FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2079253449> ?p ?o ?g. }

• W2079253449 endingPage "1469" @default.
• W2079253449 startingPage "1459" @default.
• W2079253449 abstract "Objectives Clinical algorithms can be helpful in decisions about treatment and feeding options in infancy, but have had limited exposure to real data. This analysis uses data from a large clinical trial to test such algorithms, and thereby develop a successor which performs usefully in poor countries with high HIV-prevalence. Methods The ZVITAMBO trial followed 14 110 mother-baby pairs through infancy. 32% of mothers were HIV-positive. Infants were HIV tested regularly using DNA PCR. Clinical signs were evaluated in terms of identifying HIV-infection at 6 weeks, 6 and 12 months, using Zimbabwean, South African, and WHO generic adaptations of the WHO integrated management of childhood illness HIV algorithm. A modification, in which HIV-exposed infants are first divided into being at least or less than median weight-for-age, was derived and evaluated. Results At 6 weeks 65% of all infected babies are less than median weight-for-age. Adding at least two clinical signs reduces sensitivity to 20% but those identified are 1.5 (95% CI 1.1–2.1) times more likely to die by 6 months than other infected infants. At 6 months, 86% of uninfected babies of HIV-infected mothers can be identified by selecting those whose weight is greater than median or, if less than median, who exhibit <2 clinical signs. Conclusions In poor settings a simple clinical algorithm can identify children with probable HIV infection, especially those at risk of early death, who can then be referred for further testing and care, including highly active antiretroviral therapy. Most infants who are HIV-uninfected can be identified at 6 months and provided with support to maintain infection-free survival, including focussed infant-feeding counselling. Objectifs: Développer un nouvel algorithme clinique d’aide aux décisions sur les traitements et les options d’alimentation chez les nourrissons, qui fonctionne effectivement dans les conditions à pauvres ressources avec une haute prévalence du VIH. Méthodes: Cette analyse utilise des données provenant de l’étude ZVITAMBO, qui a suivi 14110 paires mère-enfant durant la petite enfance. 32% des mères étaient VIH séropositives. Les nourrissons ont été testés régulièrement pour le VIH en utilisant la PCR de l’ADN. Les signes cliniques ont étéévalués en termes d’identification de l’infection VIH à 6 semaines, 6 et 12 mois, à l’aide des adaptations zimbabwéennes, sud africaines et génériques OMS de l’algorithme VIH de la stratégie PCIME de l’OMS. Une modification dans laquelle les nourrissons exposés au VIH sont d’abord divisés en ÿ ou en < poids pour l’âge médian, a été dérivée et évaluée. Résultats: A 6 semaines, 65% de tous les bébés infectés sont de poids < poids pour l’âge médian. L’ajout d’au moins 2 signes cliniques réduit la sensibilitéà 20%, mais ceux identifiés sont 1,5 (IC95%: 1,1-2,1) fois plus susceptibles de mourir à 6 mois comparés aux autres nourrissons infectés. A 6 mois, 86% des bébés non infectés de mères infectées par le VIH peuvent être identifiés en sélectionnant ceux dont le poids est > poids médian ou, dont le poids < poids médian mais qui présentent moins de 2 signes cliniques. Conclusions: Dans les circonstances à pauvres ressources un algorithme clinique simple peut identifier les enfants ayant une infection probable par le VIH, en particulier ceux à risque de décès prématuré, qui peuvent alors être référés pour d’autres tests et soins, y compris le traitement HAART. La plupart des enfants non infectés par le VIH peuvent être identifiés à 6 mois et offert un soutien pour maintenir une survie dénuée d’infection, y compris un conseil ciblé d’alimentation du nourrisson. Objetivos: Desarrollar un nuevo algoritmo clínico que ayude a tomar decisiones sobre el tratamiento y las opciones de alimentación en bebés y niños pequeños que sea útil en condiciones de pocos recursos y con una alta prevalencia de VIH. Métodos: Este análisis utiliza datos del ensayo de ZVITAMBO, durante el que se siguieron 14,110 pares de madre-hijo durante la niñez temprana. Un 32% de las madres eran VIH positivas. Los niños fueron testados regularmente para VIH mediante PCR de ADN. Los signos clínicos se evaluaron en términos de identificar la infección por VIH a las 6 semanas, los 6 y 12 meses, utilizando las adaptaciones de Zimbabwe, Sudáfrica y de la OMS del algoritmo WHO IMCI HIV. Se diseñó y evaluó una modificación, en la cual lo niños expuestos al VIH se dividen primero en ÿ o <mediana peso-por-edad. Resultados: A las 6 semanas, un 65% de