NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

WHO Recommendations on the Diagnosis of HIV Infection in Infants and Children. Geneva: World Health Organization; 2010.

Cover of WHO Recommendations on the Diagnosis of HIV Infection in Infants and Children

WHO Recommendations on the Diagnosis of HIV Infection in Infants and Children.

Show details

ANNEX 3Summary of guideline group decisions regarding evidence and risk–benefit analyses

Risk–benefit analyses for recommendations 1 and 2 are in the body of the report and tables. Recommendation 3 is sourced already.

Recommendation 4
In infants and children undergoing viral testing, the following assays are strongly recommended for use:
  • HIV DNA in whole blood
  • HIV DNA in DBS
  • HIV RNA in plasma (except for infants on ART)
  • HIV RNA in DBS (except for infants on ART)
  • Us p24 HIV Ag in plasma (except for infants on ART or where subtype D is common)
  • Us p24 HIV Ag in DBS (except for infants on ART or where subtype D is common)
Population:
Infants (<18 months) exposed to HIV
Intervention:
Virological testing with assays of at least 98% Sn and 99% Sp
Action:
If reactive and <12 months, start treatment and perform confirmatory test at the initiation of ART.
If not reactive, HIV infection is unlikely unless still breastfeeding.
FactorDecisionExplanation
Quality of evidenceStrongPublished evidence:
HIV DNA in whole blood: GRADE evidence HIGH
HIV DNA in DBS: GRADE evidence HIGH
HIV RNA in plasma: GRADE evidence HIGH
HIV RNA in DBS: GRADE evidence HIGH
HIV p24 Ag in plasma: GRADE evidence: HIGH
HIV p24 Ag in DBS: GRADE evidence: HIGH
Concerns exist about the performance of RNA and Us p24 Ag:
  • in the context of PMTCT-related ARV exposure, however, currently there is no evidence to suggest it is a problem
  • regarding thresholds for the use of the quantitative assay, particularly for DBS samples
  • where subtype D is prevalent for p24 Ag
  • b-DNA test is not recommended as Sp is less than 98%
  • as most DBS RNA assays also detect DNA
  • as quantitative RNA is also useful for clinical management.
HIV DNA testing will be needed to resolve discordant HIV DNA and HIV RNA test results once on ART.
Benefits or desired effectsStrong (benefits outweigh risks)Early diagnosis
Use of DBS facilitates early diagnosis in remote settings.
Risks or undesired effectsLack of standardization and quality assurance could affect assay performance.
The performance of some assays might be subtype dependent.
Laboratories will not know whether patient has been exposed to ARVs.
Values and preferencesStrongUs p24 Ag: currently kits are not commercially packed together affecting laboratory supplies and delivery.
The use of DBS sample allows easy collection and transport.
Costs and feasibilityWeakConcerns about sustainability of existing RNA commercial platforms
May be more easily available and more useful in the long term
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 5
It is strongly recommended that all HIV-exposed infants have HIV virological testing at 4–6 weeks of age or at the earliest opportunity thereafter.
Population:
All HIV-exposed infants (<18 months of age)
Intervention:
Virological testing at 4–6 weeks
Action:
If reactive and <12 months, start treatment and send confirmatory test.
If non-reactive, HIV infection unlikely, continue follow up and perform HIV serological test at 9–18 months of age.
FactorDecisionExplanation
Quality of evidenceStrongNatural history and mortality data support early ART initiation.
The CHER study1 reported 75% reduction in mortality with early diagnosis and immediate treatment regardless of clinical, immunological and virological stage.
Benefits or desired effectsStrongEarly diagnosis improves infant follow up and HIV care.
Early treatment reduces infant mortality and morbidity.
Modelling suggests that 4–6 weeks is optimal to increase detection rates and avoid increasing HIV-attributable mortality.
Risks or undesired effectsConcerns about mislabelling the specimen and other human errors (>5%)
Infants may be tested but unable to access ART.
Other aspects of the child health programme may suffer.
Values and preferencesStrongEthical obligation to identify and treat early if prevention failed Denial, neglect of child and stigma enhanced by identification of HIV in infant
Costs and feasibilityWeakIncreased by:
Health system constraints in collecting specimen and returning results
