Simon's first thought after learning the results of his HIV test was of ending his own life. "After I was told that I was positive, I was frightened," he said. (Simon's name has been changed to preserve confidentiality)
But gradually, with the help of counselor Geoffrey Wathome of the Kenya Association of Professional Counselors (KAPC), Simon started looking to the future. He received treatment for a sexually transmitted infection (STI). He convinced his wife to go to counseling with him and started using condoms. "Now I do not infect anybody or reinfect myself," he said.
Making HIV counseling and testing available to men like Simon is a cost-effective way of preventing further spread of the virus, according to a three-country study sponsored by the AIDS Control and Prevention (AIDSCAP) Project of Family Health International (FHI), the Joint United Nations Programme on HIV/AIDS (UNAIDS) and the World Health Organization. This randomized, controlled trial conducted at counseling centers in Kenya, Tanzania and Trinidad found that voluntary testing, combined with professional pre- and post-test counseling, encourages people to change their behavior in order to prevent HIV transmission.
"What we have for the first time is unequivocal data using the strongest possible design that counseling and testing is wanted by people, attracts high-risk people, and results in risk reduction," said Dr. Thomas Coates, director of the Center for AIDS Prevention Studies (CAPS) at the University of California at San Francisco, which was the coordinating center for the study.
Client-centered Counseling
A neighbor told Simon about the counseling center at Kariobangi, a low-income neighborhood outside of Nairobi, where he could get a free HIV test. During his first visit, he learned about the study and agreed to participate.
Assigned through random assignment procedures to the counseling and testing arm of the study, Simon met with Wathome for a pretest counseling session and received 25 condoms, then had blood drawn for an HIV test. He would return for the results in two weeks.
Others who were assigned at random to the health information arm of the study were offered the option of receiving counseling and being tested for HIV six months after their enrollment. In the meantime, each member of this group attended a health information session that included a video about HIV/AIDS prevention and a group discussion led by a health educator. At the end of these sessions, the health educator gave each person 25 condoms and invited everyone to come back for more at any time.
Dr. Claudes Kamenga, FHI's technical monitor for the study, explained that the comparison group was given health information for ethical reasons. "These people were concerned about HIV, so we had to make sure we were not denying them crucial information," he said.
A separate video was produced for each site, using local languages, actors and settings. These videos were shown during the health information sessions in an effort to standardize the content and to ensure that the two groups received distinctly different services.
The line between health information and counseling can be a fine one, explained Dr. Kamenga, a technical officer in FHI's HIV/AIDS Prevention and Care Department. "If it's given one-on-one, health information can become counseling, depending on the background of the person giving the health information."
In fact, many health information sessions were conducted with one individual or couple because study participants often came to the counseling centers at different times.
What distinguished the counseling sessions from these one-on-one discussions, however, was the counselors' use of a culturally appropriate, client-centered counseling model. Counselors worked with each client to assess his or her own risk of acquiring or transmitting HIV and to develop a personal risk reduction plan based on the client's level of knowledge, relationships with partners, practice of HIV-risk behaviors and readiness to change.
In developing risk reduction plans, clients chose the method of prevention they believed would be most effective for them.
"A client may say, 'I have more than five sexual partners,'" Wathome explained. "After counseling, and exploring how high risk he is with more than five sexual partners, he may decide himself, 'I'm going to reduce my five sexual partners to one.' Another client may decide, 'With my five sexual partners, I will no longer have sex without condoms.' It is the client himself who makes that decision."
Clients also decided how many counseling visits they needed. Most returned for only one post-test counseling session, but others visited the counselors more often. Those returning for more than one visit usually had tested positive.
Clients at Risk
Simon was one of 4,293 people who agreed to participate in the study at the three sites. Roughly equal numbers of men and women participated, and 27 percent enrolled as couples.
All of the sites reported some difficulty in recruiting people to participate in the study as couples. In Trinidad, only 107 couples were recruited. Many people who enrolled in the study as individuals, however, later brought partners to the center for counseling and testing.
"I think that people were testing the waters for the quality of the service," explained Dr. Colin Furlonge, principal investigator of the study in Trinidad. "If one partner felt it was good, and felt safe and comfortable, then we had many who would bring their partners along."
