Multi-country survey of attitudes towards the treatment of HIV / AIDS in the workplace

The Academy for Educational Development (AED) held a long-term, multi-country, multi-continent program funded by the United States Department of Labor (USDOL). It aimed to help both employees and employers deal with the impacts of the increase in the spread of human immunodeficiency virus infection/acquired immunodeficiency syndrome (HIV/AIDS) in the various parts of the world. The focus of this program was to work in the private sector to ameliorate the negative impacts of HIV/AIDS in the workplace. Measuring employees’ and employers’ attitudes towards people working with HIV/AIDS as part of early interventions was the key to the project success. AED has subsequently been acquired by Family Health International, to create FHI 360, which also does a significant amount of work with people living with HIV/AIDS. The project has ended but the survey results, while analyzed in 2006, continue to be relevant for practitioners planning to implement similar activities through corporateand donorfunded health in the workplace initiatives. Both co-authors now work at Creative Associates International.


INTRODUCTION
With HIV/AIDS most often affecting adults in their productive and reproductive prime, the business community has recognized its ethical responsibility and economic interest in protecting its most valued asset -its workforce.Employers, especially those in countries where the disease has been most devastating, are acutely aware of the direct financial and indirect social costs of the epidemic; at a minimum the business community has recognized the common good of having a healthy workforce.AIDS-related illnesses and deaths result in increased absenteeism, labor turnover, discrimination, stigmatization, and costly iterations of recruitment and training.Moreover, with a reduction in the number of skilled workers, employers have seen expertise decline and workers' productivity reduced.Some early projects led in the effort to get the corporate sector involved in helping to change peoples' behaviors through workplace initiatives -and had success doing so.In particular, the international HIV/AIDS workplace program SMARTWork -Strategically Managing AIDS Responses Together in the Workplace -was a pioneer in this area and contributed to ushering in an era wherein *Corresponding author.Author agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License policy makers and researchers began to view the workplace as a setting for HIV testing (Arimoto et al., 2012).The U.S. Department of Labor funded SMARTWork in 2001 in the Dominican Republic, Haiti, Nigeria, Zimbabwe, Ukraine, and Vietnam, and the United States Agency for International Development (USAID) provided additional funds through 2011 in Nigeria and Vietnam.The program was assessed to determine the degree to which it had brought about positive change in HIV-related knowledge, attitudes, beliefs and practices (KABP) among its main beneficiaries, that is, workplace employees.While this assessment was conducted in the program's early stages, it continues to be relevant for practitioners planning to implement similar activities through corporate-and donor-funded health in the workplace initiatives.

OVERVIEW OF THE SMARTWORK PROGRAM
The major objectives of the global SMARTWork program were to reduce employment related stigma and discrimination against people living with HIV/AIDS (PLHIV) and HIV/AIDS-related risk behaviors among employees.To accomplish these objectives, technical support was provided rather than direct funding to allow committees of workplace managers and workers to design, implement, and sustain their own HIV/AIDS workplace policies and programs.Program components included workplace policy development and dissemination; formal and informal HIV/AIDS prevention education; condom accessibility; services to address sexually transmitted diseases (STIs) and opportunistic infections; voluntary HIV counseling and testing; and HIV/AIDS counseling, care, support, and treatment services for PLHIV including employees and their partners and families.Health services administered through the program were on site or off site or through referral.
SMARTWork operated at multiple levels and with multiple sectors to build an environment to support HIV/AIDS policies and programs within individual enterprises.At the national and provincial levels, it took a tripartite approach fostering partnerships among government, labor, and business sectors to promote a national HIV/AIDS workplace policy and program framework.At the enterprise level, it took a bipartite approach fostering partnerships between business and labor to gain the commitment of employers and managers and to provide technical support to workplace HIV/AIDS committees in their design and implementation of workplace HIV/AIDS policies and programs for employees, their partners, and family members.The ultimate goal of the program was to impact employees of enterprises targeted by SMART-Work.More specifically, positive change among employees was expected in the following areas: 1. Increased knowledge of HIV transmission and prevention 2. Improved attitudes toward condom use 3. Improved attitudes toward persons living with HIV/ AIDS (PLWHA) 4. Reduced beliefs about workplace discrimination against PLWHA 5. Increased knowledge of national and workplace HIV/ AIDS policies 6. Decreased HIV/AIDS risk behaviors

