Making a working clinical diagnosis of HIV infection in infants in Zimbabwe

Addition of Questions on Parental Factors to the WHO (Integrated Management of Childhood Illnesses) IMCI-HIV Algorithm Improves the Utility of the Algorithm for Diagnosis of HIV Infection in Children

The WHO Integrated Management of Childhood Illnesses-HIV (IMCI-HIV) algorithm and its regional adaptation have shown variable performance in clinically identifying HIV-infected children with lack of validation in low prevalence areas. Addition of certain 'parental factors' (proxy indicators of parental HIV) may improve its utility. In this study, children aged 2 months to 5 years were enrolled into Group A (n = 1000, 'suspected symptomatic HIV infected' children as per the IMNCI-HIV algorithm) and group B (n = 50, children newly diagnosed with HIV infection). Parental factors were asked and HIV infection was tested for in Group A. For Group B, retrospective data were collected regarding IMNCI-HIV algorithm signs and parental factors. Utility of individual and various combinations of IMNCI-HIV signs and parental factors to predict HIV status was evaluated. Results showed that incorporating parental factors to IMNCI-HIV algorithm improved its sensitivity and positive predictive value in identifying HIV-infected children while maintaining the same sensitivity.

Temporal trends in the characteristics of children at antiretroviral therapy initiation in southern Africa: the IeDEA-SA Collaboration

BACKGROUND

Since 2005, increasing numbers of children have started antiretroviral therapy (ART) in sub-Saharan Africa and, in recent years, WHO and country treatment guidelines have recommended ART initiation for all infants and very young children, and at higher CD4 thresholds for older children. We examined temporal changes in patient and regimen characteristics at ART start using data from 12 cohorts in 4 countries participating in the IeDEA-SA collaboration.

METHODOLOGY/PRINCIPAL FINDINGS

Data from 30,300 ART-naïve children aged <16 years at ART initiation who started therapy between 2005 and 2010 were analysed. We examined changes in median values for continuous variables using the Cuzick's test for trend over time. We also examined changes in the proportions of patients with particular disease severity characteristics (expressed as a binary variable e.g. WHO Stage III/IV vs I/II) using logistic regression. Between 2005 and 2010 the number of children starting ART each year increased and median age declined from 63 months (2006) to 56 months (2010). Both the proportion of children <1 year and ≥10 years of age increased from 12 to 19% and 18 to 22% respectively. Children had less severe disease at ART initiation in later years with significant declines in the percentage with severe immunosuppression (81 to 63%), WHO Stage III/IV disease (75 to 62%), severe anemia (12 to 7%) and weight-for-age z-score<-3 (31 to 28%). Similar results were seen when restricting to infants with significant declines in the proportion with severe immunodeficiency (98 to 82%) and Stage III/IV disease (81 to 63%). First-line regimen use followed country guidelines.

CONCLUSIONS/SIGNIFICANCE

Between 2005 and 2010 increasing numbers of children have initiated ART with a decline in disease severity at start of therapy. However, even in 2010, a substantial number of infants and children started ART with advanced disease. These results highlight the importance of efforts to improve access to HIV diagnostic testing and ART in children.

Cost-effectiveness of World Health Organization 2010 guidelines for prevention of mother-to-child HIV transmission in Zimbabwe

BACKGROUND

In 2010, the World Health Organization (WHO) released revised guidelines for prevention of mother-to-child human immunodeficiency virus (HIV) transmission (PMTCT). We projected clinical impacts, costs, and cost-effectiveness of WHO-recommended PMTCT strategies in Zimbabwe.

METHODS

We used Zimbabwean data in a validated computer model to simulate a cohort of pregnant, HIV-infected women (mean age, 24 years; mean CD4 count, 451 cells/µL; subsequent 18 months of breastfeeding). We simulated guideline-concordant care for 4 PMTCT regimens: single-dose nevirapine (sdNVP); WHO-recommended Option A, WHO-recommended Option B, and Option B+ (lifelong maternal 3-drug antiretroviral therapy regardless of CD4). Outcomes included maternal and infant life expectancy (LE) and lifetime healthcare costs (2008 US dollars [USD]). Incremental cost-effectiveness ratios (ICERs, in USD per year of life saved [YLS]) were calculated from combined (maternal + infant) discounted costs and LE.

