Non-vertical HIV transmission to children in sub-Saharan Africa

Abstract

This review considers whether HIV prevalence data on children in sub-Saharan Africa support the hypothesis that blood exposures account for a large proportion of HIV infections in Africa. Data from a systematic search on HIV-infected children support two analyses. In 25 studies where the mothers' HIV status was not matched with data on each child (excluding non-representative samples of children), the observed prevalence in children in 20 studies was greater than expected from vertical transmission. The population-weighted difference – 1.3% – was approximately one-third of observed prevalence in children. In 32 studies that match HIV-positive children with HIV-negative mothers, 406 discordant mother–child pairs were identified, and in studies identifying at least five non-vertical infections in children, 17.5% of HIV-positive children had HIV-negative mothers. In discussing an important role for unsafe health care in exceptionally rapid HIV transmission in Africa, leading AIDS researchers cite low HIV prevalence in children not yet sexually active. The assumption that childhood HIV prevalence would increase with age in children, if injections transmitted HIV is shown to be erroneous; it fails to account for early mortality in HIV-positive children. Evidence of child-to-child HIV transmission supports the theory that nosocomial infections are important to the AIDS pandemic, and procedures more prevalent than blood transfusions, such as injections, are likely involved.

Keywords

children HIV Africa iatrogenic

INTRODUCTION

Researchers providing evidence that blood exposures from unsterilized instruments are a major HIV transmission route in Africa have drawn attention to studies reporting HIV prevalence in children too high to be explained by mother-to-child transmission alone, and to hundreds of mother–child pairs in which only the child is HIV positive.¹ These data have been contested as unrepresentative or inaccurate.² Using selected data, others have argued that observed HIV prevalence in children is lower than would be expected if unsafe injections were making an important contribution to HIV prevalence in Africa.³

The purpose of this review is to assemble and present data on HIV infection and medical injections in children, and to assess whether these data support a major role for iatrogenic (in this case, patient to patient) HIV transmission in Africa. The role of transfusion medicine in the early history of Africa's heterosexual AIDS pandemic has been recently examined.⁴ The potential for more prevalent medical procedures such as injections administered with reused needles or syringes is a controversy of ongoing practical importance.

METHODS

This review analyses data from a systematic search for data on HIV prevalence in children in Africa from the US Census Bureau HIV/AIDS Surveillance Data Base, surveys identified through the Google Scholar web search engine (search terms: HIV, prevalence, children, Africa), abstracts from International AIDS Conferences XIV–XVII (search term: children) and personal files.^5,6 Electronic abstracts are not available for AIDS Conferences I–XIII. These data are analysed in two groups, depending on whether studies report the HIV status of the mothers of HIV-positive children, or report the prevalence among adults and children independently.

In the first analysis, of children for which only the prevalence among adults or mothers is known and mother–child pairs are not identified, the exclusion criteria are (1) any non-representative sample, such as prevalence in street children or in-patients; (2) lack of confirmatory test, unless the single test is a polymerase chain reaction (PCR) test; (3) absence of maternal or adult prevalence data for the same study population; or (4) sample sizes under 50 persons. These data are compared with the HIV prevalence that could be expected from mother-to-child transmission based on maternal or adult HIV prevalence (the Appendix details the assumptions underlying estimates of children's HIV prevalence from vertical transmission).

In the second analysis, of HIV-positive children matched with their biological mothers, the inclusion criteria for meta-analysis are (1) an unbiased sample of children tested for HIV regardless of the mother's serostatus; (2) a confirmatory HIV test, unless the single test is a PCR test; and (3) at least five discordant mother–child pairs.

Following these analyses, two themes in the interpretation of data on non-vertical HIV transmission are reviewed. The expectation that HIV prevalence increases with age in children where medical injections are transmitting HIV is assessed by modelling non-vertical HIV incidence and survival alongside mother-to-child transmission and survival. The expectation that injection risks to children are similar to or greater than risks to adults is also evaluated.

RESULTS

Prevalence data meeting the inclusion criteria for the first analysis (in which the HIV status of the mothers of HIV-positive children is not reported) are available for 28 groups of children – totaling almost 40,000 children – in 12 countries over three decades (Table 1). Children range up to 18 years old, and many studies combine data for a wide range of ages (e.g. children 2–12 years old). The observed prevalence is too high to be explained by vertical transmission alone in 20 of 25 groups of children, including seven of nine large (>1000 children) studies. The observed HIV prevalence was 1.3% greater than expected (population weighted mean), and the difference is statistically significant (P < 0.05 in 15 of 25 studies, cumulative z-score = 15.4 for the 25 studies combined). The average observed prevalence in children of all ages is 3.7%, meaning the excess prevalence is relatively large. Excluding groups of children under age 18 months that are not confirmed by PCR, the excess prevalence is reduced to 1–1%, but remains statistically significant (z score = 4.6).