Overall demand for counseling and testing proved high even after the study was over. "The study sites had to continue providing counseling and testing, given that clients continued to show up at the sites beyond the recruitment period," Dr. Kamenga said.
The counseling centers used various methods to recruit participants, from television in Trinidad to word of mouth in Kenya, but all attracted people at high risk of HIV infection. HIV prevalence rates among those randomly assigned to the counseling and testing group, ranging from 21 percent in Kenya and Tanzania to 4 percent in Trinidad, were higher than those among the general population in each of the three communities.
Confidentiality and Trust
Most participants said they had enrolled in the study because they wanted to know their HIV status, and 82 percent of those tested returned for their results. The same percentage of all the participants returned at six months.
Although free HIV testing and STI treatment attracted participants to the study, they were not necessarily what motivated clients to return, according to Francis Kihuho, counseling director at the KAPC site. "People came for the testing at first, but they really valued the counseling," he said.
Researchers believe the intervention succeeded in changing behavior because of the quality of the counseling. Quality assurance measures included ensuring sufficient space to guarantee privacy to counselors and clients, careful selection and training of counselors, rigorous monitoring and evaluation of the service, supportive supervision of staff and early identification of appropriate referral systems.
"Unfortunately, in most other places, due to limited resources, all of these important aspects of HIV voluntary counseling and testing are not addressed," Dr. Kamenga said. "The results are poor quality of service, low retention of trained counselors, burnout among retained counselors, and lack of confidence in the service by the clients and the community."
Confidentiality was critical in building that confidence among study participants, according to the findings of a qualitative interview study conducted in Kenya and Tanzania as part of the counseling and testing trial. It was also what set the counseling apart from the kind of counseling participants were used to receiving from elders or health care workers.
"At every stage, staff members took time to show clients how they ensured confidentiality," Kihuho noted. Each participant was assigned a unique number, which was used to identify all samples, test requests and interview forms. The files linking names and identification numbers were always kept in locked file cabinets, and staff members were trained in confidentiality procedures.
Counselors were able to overcome clients' initial reluctance to trust the promise of confidentiality. "Because their personal information was not released into the community, clients gained confidence in the counselor and in the counseling center," said Dr. Olga Grinstead of CAPS. "This facilitated disclosure of risk behavior, which in turn facilitated the effectiveness of counseling for risk reduction."
Changing Behavior
Simon did not change his behavior overnight. After learning that he was HIV-positive, he decided to be monogamous and to use condoms. But at first his wife refused to use condoms because they had only one child and she wanted to have more children. It was months before Simon could muster the courage to ask his wife to go to the counseling center with him.
After Simon's wife also tested HIV-positive, they talked with counselor Wathome about how they could reduce their risk of reinfection. "Now I always use condoms with my wife," Simon said.
Like Simon, members of both the counseling and testing and the health information groups reported changing their behavior to reduce HIV risk. Counseling and testing, however, produced more than a 50 percent greater reduction in risk behavior.
Counseling and testing and health information led to similar declines in unprotected sex with a spouse or other "primary" partner. But those who received counseling and testing were significantly more likely to report decreases in unprotected sex with any non-primary partner and with partners in commercial sex.
Counselors cite less scientific but still powerful anecdotal evidence of behavior change. Counselor Rose Kairuthi of the KAPC in Kariobangi, Nairobi, remembers one client at high risk of infection when he enrolled in the study, a young man who said he had had 20 sex partners.
"When he came negative, he cried," she said. "When he came back after six months, he had reduced the number of sexual partners and was still negative. At twelve months--still negative. He went around recruiting other young men to put in the study."
Behavior change was assessed by comparing the behaviors reported by participants during interviews at enrollment and at six months. STI tests on urine samples collected at baseline and at six months confirmed the validity of this self-reported behavior. Those who reported unprotected sex with a non-primary partner were twice as likely to have a new case of STI as those who did not.
Researchers are analyzing the data collected at twelve months. Since all study participants were offered counseling and testing at six months and most chose to be tested, Dr. Kamenga noted, "the information 12 months can provide is more in terms of sustainability of the behavior change accomplished by counseling and testing."