Supporting surveys and studies
Businesses have joined more traditional sectors, that is, government and civil society, in the fight against HIV/AIDS with increasing numbers of employers by investing in the provision of prevention, care, and treatment of their workforce.In South Africa, the South African Business Coalition on HIV/AIDS (SABCOHA), a member-driven organization comprises 173 local businesses, aims to engage businesses in the national response to HIV/AIDS and coordinate a private sector response to the epidemic.
Workplaces are indeed strategic venues for implementing HIV/AIDS programs and policies, as they provide structured environments conducive to sharing information, accessing training and education, and implementing diverse interventions through a variety of communication methods.With employers' support for such initiatives, there is potential for increased awareness and action to prevent the disease among employees and for this knowledge to be transferred to employees' families.The following short descriptions illustrate the point.
1. Numerous studies focus on the costs, benefits, and best practices of workplace interventions (Arimoto et al., 2012;Cohen, 2002;Corebett et al., 2006;Dickinson, 2005;Lisk, 2002;Nattrass et al., 2004;Rodgers, 2010;Rosen et al., 2004;Sequeira et al., 2002).There are a series of issues these articles bring to light, but some common themes include: the workplace as a gathering place where it is common for people to meet, information to be distributed, and services to be provided at low costs; measurement in the workplace can be easier than in other environments because participant populations tend to be relatively stable; and workplaces engender environments where both employees and employers have incentives to participate in services provision and directly change behaviors.Moreover, workplaces can serve as gathering places for additional populationsfamilies, customers, and suppliers -that can be reached and influenced.The evidence is mixed on how much behaviors actually change in, for instance, not participating in extramarital sex.However, the evidence is consistent that in workplaces more information is disseminated and absorbed and that some key behavior changes, such as using a condom while having sex, are more positive in workplace settings.
2. The Investment Framework Study Group asserted that HIV/AIDS workplace programs need to focus on behavior change via group communication delivered through civil society and faith-based organizations targeting the workplace (Schwartländer et al., 2011).3. Bakuwa's research on private company-sponsored HIV/AIDS programs initiated in the workplace showed that bottom-up programs work better than top-down, and that programs with more staff involvement work better than those with less because peer group communications and pressure to conform are greater when large networks of people are involved.However, variations in responses by sector were noticeable.For example, companies in the service sector responded more often than those in trading (2010).4. A 2009-2010 study examined how South African workers' relational value with colleagues affected their willingness to participate in workplace HIV testing and counseling (HCT) programs.It showed that in one large enterprise, workers with strong relational value with colleagues were less likely to test for HIV as compared to colleagues with weak relational connections in the workplace (Arimoto et al., 2012).While it would appear to be a surprising finding, the cause is probably due to workplace stigma where workers who have closer affiliations are more concerned about others' judgment when being tested than those other groups who have more tangential relational affiliation.5. Studies of peer education programs in two different workplaces in South Africa and Botswana found there were conflicting results regarding the interventions' impact on employees' HIV/AIDS-related knowledge, attitudes, beliefs, and behaviors (KABP) (Hope, 2003;Sloan and Myers, 2005).Another study of a workplacebased HIV care program in South African demonstrated the effec-tiveness of an anti-retroviral therapy (ART) counseling intervention on patients' increased HIV/ART knowledge (Stenson et al., 2005).6. Bärnighausen et al. (2012) use data from ART programs to show the economic value of HIV treatment as prevention (TasP) where the economic situation of patients improves as they get ART treatment and the economic situation of patients who receive preventative TasP therapy avoid drops in economic productivity.This finding has direct implications for employers to invest in HIV-infected workers' early treatment to maintain their workforce contributions.7. Kortum et al. (2011), who explore the nature of psychosocial hazards in the workplace and work-related stress, find that HIV/AIDS-related absenteeism can increase and create stress for coworkers picking up theextra workload, which is beyond the predictable stress PLWHA would face., 2000).Once the questionnaire was finalized, it was distributed to SMARTWork staff in the related countries where it was reviewed for cultural appropriateness.Minor adjustments were made in the wording of some items to better fit the cultural contexts.This contributed to the instrument's in-country reliability.The questionnaire included items concerning respondents' socio-demographic characteristics, knowledge of HIV prevention and transmission, attitudes toward condom use, attitudes toward PLWHA, beliefs about workplace discrimination against PLWHA, knowledge of national and workplace policies, and HIV/AIDS risk behaviors.
After receiving written consent, the questionnaire was administered to most employees by trained interviewers in a private setting taking approximately 30 min per interview.The exception was in Ukraine where the questionnaire was administered through touch screen computers.Interviewers received good cooperation from the employees with an approximate 95% acceptance rate of those who were invited to participate in the study.
The cross-sectional study recruited independent samples of employees from the same enterprises at baseline and postintervention in Haiti, the Dominican Republic, Nigeria, Ukraine, and Vietnam.The study was not completed in Zimbabwe due to reduced program funding.Although SMARTWork worked in approximately 10-20 enterprises per country, based on budget exercises, the study was implemented in 1-2 enterprises per country.A purposeful sampling approach was used to invite and recruit enterprise participation in the study.Recruited enterprises were representative of the larger group of enterprises, committed to SMARTWork, agreeable to study participation, and targeted for substantive technical assistance from SMARTWork.Working with employers, the study team obtained lists of employees by department and by whether the employee was non-management or a manager.Names of employees were then randomly selected across departments so that a stratified sample of approximately 100 employees, 75 nonmanagement employees, and 25 managers could be obtained.
The baseline study was completed in the spring of 2004 with a sample of 686 employees from the five countries and the post study in the spring/summer of 2005 in the Dominican Republic and in 2006 in all other countries with a sample of 706.Table 1 summarizes the sample, including adjustments after weighting the data.