Table 1

Child HIV prevalence and expected prevalence from vertical transmission

Country, year, reference	Adult HIV prevalence (%)			Age of children	HIV in children (%)	Expected HIV in children (%)	SE	z	No.
Country, year, reference	Current*	Five years previous	Ten years previous	Age of children	HIV in children (%)	Expected HIV in children (%)	SE	z	No.
DRC, 1985⁷	10.3	Increasing	Increasing	0–1 year	8.1	2.5	0.280	5.2	136
				1–2 years	1.3	1.7			224
				2–14 years	1.3	0.5			230
				All	2.9	1.4			590
Rwanda, 1986^8,9	15.4	Increasing	Increasing	<5 years	10.1	2.1	0.502	15.9	>500
Tanzania, 1987¹⁰	6.86	Increasing	Increasing	0–1 year	3.1	1.6	0.203	1.1	130
				1–4 years	0.9	0.9			422
				5–9 years	0.4	0.5			531
				All	0.9	0.8			1083
Tanzania, 1987⁸	4.47	Increasing	Increasing	0–14 years	0.8	0.3	0.050	10	565
Uganda, 1987¹¹	17	Increasing	Increasing	0–14 years	0.0	1.0	0.470	−2.1	131
Tanzania, 1988⁹	20.7	Increasing	Increasing	1–4 years	3.5	2.8	0.593	1.7	112
				10–14 years	2.0	0.7			99
				All	2.8	1.8			211
Cameroon, 1988¹¹	0.5	Increasing	Increasing	0–13 years	0.3	0.0	0.014	21.4	667
Cameroon, 1988¹²	0.42	Increasing	Increasing	0–14 years	1.1	0.0	0.014	75.6	178
Cameroon, 1988¹³	0.62	Increasing	Increasing	0–14 years	0.0	0.0	0.014	0.0	319
Cameroon, 1989¹⁴	0.43	Increasing	Increasing	0–14 years	0.8	0.0	0.013	61.5	337
Uganda, 1991¹⁵	21	Increasing	Increasing	2–14 years	2.6	1.1	0.580	1.2	694
	11.8	Increasing	Increasing	2–14 years	0.8	0.6			1175
	35	Increasing	Increasing	2–14 years	1.8	1.8			465
	All	Increasing	Increasing	All	1.5	1.0			2334
Uganda, 1993¹⁶	11	Stable	Increasing	11–14 years	0.7	0.3	0.278	1.4	139
Ethiopia, 1994¹⁷	12.61	Increasing	Increasing	0–4 years	0.8	1.9	0.353	−0.8	121
				5–9 years	0.2	0.9			250
				10–14 years	0.7	0.4			371
				All	0.2	0.2			2601
Uganda, 1995¹⁸	37.6	Stable	Increasing	<18 months	19.2	8.0	1.651	6.8	697
Zambia, 1995¹⁹	23.3	Increasing	Increasing	1–4 years	4.1	3.2	0.766	1.2	268
South Africa, 2002²⁰	26.2	Increasing	Increasing	20 months	24.9	4.2	1.132	18.3	1448
South Africa, 2002²¹	26.5	Increasing	Increasing	2–9 years	6.2	1.9	0.759	6.4	1377
				10–14 years	4.7	0.9			973
				All	5.6	1.5			2350
Tanzania, 2002²²	1.4	Increasing	Increasing	0–18 months	3.4	0.3	0.027	99.9	175
				2–5 years	1.8	0.2			113
				6–10 years	3.4	0.1			89
				11–15 years	0.0	0.0			84
				All	2.4	0.2			461
Senegal, 2003²³	0.8	Stable	Increasing	0–14 years	1.9	0.0	0.027	70.4	369
South Africa, 2005²⁴	30.2	Stable	Increasing	2–4 years	5.1	3.7	0.811	1.7	729
				5–9 years	4.4	2.1			1341
				10–14 years	1.7	1.0			1745
				All	3.3	1.9			3815
Uganda, 2005²⁵	12.1	Stable	Decreasing	0–18 months	1.0	2.6	0.348	−3.4	2666
				1–5 years	0.5	1.5			5689
				All	0.7	1.9			8355
Zimbabwe, 2005²⁶	21.1	Decreasing	Stable	2–5 years	2.7	2.4	0.605	2.4	330
				6–8 years	5.8	1.4			189
				9–11 years	1.3	1.0			170
				All	3.2	1.8			689
South Africa, 2006²⁷	37.6	Stable	Increasing	4–8 weeks	7.6	11.7	2.469	−1.7	2439
Uganda, 2006²⁸	19	Stable	Decreasing	0–5 years	20.0	2.6	0.573	31.7	2246
Swaziland, 2007²⁹	39	Increasing	Increasing	2–4 years	5.1	4.8	1.084	1.1	820
				5–9 years	4.2	2.7			1367
				10–14 years	2.6	1.3			1392
				All	3.8	2.6			3579
Cumulative z-score	22.9	36/139						15.4

*Current – adult prevalence at time of data collection on children in the same study population

To explain the observed excess prevalence by mother-to-child transmission, the product of the survival rate and the vertical transmission rate would have to be increased by +54%. This would involve adjustments equivalent to substituting a 38% rate of vertical transmission (compared with 30% observed, 95% confidence interval [CI] 20–40%) and a 39% five-year survival rate (compared with 31% observed, 95% CI 23–39%) (see the Appendix).

