Scale-up of Voluntary HIV Counseling and Testing in Kenya

Overburdened health care systems in resource-poor countries are unable to cope with the impact of human immunodeficiency virus (HIV) and the increased responses to treatment that have been proposed internationally. There is an urgent need for robust, replicable approaches that meet the need for rapid expansion of HIV testing services. Provision and utilization of voluntary HIV counseling and testing (VCT) services in Kenya expanded rapidly between 2000 and 2005. Lessons learned from increasing access to VCT are applicable to global efforts to increase access to HIV prevention, care, and antiretroviral therapy (ART).

In 2000, Kenya had a population of approximately 30 million, with an estimated HIV prevalence of 9% in adults and an estimated 110 000 deaths from HIV/AIDS annually.^{1 - 2} Earlier HIV prevention campaigns had focused on AIDS awareness, behavioral change, and condom use.¹ Knowledge of HIV status was rarely mentioned¹ and only an estimated 14% of the adult population had an HIV test in any context,³ such as clinical care, mandatory testing for preemployment, blood donation, or research. Human immunodeficiency virus testing was largely conducted in hospital laboratories; in some cases, patients were not told their test results.^{4 - 5}

Voluntary HIV counseling and testing is an intervention initiated by persons who wish to ascertain their HIV status and is primarily intended for asymptomatic adults. Traditionally, VCT has been provided in health care facilities or in nonmedical settings, such as community sites and stand-alone VCT centers in nonclinical settings.⁶ Research had shown that VCT was effective in reducing HIV risk behavior in Kenya,⁷ and pilot projects indicated that it was both acceptable to Kenyans and feasible to implement in public health care settings.^{5 ,8} The government of Kenya made a commitment in 2000 to the rapid extension of VCT services, declaring that it would open 350 sites by the end of 2004, 5 in each of the 70 districts in Kenya.⁹ Donors and international organizations pledged support for increasing VCT in both public sector and community sites, based on the concept that knowledge of HIV status is critical in a generalized epidemic for prevention and access to care.¹⁰

Development of national guidelines to regulate VCT was an essential component of the approach by Kenya.⁹ A committee was formed with representatives from the Ministry of Health's National AIDS and STD (sexually transmitted disease) Control Program (NASCOP), counselor and laboratory organizations, donors and international organizations, persons living with HIV, and other stakeholders. Guiding principles articulated by the committee were that VCT services should be private, confidential, accessible, affordable (free when possible), and convenient.¹⁰

The committee membership included VCT counselors who contributed practical information derived from their experience. After wide debate, consensus was achieved on issues such as supervision, optimum number of persons to receive VCT daily, minimum age for consent for VCT, and ethical standards. The guidelines recommended anonymous testing, although name-based confidential testing is permitted, particularly in health care facilities.¹⁰ The guidelines discourage sites from providing written results because certificates with HIV-negative test results could be abused. Counselors are trained to provide education about condom use; the guidelines encourage but do not require the provision of free condoms to all persons receiving VCT.

Dissemination of the guidelines throughout Kenya meant that all VCT sites were, and still are, operating under the same standards. A national data form was developed, and registered sites complete a form for each individual; however, some sites are not able to enter these data on computers but maintain onsite records and log books. The committee oversaw the development of a standard training curriculum and counseling protocols. Minimum standards were set for site registration, a requirement for receiving free test kits from NASCOP. The registration system formed the basis of a comprehensive quality assurance program, which included supervision of both counseling and testing, continuing education, laboratory-based testing using dried blood spots, and inspection of sites.¹¹ In late 2003, a National Quality Assurance Team was established, including technical and supervisory staff who visit all sites annually.

The issues generating the most debate were the testing protocol and who should conduct the counseling and testing. Simple, whole-blood rapid tests were recommended, which allowed for testing to be performed immediately, reducing waiting time.¹² Using a fingerprick sample was acceptable to a public accustomed to undergoing fingerprick to obtain blood for malaria testing, and use of fingerprick samples reduces medical waste and occupational exposure risk. During the policy development phase, 9 pilot sites collected additional venous samples for all persons receiving VCT for external quality assurance and showed 100% concordance in results.¹³ An initial decision was made to recommend parallel testing (2 different enzyme-linked immunosorbent assay [ELISA]−based rapid tests used simultaneously). Serial testing (1 rapid ELISA screening test with a second, different rapid ELISA confirmatory test, if the first test is positive) is permitted and has over time become the dominant protocol.

The second area of debate centered on whether health care workers (primarily nurses) or counselors should provide the counseling and testing. Some committee members argued that it was wrong to divert health care workers from clinical duties; others argued that medical diagnoses should remain the prerogative of clinicians. The guidelines permit both approaches. Currently, VCT services delivered in health care facilities are primarily provided by health care workers, and those services in stand-alone or community sites are primarily provided by lay counselors, professional counselors, or both.