todos los bebes infectados están por debajo de la mediana de peso-por-edad. Añadir ÿ 2 signos clínicos reduce la sensibilidad a un 20%, pero aquellos identificados tienen 1.5 (95%CI 1.1-2.1) veces más posibilidades de morir a la edad de 6 meses, comparados con otros niños infectados. A los 6 meses, un 86% de los niños no infectados de madres VIH positivas podrían identificarse, seleccionando aquellos cuyo peso está por debajo de la mediana o, si está por encima, a aquellos que presentan dos o más signos clínicos. Conclusiones: En circunstancias con pocos recursos, un simple algoritmo clínico puede identificar a niños con una probable infección por VIH, especialmente aquellos con riesgo de muerte prematura, quienes pueden entonces ser referidos para testaje y cuidados especiales, incluyendo el HAART. La mayoría de los niños que no están infectados por VIH pueden identificarse a los 6 meses y ser provistos con apoyo para mantener su status, incluyendo una alimentación focalizada. Programs for children infected with HIV are being ‘scaled up’ in Zimbabwe and elsewhere. Effective early interventions include cotrimoxazole prophylaxis, sustained breastfeeding, and early initiation of anti-retroviral (ARV) therapy (Tozzi et al. 1990; Coutsoudis et al. 2003, 2005; HIV Paediatric Prognostic Markers Collaborative Study Group 2003; Chintu et al. 2004; Chiappini et al. 2006; Violari et al. 2007). Also the chances of uninfected infants born to HIV positive mothers remaining uninfected and healthy can be maximised by focussed infant-feeding counselling. Specifically, where alternatives to breastfeeding are not ‘AFASS’ (acceptable, feasible, affordable, safe, sustainable), support for exclusive breastfeeding to 6 months is highly effective (WHO et al. 2003; Iliff et al. 2005; Bulteel & Henderson 2007). After 6 months the best feeding method for HIV-exposed infants is unknown, but a family’s decision can be better informed if it is known whether the child is infected. For infants <18 months, formal diagnosis of HIV infection requires virological tests (Moodley et al. 1995). Alternatives (total lymphocyte count, CD4/CD8 ratio, serum albumin) have been explored, either to assist in diagnosis or as proxies of disease progression, but these methods remain imprecise and may be unavailable (Mofenson et al. 2003; Zijenah et al. 2005). Selective referral of infants or their blood specimens is therefore crucial to maximise the efficiency of paediatric ARV programs. Diagnosis of illness using clinical algorithms is the basis of the integrated management of childhood illness (IMCI) initiative (Oluwole et al. 2000; Gouws et al. 2004; Armstrong Schellenberg et al. 2004a,b; El Arifeen et al. 2004; Adam et al. 2005; Chopra et al. 2005). Guidelines for identifying suspected HIV infection in IMCI have been proposed and recently elaborated by WHO (Qazi & Muhe 2006; WHO-CAH & UNICEF 2006) and two validation studies have been reported. In South Africa the sensitivity (Se) of the generic IMCI HIV algorithm among 690 children aged 2 months to 5 years presenting to a paediatric outpatient clinic was 56% and the specificity (Sp) was 85%, using RNA PCR as the gold standard (Horwood et al. 2003). Modification of the algorithm increased Se to 70% and Sp to 80%, but subsequent validation using clinical data from another hospital was less successful (Jones et al. 2005). In Ethiopia, Se of the generic IMCI algorithm in detecting ELISA or RNA PCR-confirmed HIV among 1777 children aged 2–59 months was 16% and Sp was 98% (Lulseged et al. 2004). In Zimbabwe, the Ministry of Health and Child Welfare (MoHCW) adopted a modified HIV algorithm which has not yet been evaluated. In this paper we explore the sensitivity and specificity of the WHO generic, South African, and Zimbabwean IMCI algorithms (Table 1) for identifying infants with suspected and unlikely HIV-infection at 6 weeks, 6 months, and 12 months, using data from the ZVITAMBO trial (Iliff et al. 2005; Malaba et al. 2005; Humphrey et al. 2006). Clinical data obtained during study follow-up visits were compared with HIV-PCR data. A modified HIV algorithm was developed and tested. We focussed on these three age points because of their programmatic implications: 6 weeks when decisions regarding early initiation of ARV drugs might be entertained and prophylaxis with cotrimoxazole optimised; 6 months when results could inform infant feeding decisions and counselling; and 12 months when clinical signs of HIV are likely to become more apparent. Details of the ZVITAMBO trial have been described in detail by Iliff et al. (2005); Malaba et al. (2005) and Humphrey et al. (2006). Briefly, 14 110 mother and child pairs were recruited