Participants' demographics
Table 2 summarizes the demographic characteristics of the participants in the study.Slightly fewer males than females participated in the study at baseline (47 vs. 53%) than at post (53 vs. 48%); they were on average the same age (32 years) at both times of the study and somewhat less educated at baseline than at post (43% with higher education vs. 50%).Regarding marital status, baseline participants were less often married than those at post (56% vs. 62%) and more likely to be unmarried but living with a sexual partner at baseline than at post (22% vs. 18%).There were significant differences regarding participants' marital status at baseline and post by country.In Vietnam, only 35% of all participants were married at baseline vs. 60% at post and in Haiti only 24% at baseline vs. 16% at post.With respect to the employees' position at work, at both baseline and post, most participants were non-management employees (83%) vs. management (17%).Participants at baseline generally had not been employed at their workplace as long as those at post (70% > one year vs. 75%).

Survey results
The results of the KABP survey are shown in Table 3 including baseline and post findings by survey item.Significant findings were reported at the p<.001 level.A more detailed discussion of these results is provided below, including findings by country.

Source of information and participation in the HIV/AIDS program
Having access to HIV/AIDS information is critical to the advancement of employees' HIV/AIDS knowledge.Surveyed employees reported a significant increase from baseline to post in getting information from a manager or professional at the worksite (46.7 vs. 79.2%), a labor union representative (39.2 vs. 57.9%),and/or a co-worker (67.7 vs. 86.7%)-three major categories for HIV/AIDS dissemination in the workplace targeted by the SMART-Work program.These positive changes were most pronounced in the Dominican Republic, Vietnam, and Haiti.Moreover, other findings showed that program implementation efforts had reached the target population, with a variation in implementation by country.Significantly more employees at post than at baseline had received: formal HIV/AIDS education (41.3% vs. 77.4%);informal HIV/AIDS education from a co-worker (41.2% vs. 81.9%,with the most limited change shown in the Ukraine enterprises); and free or low cost condoms (32% vs. 54.6%, with the strongest change shown in the Dominican Republic and Vietnam enterprises and no change shown in the Ukraine enterprise).In addition, significantly more workers, their partners, and family members at post vs. baseline had received STI treatment (2.7 vs. 7%, with most of this change being accounted for in the Haiti enterprise); and voluntary HIV counseling and testing (17% vs. 30.6%,with the strongest change seen in the Dominican Republic and Haiti enterprises and the most limited in the Nigeria, Vietnam, and Ukraine enterprises).Surprisingly, it was reported that the receipt of drug treatment by workers, their partners, and/or family members decreased from baseline to post (22% vs. 8%).