Early experience revealed an unexpected benefit of fingerprick tests. Counselors reported that persons receiving VCT appreciated seeing their own test strips and being engaged in the interpretation of the test results (using diagrams). This approach, which allows the counseling and testing to be delivered in 1 session, was preferred by persons receiving VCT,¹⁴ enhanced confidence in the test results, reduced waiting time, and has virtually eliminated loss to follow-up. Persons receiving VCT are offered follow-up and approximately 12% of sessions are return visits (E.M., unpublished data, January 2006). Ongoing external quality assurance using dried blood spots, with testing conducted in central laboratories, has documented that in-room testing conducted by counselors has a high rate of concordance (range, 97%-100%) with laboratory-based testing.^{15 - 16} Testing conducted in front of the individual receiving VCT virtually eliminates the potential for clerical errors, such as inaccurate recording.¹⁷

Services in Kenya were delivered through a variety of models with stand-alone sites accounting for 25% of the total.¹⁸ The use of trained counselors who were not health care workers allowed for extension of VCT to nonmedical sites, and numerous organizations with mandates in areas such as HIV education and home-based care became involved. By September 2005, a total of 41 community-based groups and 17 faith-based organizations operated 125 registered sites.¹⁸ Kenya has also pioneered a large-scale mobile VCT program to meet the demand among remote and rural populations. Regardless of the site type, all services are under the jurisdiction of the National Quality Assurance System, and all counselors are trained using the same curriculum. When testing is conducted in health care facilities, referral for care of those persons who are HIV positive is usually to services within the facility; in the case of stand-alone, community-based, or mobile VCT, counselors refer persons who are HIV positive to the nearest site delivering care. Due to anonymity of services, no data are available on rates of referral uptake. An increasing number of VCT sites are evolving into comprehensive care centers in which follow-up care is provided.

Information from selected stand-alone sites in Kenya suggests that the cost for VCT services is approximately $15 per individual, including rent, salaries, and quality assurance. For community-based groups that integrate VCT into their other services and often use volunteer staff, the cost is approximately $10 per individual. In health care facilities, costs range from approximately $11 to $13.¹⁹ The VCT program has been supported by the government of Kenya and multiple donors. Approximately $35 million was devoted to the scale-up, including the direct costs of providing VCT services and indirect costs, such as the promotion of VCT, training of health care workers and counselors, and technical assistance provided by both national and international organizations. Although official, publicly verified information is not available, the estimated annual expenditures in 2000 were approximately $320 000; for 2001, $1.2 million; for 2002, $3.8 million; for 2003, $4.5 million; and increasing to approximately $10 million in 2004 and $15 million in 2005 (E.M., unpublished data, January 2006).

A mass media campaign to address fears, misconceptions, and stigma surrounding testing was conducted in 2002. Registered sites and promotional materials were linked through the same, easily recognizable logo. A youth-focused campaign was broadcast in 2003; campaigns in 2004 and 2005 concentrated on the benefits of couples getting tested together.

In 2000, 3 sites provided VCT services in Kenya. Since then, new sites opened at an increasing rate from an average of 10 new sites per quarter in 2001 to 75 new sites per quarter in 2005, reaching 680 VCT sites by the end of 2005, with 75% of sites being in health care facilities.¹⁸ Based on data submitted to the Ministry of Health, approximately 1100 persons received VCT in 2000, 23 000 in 2001, 110 000 in 2002, 200 000 in 2003, 400 000 in 2004, and 545 000 in 2005 (National AIDS and STD Control Programme, unpublished data, January 2006). The national target of 350 sites has been exceeded, although the target of 5 sites has not been achieved in the districts in remote or hard-to-reach areas. Kenya's 3 largest cities, Nairobi, Mombasa, and Kisumu, have more than 5 registered sites per 100 000 population, and there are few towns in Kenya without at least 1 site available.

Despite these increases in availability and use of VCT services, access to testing in rural areas remains limited, and 75% of persons receiving VCT nationwide have been served in urban and periurban sites. When service use is analyzed as the average number of persons receiving VCT per site type per year, stand-alone sites (mostly in urban areas) serve 3 times as many persons as VCT sites in health care facilities, which have far fewer staff dedicated to VCT (3.5 full-time equivalent staff vs 1.0). However, these sites continue to provide an efficient and sustainable service, which increases coverage and normalizes the concept of HIV testing as part of health care.