within 96 h of delivery. Acutely ill mothers or babies, babies of birth weight <1500 g, and twins were excluded. Written informed consent was obtained. Follow-up was at 6 weeks, 3 months, and three monthly intervals thereafter to 12–24 months. Mothers were tested for HIV antibodies (ELISA +/− Western Blot) at recruitment and, if positive, retested at the next visit. Plasma and cell pellets from HIV-exposed infants were archived at all visits and the last available specimen tested [plasma by GeneScreen ELISA for samples collected ≥18 months; cell pellets by Amplicor HIV-1 DNA test version 1.5 (Roche Diagnostic Systems, Branchburg, NJ, USA) for samples collected <18 months]. If this sample was HIV-negative, the infant was classified as negative; if it was HIV positive, earlier samples were tested to determine timing of infection. RNA HIV viral load was determined by Roche Amplicor HIV-1 Monitor test, version 1.5. Analysis at each time point utilised the immediately previous classification of infant infection status. At each follow-up visit, a history was elicited by a study nurse using a structured questionnaire, for diarrhoea, cough/rapid breathing, ear discharge and exposure to tuberculosis (TB). Weight was measured using an electronic scale (Seca Model 727; Hanover, MD, USA). Enlarged lymph nodes (>0.5 cm) were identified by palpation of occipital, post-auricular, and cervical (grouped as ‘neck’), axillary, and inguinal sites. Oral thrush was noted. Zimbabwean nurses are trained and familiar with IMCI but no specific additional training in this approach was given to study staff. Study staff were blind to the mothers’ HIV status, except in the case of a counsellor discussing a mother’s test result with her, on her request and in confidence. All data were managed in dBase PLUS Version 2.01 (dataBased Intelligence, Inc., 2548 Vestal Parkway East, Vestal, NY 13850, USA) and statistical analyses performed in Stata Version 8.2 (Stata Corp LP, College Station, TX, USA). Sensitivity, specificity and predictive values were calculated using diagt package (package ID sbe36_2) for each follow-up time using 2, 3, and ≥4 of the diagnostic criteria defined by each algorithm (see http://www.stata-journal.com/software/sj4-4). Logistic regression was used to estimate univariate and multivariate odds ratios (95% CI) of HIV infection at each follow-up time, given the presence of each criterion. A new algorithm was derived: Bootstrap stepwise logistic regression modelling using the swboot package for Stata 8 was used to select predictors (Efron & Tibshirani 1986). Using simple random sampling with replacement, 100 samples of 1000 infants each were selected from the full dataset and a stepwise logistic regression model (exclusion α = 0.1 and retention α = 0.05) was run, offering all diagnostic criteria. A summary of the frequency with which each criterion remained significant in the model was used to select the group of criteria that was most predictive of HIV infection at each follow up time. Logistic models for predicting infant HIV infection were thereby selected for infants aged: (i) throughout infancy; (ii) in the first half of infancy, and (iii) in the second half of infancy. The age-specific models were then tested against the all-ages model using Receiver Operating Characteristic curves. The age-specific models performed slightly better but not statistically significantly so. For simplicity, the all-ages model was selected for developing the final algorithm. The model was validated using cross-validation techniques whereby half the data were used to train the model that was then applied to the other half. The process was then switched (Dahyot 2005). Correct classification varied from 91% (95% CI: 83.6%–95.8%) at 6 weeks to 95% (88.7%–98.4%) at 12 months. The final proposed algorithm (see Table 1) was tested and performed with little difference at all ages, with or without screening signs. Hence, in the final version, screening questions were not used. Z-scores of weight-for-age, length-for-age, and weight-for-length were investigated as additional criteria for predicting infant HIV infection using the new WHO reference standards (de Onis et al. 2006). Since parameters are given for whole months only, infant age was first rounded to the nearest month and corresponding parameters selected from the tables. The 3rd and 50th weight centiles, and weight-for-age and length-for-age Z-scores were then calculated. Because tests may perform differently in populations with different prevalences, the proposed algorithm was evaluated in three groups: infants of HIV positive mothers (high prevalence: 17.0–25.4%), infants of all mothers (medium