RESULTS

Replacing sdNVP with Option A increased combined maternal and infant LE from 36.97 to 37.89 years and would reduce lifetime costs from $5760 to $5710 per mother-infant pair. Compared with Option A, Option B further improved LE (38.32 years), and saved money within 4 years after delivery ($5630 per mother-infant pair). Option B+ (LE, 39.04 years; lifetime cost, $6620 per mother-infant pair) improved maternal and infant health, with an ICER of $1370 per YLS compared with Option B.

CONCLUSIONS

Replacing sdNVP with Option A or Option B will improve maternal and infant outcomes and save money; Option B increases health benefits and decreases costs compared with Option A. Option B+ further improves maternal outcomes, with an ICER (compared with Option B) similar to many current HIV-related healthcare interventions.

Nonvirologic algorithms for predicting HIV infection among HIV-exposed infants younger than 12 weeks of age

BACKGROUND

Early initiation of antiretroviral therapy has been shown to reduce mortality among perinatally HIV-infected infants, but availability of virologic testing remains limited in many settings.

METHODS

We collected cross-sectional data from mother-infant pairs in three primary care clinics in Lusaka, Zambia, to develop predictive models for HIV infection among infants younger than 12 weeks of age. We evaluated algorithm performance for all possible combinations of selected characteristics using an iterative approach. In primary analysis, we identified the model with the highest combined sensitivity and specificity.

RESULTS

Between July 2009 and May 2011, 822 eligible HIV-infected mothers and their HIV-exposed infants were enrolled; of these, 44 (5.4%) infants had HIV diagnosed. We evaluated 382,155,260 different characteristic combinations for predicting infant HIV infection. The algorithm with the highest combined sensitivity and specificity required 5 of the following 7 characteristic thresholds: infant CD8 percentage >22; infant CD4 percentage ≤44; infant weight-for-age Z score ≤0; infant CD4 ≤1600 cells/µL; infant CD8 >2200 cells/µL; maternal CD4 ≤600 cells/µL; and mother not currently using antiretroviral therapy for HIV treatment. This combination had a sensitivity of 90.3%, specificity of 78.4%, positive predictive value of 22.4%, negative predictive value of 99.2% and area under the curve of 0.844.

CONCLUSION

Predicting HIV infection in HIV-exposed infants in this age group is difficult using clinical and immunologic characteristics. Expansion of polymerase chain reaction capacity in resource-limited settings remains urgently needed.

Performance of the integrated management of childhood illness algorithm for diagnosis of HIV-1 infection among African infants

OBJECTIVES

Early infant HIV-1 diagnosis and treatment substantially improve survival. Where virologic HIV-1 testing is unavailable, integrated management of childhood illness (IMCI) clinical algorithms may be used for infant HIV-1 screening. We evaluated the performance of the 2008 WHO IMCI HIV algorithm in a cohort of HIV-exposed Kenyan infants.

METHODS

From 1999 to 2003, 444 infants had monthly clinical assessments and quarterly virologic HIV-1 testing. Using archived clinical data, IMCI sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated using virologic testing as a gold standard. Linear regression and survival analyses were used to determine the effect of age on IMCI performance and timing of diagnosis.

RESULTS

Overall IMCI sensitivity, specificity, PPV, and NPV value were 58, 87, 52, and 90%, respectively. Sensitivity (1.4%) and PPV (14%) were lowest at 1 month of age, when 81% of HIV infections already had occurred. Sensitivity increased with age (P < 0.0001), but remained low throughout infancy (range 1.4-35%). Specificity (range 97-100%) was high at each time point and was not associated with age. Fifty-eight percent of HIV-1-infected infants (50 of 86) were eventually diagnosed by IMCI, and use of IMCI was estimated to delay diagnosis in HIV-infected infants by a median of 5.9 months (P < 0.0001).

