University of Zambia Medical Library

Epidemiology: Overview Reports (Aids in Zambia Bibliography #1-4)

(#1) "Current state of HIV/AIDS Epidemic In Zambia and How Research Findings Could Aid Policy on Health Interventions and HIV Prevention."
Msiska, R. (1993) Unpublished paper
Geographical area: National; Keywords: Surveillance, policy interventions; Location: NASTLP

A commentary paper which provides a profile on a situation analysis of the HIV/AIDS epidemic in Zambia and considers measures of interventions for HIV prevention. The information is based on a literature review and hospital data for the period from l989 to 1992. Explorative evidence reveals two possible phases of HIV epidemic behaviour in Zambia - the silent epidemic with no symptoms, estimated to have started in mid to late 1970s, and the visible epidemic with signs and symptoms of AIDS mortality.

The first AIDS case was diagnosed in Zambia in 1985, followed by the wave of the epidemic related to social and economic consequences of AIDS. As sources of information are hospital notification systems and sentinel surveillance, the use of information on the actual epidemiological situation in the country is still problematic due to the high incidence of the under-reporting of cases. The main factors have been poor management information systems and inadequate equipment for diagnostic purposes and accessibility. However, the cumulative total number of notified persons with AIDS and AIDS-related complex (ARC) has increased to 27,901 in 1992. The observed cases emerge from the age-group 0-4 years through perinatal transmission, a high proportion of untreated STD in women as compared to males, the majority of reported cases of AIDS fall in age-group of 20-29 years, and the male to female ratio of 1:1 suggests a predominantly heterosexual transmission.

Highest HIV prevalence is in urban antenatal clinic attenders. These results suggest that the behaviour of the HIV epidemic is not uniform in Zambia. The HIV-seropositive blood units have recorded a significant increase in HIV seroprevalence in blood donors. The analysis still indicates that the main current and future determinants of the HIV/AIDS problem in Zambia are demographic, medical, economical, sexual, gendered, cultural, political and social. To alleviate some of the prevailing factors, there is a need for continued development of projections for various modelling methods to assist planners with the possible paths that HIV/AIDS will take under different underlying assumptions.

(#2) "The Socio-Economic Impact of AIDS: Zambia: The Current HIV/AIDS Situation and Future Demographic Impact."
Fylkesnes, K., Brunborg, H., Msiska, R. (1994) National AIDS/STD/TB/Leprosy Programme, Ministry of Health
Geographical Area: National; Keywords: Surveillance, orphans; Location: NASTLP, UNICEF, Family Health Trust

The aim of this report is describe the current HIV/AIDS situation in Zambia and provide an overview of the most likely short term impacts of the disease on mortality and the situation of children (orphans). Furthermore, the intention is to provide some scenarios indicating potential long term demographic changes.

Different sources of data have been reviewed or analysed for this report. Age-specific HIV seroprevalence estimates for rural versus urban residents have been provided from the sentinel surveillance system (SSS) on pregnant women attending antenatal clinics (age-group 15-39). Prevalence estimates for the adult (15+) population have been arrived at by complementing SSS data with data from blood donors for the older age-groups and standardising them to the given age distribution.

The data from the HIV sentinel surveillance system (SSS) indicate that Zambia is among the countries most seriously affected by the HIV/AIDS pandemic. The information on the geographical variation in HIV prevalence indicates a marked difference between the urban and rural areas. In the age-group 15-39, based on the 1992 HIV SSS on women attending antenatal clinics, the HIV prevalence is estimated to be 34 per cent in the urban and 13 per cent in the rural areas [editor's note- we now know these figures to be approximately 10 per cent overestimates due to the nature of the testing kits used]. The figures for the "high" estimate of the adult HIV prevalence in 1992 are 27 per cent and 10 per cent in urban and rural areas respectively. The respective figures for the "low" estimate are 21 per cent and 10 per cent. Accordingly, the total number of HIV infected adults in the whole country is between 600,000 and 700,000 or within the range of 14.4 and 16.5 per cent of the adult population.

Short term projections show that the number of annual AIDS deaths is likely to be doubled during a four year period from 40-50,000 in 1993 to 80-100,000 in 1997. About 30 per cent of all deaths are likely to be children infected by their mothers. The steep rise in adult mortality in specific age-groups (for instance, the majority of AIDS deaths in adult population will occur in the age-group 20-44 years) will have immense impact on the situation of children. From the assumed 70,000 AIDS orphans in 1993, by the end of the decade Zambia will have between 530,000 and 600,000 AIDS orphans. Long term demographic impact shows that assuming no change in the total fertility rate, the population growth rate in Zambia might drop from 3.5 per cent in 1990 to 2.6 per cent which is about 1.3 per cent in urban and 2.2 per cent in rural Zambia. The effect on the growth rate when assuming a decline in fertility is clearly more significant, and the results show that by the end of the period (in 2030) the growth rate could be below 1 per cent.

