HIV biggest cause of adolescent death and hospitalisation in Zimbabwe

Among adolescents in Harare, Zimbabwe, HIV is now the single most common cause of acute admission and in-hospital death, Rashida A Ferrand and colleagues reported in a study published in the February online edition of PLoS Medicine.

Close to fifty percent of all hospitalised adolescents were HIV-infected. Most were infected perinatally. All were severely immunosuppressed with a median CD4 cell count of 51 cells/mm³. These findings corresponded to similar studies reported in African adults before antiretroviral treatment was available.

HIV-infected adolescents were four times more likely to die in hospital than their HIV-uninfected counterparts.

In an accompanying Perspective article, Dr Gray cited the 2006 Society for Adolescent Medicine’s second position paper on the growing HIV crisis among adolescents, and noted that, according to UNAIDS, adolescents and young adults now represent half of all new infections in the developing world. In addition, a considerable number in this age group are already living with HIV.

An estimated 21% of pregnant women received an HIV test in 2008 and 45% received drugs to prevent mother-to-child transmission, of whom around one third received nevirapine (Viramune), the least effective form of preventive treatment. With an estimated 1.4 million pregnant women living with HIV in the developing world, it is not surprising that approximately 500,000 children are infected through vertical transmission each year.

HIV-infection in infants progresses rapidly leaving approximately half without access to antiretroviral treatment to die before they reach two years of age. Yet as the HIV epidemic matures, notably in sub-Saharan Africa, more children with perinatally acquired infection survive without treatment to adolescence.

Sexually acquired HIV among 15 to 24 years old in Africa is well documented. Yet little is known about the prevalence of perinatally acquired HIV and how it affects illness and death in older children and adolescents.

Dr Gray notes an earlier study of the author assessing the emerging and growing epidemic among child and adolescent survivors of mother-to-child transmission, the consequent failure to recognise its development and address the clinical needs of this population.

Estimated prevalence in South Africa in 2008 among 2 to 14 year-olds was 2.5 % (CI 95% 1.9 to 3.5) and is anticipated to grow to 3.3% in 2020. Deaths among untreated slow progressors are expected to increase from 7000 per year in 2008 in South Africa to 23,000 per year in 2030, and in Zimbabwe deaths will peak in 2014 at 9700 from 8000 in 2008.

While HIV disease varies considerably between adults and children, it is not well understood in adolescents. The authors note clinical studies of HIV infection usually break children and adults into 0 to 14- and 15 to 24-year age groups respectively, potentially missing the unique features of HIV-associated illnesses among adolescents.

The authors recruited 301 children and adolescents between the ages of 10 and 18 who were admitted for any acute illnesses to two public hospitals in Harare, Zimbabwe. All answered a questionnaire about themselves and their health. Standard tests including HIV with consent were performed.

The HIV-infected adolescents, unlike their HIV negative counterparts, were more likely to be admitted with chronic complications including stunted growth or pubertal delay, and to be a maternal orphan or have an HIV-infected mother.

70% of those infected with HIV were admitted because of tuberculosis, pneumonia, cryptococcus and septicaemia (infections of the blood stream), compared to 20% of the HIV-negative group. Among the HIV-positive group these underlying chronic conditions were associated with a four-fold increased risk of in-hospital death.

Dr Gray highlights a study in Zambia that supports these findings. Among HIV-positive children aged 1 to 14 years, a single hospitalisation for a bacterial infection increased the risk of death by 42%. A further hospitalisation doubled the risk again. She stressed the urgent need to find strategies to prevent these HIV-related deaths.

Twenty-five percent learned of their diagnosis during the study. Late diagnosis meant that most presented at an advanced stage of illness. Median age at diagnosis was 12 years. Dr Gray notes other studies that support these findings and show a delay of 3.5 years (interquartile range: 1 to 6 years) between first serious illness and diagnosis.

She adds that, as in this study, when these children are diagnosed they are already below average height and weight, are moderately or severely immunodeficient, and have often had recurrent infections as well as tuberculosis.

In resource-rich settings studies show high rates of admission among this population for psychiatric reasons, Dr Gray notes. Yet, she adds, in resource-poor settings there is little or no evidence of the psychological impact, depressive symptoms and psychiatric admissions for HIV-positive children and adolescents.

She refers to a study in Harare among HIV-infected adolescents that nonetheless showed the vast burden HIV has on the family. She highlights the need to determine the effect of death and chronic illness of a caregiver and/or siblings on the mental health of HIV-infected adolescents in southern Africa.

The benefits of early diagnosis and access to antiretroviral treatment are well-documented. The authors note that late diagnosis and delay in starting ART among children and adolescents may result in poor immune response, stunted growth as well as delayed sexual development.

Dr Gray refers to another study that compared initial responses to antiretroviral treatment in the United Kingdom, Ireland and Uganda. While early death rates after starting ART were three times higher in Uganda compared with those from the UK or Ireland, older children and adolescents showed a better virological response in Uganda. This difference was believed to be due to the successful adolescent support programme at the Mulago Hospital in Kampala, Uganda. Other data showed improved CD4 cell counts, viral loads and growth and – to a lesser degree – sexual development.

The authors note that the long-term survival in HIV-infected infants in Zimbabwe is a known fact. Yet they believe there are factors specific to this age group that are barriers to diagnostic testing. No free-standing counselling and testing sites for those under 16 years of age exist. Testing cannot be done without the permission of the legal guardian who may be unable or absent. In addition the authors found that a few of the guardians were reluctant to let the children know the true nature of their illness.

The authors suggest that health professionals be advised to offer provider-initiated testing and counselling (PITC) and to assist guardians with disclosure as a way to improve early diagnosis and adherence to ART.

Dr Gray mentions several studies that support disclosure as well as the the American Academy of Pediatrics’ recommendation of disclosure to adolescents so that they are fully informed about all aspects of their health including sexual behaviour.

Dr Gray notes, as shown in the study, sexual development among HIV-infected adolescents is delayed yet the median age of sexual debut of adolescents infected perinatally in southern Africa is unknown. Studies in the United States looking at the sexual behaviour of adolescents infected with HIV perinatally showed sexual activity ranging from 18% (mean age of 15.5 years) to 59% (mean age 18.5 years).

What is of concern, she notes, is that while 63 to 80% of perinatally infected adolescents report using condoms, knowledge about HIV transmission is low. This highlights, she adds, the need to provide risk-reduction counselling to adolescents who acquire HIV early in life.

The authors note a possible bias toward sicker patients as the study was hospital-based. One of the sites, based at a referral hospital, could have resulted in a higher number of specialist referrals for conditions such as cancer.

However, the authors state that 87% of the participants were referred from primary health care clinics or through hospital casualty departments. Since the alternative to the clinics would have been private facilities, they believe their results to be representative of the pattern of acute severe morbidity and mortality in Harare.

Zimbabwe, they note, unlike other countries in the region, has had high HIV prevalence among those attending antenatal clinics from the early 1990s. As such these numbers, among adolescents, may signal what is to come for other countries in the region.

As Dr Gray concludes, given the substantial burden among this population and evidence of significant hospital admissions and in-hospital death, “there is an urgent need for services that will be able to provide accessible and appropriate HIV testing, counselling and support, as well as facilitate access to ART and appropriate sexual risk-reduction interventions.”