prevalence: 6.0–8.4%), and an artificially constructed group comprising all infants of HIV negative mothers plus a 15% random sample of those born to HIV+ mothers (low prevalence: 1.3–2.3%). The WHO generic algorithm includes maternal HIV status as a clinical criterion, unlike the South African and Zimbabwean versions. With national roll-outs of PMTCT and Voluntary Counselling and Testing an increasing proportion of mothers visiting peripheral health care facilities will have had, or will be able to have, their HIV-status determined. We therefore report on the scenario in which the mother is known to be HIV positive, the algorithms being applied to HIV-exposed infants only. The Medical Research Council of Zimbabwe, the Medicines Control Authority of Zimbabwe, the Johns Hopkins Bloomberg School of Public Health Committee on Human Research, and the Montreal General Hospital Research Ethics Committee approved the ZVITAMBO trial protocol. Of the 14 110 mothers enroled into the ZVITAMBO trial, 4495 tested HIV-positive, 9560 tested HIV-negative, and 55 tested indeterminate at baseline (see Figure 1 which also details data missing at each time point). For this analysis, the 293 baseline-negative mothers who sero-converted during the first year post-partum were included with the other baseline-negative mothers, as were the 53 baseline-indeterminate mothers. 12.9% (580) of the HIV-exposed infants died during the first year; 1.7% (164) of the infants born to baseline HIV-negative or indeterminate mothers died. Approximately 70% (95% CI: 67.5–74.2) of all infants not known to have died attended at each visit. Flow diagram of infants assessed for human immunodeficiency virus (HIV) at 6 weeks, 6, and 12 months. Flow chart of infants assessed for HIV using clinical signs according to WHO generic, RSA and Zimbabwean adaptations of the integrated management of childhood illness algorithms for HIV diagnosis. Assessments were made at age 6 weeks, 3 months and at three monthly intervals thereafter up to 1 year. The study population comprised women who were urban/peri-urban, mostly young (mean age, SD: 24.5, 5.3) and married. They were well educated (82% with some secondary schooling), many were poor but there was a wide spread of family income (interquartile range: 50–137 US$/month). Table 2 summarises the performance of each algorithm at different ages. This table was constructed as if HIV-infected mothers, tested either previously (e.g. in a PMTCT program) or on the spot by ‘Rapid’ testing, had presented with their babies to a health centre for ‘well-baby’ care, the infants being then assessed clinically according to the IMCI algorithms. The sensitivity of the South African algorithm was higher than either the WHO or Zimbabwean versions; specificity and positive predictive value (PPV) were similar and higher in WHO and Zimbabwean versions. Using the South African algorithm (the most sensitive of these three), only 23% of HIV-infected infants would be identified at 6 weeks when decisions for early initiation of ARV therapy might be made, although 73% would be picked up at 12 months. Using a cut-off of two clinical signs in the more specific algorithms, nearly 90% of HIV-negative children would be identified at 6 months and one could be 86% sure that a child classified as ‘negative’ was in fact not infected (NPV). When explored with logistic regression, all signs except for ‘Persistent diarrhoea’ showed significant odds ratios (ORs) for infant HIV infection (Table 3). The utility of ‘Pneumonia’ declined after 3 months. ‘Diarrhoea’ was the sign that gave the South African algorithm its higher sensitivity. ‘Ear discharge’, ‘enlarged lymph nodes’ and ‘oral thrush’ all returned high ORs, as did ‘TB’ (defined as a positive answer to either ‘Has the child ever been on treatment for TB?’ or ‘Has the child been in contact with anyone with TB?’). ‘Low weight’, i.e. weight <3rd centile with or without growth faltering (zero/negative weight gain over ≥3 months) was also associated with high ORs. Our proposed new algorithm is shown in Table 1. All three pre-existing algorithms use screening questions to initiate the process of counting signs and symptoms. We tested ours with and without screening questions and found little difference in performance (data not shown). However, to minimise the numbers of children to whom a health-worker would need to apply the full clinical algorithm, a simple screening approach using median weight-for-age (Phase 1) as a threshold (Z-score = 0 using the new WHO standards) was applied. Only those infants below the median are subjected to the clinical algorithm (Phase 2). Other Z-score values were explored as potential screening