Prevalence data meeting the inclusion criteria for the second analysis (in which HIV status is known for child–mother pairs) are available for 15 of 33 studies reporting any discordant mother–child pairs (Table 2). These cases are rarely isolated, with a median of seven cases in every such study. More detailed descriptions of 16 of these studies are presented in an earlier review.⁶⁰ In 14 of these studies, the number of children who received blood transfusions is reported, an HIV transmission risk that can be ruled out for 148 of 312 cases. For studies included in the meta-analysis, a median of 17.5% of HIV-positive children have HIV-negative mothers.

Table 2

HIV-positive children with HIV-negative mothers in sub-Saharan Africa, 1985–2009

Country, year, reference	HIV-positive children with HIV-negative mothers	Number of HIV-positive children	% HIV-positive children with HIV-negative mothers	Confirmatory test or PCR
DRC, 1985³⁰	17	44	38.6	No
Gabon, 1986³¹	2	315	0.6	No
Rwanda, 1986³²	15	75	20.0	Yes
Cote d'Ivoire, 1988³³	26	192	13.5	Yes
Tanzania, 1988³⁴	7	47	14.9	No
Cote d'Ivoire, 1989³⁵	5	22	22.7	No
Ethiopia, 1989³⁶	2	11	18.2	Yes
Guinea-Bissau, 1989³⁷	1	2	50.0	No
Rwanda, 1990³⁸	39	684	5.7	No
Uganda, 1990³⁹	4	200	2.0	No
Uganda, 1990⁴⁰	5	28	17.9	No
Uganda, 1990⁴¹	4	17	23.5	Yes
Zambia, 1991⁴²	8	247	3.7	No
DRC, 1991⁴³	21	60	35.0	Yes
Burkina Faso, 1993⁴⁴	14	60	23.3	Yes
Rwanda, 1993⁴⁵	61	NA	NA	No
Ethiopia, 1994⁴⁶	1	3	33.3	Yes
Uganda, 1994⁴⁷	7	34	20.6	No
Zambia, 1994⁴⁸	4	32	12.5	No
Nigeria, 1995⁴⁹	30	44	68	Yes
Sudan, 1995⁵⁰	5	15	33.3	No
Nigeria, 1996⁵¹	7	23	30.4	No
Nigeria, 1997⁵²	8	10	80.0	No
Senegal, 2000⁵³	1	98	1.0	Yes
Nigeria, 2001⁵⁴	14	80	17.5	Yes
South Africa, 2003⁵⁵	15	NA	NA	Yes
Uganda, 2004⁵⁶	1	79	1.3	Yes
Nigeria, 2005⁵⁷	46	840	5.5	Yes
South Africa, 2005⁵⁸	7	599	1.2	Yes
Uganda, 2005²⁵	7	52	13.5	Yes
Swaziland, 2006^25,59	11	50	22.0	Yes
Kenya, 2009⁵⁹	11	NA	NA	Yes
Total	406	4122	17.5*	17/32

PCR = polymerase chain reaction

*Median excluding studies with fewer than five discordant pairs and any with neither a confirmatory test nor a PCR test

Age distribution of HIV in children

In expert testimony and an authoritative review in Lancet, leading researchers in HIV epidemiology have pointed to low HIV prevalence in African children as evidence that unsafe health care cannot be a major HIV transmission route.³ This view has been strikingly mangled in a report contracted to RTI by the Institute of Medicine for the US Senate, a document which proposes that HIV prevalence would increase with age if unsafe medical injections were infecting children, ignoring the more rapid progression to AIDS and death in the infected child and supposing that vertical transmission might not also occur.⁶¹ An increasing trend could only be expected if children infected via vertical transmission were excluded from analyses. This can be shown by modelling non-vertical HIV incidence, similar to the incidence observed among children 2–14 years old in South Africa, alongside mother-to-child transmission.⁶²

Table 3 lists the HIV incidence from mother-to-child transmission on the diagonal and lists non-vertical incidence every year off the diagonal. For simplicity all vertical transmission is assumed to occur at birth. Representing a stable epidemic with 20% adult HIV prevalence, Table 3 predicts 7% incidence from mother-to-child transmission in every birth cohort. The difference from 7% in each cell on the diagonal is the difference in survival to year 10 for each birth cohort, depending on when the children were infected. The non-vertical transmission rate represented on the horizontal is 5% of the vertical transmission rate, i.e. 0.35% non-vertical HIV incidence in every subsequent year. Again, the difference from 0.35% in each cell on the horizontal is the difference in survival to year 10 for each age group, depending on when the children were infected. The estimated survival rates for vertically and non-vertically infected children are detailed in the Appendix.