Knowledge of HIV transmission and prevention
While findings showed that the program had reached many of the employees in the study, results showed that the intervention's impact on knowledge related to correct routes of HIV transmission and its prevention was mixed.Significantly more employees at post than at baseline knew that a healthy looking person could be infected with the virus that causes AIDS (94 vs. 98%) and that you could not become infected by sharing food with a person who has AIDS (78.5 vs. 86.2%).However, there was no notable change in the portion of employees at baseline vs. post who knew that HIV transmission could be prevented by having one faithful sexual partner who is not infected with HIV (74.3 vs. 76.9%), by using condoms during penetrative sex (86.8 vs. 88.5%), by abstaining from sex (60.3 vs. 62.9%), or, for drug users, by not sharing injection equipment (79.6 vs. 80.9%).Of notable concern was the fact that at post survey over a quarter of employees (28.5%) still believed that one could prevent infection by avoiding toilets, including 59% of the Nigerian employees.

Attitudes toward condom use
Results showed that the intervention generally had a minimal impact on employee's attitudes toward condom use.At both baseline and post, most had a favourable attitude toward usage (that is, it is 'good', 'wise', and/or 'easy' to use a condom if having penetrative sex with a person other than your spouse or live-in partner).The strongest impact of the intervention was shown in employees' intention to use a condom with significantly more at post than baseline 'very sure' that they could use a condom if having penetrative sex with a person other than their spouse or live-in partner (74.7 vs. 88.3%).Regarding country differences, the intervention had the strongest impact on employees' attitudes toward condom use in the Dominican Republic and Haiti enterprises.

Attitudes toward persons living with HIV/AIDS
The strongest consistent impact of the intervention across survey items in all countries was seen in employees' increased favourable attitudes toward PLWHA, that is, reduced stigma attitudes.Significantly more employees at post than baseline were comfortable concerning the following behaviours with a person/employee with HIV/ AIDS: eating lunch alongside (69.3 vs. 79.6%);traveling to work with (82.5 vs. 91.8%);working alongside (67.3 vs. 81.2%); and visiting (74.3 vs. 86.3%).While significantly more employees at post than baseline felt comfortable buying food from a shopkeeper or food seller who has HIV/AIDS (32.4 vs. 44.3%),considerable stigma still remained related to this behavior.Regarding findings by country, the Ukrainian employees made the smallest gains in reducing HIV/AIDS stigma in the workforce.

Knowledge of HIV/AIDS policy and beliefs about workplace discrimination against PLWHA
As a result of the intervention, all enterprises that participated in the survey had an HIV/AIDS workplace policy in place at the time of the post survey.As a result, more employees at post than at baseline knew that their workplace had an HIV/AIDS policy (42.7 vs. 71.1%),although approximately two out of five (42.8%) at post reported that they had not actually seen the policy.Changes in employees' awareness of their workplace policy were strongest in the Dominican Republic and in Vietnam.
Overall, results showed that the intervention had a positive impact on employees' beliefs about workplace discrimination against PLWHA.Significantly fewer employees at post than at baseline believed that people with HIV/AIDS who worked at the worksite should worry about losing their job (64 vs. 46.5%)or that people at the worksite should worry that their employer might find out that they got tested for HIV/AIDS (54.8 vs. 32%).And significantly more employees at post vs. baseline believed that their employer would keep people who have HIV/AIDS employed as long as they can perform their duties (50.2% vs. 73.3%).While these results demonstrate a positive change, they also demonstrate that a sizable portion of employees still believe that workplace discrimination against PLWHA is possible despite a new non-discriminatory policy.Across the enterprises in the study, this was especially evident in the Ukraine enterprise.