Counseling Couples
Many of the couples enrolled in the study made some changes in their sex lives. Couples from both the counseling and testing and the health information groups reported less unprotected intercourse with each other and more abstinence from sex.
Among couples, as among individuals, greater change was seen in the counseling and testing group. The difference between the two groups of couples, however, was not statistically significant.
FHI's Dr. Kamenga believes that this finding may not be as contradictory as it seems. "If you come as a couple to a health information session, that may be a starting point for you as a couple to take some action," he explained. "Whereas if you are a single person watching a video and learning how to use condoms, it's still a challenge to initiate a discussion with a partner. So this may be one of the possible explanations why, among couples, we didn't notice a significant difference between the two study arms."
Dr. Grinstead also points out that the couples had a lower baseline rate of risk behavior than the people enrolled as individuals. "They had less change to make," she said.
Counselors, researchers and many study participants believe that counseling couples is even more effective than counseling individuals. Couples' counseling was particularly effective in reducing unprotected sex when one or both partners had tested positive for HIV.
"A discordant couple may be more likely to use condoms than a couple that is concordantly HIV-negative or HIV-positive, so it's important to provide couples with that information," Dr. Kamenga said. "And it's even more crucial for couples because there are so many decisions to be made about children and whether to have children."
Positive and Negative Experiences
After Simon learned that he was HIV-positive, he told his boss--a Catholic priest--in confidence. But the priest revealed his secret to all Simon's coworkers, who shunned and ridiculed him. When he could bear it no longer, Simon quit his job.
The study results suggest that Simon's experience was not typical of most participants. When asked at six months, few said they had been estranged by peers, discriminated against by employers or neglected by family members.
On the other hand, many participants reported improvements in their lives, such as strengthening of sexual relationships and increased emotional support from family, peers, employers and health professionals. Participants enrolled as couples were more likely to say that their relationships had grown stronger.
"In general, the bad experiences tend to make the news," Dr. Furlonge noted. "But one thing the study does show is that there are far more positive experiences than negative experiences."
In all three sites, participants randomly assigned to the counseling and testing group were no more likely to report negative experiences than those who had been assigned to the health information group. "Counseling and testing didn't seem to be associated with negative effects," Dr. Kamenga said.
Dr. Gloria Sangiwa, principal investigator at the Muhimbili center in Tanzania, attributes the low frequency of adverse effects among those in the counseling and testing group to the quality of the counseling and the strict confidentiality the study maintained. "Confidentiality enabled clients to control negative life events," she said.
Other researchers agree that lack of disclosure of HIV status may have helped reduce the risk of negative experiences. They are analyzing study data to determine the impact of disclosure on participants' lives.
Negative experiences were more common among those who tested positive for HIV. Of those in the counseling and testing group, HIV-positive people were more likely to report estrangement from peers, discrimination by employers and neglect by families.
HIV-positive women enrolled as couples were five times more likely than those in the counseling and testing group as a whole to say that their marriages had broken up and more than three times more likely to have experienced physical abuse. HIV-positive women with HIV-negative spouses were at greatest risk of adverse effects, with 23 percent reporting physical abuse and 18 percent reporting marital breakup.
Dr. Sangiwa has vivid memories of the first women who told counselors they were experiencing these problems. She and her staff quickly identified an organization, the Tanzania Media Women's Association, where they could refer the two women and others in need of more specialized counseling and legal assistance.
In addition to providing such support for clients, either directly or through referrals, Dr. Sangiwa recommends alerting counseling staff to the potential for abuse and abandonment among women who may be too ashamed to broach these subjects. "It's important that counselors are aware so they can look for this and talk to people about it," she said.
A Valued Service
As participants in the Voluntary HIV Counseling and Testing Study, Simon and his wife received free HIV tests, counseling and STI treatment. But the time they spent receiving these services was still costly for a family with a small income.
Clients in Kenya and Tanzania reported that their participation in the study cost them an average of $U.S.2.57 to $7.75. Lost wages were the main expense in both countries, with clients at both sites giving up a day's wages to travel to and from the site, receive services and participate in the interviews.