Table 3

Survival to ages 1–10 of HIV-infected children: 0.35% incidence, 7% MTC

Age in year 10	0	1	2	3	4	5	6	7	8	9	10
Newly infected in year 10	7.0	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35	0.35
Newly infected in year 9		4.4	0.34	0.34	0.34	0.34	0.34	0.34	0.34	0.34	0.34
Newly infected in year 8			3.4	0.32	0.32	0.32	0.32	0.32	0.32	0.32	0.32
Newly infected in year 7				2.8	0.29	0.29	0.29	0.29	0.29	0.29	0.29
Newly infected in year 6					2.3	0.27	0.27	0.27	0.27	0.27	0.27
Newly infected in year 5						2.0	0.25	0.25	0.25	0.25	0.25
Newly infected in year 4							1.8	0.22	0.22	0.22	0.22
Newly infected in year 3								1.5	0.19	0.19	0.19
Newly infected in year 2									1.4	0.17	0.17
Newly infected in year 1										1.2	0.15
Newly infected in year 0											1.0
Prevalent infections (%)	7.0	4.8	4.1	3.8	3.6	3.6	3.6	3.5	3.6	3.6	3.5

MTC – mother-to-child transmission. Children born in year 8 (sample calculation in bold) have reached age 2 in year 10 (see top row). Of these 7% were infected vertically in year 8, another 0.35% were infected non-vertically in year 9, and 0.35% were infected non-vertically in year 10. While 94% of the 0.35% infected in year 9 (0.94 × 0.35 = 0.33) survive to year 10, only 48% of those infected vertically in year 8 (0.48 × 7 = 3.4) survive to be counted. For this reason there are fewer surviving HIV-positive children at age 2 (4.1%) than there are at age 1 (4.8%) in year 10

A 1.4% excess prevalence in children under age 10 is predicted, one-third of all infections. Thus even if one in three prevalent infections in children were from non-vertical HIV transmission, the prevalence of HIV in children would not increase with age. The modelled prevalence of non-vertical HIV infection in children aged 0–5 is 0.8%, out of a prevalence of 4.5% in this age group. In older children, the proportion infected non-vertically is 2.0% out of a prevalence of 3.4% among 6–10 year olds.

Implications for adults

The incidence of non-vertical HIV transmission to children has been proposed as an estimate of iatrogenic risks to African adults.³ This assumes that children and adults have the same iatrogenic risks. In reality, invasive medical procedures for children and adults differ in nature and occur in separate clinical settings, allowing adult-to-adult transmission among patients and child-to-child transmission in paediatric wards, but not adult-to-child or child-to-adult HIV transmission. Adults share some blood exposures with children as they also receive injections, but adults have additional exposures (e.g. dental care, shaving and the invasive care involved in assisted delivery).^63,64

Adults in Africa receive an estimated 2.1 injections per year.⁶⁵ The World Health Organization (WHO) has estimated injection risks to children as being identical to adults, except for the addition of immunization injections.⁶⁶ Excluding immunizations, most of which are now performed with auto-disable (non-reusable) syringes in Africa, children in Uganda receive fewer injections with age: children under two years old receive 3.1 therapeutic injections per year, while children 9–12 years old receive 0.6.⁶⁷ In Rwanda, well children receive on average 4.0 injections per year.⁶⁸

The risk from an individual injection, however, depends on the background clinical prevalence of HIV among patients. In paediatric settings fewer patients are HIV positive, meaning that an injection with a needle used on another patient in the paediatric ward is less likely to have been contaminated with HIV. No more than one in five HIV-positive mothers in Africa will have an HIV-positive child who survives to age two without antiretroviral (ARV) therapy. Adjusting for a higher HIV prevalence among patients, the adult annual HIV incidence from unsafe medical injections alone should be greater than the non-vertical incidence in children.

DISCUSSION

This review demonstrates that the available data on HIV prevalence in children in Africa support an important role for iatrogenic transmission in Africa's generalized HIV epidemics. In 21 of 28 studies with information on prevalence in children and relevant adults, the HIV prevalence in children is too high to be explained by vertical transmission, and this difference is statistically significant. In studies matching HIV-positive children to HIV-negative mothers, a median 17.5% of HIV infections in children are non-vertical.

Importantly, attributing over half of prevalent HIV infections in children to non-vertical transmission does not mean that 50% or more of incident HIV infections result from unsafe health care. Non-vertically infected children survive several times longer to be counted, and some will have been infected by other means, including traditional scarification, circumcision, sexual abuse and early sexual debut. That such a large number of children would be subject to these other risk factors, more so than to unsafe injections, seems unlikely, however.

The data presented in Tables 1 and 2 underestimate the proportion of HIV infections in children from non-vertical transmission. The expected HIV prevalence from mother-to-child transmission reported in Table 1 and estimated using assumptions detailed in the Appendix is likely to be an overestimate because: (1) children in longitudinal survival studies probably lived longer than other HIV-positive African children would, receiving free health care over the course of these studies; (2) the HIV prevalence in adults in increasing epidemics will have been lower at the time when the sampled children were born; and (3) in recent studies the mother-to-child transmission rate is an overestimate, as treatment for the prevention of mother-to-child transmission has increasingly been made available. As a result, the excess prevalence from non-vertical HIV transmission to children has likely been underestimated.