Recommendations are made on issues relevant to informational needs. The main emphasis is on the establishment of a core data collection system with the potential of providing timely and relevant information for decision making within the national programme.

(#3) "HIV in Zambia : Myth or Monster?"
Godfrey-Faussett, P., Baggaley, R., Scott, G., Sichone, M. (1994) Nature Vol 368 pp 183-8, 17-3-1994.
Geographical area: National; Key words: Surveillance, mortality, awareness; Location: UNZA Medical Library

[Editor's note: This paper was written in response to a series of articles in the British Sunday Times which questioned the existence of AIDS in Africa and its aetiology. Claims were that if AIDS actually exists, it was an old disease under a new name, and that HIV could not be its aetiological agent].

Ten years ago, Anne Bayley, working at the University Teaching Hospital in Lusaka, Zambia, noticed a change in the patients she was seeing in her Kaposi's sarcoma clinic. Not only did the number of cases begin to rise rapidly but the characteristics of the disease also changed. Previously patients with Kaposi's sarcoma had been older men with disease predominantly affecting the legs and responding satisfactorily to treatment. Now cases were also being seen in young women, who had widespread lesions on the trunk, in the lymph nodes, in the mouth and in the lungs. Treatment was much less satisfactory and patients were dying with generalised oedema. Over the next years, the number of atypical African Kaposi's sarcoma cases continued to increase and was shown to be associated with antibodies to HTLV-III (as HIV-1 was then known). Any physician or surgeon who was working through the 70's and 80's will confirm that a new pattern of disease had emerged. AIDS is not a syndrome that has been present for decades in Africa, unsuspected until clinicians in the USA drew attention to it.

In contrast to the initial studies in the USA, serological surveys in Zambia and elsewhere in Africa showed that HIV seropositivity was equally distributed between the sexes. Both men and women showed the highest seroprevalence in young adults, with the peak in women being at a younger age than that in men. Seropositivity was strongly associated with previous sexual exposure, both the number of reported partners and previous episodes of sexually transmitted diseases. Since then repeated surveys have demonstrated rising rates in the general population but seropositivity is still associated with the sexually active age group, their infants or those who have received blood transfusions.

Deaths in early adult life have become an accepted part of Lusakan life. In surveys carried out among 767 university students, 397 employees of companies and 500 of those attending voluntary HIV counselling and testing centres, 39 per cent, 23 per cent and 29 per cent of those interviewed had lost relatives to what they believed was AIDS and a further 12 per cent, 8.4 per cent and 11 per cent had a relative known to be HIV-positive.

Despite being a predominantly Christian society, traditional beliefs are still widely held and nowhere more so than beliefs about health. More than 75 per cent of people will consult a traditional healer before attending a clinic or hospital and anthropological studies have shown that deaths that might have been attributed to AIDS are often assumed to be caused by witchcraft. HIV and AIDS still carry a powerful social stigma, which leads to deaths being ascribed to pneumonia, tuberculosis or diarrhoea rather than HIV infection, and rather than cause distress in the families of the deceased, doctors collude with these attributions and few death certificates mention HIV or AIDS.

In many areas, forms for the notification of AIDS to the Ministry of Health are not available and resources are inadequate to perform HIV tests on most patients with AIDS-like illnesses. This combination of stigma and poor infrastructure has led to 7,761 cases of AIDS and a further 21,973 cases of AIDS-related complex being reported to the National AIDS Prevention and Control Programme by October 1993 and Dr. Roland Msiska, the head of that programme, is confidant that the numbers are hopelessly inaccurate.

Since the first tests were developed, Zambia has been one of the countries reporting alarming seroprevalences of anti-HIV antibodies. Sequential surveys among women attending antenatal clinics in a variety of rural and urban sites throughout the country have shown rising rates with 33 per cent (95 per cent CI 30 per cent - 36 per cent) of such women in the high density areas of Lusaka being infected in the most recent study (Ministry of Health figures 1992).