Records from more than 370 000 persons receiving VCT indicate that overall HIV prevalence between 2001 and 2005 was 11% in men and 23% in women (HIV prevalence in persons receiving VCT is based on the percentages observed at sites managed by the Centers for Disease Control and Prevention−funded partners and Liverpool VCT and Care, which account for >30% of the national total [E.M., unpublished data, January 2006]), more than twice the prevalence in the general adult population (approximately 5% in men and 9% in women).²⁰ The increasing availability of ART may encourage persons with symptoms to get tested; however, the majority of persons who receive VCT continue to express social or personal reasons, such as “just to know my status” and “I want to plan my future.”²

Despite funding for test kits and logistic systems, national systems for procurement and distribution of test kits have not kept pace with the scale-up. An interruption in test kit supply during the last quarter of 2003 brought the new mobile VCT program to a virtual standstill and resulted in a reduction in numbers of persons receiving VCT at established sites, despite emergency procurements. There are divisions within the laboratory community about whether counselors should perform simple rapid tests. Participation in the National Quality Assurance System is neither universal nor consistent. Responding to the popularity of VCT, some private practitioners and community groups have opened nonregistered sites, often using a handmade version of the national VCT logo. Although it is the prerogative of the Ministry of Health to close such sites, this has rarely been done. Access to VCT in rural areas is still limited, although a program to increase mobile VCT has expanded since 2004 and now serves approximately 4500 persons monthly (E.M., unpublished data, January 2006).

International efforts to increase access to HIV treatment depend on the capacity of local programs to identify persons eligible for treatment. Despite the expansion of individual-initiated VCT services, further expansion of diagnostic testing is needed to achieve global treatment targets, such as the World Health Organization “3 by 5” Initiative to treat 3 million people by 2005.²¹ Voluntary HIV counseling and testing sites, which typically serve the “worried well,” are not the ideal settings in which to identify large numbers of persons with advanced HIV infection; health care worker−initiated testing of hospitalized patients and persons suspected of having tuberculosis is more appropriate for this purpose. Regardless of the testing venue, 2 policy elements of Kenya's successful approach, if replicated elsewhere, may help increase access to testing: (1) the use of counselors who are not clinical health care workers, and (2) allowing these counselors to conduct in-room testing.

Kenya's guidelines have set high standards while allowing VCT in a variety of sites. This has allowed flexible, locally designed models to expand, including services for previously neglected groups, such as the deaf, refugees, prisoners, and nomads. Data collected at VCT sites have informed promotional campaigns and highlighted the need to extend services to youth, couples, and rural areas. Since 2004, there has been an increase in testing of pregnant women to prevent mother-to-child transmission in Kenya, and VCT sites now attract more men than women.² Innovative programming responses to attract women to VCT are needed, given the higher rates of HIV prevalence in women.^{2 ,22}

Lessons from the VCT scale-up in Kenya may be applicable to VCT scale-up elsewhere, as well as scale-up of diagnostic testing and provision of ART in Kenya and in other high HIV burden countries that have not yet met the goals of the World Health Organization “3 by 5” Initiative.²¹ Supportive government-coordinated policies informed by pilot projects provided an essential backdrop to the scale-up of VCT. Clear national ART policies are needed; articulation of guiding principles, such as equity of access, is important to guide policies and practice.

Flexibility regarding personnel, protocols, and models of service delivery, responsiveness to preferences of persons receiving VCT, continual data review, mass media campaigns, government support, and generous donor funding have all contributed to the rapid scale-up of VCT in Kenya, and are all applicable to ART scale-up. The current inadequate numbers of health care workers trained in ART delivery can be partially addressed by creative use of auxiliary workers for nonclinical duties, such as simple bedside testing, adherence counseling, and family outreach. Mass media campaigns promoting ART services may serve to increase both provision and utilization of ART-related care. Quality assurance systems, including registration of approved ART sites and supervision methods, are urgently needed. Careful attention to the procurement and distribution of antiretroviral drugs is essential to prevent stock-outs, such as occurred with test kits in Kenya. Perhaps most important, responsiveness to patients' needs and preferences is essential and may lead to innovations through engaging patients as active participants in their own care.

Rapid scale-up of VCT services was achieved in Kenya between 2000 and 2005. Despite substantial progress, an estimated 75% of Kenyan adults still do not know their HIV status, so continued increases in sites and numbers served are needed. There is increasing pressure to make HIV testing routine in health care facilities and increasing emphasis on universal knowledge of serostatus.^{23 - 24} Similar approaches may result in the rapid scale-up of both individual-initiated and health care worker−initiated testing elsewhere. Experience from increasing VCT also provides lessons applicable for the expansion of ART-related care; in particular, the need for quality assurance systems, supportive policies, and innovative practices responsive to patients' needs.