KENYA

Monitoting and Evaluation of HIV/AIDS programs

Godfrey Balthazar, Bilha Hagembe

Background

The majority of Kenya's estimated 28 million people live in rural areas and close to 50 percent are under15 years of age. The GNP per capita in Kenya is around $280.

HIV was likely introduced in Kenya in the late 1970s or early 1980s, the first AIDS case was officially noted in 1984, and current estimates of national HIV prevalence are 9% of the adult population 15-49, with considerable regional variation.

National response

Government response in the early years was inadequate, partly because of the perceived threat to the tourist economy. With awareness that HIV had moved into the heterosexual population, small efforts through a national AIDS committee attempted to addre ss prevention and control issues. In 1985, an AIDS Programme Secretariat was established with technical and financial support from WHO-GPA, and renamed the National AIDS Control Programme in 1988. In 1987, the first medium-term plan was developed in which AIDS became notifiable and all hospital cases were reported. The STD control programme was begun in 1988 and integrated under the national programme in 1995; a TB and Leprosy control programme was also integrated in 1997. Both programmes operate indepen dent of the national programme.

Thus far, two medium-terms and a strategic plan have been formulated, focusing on public awareness campaigns, strengthening laboratory services and establishing an HIV/AIDS surveillance system. The UNAIDS Theme Group was formulated in 1996.

Funding for the second medium-term plan was budgeted for 22 million over two years, but did not materialize. The new Sexually Transmitted Infections (STI) project has drawn donor support of approximately US $65 million from various sources.

Monitoring and evaluation was allocated US $1.2 million under the national programme, which was an increase from previous years.

District committees were established in 1995 with the launching of a programme targeting sexually transmitted infections (STI), although no monitoring and evaluation was addresses though district-level projects. Because the STI project commanded such g reat donor resources, it has somewhat eclipsed the national programme.

AIDS care is limited, with patient support centers established at around 15% of government health facilities. No legislation exists on human rights and discrimination, although the government has approved policy papers. A major concern in Kenya is the numerous NGO and community-based organizations through which many HIV/AIDS control activities are carried out. The national programme does little to monitor the NGO activities.

Monitoring and evaluation

There is no specific unit for M&E within the national programme. The epidemiology and research unit only monitors the HIV/AIDS trends. The national AIDS programme incorporated monitoring and evaluation in the second mid-term plan and today includes data from hospital records, HIV sentinel surveillance in antenatal clinic sites, HIV surveillance among STD patients and notification of AIDS cases.

From 1990 HIV prevalence from the national HIV sentinel surveillance system was conducted in 13 urban antenatal clinics around the country. In 1994, 6 peri-urban and rural sites were added. Additional research and intervention studies provide some inf ormation for monitoring and evaluation purposes. These include a randomized control trial to test the impact of HIV voluntary counseling and testing was conducted in 1997 and cohort studies among high and low risk groups in Nairobi.

Numerous studies have been conducted on knowledge and behaviour in regards to HIV/AIDS, although the national programme does not centrally collect information. The Kenya DHS in 1989 (knowledge only), 1993 and 1998, a national prevention indicator surve y in 1989/90 and numerous smaller studies conducted at STD and other health clinics provide information on knowledge, condom use and sexual behaviour. In association with the World Bank loan the national programme was requested to develop a set of benchma rk indicators, although use of these indicators has been piecemeal.

Input, output and context

Programme input data are available for government and some donor programmes through national programme and project evaluation reports from 1990 to the present. Other input information is available through the many registered and non-registered NGOs, b ut information is not collected regularly nor is easily accessible. National programme evaluations have been carried out in conjunction with the end of each mid-term plan and with specific donor projects.

Condom distribution data from the MOH and donors are available although there is no national tracking system. In 1997, condom distribution reached a peak at 73 million, including 8 million by the social marketing programme and 33 million condoms distri buted through family planning clinics.

The national programme does not have data on the number of trained health personnel, partly because of the great number of agencies involved.

AIDS knowledge and attitudes have assessed through many studies, e.g. school education studies a UNICEF schoolgirls study, and the DHS surveys. There are national blood safety guidelines at the 78 screening centers. Blood transfusion safety is monito red and is thought to be monitored effectively at the 78 screening centers. There is no quality control system in place but an in-depth study in 1997 showed that a significant number of blood units were infected, even after going through testing procedure s.

Proximate determinants

Data on non-regular partners are available from a GPA KABP Behavioural Study (1992); data on condom use with last act with non-regular partner are available from a workplace study and the DHS surveys. In the DHS 1993 32% of men reported more than one s exual partner in the last 6 months. Condom use with a non-regular partner in the last 6 months did not increase between 1993 and 1998, according to the DHS surveys (20 and 21% of men respectively).

HIV/AIDS and STD

Reporting by HIV sentinel sites has been fairly good with 16 of the 19 sites providing reports at least 3 times in the last 4 years. There is however no quality control system. HIV trends seem to indicate that HIV prevalence was still rising in 1997, e specially in the peri-urban sites. STD morbidity data are available from the regular health information system, but underreporting and misclassification are common. In-depth studies in urban areas have shown high prevalence of most STDs, while some anten atal clinic sentinel sites provide data on syphilis (positive RPR test), showing about 2-6% prevalence in 7 sites in 1997.

AIDS impact mitigation

A cumulative total of 82,000 AIDS cases had been reported to NASCOP by 1998 and it is believed that AIDS cases are heavily underreported. Few studies have addressed stigma and discrimination toward people living with HIV/AIDS and there is no mechanism by which to measure impact. The national programme has established patient support centers at 15% of government health facilities.

Conclusion

The initial response to the epidemic was slow, but local HIV trend data and the AIDS impact (AIM) simulations have helped break the silence. Until the recent loan was assumed by Kenya, external donors and NGOs funded most of Kenya's AIDS control effor ts which led to fragmentation of M&E. The national programme itself has only limited control over M&E of the epidemic and interventions, with the exception of HIV monitoring. No clear benefits can be seen from the integration of HIV/AIDS, STD and TB programmes, as all operate with separate programme, separate donors and separate M&E.