A major concern has been with the accuracy of tests for anti-HIV antibodies used in Africa. Initial observations were made using crude antigen preparations from HTLV-III but sera containing antibodies to malaria or immune complexes were also shown to react causing many false positive reactions on African samples. Subsequently, test systems have been produced that either use recombinant proteins or competitive inhibition and these have been shown to be much more specific. Even a specificity of 99.5 per cent is disastrous if a test is to be used for surveillance in a population with little disease, despite the fact that the risk of an individual being given the wrong result is 1 in 200. If the prevalence is 1:1000, there will be 5 false positives for each true positive recorded. On the other hand a test that only has a specificity of 95 per cent will not be useful as a diagnostic test since 1 in 20 people without disease will receive positive test results. However if the same test is used for surveillance in a population with a 25 per cent prevalence of disease, the measured seroprevalence will be 28.5 per cent (assuming a sensitivity of 99 per cent). Thus obsession with the specificity of HIV tests is highly appropriate in the context of low prevalence areas but matters less when the prevalence rises. The distinction between 25 per cent and 28.5 per cent destroys few myths.

The past decade has seen tuberculosis notification rates in Zambia climb from 120 to 310/100,000 population/ year (Ministry of Health figures 1992). In 1989, a cohort of patients was recruited and followed at the Chest Clinic in Lusaka. 73 per cent were HIV seropositive. This study and those from other African states showed major differences in the course of tuberculosis, the reaction to drugs and the outcome of treatment in those patients who happen to have a positive HIV test (whatever it may signify). Such individuals are up to six times as likely to die within a year of starting treatment, up to six times as likely to develop dangerous skin reactions to thiacetazone, (an anti-tuberculosis drug widely used in Africa), and up to thirty times more likely to develop recurrent disease after completing treatment as patients whose HIV test is negative. Nor is this because those with a positive HIV test were the most impoverished - on the contrary they were better educated and lived in less crowded conditions than those with negative HIV tests.

Predicting the number of deaths expected from observed seroprevalence data is speculation. Nationwide mortality statistics are difficult to interpret because of delays and incomplete records. One place that mortality figures are well recorded is in industry. Zambian employers traditionally offer considerable assistance to the families of employees who die. In addition to paying costs for the funeral arrangements, they usually provide transport for the coffin and mourners to the cemetery. Statistics are therefore necessary to keep track of these expenditures. A survey of 33 businesses in Lusaka and the Copperbelt (the other major urban area of the country) which included a workforce of over 10,000 has shown that from 1987 to 1993 mortality rose from 0.25 to 1.83 per cent per annum. Similarly a survey of nurses employed in two hospitals in the Southern province has shown mortality rising from 0.2 to 2.7 per cent per annum.

We remain convinced that there is a disaster unfolding in Africa. The most productive members of society are being destroyed. HIV antibodies predict those who will die but most deaths are in those who have not been tested and are unlikely to be recorded officially as HIV-related. The natural history of HIV infection in Africa remains obscure. What is clear is that a collective denial is occurring among Zambian society. Everybody "knows" that HIV causes AIDS and is passed between sexual partners. Everybody "knows" that reducing their number of partners and using condoms are the strategies to reduce their risk. Yet few people translate knowledge into behaviour. Even counsellors who are listening to family tragedies every working day fail to protect themselves from the risk of infection. It remains a Herculean task to promote behavioural change. There is considerable ambivalence about discussions of people's sexual lives and of condoms. Christian moralists argue that fidelity should render condoms irrelevant and condoms do not fit easily into Zambian culture.

The "AIDS-is-a-myth" myth flies in the face of every explanation of current events which could conceivably be true, given the easily accessible facts. And it is downright dangerous. The "awareness raising" that has been going on in Zambia, and in other African countries, for a decade is beginning to pay dividends in the form of safer sexual behaviour. It is irresponsible to offer people a reason to ignore the truth.

(#4) "Overview of the HIV/AIDS Situation in Zambia: Patterns and Trends."
Fylkesnes, K. (1995) NASTLP, Ministry of Health, paper presented at the 5th National AIDS Conference, Lusaka, May 1995
Geographical area: National; Key words: Surveillance; Location: NASTLP, Family Health Trust

During the last few years the National AIDS/STD/TB & Leprosy Programme (NASTLP) has managed to establish a comprehensive information system on HIV and STD covering at least two areas in each province. The main objective of this system is to provide information on the geographical patterns, socio-economic group differentials of HIV/STD, and to study trends of infection. The system might be seen as the "national watch" of the course of the epidemic. The main focus will be on recent findings indicative of changes in trends of infection. If favourable behavioural changes are taking place in various subgroups of the population, this is likely to be indicated in our data.

The main indicator group used to monitor patterns and trends of HIV infection in most countries are women being first attenders of antenatal care. Studies from other African countries have shown that this indicator group, when estimating HIV prevalence, is surprisingly representative of the general population 15-44 years of age. At the moment we are on the way to doing a population-based survey in order to study if this is the case in the particular Zambian setting. Especially since we are uncertain about the male to female ratio of infection, the data must be viewed with some caution.

