Health Education Research, Vol. 15, No. 4, 463-467,
August 2000
© 2000 Oxford University Press
Orginal Articles |
A National Institute for Occupational Safety and Health Alert sent to hospitals and the intentions of hospital decision makers to advocate for latex allergy control measures
1 National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention, HHH Building, 200 Independence Avenue SW, Washington, DC 20201 and
2Department of Community Medicine, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piscataway, NJ 08854, USA
Abstract
This study evaluated a National Institute for Occupational Safety and Health (NIOSH), Centers for Disease Control and Prevention, Alert concerning the risk and prevention of latex allergy among health care workers. It has been estimated that 8–12% of health care workers are sensitized to latex. NIOSH Alerts are publications that are intended to educate stakeholders about risks in the workplace; this Alert contained four recommendations for administrative control measures that hospital decision makers could adopt to reduce the risk of latex allergy to employees. The Alert was mailed to a random selection of Directors of Infection Control and Directors of Nursing in hospitals in the US. A random sample of these targeted recipients and a control group were surveyed by telephone (N = 298). Although nearly all of the respondents were concerned about latex allergy (96%), those reporting having seen the Alert were significantly more likely to report an intention to advocate for one or more of the control measures.
Introduction
The National Institute for Occupational Health and Safety (NIOSH), part of the Centers for Disease Control and Prevention, has estimated that 8–12% of health care workers versus 1–6% of the general population are sensitized to latex (National Institute for Occupational Safety and Health, 1997
). Latex allergy is caused by a reaction to proteins in natural rubber latex (NRL) which is contained in a variety of products commonly used in hospitals. The frequent and prolonged use of latex gloves by health care workers for blood-borne pathogen barrier protection results in a high degree of exposure to the NRL antigens. The cornstarch powder used in most powdered latex gloves is also a risk factor. When powdered gloves are removed, the powder and allergens become airborne and can be inhaled. Allergic reactions to NRL can range from mild skin irritation to much more severe, but rare, anaphylactic shock.
NIOSH Alerts are educational documents concerning newly identified health and safety risks in the workplace. The NIOSH Alert: Preventing Allergic Reactions to Natural Rubber Latex in the Workplace described the causes of latex allergy, the symptoms, tests available for screening employees and recommendations for preventing latex allergy (National Institute for Occupational Safety and Health, 1997). In July of 1997, NIOSH disseminated this Alert to the occupational health and safety community and a number of targeted stakeholders. NIOSH also made the document available to those requesting it through a 1-800 number.
In January of 1998, NIOSH mailed the Alert directly to hospitals. In order to evaluate the effects of this mailing, the Alert was first sent to a randomly assigned group of hospitals and, specifically, the Director of Infection Control or the Director of Nursing. Hospitals not receiving this initial mailing were later sent the Alert.
The Alert made six voluntary recommendations to employers to reduce the risk of latex allergy, four of which were the focus of this study (National Institute for Occupational Safety and Health, 1997).
- Provide workers with non-latex gloves to use when there is little potential contact with infectious materials (e.g. in the food service industry).
- If latex gloves are chosen, provide reduced protein, powder-free gloves to protect workers from infectious materials.
- Provide workers with education programs and training materials about latex allergy.
- Ensure that workers use good housekeeping practices to remove latex-containing dust from the workplace.
As should be clear from the recommendations, reducing the risk of latex allergy to health care workers is largely contingent on the adoption of these control measures at the level of the hospital. Yet, only one copy of the Alert was sent to a single recipient in each hospital. In order for the mailing to have any effect on hospital policy, these recipients would have to see the Alert and either act on the Alert themselves or forward the Alert to others who will act on the recommendations.
In diffusion theory, `change agents' are the link between the source of an `innovation' and the potential adopter of that innovation (Rogers, 1995). In this case, NIOSH and the Alert are the source, the `innovation' is a set of recommended administrative policies to control for the risk of latex allergy, and the potential adopters are the decision-making bodies in hospitals that set policy. The Alert has to create a change agent in the hospital, or assist a change agent who is already advocating or intending to advocate for the recommendations.
The selection of the recipient for the Alert is important. Many in hospitals may need to be persuaded that latex allergy needs to be addressed. The recommended control measures may be particularly difficult to implement. However, if recipients are already concerned about the risk of latex allergy to employees in their facility, the NIOSH Alert by providing credible evidence for the risk of latex allergy that recipients could use to advocate for addressing the problem.
In other words, if the Alert is sent to a recipient who is already concerned about latex allergy, the Alert may increase their ability to advocate for policy changes and thus their willingness to be proactive change agents.