CONCLUSION

IMCI had low sensitivity during the first month of life, when the majority of HIV-1 infections had already occurred and initiation of treatment is most critical. Although sensitivity increased with age, the substantial delay in HIV-1 diagnosis using IMCI limits its utility in early infant HIV-1 diagnosis.

Discrepant test findings in early infant diagnosis of HIV in a national reference laboratory in Kenya: challenges and opportunities for programs

BACKGROUND

In Kenya, the availability of a cheap diagnostic service for HIV-exposed infants has helped scale-up access to treatment, and provided a means by which programs that support Prevention of Mother to Child Transmission of HIV can be evaluated. As expected for any large testing program, discrepant and indeterminate results present a significant challenge.

METHODS

Dried Blood Spots were collected from health centers countrywide and couriered to four laboratories for tests. Results were dispatched either by email, telephone, GSM SMS printer or courier. Between 2006 and 2009, tests were conducted with the Manual Roche v. 1.5 Assay. In 2010 the labs switched fully to the Cobas® AmpliPrep/ Cobas® TaqMan® HIV-1 Qual automated Roche Test.

RESULTS

Between 2006 and 2010, the KEMRI CVR EID Lab conducted 64 591 HIV tests in on children <18 months of age. HIV tests (38 834) used the manual assay, while 17 133 tests used the automated assay. Overall, 10.7% (6915) of the samples tested positive, while 86.6% (55 967) tested negative. A total of 1.6% (1041) tested indeterminate and required a re-bleed of the infant. Two hundred positive tests by the manual assay were retrieved randomly and retested using the automated assay. Among them, 192 (96%) remained positive, 5 (2.5%) were negative while 3 (1.5%) failed. A total of 160 negative samples by the manual assay were retrieved and retested with the automated assay. Among them, 154 (96.24%) remained negative, 3 (1.88%) tested positive while 3 (1.88%) failed. A total of 215 samples that gave indeterminate results by the manual assay were retested using the automated system. Among them, 62 (28.8%) gave positive results, 144 (66.97%) negative and 6 (2.8%) samples still gave discrepant results. Three (1.4%) did not amplify successfully. A few infants who were apparently positive appeared to test HIV negative with age.

CONCLUSIONS

Indeterminate results are a significant challenge for HIV diagnostic services, as seen in the Kenyan EID Program. In our experience, they are more often negative than they are positive. False positive and false negative results can arise from clerical error, contamination and limitations of the technologies available. To forestall the consequences of such outcomes, the sensitivity and specificity of available assays must be further improved. All HIV positive samples should be retested for confirmation, and if confirmed, a new sample must be drawn and tested for DNA at the time the infant receives their initial results or starts antiretroviral therapy. Viral clearance is a phenomenon that requires further studies.

Utility of clinical parameters to identify HIV infection in infants below ten weeks of age in South Africa: a prospective cohort study

BACKGROUND

As HIV-infected infants have high mortality, the World Health Organization now recommends initiating antiretroviral therapy as early as possible in the first year of life. However, in many settings, laboratory diagnosis of HIV in infants is not readily available. We aimed to develop a clinical algorithm for HIV presumptive diagnosis in infants < 10 weeks old using screening data from the Children with HIV Early Antiretroviral therapy (CHER) study in South Africa.HIV-infected and HIV-uninfected exposed infants < 10 weeks of age were identified through Vertical Transmission Prevention programs. Clinical and laboratory data were systematically recorded, groups were compared using Kruskal-Wallis, analysis of variance (ANOVA), and Fisher's exact tests. Receiver Operating Characteristic (ROC) curves were compiled using combinations of clinical findings.

