National HIV surveillance in South Africa--1993-1995.

OBJECTIVE: To determine the point prevalence of HIV infection by serological examination which, performed consistently over several years, serves to monitor the distribution and trend of the HIV epidemic in South Africa. DESIGN: Annual cross-sectional surveys conducted nationally, in October/November. SETTING: South Africa, including areas that used to be known as self-governing and independent national states. SUBJECTS: Pregnant women in the age group 15-49 years who attend antenatal clinic services provided by the public health services. OUTCOME MEASURES: HIV positivity as determined serologically. RESULTS: The rising trend found previously continues: HIV positivity in South Africa was found to be 4.25% in 1993, 7.57% in 1994 and 10.44% in 1995. In 1995 the highest rate was recorded in KwaZulu-Natal (18.23%) and the lowest in the Western Cape (1.66%). On the basis of certain assumptions it is estimated that these rates are indicative of 1.7 million sexually active adults having been infected with HIV by October/November 1995, plus a cumulative total of 40,000 infants. Women in their 20s had the highest age-specific prevalence rates, viz. 13.12% (20-24 years) and 11.03% (25-29 years). HIV positivity in pregnant teenagers was 9.5%. CONCLUSIONS: The HIV epidemic is firmly established in South Africa with a wide variation in provincial prevalence rates. Of particular public health significance is the finding of a high and rising prevalence rate in pregnant teenagers. The observed prevalence rate for 1995 is lower than expected, giving rise to guarded hope that provinces with the highest recorded rates are moving away from the exponential growth found hitherto.

The South African HIV epidemic, reflected by nine provincial epidemics, 1990-1996.

OBJECTIVES: To determine by serological examination the annual point prevalence rates of infection with the human immunodeficiency virus (HIV) in representative samples of subjects in the nine provinces of South Africa, 1990-1996. DESIGN: Annual cross-sectional point prevalence surveys conducted in October/November of each year. SETTING: South Africa, including areas that used to be known as self-governing and independent National States. SUBJECTS: Pregnant women in the age group 15-49 years who attend antenatal clinic services provided by the public health services, and who act as an indicator group of the HIV epidemic among the heterosexually active population. OUTCOME MEASURES: HIV positivity as determined serologically; done consistently over several years, this serves to monitor the distribution and trend of the HIV epidemic in each of the nine provinces of South Africa. RESULTS: Empirical data gained from seven annual, consecutive countrywide surveys demonstrate a wide geographical variation in the point prevalence rates of HIV infection. In October/November 1996 the point prevalence rates (%) were as follows: Western Cape 3.09, Northern Cape 6.57, Northern Province 7.96, Eastern Cape 8.10, Gauteng 15.49, Mpumalanga 15.77, Free State 17.49, KwaZulu-Natal 19.90 and North West 25.13. The weighted national average was 14.17%. There are indications that some of the provinces (KwaZulu-Natal and possibly Mpumalanga) might have passed a point of inflection suggesting deceleration in their specific rates of increase. These results are, however, counterbalanced by the exponential growth still being experienced in provinces with large populations such as Gauteng and the Eastern Cape. It is concluded that the net effect of these divergent trends currently affects the national figure only marginally. A major deflection from the exponential growth patterns seen hitherto can be anticipated only once all or most of the highly populated provinces have traversed their respective points of inflection. The exponential model significantly explains the HIV epidemics in the provinces. The combination of these provincial epidemics describes the initial exponential phase of the epidemic.

National HIV surveillance--South Africa, 1990-1992.

The findings of three annual surveys of women attending antenatal clinics (at the end of 1990, 1991 and 1992) are presented here. These surveys form part of the National HIV Surveillance Programme. This programme is probably the most useful means of monitoring the trend and distribution of the epidemic. In all strata, a consistent rise in the HIV prevalence rate was found; it doubled almost every 12 months. The point prevalence rate in antenatal clinic attenders in South Africa increased from 0.76% in 1990 to 1.49% in 1991 and 2.69% in 1992. The prevalence rate was found to vary widely geographically: Natal/KwaZulu formed the spearhead of the epidemic with a 4.77% rate of HIV infection in 1992. Venda and the Cape appeared to be the least affected with rates of 0.64% and 0.66% respectively.

Changes in total death rate and in ischaemic heart disease death rate in interethnic South African populations, 1978-1989.