Methods
Study design and population
The population was defined as inclusion in the 1997 American Hospital Association (AHA) list. The AHA list includes registered hospitals and US Veterans Administration facilities. Hospitals in Massachusetts, which had already received copies of the Alert from their State Department of Health, were excluded from the study.
Hospitals were randomly assigned to one of four conditions in a 2x2 factorial design. The two independent variables were sending or not sending the Alert and mailing the Alert to either the Director of Infection Control or the Director of Nursing. Results of a series of open-ended interviews conducted by the University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School (UMDNJ-RWJMS) indicated that the Director of Nursing or the Director of Infection Control would be the most appropriate recipients for the Alert. They were judged as having (1) influence in hospitals, (2) formal responsibility for the health and safety of hospital employees, and (3) access to the decision-making bodies regarding latex control measures.
The SPSS 8.0 pseudo-random number generator was used in all randomization procedures (SPSS, 1988).
The Alert was inserted in a Department of Human Health and Services manila envelope, the title of the target recipient was added to the mailing label and the packages were sent on January 5, 1998.
Data collection
UMDNJ conducted a telephone survey for the post-treatment observation. The samples were randomly selected by randomly sorting the telephone numbers included in the AHA list. The instrument was pre-tested with a small sample and the survey was conducted 2–6 weeks after the Alert was mailed.
For each of the four recommendations, hospitals were classified as either: (1) already having adopted the policy, (2) evaluating the policy or procedure, or (3) inactive, i.e. not already acted or begun to act on the policy. If hospitals were `inactive', respondents were asked about their intentions, `Do you intend to advocate for [the policy]'. In a Likert scale item with no mid-point, respondents were asked to strongly agree, agree, strongly disagree or disagree with the statement, `Hospitals should immediately address the risk of occupational latex allergy'. Respondents were also asked, `Have you talked to a colleague about latex allergy in the last 2 weeks'.
In order to reduce response bias, respondents were not asked about the Alert itself until the end of the interview when they were asked whether they recalled the Alert. If they did recall the Alert, they were asked whether (1) they intended to use the Alert to educate administrators and (2) whether they forwarded the Alert within the hospital.
Hypotheses and data analysis
It was hypothesized that those sent the Alert would be significantly more likely than those not sent the Alert to intend to advocate for control measures consistent with the recommendations. The test of significance was a 
2 statistic used with 2 (sent the Alert/not sent the Alert)x2 (intention to advocate/no intention to advocate) contingency tables. The sample for each of the contingency analyses was restricted to hospitals that were classified as `inactive'.
However, the power of the study was reduced by two factors. There was (1) a ceiling on the number of inactive hospitals and (2) a diffusion of the treatment, i.e. the Alert, across conditions. As a result, the study was also analyzed as a cross-sectional survey. The intention to advocate measure was used as the dependent variable in a series of logistic regression analyses to determine whether the respondent (Director of Nursing versus the Director of Infection Control), the size of the hospital and the incidence of latex allergy were independent predictors or covariates of an intention to advocate for the policy measures. Then, the intentions of those that reportedly saw the Alert in both the treatment and control conditions were compared to those that reportedly did not see the Alert. Again, contingency analyses with the 2 statistic were used to test for differences.
Samples
Sample sizes were determined by estimating the statistical power necessary to assess differences between the sent Alert treatment condition and no Alert control condition. Power analysis was conducted anticipating the use of the Pearson 2 as the test of significant differences. Assuming that 60% of the respondents receiving the Alert versus 40% of the control group reported an intention to advocate for a specific control measure, this would be an effect size of 0.2; with N = 200, the power would be 0.81 (Cohen, 1988).
There were a total of 298 respondents, 101 in the control condition and 197 in the treatment condition. (Twenty-five cases were dropped from the analysis because the wrong individual was interviewed or the respondent was not representing a hospital.) The response rate was 44%.
The treatment and control condition samples were compared in terms of the number of beds and incidence of latex. There is a greater proportion of smaller hospitals (under 50 beds) in the treatment group (36 versus 26% for the control group) and a larger proportion of mid-size hospitals (between 50 and 200) in the control group, but this difference is not significant at P < 0.10. Both the treatment and control groups have approximately the same proportion of larger hospitals (29 and 27%, respectively, had 200 or more beds). The samples are nearly identical in terms of the respondents' reported suspected or confirmed incidence of latex allergy among employees in their hospitals. Approximately 36% of the respondents reported no cases of latex allergy and 19% reported having five or more suspected or confirmed incidents of latex allergy.