criteria (as were ‘length-for-age’ and ‘weight-for-length’) but none provided added benefit (data not shown). Table 2 demonstrates the performance of our proposed algorithm in comparison with the other three. Greatly increased sensitivity can be achieved with the ‘Phase 1’ screen (dividing infants into those above and below the median weight-for-age) although inevitably at the expense of specificity and PPV. Application of the modified clinical algorithm (Phase 2) then brings specificity back to levels similar to those achieved by the other systems. Referring all HIV-exposed infants for virological testing will achieve 100% sensitivity, but this entails referral of nearly a quarter of all babies (depending on the HIV prevalence among mothers). This proportion can be reduced by approximately half if <median weight-for age is used as a threshold for confirmatory testing, while retaining 65, 79 and 86% sensitivity at 6 weeks, 6 months and 12 months respectively. Adding the two clinical signs criterion drops the percentage of all infants selected for referral to between 2 and 13%. Of infants who were PCR+ at 6 weeks, selection using our algorithm (set at <median weight-for-age +2 clinical signs) and viral load are independently associated with mortality at 6 months. Using Cox models, the univariate hazard ratio for mortality associated with being ‘algorithm positive’ is 1.51 (95% CI 1.08–2.10, P = 0.015). If viral load is included the hazard ratio remains at 1.49 (1.07–2.08, P = 0.017). Differences among the three groups of infants with different prevalences of HIV were unremarkable (data not shown), implying that the algorithms would perform similarly in differing settings, e.g. both well-baby and outpatient or casualty departments. All families in Zimbabwe are affected by HIV. The MoHCW estimated that 18% of mothers attending for maternity care in 2006 were infected with HIV (Ministry of Health and Child Welfare of Zimbabwe: AIDS & TB Programmes 2006). PMTCT programs have spread throughout the country and antiretroviral treatment is underway at several sites, urban and rural but management of children presents many challenges. Unlike previous assessments of clinical algorithms for the identification of HIV infection in children, this analysis specifically addresses infancy, the time when formal laboratory diagnosis of HIV is difficult, and yet there are critical decisions to be made by families and by those caring for them. Women who test HIV positive in PMTCT programs are faced with a dilemma about how to feed their babies (Humphrey & Iliff 2001). Now that the dangers of early introduction of mixed feeds have been established, coupled with the difficulties of making an ‘AFASS’ choice for exclusive replacement feeding, most mothers in Zimbabwe choose exclusive breastfeeding. However, at 6 months breastmilk alone is insufficient and a second decision is required which will be greatly influenced by whether or not the baby is infected with HIV. The mother of an uninfected infant may decide that exclusive replacement feeding is more ‘AFASS’ than at the time of delivery. A 6 month-old HIV-exposed baby who is either above median weight-for-age or, if below, exhibits <2 signs of the proposed clinical algorithm, will be very likely to be uninfected (NPV: 86%). Most uninfected babies can be identified in this way (Sp: 84%). HIV-infected infants sustain very high early mortality, probably largely from Pneumocystis pneumonia, much of which can be prevented by cotrimoxazole prophylaxis (Chintu et al. 2002, 2004; Ruffini & Madhi 2002). WHO and MoHCW guidelines recommend giving cotrimoxazole to all ‘exposed’ infants (those with HIV infected mothers) from the age of 4–6 weeks until HIV infection has been definitely ruled out and the mother is no longer breastfeeding (WHO/UNICEF/UNAIDS 2004). Identification on clinical grounds of children who are likely to be infected with HIV and at high risk of early mortality, enables targeted, more intense (and expensive) follow-up (e.g. using home visits) to ensure maintenance of cotrimoxazole supplies and adherence to prophylaxis. Access for children to HIV treatment remains limited. Identifying infected children early and initiating treatment before severe clinical deterioration has occurred is important (Violari et al. 2007), but the need for virological testing in children <18 months makes this difficult and expensive. However it is possible for programs to use clinical criteria, as assessed in this paper, for selective referral of babies or their blood specimens to central units that have the necessary capacity. Our proposed clinical algorithm preferentially selects HIV-infected babies at higher