Ischaemic heart disease (IHD) in prosperous Western populations rose markedly in the 1940s, peaked between 1970 and 1975 and then fell variably--by about 50% in the USA and Australia. Undoubtedly, decreases in serum cholesterol levels, in the incidence of hypertension and in smoking frequencies are largely responsible. In South Africa, in all population groups other than blacks, IHD rates rose analogously, with Asians and whites attaining very high rates. However, from 1978 to 1989, the total death rate among white males (per 100,000 world population) fell from 1,002 to 631 (37%), and the IHD mortality rate from 312 to 139 (56%). Rates for Asians were 1,306-1,130 (14%) and 355-226 (36%), respectively, and for coloureds 1,691 to 1,392 (18%) and 171 to 110 (36%). For blacks, the total mortality rate remained unchanged; IHD rates were low, but these data are unreliable. Percentage falls in the IHD rate exceed those in the total death rate, especially among whites, indicating true decreases in IHD rate. Understandably, the accuracy of the data, as with such data overseas, is questionable. Local falls are none the less in line with those noted in Western countries. Rates for whites remain very high, and are even higher for Asians. While knowledge of the reasons for the rises and falls in IHD rates remains incomplete, whites have none the less taken some preventive action, although Asians and coloureds apparently little.

Risk of human infections with Crimean-Congo hemorrhagic fever virus in a South African rural community.

Crimean-Congo hemorrhagic fever (CCHF) virus is widely distributed in wild and domestic mammals, birds, and ticks throughout many regions of Africa, Europe, and Asia. Interviews were conducted with 484 individuals from nine farms in the Republic of South Africa from which recent human CCHF cases had originated and with individuals from 27 farms without recognized cases. Serum samples were obtained from all consenting individuals. Blood was also drawn from 2,212 farm animals. Human infection with CCHF virus was uncommon (point prevalence 12.6/1,000). Antibody prevalence in humans on farms increased with age (P less than 0.001), and was correlated with handling lambs. Overall, a greater number of older animals were antibody positive than animals less than one year of age (P less than 0.001), but 12.7% of young animals on farms with human were antibody positive compared with 5.8% on those farms without human infection (P less than 0.05). Physical contact with ticks or tick bite was also found to be a risk factor, but contact with animal blood or fresh meat was not. The risk of CCHF virus infection in the community increased seven-fold following contact with a recognized CCHF case, even when other risk factors were taken into account (point prevalence rate 4.7%). In contrast, antibody prevalence was less than 1% (1 of 128) in the local hospital staff who had cared for patients with CCHF. Prevention is best achieved by education of the farming community and establishing and maintaining awareness in the hospital staff.

Poliomyelitis outbreak in Natal/KwaZulu, South Africa, 1987-1988. 2. Immunity aspects.

An extensive poliomyelitis outbreak due to type 1 poliovirus took place in Natal/KwaZulu, South Africa, in 1987-1988, causing 412 paralytic cases. This epidemic differed from a previously described outbreak in Gazankulu, South Africa, in 1982 in that it occurred against a background of relatively good immunity. Thus, only 12% of patients lacked antibodies to types 2 and 3, indicating lack of previous immunization, and 76% of healthy children sampled in the epidemic area had serological immunity to all 3 types of poliovirus. The occurrence of extensive outbreaks in relatively well-immunized communities emphasizes the need to maximize herd immunity and reduce reservoirs of infection in the gut and in the environment, which can be achieved only with oral polio vaccine.

Poliomyelitis outbreak in Natal/KwaZulu, South Africa, 1987-1988. 1. Epidemiology.

An epidemic of type 1 poliomyelitis occurred in Natal/KwaZulu in the eastern part of South Africa between December 1987 and November 1988. 412 poliomyelitis cases were reported, of whom 74% were younger than 5 years. The case-fatality rate was 8%. It is suggested that massive floods, experienced in the area 2 months earlier, triggered the outbreak.

The cholera epidemic in South Africa, 1980-1987. Epidemiological features.

During the cholera epidemic in South Africa, 1980-1987, 25,251 cases of cholera were bacteriologically proven. The case-fatality rate was 1.4%. Outbreaks occurred in the summer rainfall season. Age-specific attack rates followed the pattern typically found during the 'epidemic phase' of the disease in most years. The vast majority of patients were black South Africans living in rural areas with an average annual rainfall in excess of 600 mm. The containment strategy employed is summarised. Despite the apparent eradication of the disease, it is strongly recommended that vigilance should be maintained and investigations of all possible sources of infection and all human contacts of any new proven case should be carried out speedily and thoroughly.

The Tuberculosis Control Programme, 1985-1986. Results of the data-capturing programme instituted by the Department of National Health and Population Development and the Standing Committee on Epidemiology of the Regional Health Organisation for southern Africa.