Results
The results indicate that the majority of hospitals were already moving to address latex allergy. Approximately 63% of the respondents reported that their facilities were either intending to eliminate the use of powdered latex gloves or they had already done so; 57% said that their facility had educated their employees about the risks of latex allergy and an additional 24% said they were intending to do so. Nearly all of the respondents, 96%, either agreed or strongly agreed that hospitals needed to immediately address the risk of latex allergy.
The intentions of those in the treatment (sent Alert) condition and those in the control (not sent Alert) condition in hospitals that had not yet acted (the `inactive' hospitals) on the specific policy measures are shown in Table I. Except with regard to employee health education programs, the majority of respondents in hospitals classified as `inactive' stated that they did not intend to advocate for the control measures. Those who were sent the Alert were not significantly different than those not sent the Alert in terms of intentions to advocate for any of the recommended measures.
|
Of those who were sent the Alert, only 55% recalled having seen it. While it was not expected that a direct mail piece would have 100% recall, more problematic, from the standpoint of the evaluation, was that 29% of the respondents in the control condition also reported having seen the Alert. Respondents in the control group may have reported seeing the Alert because they felt they should have seen it, a social desirability bias, but it is likely that many received a copy from other sources. In particular, the American Nurses Association had received the Alert in the earlier dissemination to stakeholders and had conducted their own mailing.
These factors reduced the power of the study. As a result, the treatment and control groups were collapsed and the overall sample was examined as a cross-sectional survey. In a multiple logistic regression analysis, the size of the hospital, the respondent and the number of incidences of latex allergy were not significantly associated with an intention to advocate for any of the recommended measures. The results of the contingency analyses comparing the intentions of those who reportedly saw the Alert to those who reportedly did not see the Alert are indicated in Table I
.
Having seen the Alert is positively and strongly associated with an intention to advocate for policies consistent with three of the four recommendations: employee health education programs, eliminating powdered latex gloves and restricting latex glove use. There was also a significant association between recall of the Alert and having recently talked to a colleague about latex allergy. Fifty-seven percent of those who reported seeing the Alert said that they had talked to a colleague in the last 2 weeks about latex allergy compared only 43% of those respondents who reportedly did not see the Alert (2 = 16.8, P < 0.001, N = 289). Of those who reportedly saw the Alert, 96% stated that they would use the document to educate administrators in their facility.
Discussion
The use of single-item indicators of an intention to advocate for the recommendations is a limitation in this study, as is the lack of precise behavioral outcome measures. It is unclear what was entailed in their intentions to advocate and single item measures lack reliability. Those recalling the Alert were more likely to have recently talked to a colleague about latex allergy, but what other behaviors may be involved in `advocacy' is unknown. In general, very little is known about what makes change agents within organizations and under what circumstances they are effective.
The low response rate, 44%, limits the degree to which the results can be generalized to the larger study population. It may be that some respondents participated in the survey because their hospital had already addressed latex allergy. It is also possible that respondents who reported seeing the Alert, noticed it, and perhaps in some cases, procured it (e.g. those in the control condition) because they already intended on advocating for the control measures.
Finally, due to the reduced power of the field experiment, it is unclear whether the mailing of the Alert to hospitals was successful at creating `advocates' for the control measures. However, evaluating the mailing of the Alert is not a test of the influence of the Alert. Seeing the Alert was strongly and positively associated with an intention to advocate for three of the four control measures.
Given that 96% of the respondents, including those that did not recall the Alert, agreed that hospitals should immediately address the risk of latex allergy, the Alert probably did not persuade these decision makers that latex allergy should be a concern. Rather, the Alert appears to have increased their willingness to act on their concerns. It is instructive that nearly all of those who received the Alert said that they would use the document to educate administrators.
The Alert may have provided recipients with support in the form of expertise and credibility with which to approach their colleagues about the potentially difficult changes required to adopt control measures for latex allergy.
References
Cohen J. (1988) Statistical Power Analysis for the Behavioral Sciences, 2nd edn. Lawrence Erlbaum, Hillsdale, NJ.
National Institute for Occupational Safety and Health (1997) NIOSH Alert: Preventing Allergic Reactions to Natural Rubber Latex in the Workplace. DHHS (NIOSH) publ. no. 97-135. NIOSH, CDC, Washington, DC
Rogers, E. M. (1995) Diffusion of Innovations, 4th edn. Free Press, New York.
SPSS (1988) SPSS Base 8.0: Users Guide. SPSS, Chicago, IL.
Received on September 9, 1999; accepted on January 12, 2000
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||