-reviewed scientific literature and government publications. Health problems were assessed statistically only if a particular exposure was known or suspected to be its cause. Exposures were examined independently; no attempt was made to assess patterns in exposure to mixed and multiple hazards, although this is an area in need of further research.
EWG researchers assessed health problems among surveyed nurses as a function of exposure duration, comparing nurses reporting exposures at least once a week for at least ten years (high exposure) against all other nurses responding to the survey (low or no exposure). Health problems among the children of surveyed nurses exposed to occupational hazards were assessed as a function of timing of exposure. The prevalence of health effects in the children of nurses exposed at least once per week during pregnancy was compared with the prevalence of health effects in the children of nurses with less frequent or no exposure during pregnancy. This website provides health effects that show an increased prevalence of greater than one percent among highly exposed nurses or their children. In addition, 90 percent confidence intervals were calculated for summary statistics reported in the study summary.
The pool of participants that responded does not necessarily represent a typical cross-section of American nurses, and the survey is subject to recall bias and other common forms of bias. Despite the limitations, the survey does show that many nurses exposed to hazards on the job are experiencing health problems. Remarkably, the government has never conducted a national survey of nurses' health and occupational and environmental workplace hazards, despite the diverse and chronic exposures nurses face. Furthermore, many hospitals have been slow to implement plans to reduce nurses' exposures to hazards. More action and more research are critically needed.
Surveys are a well-established epidemiological tool for identifying health risks from occupational exposures to hazards. Several studies of hazardous exposures and the health of health care workers have employed mail-in surveys to collect data (e.g. Hemminki et al. 1982, Valanis et al. 1999, Mohan et al. 2003, Dimich-Ward et al. 2004, Yoshinaga et al. 2005); some also allow for online response to the study questionnaire (e.g. Delclos et al. 2006). Traditional epidemiological surveys are sent to a select group of individuals. Typically, epidemiologists report rates of response to their questionnaires, and often perform statistical adjustments to the data to ensure that it is representative of the entire population of concern.
To reach a broader audience, EWG created an online survey open to anyone interested. While this online format enabled dissemination of the survey to nurses all over the United States, it is not possible to calculate or estimate the response rate. Comparisons to preliminary findings of the National Sample Survey of Registered Nurses (DHHS 2004) suggest that the collection of nurses that chose to respond to the EWG survey may well-represent others in their profession. For example, the average age of nurses in this survey was 47, just as in the national survey; nurses under 40 made up 23.9 percent of the respondents in this survey, while they were 26.6 percent of the national survey. A review of studies on human psychology reveals that internet-based research methods typically produce the same results as traditional methods, providing no evidence for unusual bias among study populations (Gosling et al. 2004).
Some research suggests that those that choose to respond to health surveys like the one presented here may be more likely to have experienced the hazardous exposures and health problems described, or be more aware of the hazardous nature of their work. For example, Axelsson and Rylander (1982) noted that women that were occupationally exposed to anesthetic gas and suffered a miscarriage always responded to a postal survey on the subject, while one-third of women that suffered a miscarriage without being exposed to anesthetic gas did not.
Another issue common to survey-based health data is the accuracy with which respondents report medical information. Nurses are typically better educated about health than the general population, and therefore may provide more accurate information via survey, but this type of "reporting bias" could influence survey findings. Analytical decisions regarding classification of the nurses into high and low exposure categories may also affect survey interpretation. In addition, analyses did not take into consideration confounding factors like exposure to multiple chemicals associated with the same health effect, or other demographic or lifestyle factors that often affect health, such as age, smoking and drinking habits, and day or night shift work; such factors could influence survey results.
While the basic statistical associations reported were not adjusted to account for the potential effects of these types of bias, the associations between hazardous exposures and health outcomes reported demonstrate the need for increased research on nurses' health, and more protective occupational regulations in the health care setting.
About the Survey Partners
Environmental Working Group
Environmental Working Group's research (like Skin Deep and the Shoppers' Guide to Pesticides in Produce) brings to light unsettling facts that you have a right to know. It shames and shakes up polluters and their lobbyists. It rattles politicians and shapes policy. It persuades bureaucracies to rethink science and strengthen regulation. It provides practical information you can use to protect your family and community.
University of Maryland School of Nursing (SON)
The University of Maryland School of Nursing (SON) founded in 1889, is one of the oldest and largest nursing schools, and is ranked seventh nationally. Enrolling more than 1,600 students in its baccalaureate, master's, and doctoral programs, the School develops leaders who shape the profession of nursing and impact the health care environment.
The School of Nursing's Environmental Health Education Center addresses environmental health issues by engaging in education, research, practice, policy and advocacy. The Center's primary focus is on the nursing profession and working collaboratively with community groups, health professionals, advocacy organizations, and government agencies in addressing environmental health challenges.
Health Care Without Harm
Health Care Without Harm is an international coalition of hospitals and health care systems, medical professionals, community groups, health-affected constituencies, labor unions, environmental and environmental health organizations and religious groups.
American Nurses Association
The American Nurses Association is the only full-service professional organization representing the interests of the nation's 2.9 million registered nurses through its 54 constituent member nurses associations. The ANA advances the nursing profession by fostering high standards of nursing practice, promoting the rights of nurses in the workplace, projecting a positive and realistic view of nursing, and by lobbying the Congress and regulatory agencies on health care issues affecting nurses and the public.