The Tuberculosis Control Programme during 1985 and 1986 in the seven health regions of the RSA is described. This information has not previously been available. In 1986 the total tuberculosis case load exceeded 90,000. Of about 56,000 patients in whom the outcome was known, nearly 44,000 were discharged from the programme cured, about 9,500 absconded from therapy and 3,000 died. About 222,000 suspects or contacts of cases were investigated. Analysis of the data by health region revealed regional differences of prevalence rates, predominance of inpatient or outpatient treatment, cure rates, the ratio of new tuberculosis suspects or contacts found for every new confirmed case, and the cost of the programme per case and per capita of the population. Although these data must be interpreted cautiously, it is hoped that they will be valuable to all health workers fighting this major infectious disease in South Africa.

Decline in the ischaemic heart disease mortality rates of South Africans, 1968-1985.

The age-adjusted ischaemic heart disease (IHD) mortality rates (MRs) of white, Asian and coloured South Africans aged 35-74 years were studied for the period 1968-1985. Asians have the highest IHD MR in the RSA, followed by whites, coloureds and then blacks. Asian female have much higher rates than females in the other groups, especially in the older age groups. Asian males have noticeably higher rates in the younger age groups. Coloured females aged 35-44 years have a surprisingly high rate. Declines of 36.5% (from 482 to 306/100,000) for whites between 1970 and 1985, 27.5% (from 583 to 422/100,000) for Asians between 1973 and 1985, and 19.5% (from 287 to 231/100,000) for coloureds between 1976 and 1985 were observed. Rates declined among both males and females as well as in all the age groups studied. Trends in IHD MRs for black South Africans were studied for 1978-1985. The MRs for IHD among blacks are very much lower than those for South African Asians, coloureds and whites. The age-adjusted IHD MR for all South Africans was 162/100,000 in 1978 and had dropped to 121/100,000 in 1985, a 25.3% decline.

Comparison of the decline in the ischaemic heart disease mortality rate in the RSA with that in other Western countries.

South African age-adjusted ischaemic heart disease (IHD) mortality rates (MRs) in the age group 35-74 years were compared with equivalent IHD MRs of five Western countries for the period 1968-1983. South African Asians had higher IHD MRs than all the other countries studied, and South African whites had among the highest despite the declines in rates over the period studied. Although the younger South African age groups had the greatest decline in IHD MRs, in the age group 35-44 years and 45-55 years South African rates still remain higher than those for the five other Western countries studied. Overall, the South African whites' ranking, compared with the other Western countries studied, had changed from the 2nd highest in 1970 to 4th position for males and 3rd for females in 1982.

Disease outbreak investigations--objectives, methods and importance.

In South Africa approximately 100 disease outbreaks or epidemics requiring investigation and control can be anticipated each year. Outbreak investigations play a critical role in determining effective disease control strategies. The application of modern epidemiological methods to the study of disease outbreaks is discussed and problem-solving methodology, which has similarities to the approach used in clinical diagnostics, described. These methods were applied to the investigation of an outbreak of Legionnaire's disease at a Johannesburg teaching hospital.

Legionnaires' disease in a Johannesburg teaching hospital. Investigation and control of an outbreak.

During the period 11 November 1985 - 21 February 1986, 12 cases of Legionnaires' disease were identified at a Johannesburg teaching hospital. Only 2 patients definitely acquired the disease in hospital. Both responded well to treatment. An epidemiological investigation was initiated to determine the source of infection and mode of transmission of the causative agent (Legionella pneumophila), which may be recovered from a wide variety of water sources. Although L. pneumophila was cultured from the hospital hot-water system, there was no association between the location of patients and culture-positive water sites. Cases were clustered in the medical and surgical intensive care units. Being on a ventilator was a significant risk factor for acquiring Legionnaires' disease (relative risk 18,4; 95% confidence interval 2,4 - 142,2). The potential role of ventilators in the transmission of the disease is discussed together with the infection control measures adopted to interrupt the outbreak. To our knowledge this is the first investigation of an outbreak of Legionnaires' disease in South Africa.

Notification of five of the EPI target diseases in South Africa. An assessment of disease and vaccination reporting.

The incidence of 5 of the 6 EPI target diseases, diphtheria, tetanus, poliomyelitis, measles and tuberculosis, in South Africa is examined. Incidence is linked to vaccination coverage and an assessment of municipal vaccination coverage and programme monitoring is made. Diphtheria incidence has decreased sharply and this is probably due to vaccination. A change in the cyclical pattern of polio incidence may be due to improved vaccination. No influence of vaccination could be detected with respect to measles, tetanus and tuberculosis. Vaccination programmes report results as 'total doses given' and not as coverage rates. Improvements in three major areas, i.e. vaccination coverage, programme monitoring and evaluation and programme co-ordination are proposed.