In one respect the opposite bias may apply. In 11 studies including children of a wide range of ages (e.g. 2–14 years) no data on the age distribution of children in the sample are reported, and the sampling method is not described. A left-skewed age distribution could exaggerate the difference between observed and expected prevalence in such studies. Hospital samples (largely avoided through the exclusion criteria) would probably over-represent younger children, whereas random household sampling (described in several studies) would not.

The percentage of HIV-positive children with HIV-negative mothers in Table 2 may also underestimate the percentage of HIV-positive children with non-vertical infections. Some mothers acquire HIV from breastfeeding children, following a hospital-acquired infection in the child. This has been documented in scores of cases from large iatrogenic HIV outbreaks among Libyan, Romanian and Russian children.^69–71 In high incidence settings in Africa, other mothers of children with hospital-acquired HIV infections will already have unmatched HIV infections.

Modelling the age distribution of HIV in children assuming a bimodal epidemic shows that a high rate of iatrogenic transmission (with non-vertical incidence as great as 5% of mother-to-child incidence) is consistent with observed HIV prevalence in children. Notably, even with this high rate of iatrogenic transmission, the high HIV prevalence in mothers (in this model 20%) is consistent with low HIV prevalence in children (3.5% in children more than five years old). This model is most apt in South Africa in 2005, where immunization injections are still given with disposable syringes. The risk to transmit HIV through injections has been reduced throughout Africa by the shift to using auto-disable syringes for immunization injections. This was a shift led by UNICEF, WHO, other donors and national governments beginning in 2001. Only Namibia and South Africa have not yet adopted the auto-disable (non-reusable) syringe for all immunizations.

Support for more efficacious infection control in other medical procedures remains poor in Africa, however, as evidenced in the limited sterilization equipment coverage for clinics and hospitals.²⁸ Unsafe health care has been proposed as a major HIV transmission route in Africa, encompassing very small blood exposures in unsafe medical injections, and cross-contamination in other invasive procedures such as assisted delivery and unsterile dental care. This hypothesis cannot be dismissed on the basis of systematically selected childhood HIV prevalence data.

Footnotes

Acknowledgements

The author gratefully acknowledges Devon Brewer's support in selecting and applying statistical methods appropriate to this analysis, though he is not responsible for errors of omission or commission.

APPENDIX

References

Gisselquist

, Rothenberg

, Potterat

, HIV infections in sub-Saharan Africa not explained by sexual or vertical transmission. Int J STD AIDS 2002;13:657–66

Senate Health, Education, Labor and Pensions Committee – Africa AIDS. Washington, DC: US Senate, March 27, 2003

Schmid

, Buve

, Mugyenyi

, Transmission of HIV-1 infection in sub-Saharan Africa and effect of elimination of unsafe injections. Lancet 2004;363:482–8

Schneider

, Drucker

. Blood transfusions in the early years of AIDS in Sub-Saharan Africa. Am J Public Health 2006;96:984–94

USAID. HIV/AIDS Surveillance Data Base. http://hivaidssurveillancedb.org/hivdb/RecordSelPage.aspx2008 (last accessed 9 July 2009)

International AIDS Society. Abstract Search. http://www.iasociety.org/AbstractSearch.aspx (last accessed 14 May 2009)

Quinn

, Mann

, Curran

, AIDS in Africa: an epidemiologic paradigm. Science 1986;234:955–62

Bizmungu

, Ntilivamunda

, Tahimana

, Nationwide community-based serological survey of HIV-1 and other human retrovirus infections in a central African country. Lancet 1989;333:941–3

Killewo

, Nyamuryekunge

, Sandstrom

, Prevalence of HIV-1 infection in the Kagera Region of Tanzania: a population-based study. AIDS 1990;4:1081–5

10.

Carswell

. HIV infection in healthy persons in Uganda. AIDS 1987;1:223–7

11.

Merlin

, Josse

, Threbucq

, Epidemiological monitoring of the acquired immunodeficiency syndrome (AIDS) in six Central Africa states. Med Trop 1988;48:381–9

12.

Kaptue

, Garrigue

, Merlin

, Serology Survey of HIV 1 During Three Years in Cameroon . IV International Conference on AIDS, Stockholm, 6/13–14, Poster 5031, 1988

13.

Aebischer

, Angeloni

, Costa

, HIV-1 Infection in Some Urban and Rural Areas of Cameroon . V International Conference on AIDS, Montreal, 6/4–9, Poster M.G.P. 9, 1989

14.

Durand

, Merlin

, Josse

, AIDS Survey in Cameroon . III International Conference on AIDS, Washington, DC, 6/1–5, Abstract WP.78, 1987

15.

Sewankambo

, Wawer

, Gray

, Demographic impact of HIV infection in rural Rakai District, Uganda: results of a population-based cohort study. AIDS 1994;8:1707–13

16.

Weis

, Masheisha

, Sahimuller

, Assessing Trends of the HIV Epidemic in Western Uganda . X International Conference on AIDS, Yokohama, Japan, 8/7–12, Poster P.C.0041, 1994

17.