RESULTS

417 HIV-infected and 125 HIV-exposed, uninfected infants, median age 46 days (IQR 38-55), were included. The median CD4 percentage in HIV-infected infants was 34 (IQR 28-41)%. HIV-infected infants had lower weight-for-age, more lymphadenopathy, oral thrush, and hepatomegaly than exposed uninfected infants (Adjusted Odds Ratio 0.51, 8.8, 5.6 and 23.5 respectively; p < 0.001 for all). Sensitivity of individual signs was low (< 20%) but specificity high (98-100%). If any one of oral thrush, hepatomegaly, splenomegaly, lymphadenopathy, diaper dermatitis, weight < 50(th) centile are present, sensitivity for HIV infection amongst HIV-exposed infants was 86%. These algorithms performed similarly when used to predict severe immune suppression.

CONCLUSIONS

A combination of physical findings is helpful in identifying infants most likely to be HIV-infected. This may inform management algorithms and provide guidance for focused laboratory testing in some settings, and should be further validated in these settings and elsewhere.

Early infant HIV-1 diagnosis programs in resource-limited settings: opportunities for improved outcomes and more cost-effective interventions

Early infant diagnosis (EID) of HIV-1 infection confers substantial benefits to HIV-infected and HIV-uninfected infants, to their families, and to programs providing prevention of mother-to-child transmission (PMTCT) services, but has been challenging to implement in resource-limited settings. In order to correctly inform parents/caregivers of infant infection status and link HIV-infected infants to care and treatment, a 'cascade' of events must successfully occur. A frequently cited barrier to expansion of EID programs is the cost of the required laboratory assays. However, substantial implementation barriers, as well as personnel and infrastructure requirements, exist at each step in the cascade. In this update, we review challenges to uptake at each step in the EID cascade, highlighting that even with the highest reported levels of uptake, nearly half of HIV-infected infants may not complete the cascade successfully. We next synthesize the available literature about the costs and cost effectiveness of EID programs; identify areas for future research; and place these findings within the context of the benefits and challenges to EID implementation in resource-limited settings.

Presumptive diagnosis of severe HIV infection to determine the need for antiretroviral therapy in children less than 18 months of age

OBJECTIVE

To develop a new algorithm for the presumptive diagnosis of severe disease associated with human immunodeficiency virus (HIV) infection in children less than 18 months of age for the purpose of identifying children who require antiretroviral therapy (ART).

METHODS

A conditional probability model was constructed and non-virologic parameters in various combinations were tested in a hypothetical cohort of 1000 children aged 6 weeks, 6 months and 12 months to assess the sensitivity, specificity, and positive and negative predictive values of these algorithms for identifying children in need of ART. The modelled parameters consisted of clinical criteria, rapid HIV antibody testing and CD4+ T-lymphocyte (CD4) count.

FINDINGS

In children younger than 18 months, the best-performing screening algorithm, consisting of clinical symptoms plus antibody testing plus CD4 count, showed a sensitivity ranging from 71% to 80% and a specificity ranging from 92% to 99%. Positive and negative predictive values were between 61% and 97% and between 95% and 96%, respectively. In the absence of virologic tests, this alternate algorithm for the presumptive diagnosis of severe HIV disease makes it possible to correctly initiate ART in 91% to 98% of HIV-positive children who are at highest risk of dying.

CONCLUSION

The algorithms presented in this paper have better sensitivity and specificity than clinical parameters, with or without rapid HIV testing, for the presumptive diagnosis of severe disease in HIV-positive children less than 18 months of age. If implemented, they can increase the number of HIV-positive children successfully initiated on ART.

Clinical manifestations and outcome in HIV-infected young infants presenting with acute illness in Durban, South Africa

OBJECTIVES

In young infants, early development of symptomatic HIV infection increases the risk of morbidity and mortality. A prospective study was conducted over a 1-year period in a region with a high burden of HIV in order to describe the clinical presentation of HIV infection in infants aged between 0 and 59 days on attendance at hospital and the factors associated with the need for urgent hospital management.