It should be mentioned that the analyses of the 1994 data are still preliminary. The final analyses will include more detailed geographical information (urban/rural differentials etc.). This year was the first time we included urban areas other than Lusaka. About ll,000 persons represent the material in this particular study, apparently now being one of the most comprehensive HIV/STD surveillance systems in Africa. The regional pattern is one of a very high proportion of HIV seropositives in most provincial head quarters, especially in the Western, Southern, Central, Eastern provinces and Lusaka Urban, with the head quarters in Luapula and Northern provinces appearing a bit below and North-Western and Copperbelt at about half the level of "the five at the top". The rural areas are in general about half of the urban level, although with quite significant variation from area to area.

Repeated measurements from the same communities provide information on change over time. In this regard it is unfortunate that the establishment of the surveillance system came rather late, in l990, and that only a few areas were included from the beginning. After the release of our publication (No.2) we have been validating data collected in the previous years in order to feel confident about the accuracy. The opportunity of retesting of stored sera made it possible to correct errors from the 1992 surveillance. Thorough quality control is now included as an essential element of our information system. The good news regarding trends is that the prevalence in Lusaka Urban has been stable over the last 4 years, at a level of about 26 per cent (proportion being HIV+). There might be several possible explanations of this stability in prevalence. One possibility is that a high rate of new infections, the incidence, is balancing a high mortality, and thus it can not by itself be taken as an evidence of successful interventions. It might as well be possible, however, that it represents a sign of declining incidence (new infections). The trend in those rural areas we have been observing for several years is differing a lot from area to area and thus difficult to interpret at present.

The break down of prevalence by age-group is an additional indicator examining trends. In particular, observing what is happening in the youngest age-group, 15-19, where the majority of new infections are found. Unfortunately, this type of information is only available for 1993 and 1994. The results from urban areas indicate a decline in the age-group 15-19. This is a very exciting finding, indicating a possible decline in incidence. We need some more time, however, to have it confirmed.

Our third indicator of trends of infection is social class differentials, by studying differences in HIV infection by educational attainment and by age. This type of information was collected for the first time in 1994. The findings show a steep increase in HIV infection by educational attainment both in urban and rural areas. Based on what we learned from studies during the very beginning of the epidemic, that the risk of infection was increasing with increasing educational attainment, this finding was not surprising. The social class differences in transmission is expected to have changed in the last 5 years, however, mainly since the well educated are the first to receive information on protection and thus the first to change their behaviour. Although there is not a clear association between knowledge and action, this is what usually happens in any society, and is referred to as the Innovation Theory. The proper analysis in order to explore possible change is to look at the HIV/education association by age-group. The argument goes like this. In the older age-groups the majority of individuals were infected 4-10 years ago, and importantly, before having access to information on how to protect themselves. In the youngest age-group, however, the majority of seropositives represents new infections, and thus being exposed in a period of time where most people, at least the highest educational groups, have received knowledge on protection. If a significant change has started, according to the innovation theory, we should expect to find a less prominent difference in HIV infection between educational groups in the youngest age-group. This is what we find in urban areas, a rather promising indication of change in behaviour. Unfortunately, the same nice tendency is not appearing in rural areas, offering a kind of support to the assumption that change will first start in urban areas.

Summary:
The main HIV/AIDS belt is concentrated in some few countries in sub-Saharan Africa. Zambia is one of these few countries, and the very high HIV prevalence will have a dramatic and increasingly negative impact on all sectors over the next 5 years. The exciting good news presented is that the incidence or rate of new infections most probably is on a decline, at least in main urban areas. This might indicate that particular population groups have changed their behaviour favourably. We need some more time of observation to provide better scientific evidence.

[Table of Contents] [AIDS/Zambia Index] [Bibl. Contents] [Alphabetical Index] [Zamnet] [UNZA] [UNZA Library]

Last updated February 4, 1999

UNIVERSITY OF ZAMBIA Med Guide      Index    Zamnet    UNIVERSITY OF ZAMBIA    UNIVERSITY OF ZAMBIA Library 

CONTACTING SCHOOL OF MEDICINE     OTHER INSTITUTIONS   BACK TO SOM HOME PAGE     DEPARTMENTS    Central Board of Health 

Send comments and/or suggestions to: Stan chewe 
Copyright © 1996-2001, The University of Zambia Medical Library and Stan Chewe
Guide to Medical Resources WWW site: http://www.medguide.org.zm/