HIV/AIDS risk behaviors
Results showed that the intervention had a minimal impact on employees HIV/AIDS risk behaviors, but there was considerable variation by enterprise and country.At both baseline and post survey, approximately one in four employees reported to have had penetrative sex with a person other than their spouse or live-in partner (23.4 vs. 23.7%).Of these individuals, most reported at both baseline and post that they had used a condom the last time they had sex with a non-regular partner (71.6 vs. 74.0%).The impact was more pronounced in two of the seven enterprises.Among the Haiti enterprise employees, condom use at the last time of sexual intercourse with a non-regular partner increased significantly from 50% at baseline to 72% at post; and among the Ukrainian employees condom use increased from 47.2 to 73.3%.Results also showed significant differences by age with younger employees more likely to have had extra-marital sex with condoms than older employees.

DISCUSSION
There were a number of limitations to the present study.First, without the use of control or comparison sites it was difficult to attribute the observed impact on workplace beneficiaries directly to SMARTWork.Second, while SMARTWork staff provided technical assistance in surveyed enterprises to empower workplace champions to plan and implement the program model, their secondary oversight could not guarantee fidelity to the intervention model.In addition, variations in the level and timing of employer cooperation prohibited all baseline studies from being implemented before the SMARTWork intervention commenced which impacted the study design and, potentially, the results.To evaluate the impact of this factor, analyses were conducted controlling for whether employees had taken part in formal and/or informal HIV/AIDS education at the workplace both at baseline and post assessment.
To interpret the study results, it is important to understand factors that affected SMARTWork implementation efforts in the targeted workplaces.Variation existed in the time and energy devoted to the workplace initiative by site implementers -union representatives, managers, and HIV/AIDS planning committees.This, in turn, impacted the scope and quality of the resulting HIV/AIDS program and policies in individual enterprises.At the low implementation level, some enterprises were implementing one or more of the following: formal education consisting of the delivery of a short intervention on World AIDS Day, minimal information dissemination or visible posters, inactive peer educators, a lack of dissemination of the HIV/AIDS policy, and weak provision of health services including limited condom accessibility and referrals for VCT or AIDS treatment and care.At the high implementation level, some enterprises were providing consistent prevention education messages throughout the year, highly visible prevention poster, a highly visible HIV/AIDS policy, active peer educators, accessible condoms, and delivery on site or through referral of VCT and HIV/AIDS care and treatment.
A major factor that affected implementation efforts was the extent to which employers identified the importance of the initiative and invested in it.More highly invested employers accomplished some or all of the following: established a comprehensive HIV/AIDS policy for their company; authorized the use of production time so employees could take part in educational forums and workshops; allowed peer educators to engage in prevention dialogue with employees on the shop floor; allocated HIV/AIDS planning committee members time to plan and implement the initiative; and instituted employee health plans that supported substantive health services related to HIV/AIDS prevention, care and treatment.Other employers were less convinced that the initiative was a wise investment, more concerned with 'the bottom line', and less willing to be cooperative in supporting robust implementation efforts.
Despite varied implementation efforts and employer support, preliminary results suggest that the SMARTWork program reached many employees, especially in the receipt of formal and informal HIV/AIDS education.Not surprisingly, the program appeared to have a stronger impact on those who took part in formal or informal HIV/AIDS education in the workplace vs. those who did not including: better HIV/AIDS-related knowledge of transmission and prevention; more positive attitudes toward condom use and strong intentions to use condoms with non-regular partners; less stigmatized attitudes toward PLWHA; more knowledge about their national and workplace HIV/AIDS policy; less belief that employer discrimination toward PLWHA existed in their workplaces; and more self-reported use of condoms with non-regular partners.
Overall, SMARTWork seemed to have the strongest impact on employees' increased assuredness to use a condom if having penetrative sex with a non-regular partner, reduced stigmatizing attitudes toward PLWHA, reduced beliefs that discrimination against PLWHA existed in their workplaces, and increased knowledge of their national and workplace HIV/AIDS policy.Across the countries that participated in the study, the Dominican Republic and Haiti SMARTWork programs appeared to have shown the most positive impact on employees in the surveyed enterprises.