Despite these costs, clients said they would be willing to pay an average of $1.64 in Kenya and $5.11 in Tanzania for counseling and testing.
Clients overestimated how much people would actually pay, according to Dr. Michael Sweat, assistant professor of international health at the Johns Hopkins School of Public Health, who analyzed the cost-effectiveness of the services in Kenya and Tanzania. "It seems that many people are inclined to overestimate their willingness to pay as a sign of how much they appreciate the services they received," he said.
The number of clients at the Muhimbili center in Tanzania plummeted when a fee of $3 was instituted at the end of the study. The fee was reduced to $1.50, and the client load rebounded. In Kenya, where clients were charged $0.35 for counseling and testing, the fee had little effect on client access.
Cost-effective Prevention
Counseling and testing helped Simon and thousands of other people enrolled in the study, but was its public health benefit worth the cost in countries with such limited resources? Results of the cost-effectiveness analysis suggest that the answer to this central research question is yes.
The cost of providing counseling and testing to each client--U.S.$27 in Kenya and $29 in Tanzania--was relatively high for countries where per capita health expenditures do not exceed $10. Using a probability-based model of the likely number of HIV infections that did not occur as a result of the behavior change reported by participants, however, researchers estimated that the cost of averting each new case of HIV was only $241 in Kenya and $303 in Tanzania.
These costs are comparable to the estimated $217 per infection averted through enhanced treatment of sexually transmitted infections in the first randomized controlled trial of that HIV prevention intervention, which was conducted in the Mwanza region of Tanzania.
Voluntary counseling and testing is much less cost-effective in areas with low HIV prevalence. In the United States, for example, one study estimated that the cost per infection averted in a setting with at least 1 percent prevalence was $60,000.
In the study sites, voluntary counseling and testing proved most cost-effective for those who were infected with HIV. It was also more cost-effective for couples than for individuals and for women than for men. These findings reflect differences in initial risk of HIV infection among members of these groups and in the numbers of high-risk partners they had, as well as the reductions in risk behavior that occurred after counseling and testing.
Dr. Sweat recommends that counseling and testing services target populations with high HIV prevalence by working with groups that serve those populations, such as STI clinics and camps for transport workers. This was one of the key recommendations in his report to the U.S. Agency for International Development (USAID) on the sustainability of the Kenyan counseling center.
Building a Bridge
Now that the study has ended, clients continue to receive voluntary HIV counseling and testing at the three centers. The USAID missions in Kenya and Tanzania are supporting the services in Nairobi and Dar es Salaam, and UNAIDS funds them at the Voluntary Counseling and Testing Centre, which was established for the study in Port of Spain, Trinidad.
The study results seem to have sparked greater interest in HIV counseling and testing. A new pilot project in Zimbabwe, sponsored by the USAID mission in Harare and implemented by Population Services International and FHI, will assess a social marketing approach to counseling and testing.
Dr. Sangiwa noted that the Tanzanian national AIDS control program included recommendations for district-level counseling and testing in its latest medium-term plan for HIV/AIDS control. "Implementation will be something else, but in terms of policy, we have had an impact," she said.
Despite growing recognition of the value of counseling and testing, future support for continuing and expanding such services remains uncertain. UNAIDS Director Peter Piot believes that policymakers and program managers have a responsibility to act on the study findings.
"We have enough data to indicate that voluntary counseling and testing is an essential part of the national response to AIDS," Dr. Piot said. "As the bridge between prevention and care, the investment in counseling and testing services will go a long way in making needed care available to the invisible majority which today does not even know its HIV status."
-- Kathleen Henry
FHI-UNAIDS Voluntary HIV Counseling and Testing Study
Funded by
USAID, FHI/AIDSCAP, the World Health Organization's Global Programme on AIDS, UNAIDS and the U.S. National Institute of Mental Health
Coordinating Center
The Center for AIDS Prevention Studies,
University of California, San Francisco
Study Centers
Kenya Association of Professional Counselors
Nairobi, Kenya
Muhimbili University College of the Health Sciences
Dar es Salaam, Tanzania
Queens Park Counseling Centre
Port of Spain, Trinidad