Fontanet

, Messele

, Dejene

, Age- and sex-specific HIV-1 prevalence in the urban community setting of Addis Ababa, Ethiopia. AIDS 1998;12:315–22

18.

Bakaki

, Kayita

, Machado

, Epidemiologic and clinical features of HIV-infected and HIV-uninfected Ugandan children younger than 18 months. J Acquir Immune Defic Syndr 2001;28:35–42

19.

Prevalence of HIV in children in an urban community virology laboratory, UTH. Viral Infect Zambia 1995;2:45–6

20.

Dedicoat

, Page

, Mbath

, Risk Factors for HIV Infection in Mother–Child Pairs in Rural South Africa . XIV International Conference on AIDS: Abstract no. ThPeC7427

21.

Nelson Mandela/HSRC Study of HIV/AIDS. South African National HIV Prevalence, Behavioural Risks and Mass Media Household Survey. Cape Town, South Africa: HSRC Press, 2002

22.

Croce

, Fedeli

, Dahoma

, Risk factors for HIV/AIDS in a low HIV prevalence site of sub-Saharan Africa. Trop Med Int Health 2007;12:1011–7

23.

Gaye-Diallo

, Toure

. Infection a HHV8 au Senegal: seroprevalence dans differents groupes de la population (1369 echantillons). XIII International Conference on AIDS STI, Nairobi, Kenya, 9/21–26, Abstract Wed221041, 2003

24.

Shisana

, Rehle

, Simbayi

, South African National HIV Prevalence, HIV Incidence, Behaviour and Communication Survey. Cape Town, South Africa: HSRC Press, 2005

25.

Measure

DHS

. Demographic and Health Surveys. Departments of Health and ORC Macro Calverton, Maryland, USA. http://www.measuredhs.com/countries/start.cfm (last accessed 7 June 2009)

26.

Gomo

, Rusakaniko

, Mashange

, Household Survey of HIV-Prevalence and Behavior in Chimanimani District, Zimbabwe, 2005. Cape Town, South Africa: HSRC Press, 2005

27.

Rollins

, Mzolo

, Little

, HIV Prevalence Rates Amongst 6 Week Old Infants in South Africa: The Case for Universal Screening at Immunization Clinics. XVI International Conference on AIDS, Abstract no. THAC0104

28.

Kirungi

, Musinguzi

, Opio

, Prevalence and Risk Factors for HIV Infection in Uganda: Results of a National Population Based HIV Serological Survey. International Conference on AIDS, Abstract no. MOPE0331

29.

Okinyi

, Brewer

, Potterat

. Horizontally acquired HIV infection in Kenyan and Swazi children. Int J STD AIDS 2009;20:852–7

30.

Mann

, Francis

, Davachi

, Risk factors for human immunodeficiency virus seropositivity among children 1–24 months old in Kinshasa, Zaire. Lancet 1986;ii:654–7

31.

Johns Hopkins University. AIDS – a public health crisis. Population information program, population reports, issues in world health. 1996; Series L, 6:194–228

32.

Lepage

, Van de Perre

. Nosocomial transmission of HIV in Africa: what tribute is paid to contaminated transfusions and medical injections. Infect Control Hosp Epidemiol 1988;9:200–3

33.

Schuerman

, Seynhaeve

, Doustin

, HIV-1 and HIV-2 Infection and Pediatric AIDS in Dabou Protestant Hospital – Ivory Coast. IV International Conference on AIDS, Marseille, 18–20 October, Poster 246, 1989

34.

Mgone

, Mhalu

, Shao

, Prevalence of HIV-1 infection and symptomology of AIDS in severely malnourished children in Dar Es Salaam, Tanzania. J Acquir Immune Defic Syndr 1991;4:910–3

35.

Gayle

, Gnaore

, Adjorlolo , HIV-1 and HIV-2 infection in children in Abidjan, Cote d'Ivoire. J Acquir Immune Defic Syndr 1992;5:513–7

36.

Beyene

, Moss

. Clinical and epidemiological features of HIV-1 seropositive hospitalized Ethiopian children. Ethiop Medl J 1991;29:57–61

37.

Poulsen

, Aaby

, Gottschau

, HIV-2 infection in Bissau, West Africa, 1987–89: incidence, prevalence, and routes of transmission. J Acquir Immune Defic Syndr 1993;6:941–8

38.

Commenges

, Alioum

, Lepage

, Estimating the incubation period of paediatric AIDS in Rwanda. AIDS 1992;6:1515–20

39.

Muller

, Moser

. Risk Factors for Paediatric HIV-1 Infection in Uganda (Abstract). VIII Internatioanl Conference on AIDS, Amsterdam, 19–24 July. Abstract no. PoC4733, 1992

40.

Kengeya-Kayondo

J-F

, Malamba

, Nunn

, Human immunodeficiency virus (HIV-1) seropositivity among children in a rural population of south-west Uganda: probable routes of exposures. Ann Trop Paediatr 1995;15:115–20

41.