METHODS

Sick young infants presenting to the King Edward VIII Hospital, Durban between February 2003 and January 2004 were enrolled. After systematic evaluation by a primary health worker, an experienced paediatrician determined the primary diagnosis and need for urgent hospital management. Comparisons of these assessments were stratified by HIV status. Children were classified as HIV-uninfected (HIV ELISA-negative), HIV-exposed-but-uninfected (HIV ELISA-positive and HIV RNA PCR-negative), HIV-infected (HIV ELISA-positive and HIV viral load >400 copies/ml).

RESULTS

Of 925 infants enrolled, 652 (70·5%) had their HIV status determined: 70 (10·7%) were HIV-infected, 271 (41·6%) HIV-exposed-but-uninfected, and 311 (47·7%) HIV-uninfected. Factors associated with an increased probability of being HIV-infected included if the mother had children from more than one sexual partner, if the infant had had contact with a tuberculosis-infected person or if the HIV-infected mother and/or her exposed infant failed to receive nevirapine prophylaxis. Signs of severe illness were more frequently encountered in HIV-infected than in HIV-exposed-but-uninfected infants, including the prevalence of chest in-drawing (20·3% vs 8·8%, p = 0·004) and severe skin pustules (18·6% vs 8·6%, p = 0·01). Among infants requiring urgent hospital management, observed or reported feeding difficulties and severe skin pustules were more common in HIV-infected than uninfected infants. More HIV-infected infants (12·9%) required hospitalisation than those who were HIV-exposed-but-uninfected (7·7%) or uninfected (7·4%). Primary diagnoses of pneumonia, sepsis or oral thrush were more frequently seen in HIV-infected than exposed-but-uninfected or HIV-uninfected children.

CONCLUSION

Early recognition and triaging of infants suspected of having HIV infection provides an opportunity for early diagnosis and treatment which could prevent the adverse impact of rapidly progressive HIV disease.

Paediatric HIV management at primary care level: an evaluation of the integrated management of childhood illness (IMCI) guidelines for HIV

Background

Integrated Management of Childhood Illness (IMCI) is a WHO/UNICEF strategy to improve child survival in resource poor settings. South Africa adopted IMCI in 1997, and IMCI guidelines were adapted to include identification and management of HIV infected and exposed children. This study describes the validity of the IMCI/HIV algorithm when used by IMCI experts, the use of IMCI/HIV guidelines by IMCI trained health workers in routine clinical practice, and the burden of HIV among children under 5 years attending first level health facilities.

Methods

Seventy seven randomly selected IMCI trained health workers were observed in 74 health facilities in two provinces of South Africa. Consultations were observed with 1357 sick children; each child was reassessed by an IMCI expert to confirm the correct findings. Consent was requested for HIV testing of all children who attended with a parent or legal guardian. Positive rapid HIV tests were confirmed with HIV PCR in children aged less than 18 months. HIV positive children had a CD4 count and HIV clinical staging done.

Results

Of 1064 children with HIV results available, 76 (7.1% CI: 5.7% - 8.9%) children were confirmed HIV positive. IMCI experts using the HIV algorithm classified 54/76 (71.1% CI: 59.5%-80.9%) HIV positive children as suspected symptomatic HIV, and 15/22 remaining HIV positive children were identified as HIV exposed. Therefore, 69/76 (90.8% CI: 81.9-96.2) HIV infected children were identified by IMCI experts. No classification was made for HIV by observed health workers in 899/1357(66.2%) children.

906/1243(72.9%) mothers had been tested previously for HIV, of whom 221(24.4%) reported testing positive. Of 221 children therefore identified as HIV exposed, only 78(35.3%) had been tested for HIV within routine services.

Conclusion

The HIV algorithm is a valid tool for identifying HIV infected and exposed children when correctly and comprehensively implemented. However, it is not being used by IMCI trained health workers in routine practise, leading to a failure to implement life saving interventions.