risk of early mortality, increasing the potential gains in cost-effectiveness. At 6 weeks, nearly two thirds of all infected infants would be identified by simple application of the median weight-for-age screen, halving the number of PCR tests required to diagnose HIV infection in babies born to HIV-infected mothers. Although only 29% (PPV) of these babies will actually be infected, this may be considered satisfactory for the purpose of, for example, sending dried blood spots to a central laboratory for PCR, or for focussed efforts to enhance compliance with cotrimoxazole prophylaxis through outreach visits for defaulters. At age 1 year, 68% of the infected babies of HIV positive mothers will be both ≤median weight-for-age and have ≥2 clinical signs (PPV 54%). Using this approach (rather than testing all exposed babies) would again reduce the number of PCR tests by more than 50%. Program managers can make selection more rigorous by, for example, insisting on ≥4 signs for referral which will give 73% PPV (although at the cost of reducing sensitivity to 18%). Modification and/or re-printing of the Child Health Card provides an opportunity for inscribing the median (50th centile) weight-for-age curve in addition to the traditional 3rd centile. Alternatively, copies of the 3rd and 50th centile curves (available from the corresponding author if requested) can be distributed for posting in appropriate clinical areas. The recent occurrence of diarrhoea added important sensitivity to the algorithm, but ‘Persistent diarrhoea’ was not a useful discriminator in these young infants. ‘Pneumonia’ was important at 6 weeks though less so later, but ‘ear discharge’ showed the reverse pattern with age. ‘TB’ also became more useful in the older infants. We investigated whether the axilla, rather than the conventional combination of any two sites (neck, groin or axilla) could stand alone as the significant site for lymphadenopathy in the context of identifying HIV infection (data not shown) but this resulted in loss of both sensitivity and specificity. A recent report from South Africa confirms the difficulty of achieving high sensitivity with clinical signs in young infants, although very high specificity may be obtained if signs such as gastro-oesophageal reflux disease, splenomegaly or hepatomegaly are used (Jaspan et al. 2008). Two specific programmatic scenarios are presented as decision flow-charts in Figure 2 (selection for referral for confirmation of HIV infection) and Figure 3 (selection for detailed counselling as to feeding practice at 6 months). Flow chart to select infants (or their blood specimens) for referral for formal virological diagnosis. A ‘Low cost/Low sensitivity/High specificity’ program can be converted to a ‘very low cost/very low sensitivity/very high specificity’ program by increasing the cut-off number of clinical signs to three (or four). Flow chart to select breastfed infants for counselling about rapid weaning at 6 months. This analysis was based on data derived from a clinical trial based in Harare, before the initiation of the national PMTCT program. We cannot know how applicable the findings are to other settings until appropriate field testing is undertaken. However, finding little difference between groups of babies of differing HIV prevalence gives some confidence that the proposed algorithm will perform well in different populations, although it is possible that in situations where most women receive prophylaxis against MTCT, babies who nevertheless become infected may show a different clinical course. The application of a simple clinical algorithm enables selection of probably HIV-infected infants at high risk of early mortality, for focussed attention including cotrimoxazole prophylaxis and referral for possible early initiation of ARV therapy. Infants who are probably uninfected can also be identified for appropriately targeted support, especially counselling on feeding practice. The ZVITAMBO Project was supported by the Canadian International Development Agency (CIDA) (R/C Project 690/M3688), United States Agency for International Development (USAID) (Cooperative Agreement Number HRN-A-00-97-00015-00 between Johns Hopkins University and the Office of Health and Nutrition, USAID). Additional funding was received from the SARA Project, which is operated by the Academy for Educational Development, Washington DC and is funded by the USAID Bureau for Africa, Office of Sustainable Development under the terms of Contract AOT-C-00-99-00237-00, the Rockefeller Foundation (NY, NY) and BASF (Ludwigshafen Germany)." @default.
• W2079253449 created "2016-06-24" @default.
• W2079253449 creator A5003718468 @default.