A comprehensive investigation of immunity to poliomyelitis in a developing country.

A comprehensive nationwide surveillance program of serologic immunity of two-year-old black children, combined with evaluation of vaccine quality and distribution, was carried out in South Africa during 1983-1984. Sera were randomly collected from urban and rural groups and cluster samples collected from the semi-urban group. The sample represented 0.23% of the total target population. Satisfactory levels of immunity were found in the urban (80%) and semi-urban (71%) groups but a disquietingly low level was found for the rural group (59%). Individual districts in the rural group could be singled out for directed cluster sampling at a later stage. History and documentation of immunization corresponded well to serologic findings and revealed also a fairly substantial level of natural immunization among individuals who, on history, had received no vaccine. Some 95% of random samples of vaccine recalled from the field showed satisfactory levels of potency. An immunity surveillance program such as this is ideally suited and highly cost-effective for developing countries with incomplete immunization to prevent large-scale buildup of immunity deficit. The technique, however, is too insensitive to detect localized community immunity defects.

Weight-for-height nutrition surveys in rural Kwazulu and Natal, July 1983.

South Africa experienced a severe drought in 1982/1983. In July 1983 certain rural communities were identified which had recorded the most severe drought. Children were selected at random from these areas and compared against controls with regard to their respective weights and heights. On the basis of the predetermined sample sizes employed, no significant difference between the groups could be demonstrated. When compared with certain urban children attending crèches in Pretoria, there was also no significant difference. It is concluded that in July 1983 no crisis as regards the prevalence of acute malnutrition in children aged 1-5 years existed in the communities tested.

Poliomyelitis outbreak in South Africa, 1982. I. Epidemiology.

An epidemic of type 1 poliomyelitis in Gazankulu, which is a self-governing National State in the north-eastern part of South Africa, is described. 260 cases of paralytic poliomyelitis, included 42 deaths, were recorded between 1st May and 3rd September 1982. 92% of cases, who were all admitted to hospital, were in the age group 0 to 5 years. The vaccination status of cases was found to be low: 69% said they had received no vaccination and a further 15% were unsure. Recommendations on the steps to be taken to prevent a recurrence can be gleaned from the discussion in Part II of this article, Laboratory and vaccine aspects (see pp. 24-29 below).

Poliomyelitis outbreak in South Africa, 1982. II. Laboratory and vaccine aspects.

Between May and September 1982 an extensive epidemic of poliomyelitis occurred in Gazankulu in the north-eastern region of South Africa, with a total of 260 paralysed cases and 42 deaths. A three-part study was instituted, the first of which established the aetiological agent of the epidemic to be poliovirus type 1. This was determined initially by serology and later by isolation of the virus, chiefly from stool specimens. The potency of the trivalent oral polio vaccine manufactured at the National Institute for Virology, was evaluated in the second study by in vitro potency testing of samples recalled from the epidemic areas and also in vivo by determining the serological response of seronegative children immunized with fresh vaccine. Nearly half of the vaccine samples withdrawn from the field had sub-optimal titres whereas fresh vaccine produced a very satisfactory serological response. In order to plan immunization rationally, the third study, namely the determination of the immune status of the various populations throughout the country was instituted. This involved both serology, using a neutralization test on randomly collected sera and also an analysis of patient history and health card documentation. With the exception of the urban black area (Soweto), which had high levels of immunity both serologically (75%) and on history, other areas examined had poorer levels of immunity of approximately 52% on serological testing and 57% on history. It is clear that the epidemic was multifactorial in origin, with fall-offs in the level of immunization and evidence of breaks in the cold chain being the major factors.

Acute haemorrhagic conjunctivitis in Soweto, 1982. An epidemiological investigation.

In mid-March 1982 an epidemic of acute haemorrhagic conjunctivitis (AHC) broke out in Soweto. Over 34 000 people were treated before the epidemic subsided some 13 weeks later, the overall attack rate being 34,1/1 000. The records of 3 945 patients seen at St John's Eye Hospital were analysed, and the populations at greatest risk of contracting the disease were identified. Males were at greatest risk, with a male to female attack rate ratio of 1,8:1. The age-specific attack rates for both sexes show a similar pattern, peaking in the young adult age groups.