Study implications for international HIVAIDS workplace programs
The study offered some preliminary empirical findings that can help inform other workplace HIV/AIDS programming efforts in the field.Findings showed that many employees were not receiving free or low cost condoms and VCT.And, while it was difficult to know employees' need for STD or drug treatment, there was little evidence that the participating enterprises were making provisions for these services.This implies that HIV/AIDS workplace programs should make concerted efforts to disseminate condoms, establish relationships with agencies and health facilities that can provide needed health services and utilize confidential referral processes for employee use of these services.
Results showed that more could be done to increase employees' knowledge of HIV prevention and transmission especially in the areas of understanding the effectiveness of abstinence and having one faithful partner who is not infected with HIV.Moreover, educational interventions should also address the myths of HIV transmission, such as acquiring HIV through the use of public toilets, a misperception that remains prevalent.Findings indicated that stigma still exists implying that workplace programs should broaden their teachings to help employees understand, for example, that buying food from a shopkeeper or food seller poses no risk.
While findings showed that most employees learned about their national and workplace HIV/AIDS policies, many had not actually seen their workplace policy.Moreover, despite the existence of a workplace policy, many employees felt that a PLWHA could not feel confident about job security or that other forms of discrimination may take place.These results imply that workplaces should more consistently make HIV/AIDS policies available to employees and utilize key opinion leaders such as union representatives and peer educators to reinforce companies' non-discrimination policies.
Disappointingly, the SMART Work intervention did not appear to have a strong impact on the adoption of the protective behavior of condom use with non-regular partners, with the exception of the Haiti enterprise and one of the two Ukraine enterprises.Considering that the intervention had only operated between 1-2 years in the participating enterprises, this is not so surprising.General knowledge of HIV/AIDS risk behavior change indicates that behavior change can often take a long time to adapt and sustain.This implies that workplace programs need to be supported over time to bring about behavior change and require consistent risk reduction messages targeted to specific populations, the delivery of these messages by key opinion leaders and the availability of materials to reduce harm, such as condoms.This last requirementavailability of materials to reduce harm -is supported by empirical findings in the present study.The enterprise in Haiti demonstrated significant increased condom use at last sex with a non-regular partner and was also the enterprise with a significantly higher rate of condom dissemination at post-study (91.8%) compared to all other enterprises in the study (56.8%).

Conclusion
While the SMARTWork program demonstrated that it can impact employees in several positive ways, considerably more can be done by workplace programs.With employees harboring misconceptions about how HIV is transmitted, more concentrated educational efforts should be fostered in confronting these myths and providing more quality control over what prevention messages are being delivered by educators in the workplace.SMART-Work established a successful foundation of workplace policies and positively impacted many employee attitudes toward stigma.Real behavior change for risk behaviors takes time and will need more sustained support and inputs.It is recommended that workplace HIV/AIDS programs move away from relying on general informational and educational campaigns and towards more comprehensive behavior change programs with segmentation and clear communication channels transmitting targeted messages about needed behavior change.Comprehensive behaviour change programs should also consistently ensure employees' access to harm reduction materials, such as condoms, to facilitate reduction of risk behaviors.

METHODSA
structured KABP questionnaire was designed based on SMART-Rogers and Wright Jr. 533 Work's strategic framework, a tool that outlined the project's objectives and related performance indicators.Most of the performance indicators were taken from the standardized list of USAID HIV/AIDS indicators and measures (USAID

Table 3 .
Results at baseline and post by survey item.
IV. Attitudes toward/Perceptions about condom useIt is good to use a condom if having penetrative sex with a person other than your spouse or live-

. Attitudes toward persons living with HIV/AIDS
Of those who had penetrative sex with person other than spouse or live-in partner in last 12 months) Used a condom the last time had sex with person other than spouse or live-in partner. *p<.001.