• W2079253449 creator A5008743607 @default.
• W2079253449 creator A5030554910 @default.
• W2079253449 creator A5030805938 @default.
• W2079253449 creator A5082657568 @default.
• W2079253449 creator A5083800348 @default.
• W2079253449 date "2008-12-01" @default.
• W2079253449 modified "2023-10-05" @default.
• W2079253449 title "Making a working clinical diagnosis of HIV infection in infants in Zimbabwe" @default.
• W2079253449 cites W1571154045 @default.
• W2079253449 cites W1576239543 @default.
• W2079253449 cites W1719492433 @default.
• W2079253449 cites W1969671892 @default.
• W2079253449 cites W2018659533 @default.
• W2079253449 cites W2020178995 @default.
• W2079253449 cites W2027342539 @default.
• W2079253449 cites W2035391525 @default.
• W2079253449 cites W2040151781 @default.
• W2079253449 cites W2047219499 @default.
• W2079253449 cites W2054991082 @default.
• W2079253449 cites W2076553898 @default.
• W2079253449 cites W2082229328 @default.
• W2079253449 cites W2095956016 @default.
• W2079253449 cites W2111322705 @default.
• W2079253449 cites W2112688502 @default.
• W2079253449 cites W2124919869 @default.
• W2079253449 cites W2140469094 @default.
• W2079253449 cites W2144275513 @default.
• W2079253449 cites W2155514999 @default.
• W2079253449 cites W2162881343 @default.
• W2079253449 cites W2164633309 @default.
• W2079253449 doi "https://doi.org/10.1111/j.1365-3156.2008.02178.x" @default.
• W2079253449 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/19055624" @default.
• W2079253449 hasPublicationYear "2008" @default.
• W2079253449 type Work @default.
• W2079253449 sameAs 2079253449 @default.
• W2079253449 citedByCount "14" @default.
• W2079253449 countsByYear W20792534492012 @default.
• W2079253449 countsByYear W20792534492013 @default.
• W2079253449 countsByYear W20792534492018 @default.
• W2079253449 countsByYear W20792534492020 @default.
• W2079253449 crossrefType "journal-article" @default.
• W2079253449 hasAuthorship W2079253449A5003718468 @default.
• W2079253449 hasAuthorship W2079253449A5008743607 @default.
• W2079253449 hasAuthorship W2079253449A5030554910 @default.
• W2079253449 hasAuthorship W2079253449A5030805938 @default.
• W2079253449 hasAuthorship W2079253449A5082657568 @default.
• W2079253449 hasAuthorship W2079253449A5083800348 @default.
• W2079253449 hasBestOaLocation W20792534491 @default.
• W2079253449 hasConcept C159047783 @default.
• W2079253449 hasConcept C187212893 @default.
• W2079253449 hasConcept C3013748606 @default.
• W2079253449 hasConcept C71924100 @default.
• W2079253449 hasConceptScore W2079253449C159047783 @default.
• W2079253449 hasConceptScore W2079253449C187212893 @default.
• W2079253449 hasConceptScore W2079253449C3013748606 @default.
• W2079253449 hasConceptScore W2079253449C71924100 @default.
• W2079253449 hasIssue "12" @default.
• W2079253449 hasLocation W20792534491 @default.
• W2079253449 hasLocation W20792534492 @default.
• W2079253449 hasOpenAccess W2079253449 @default.
• W2079253449 hasPrimaryLocation W20792534491 @default.
• W2079253449 hasRelatedWork W1506200166 @default.
• W2079253449 hasRelatedWork W1973460911 @default.
• W2079253449 hasRelatedWork W1995515455 @default.
• W2079253449 hasRelatedWork W2080531066 @default.
• W2079253449 hasRelatedWork W2096903645 @default.
• W2079253449 hasRelatedWork W2748952813 @default.
• W2079253449 hasRelatedWork W2899084033 @default.
• W2079253449 hasRelatedWork W3031052312 @default.
• W2079253449 hasRelatedWork W3032375762 @default.
• W2079253449 hasRelatedWork W3108674512 @default.
• W2079253449 hasVolume "13" @default.
• W2079253449 isParatext "false" @default.
• W2079253449 isRetracted "false" @default.
• W2079253449 magId "2079253449" @default.
• W2079253449 workType "article" @default.