Katongole-Mbidde

, Kazura

, Banura

, Latency Period to the Development of Childhood AIDS-Associated Kaposi's Sarcoma (KS) in African Children (Abstract). International Conference on AIDS, Florence, 16–21 June. Abstract MC3185, 1991

42.

Luo

. HIV infection among hospitalized children at the University Teaching Hospital, Lusaka: diagnosis and risk of transmission through therapeutic practices. Unpublished, UNZA Medical Library 1991

43.

Muganga

, Nkuadiolandu

, Mashako

. Clinical manifestations of AIDS in children in Kinshasa. Pediatrie 1991;46:825–9

44.

Prazuck

, Tall

, Nacro

, HIV infection and severe malnutrition: a clinical and epidemiological study in Burkina Faso. AIDS 1993;7:103–8

45.

Hitimana

, Luo-Mutti

, Madraa

, A multicentre matched case control study of possible nosocomial HIV-1 transmission in infants & children in developing countries. Int Conf AIDS 1993;9:94, abstract no. WS-C13-2

46.

Simonon

, Lepage

, Karita

, An assessment of the timing of mother-to-child transmission of Human Immunodeficiency Virus Type 1 by means of Polymerase Chain Reaction. J Acquir Immune Defic Syndr 1994;7:952–7

47.

Zeigler

, Katongole-Mbidde

. Kaposi's sarcoma in childhood: an analysis of 100 cases from Uganda and relationship to HIV infection. Int J Cancer 1996;65:200–3

48.

Muyanga

. Paediatric HIV seropositivity in an urban community of Lusaka. Unpublished, UNZA Medical Library 1994

49.

Emodi

, Okafor

. Clinical manifestations of HIV in children at Enugu, Nigeria. J Trop Pediatr 1998;44:73–6

50.

Hashim

, Salih

, El Hag

, AIDS and HIV infection in Sudanese children: a clinical and epidemiological study. AIDS Patient Care STDs 1997;11:331–7

51.

Angyo

, Okpeh

, Onah

, Pediatric AIDS in Jos, Nigeria. W Afr Med J 1998;17:268–72

52.

Odaibo

, Bamgbose

, Jegede

, Demographic and Laboratory Evidence of Non sexual Transmission of HIV in Nigeria. CD Only: XV International Conference on AIDS. Abstract no. C12079

53.

Diack Mbaye

, Signate Sy

, Diagne Gueye

, Epidemiological and clinical aspects of paediatric HIV infections in Albert-Royer Paediatric Hospital (Dakar, Senegal). Arch Pediatr 2005;12:404–9

54.

Adejuyibge

, Oyelami

, Onayemi

, Paediatric HIV/AIDS in Ile-Ife, Nigeria. Cent Afr J Med 2003;49:74–8

55.

Cotton

. Evidence for Unusual HIV 1 Transmission in Children. Poster Late Breakers: The 2nd IAS Conference on HIV Pathogenesis and Treatment: Abstract no. LB49

56.

Were

, Mermin

, Wamai

, Undiagnosed HIV infection and couple HIV discordance among household members of HIV-infected people receiving antiretroviral therapy in Uganda. J Acquir Immune Defic Syndr 2006;43:91–5

57.

Onankpa

, Airede

, Paul

, Pattern of pediatric HIV/AIDS: a five-year experience in a tertiary hospital. J Nat Med Assoc 2008;100:821–5

58.

Shisana

, Connolly

, Rehle

, HIV risk exposure among South African children in public health facilities. AIDS Care 2008;20:755–63

59.

60.

Gisselquist

, Potterat

, Brody

. HIV transmission during paediatric health care in sub-Saharan Africa – risks and evidence. S Afr Med J 2004;94:109–16

61.

Viswanathan

, Lux

, Kirmeyer

, Examination of Health care Transmission of HIV/AIDS in sub-Saharan Africa and the Caribbean Region: A Literature Review, Expert Panel, and Report. RTI Project Number: 08490.005

62.

Rehle

, Shisana

, Pillay

, National HIV incidence measures – new insights into the South African epidemic. S Afr Med J 2007;97:194–9

63.

Onyeaso

, da Costa

, Arowojolu

. Sterilization and disinfection procedures by dental practitioners in Nigeria. Niger J Health Biomed Sci 2002;1:106–13

64.

Owoyemi

. Knowledge and Attitude of Hair-care Providers to HIV/AIDS Transmission in North Central Nigeria. XVII International Conference on AIDS, Abstract no. MOPE0337

65.

Hauri

, Hutin

, Armstrong

. Contaminated injections in health care settings. In: Ezzati

, Lopez

, Rodgers

, Murray

, eds. Comparative Quantification of Health Risks. Vol. 2. Geneva: WHO, 2004:1803–50

66.

Hutin

, Hauri

, Armstrong

. Use of injections in healthcare settings worldwide, 2000: literature review and regional estimates. BMJ 2003;327:1075

67.

Biraro

, Morison

, Nakiyingi-Miiro

, The role of vertical transmission and health care-related factors in HIV infection of children: a community study in rural Uganda. J Acquir Immune Defic Syndr 2007;44:222–8

68.

Spira

, Lepage

, Msellati

, Natural history of human immunodeficiency virus type 1 infection in children: a five-year prospective study in Rwanda. Pediatrics 1999;104:e56

69.

Visco-Comandini

. Possible child-to-mother transmission of HIV by breastfeeding. JAMA 2005;294:2301–2

70.

Hersh

, Popovici

. Acquired immunodeficiency syndrome in Romania. Lancet 1991;338:645–9

71.

Pokrovsky

, Eramova

, Kuznetsova

, Transmission of HIV-1 from infant to mother during breast-feeding (in Russian). Microbiol Epidemiol Immunobiol 1990;3:23–6

72.

Katzenstein

, Mbizvo

, Zijenah

, Serum level of maternal Human Immunodeficiency Virus (HIV) RNA, infant mortality, and vertical transmission of HIV in Zimbabwe. J Infect Dis 1999;179:1382–7

73.

Bertolli

, St. Louis

, Simonds

, Estimating the timing of mother-to-child transmission of human immunodeficiency virus in a breast-feeding population in Kinshasa, Zaire. J Infect Dis 1996;174:722–6

74.

Coutsoudis

, Pillay

, Kuhn

, Method of feeding and transmission of HIV-1 from mothers to children by 15 months of age: prospective cohort study from Durban, South Africa. AIDS 2001;15:379–87

75.

Biggar

, Miotti

, Taha

, Perinatal intervention trial in Africa: effect of a birth canal cleansing intervention to prevent HIV transmission. Lancet 1996;347:1647–50

76.

Leroy

, Karon

, Alioum

, Twenty-four month efficacy of a maternal short-course zidovudine regimen to prevent mother-to-child transmission of HIV-1 in West Africa. AIDS 2002;16:631–41

77.

Kuhn

, Bobat

, Coutsoudis

, Cesarean deliveries and maternal-infant HIV transmission: results from a prospective study in South Africa. J Acquir Immune Defic Syndr Hum Retrovirol 1996;11:478–83

78.

Gray

, McIntyre

, Lyons

. The effect of breastfeeding on vertical transmission of HIV-1 in Soweto, South Africa. Int Conf AIDS 1996;11:237, Abstract no. Th.C.415

79.

Mbori-Ngacha

, Nduati

, John

, Morbidity and mortality in breastfed and formula-fed infants of HIV-1-infected women. JAMA 2001;286:2413–20

80.

Nicoll

, Timaeus

, Kigadye

, The impact of HIV-1 infection on mortality in children under 5 years of age in sub-Saharan Africa: a demographic and epidemiologic analysis. AIDS 1994;8:995–1005

81.

Kamelan-Tano

, Boni-Ouattara

, Ekpini

, Children with early HIV-infection in Abidjan, Cote d'Ivoire: immunologic and nutritional status, mortality rates, and causes of death. Int Conf AIDS 2000;13:Abstract no. MoOrB115

82.

Becquet

, Bequet

, Ekouevi

, Two-year morbidity-mortality and alternatives to prolonged breast-feeding among children born to HIV-infected mothers in Cote d'Ivoire. PLoS Med 2007;4:e17

83.

Bobat

, Coovadia

, Moodley

, Mortality in a cohort of children born to HIV-1 infected women from Durban, South Africa. S Afr Med J 1999;89:646–8

84.

Newell

, Coovadia

, Cortina-Borja

, Mortality of infected and uninfected infants born to HIV-infected mothers in Africa: a pooled analysis. Lancet 2004;364:1236–43

85.

Ota

, O'Donovan

, Alabi

, Maternal HIV-1 and HIV-2 infection and child survival in The Gambia. AIDS 2000;14:435–9

86.

Marston

, Zaba

, Solomon

, The Net Effect of HIV on Child Mortality in Generalised HIV Epidemics. XV International Conference on AIDS, Bangkok, Thailand,. Abstract no. MoPeC3399, 2004

87.

Rosenthal

. Meta-analytic Procedures for Social Science. London: Sage, 1984

88.

Mosteller

, Bush

. Selected quantitative techniques. In: Lindzey

, ed. Handbook of Social Psychology: Vol. 1. Theory and Method. Cambridge, MA: Addison-Wesley, 1954

89.

Morgan

, Malamba

, Maude

, An HIV-1 natural history cohort and survival times in rural Uganda. AIDS 1997;11:633–40

90.

Glynn

, Sonnenberg

, Nelson

, Survival from HIV-1 seroconversion in Southern Africa: a retrospective cohort study in nearly 2000 gold-miners over 10 years of follow-up. AIDS 2007;21:625–32

91.

Anglaret

, Toure

, Gourvellec

, Impact of vital status investigation procedures on estimates of survival in cohorts of HIV-1 infected patients from sub-Saharan Africa. J Acquir Immune Defic Syndr 2004;35:3

92.

Morgan

, Mahe

, Mayanja

HIV-1 infection in rural Africa: is there a difference in median time to AIDS and survival compared with that in industrialized countries?

AIDS 2002;16:597–603