Lack of sufficient capacity/infrastructure to do a different or second test in many countries
Emotional costs for families and staff
Reduced by:
Avoiding morbidity-related costs
Processing a large number of specimens as unit cost per test decreases with increasing numbers of specimens processed
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 6
In infants with an initial positive virological test result, ART should be started without delay and a second specimen collected at the same time to confirm the initial positive virological test result. Do not delay ART. Initiation of ART saves lives and ART should not be delayed while waiting for the results of the confirmatory test.
Population:
Infants and children with a first positive virological test result
Intervention:
Confirmatory test on a separate specimen preferably taken at or just before starting ART
Action:
If reactive, continue ART.
If not reactive, a third test will be required to resolve the discordance between the two earlier viral tests.
FactorDecisionExplanation
Quality of evidenceStrongSerial testing with tests that have at least 95% Sn minimize the chances of FP results
Depends on prevalence in population being tested
Natural history and mortality data support early ART initiation.
In the CHER study,1 16% versus 4% deaths were deferred with immediate ART; 75% reported reduction in mortality with early diagnosis and immediate treatment of infants who were well at diagnosis.
Benefits or desired effectsStrong (benefits outweigh risks)Using a separate sample reduces human error (e.g. mislabelling of first specimen, etc.).
Early diagnosis improves infant follow up and HIV care. Confirmatory testing reduces FP results and unnecessary ART. Early treatment reduces infant mortality and morbidity.
Risks or undesired effectsConcerns about mislabelling the specimen and other human errors (estimate may be as high as 5%)
Infants may be tested but are often unable to access ART in a timely fashion (e.g. <25% access reported by Clinton Foundation).
Other aspects of the child health programme may suffer due to repeat infant testing (human and financial resources).
Values and preferencesStrongEthical obligation to identify and treat early if prevention failed, as well as not treat unnecessarily
Denial, neglect of child and stigma enhanced by identification of HIV in infant, therefore exclusion of FP results has great value
The importance of a confirmatory test is greater in low-prevalence settings. For example, if using tests with 99% Sn and 98% Sp the following would be found:
30% prevalence: treat 311 for 297 truly infected: 14/311 treated unnecessarily
5% prevalence: treat 137 for 55 truly infected : 38/137 treated unnecessarily
1% prevalence: treat 3 for 1 truly infected: 2/3 treated unnecessarily
Costs and feasibilityStrong (despite overall increase in cost)Increased by:
Health system constraints in collecting specimen and disseminating results
Lack of sufficient capacity/infrastructure in many countries to do a different or second test
Emotional costs for families and staff
Reduced by:
Interruption of unnecessary treatment and care for those who have FP results
Unit cost for per test decreases with increasing numbers of specimens processed
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 7
Access to virological testing in infants should be accompanied by the test results being returned to the clinic and child/mother/carer as soon as possible, but at the latest within four weeks of specimen collection. Positive results should be fast tracked to the mother/baby pair as soon as possible to enable prompt initiation of ART.
Population:
All infants undergoing virological testing
Intervention:
Test result returned at the latest within four weeks of specimen collection
Action
If results are delayed, priority should be given to tracking and getting results to caregiver/infant to enable prompt initiation of ART for infected children.
FactorDecisionExplanation
Quality of evidenceStrongDelays in specimen handling and return of test results causes critical delays in confirming HIV infection and commencing life-saving ART and HIV care.
Natural history and mortality data support early ART initiation. The CHER study1 reported 75% reduction in mortality with early diagnosis and immediate treatment regardless of clinical, immunological and virological stage.
Benefits or desired effectsStrongTo avoid death and make testing useful, results need to be delivered and care started in as short a time as possible. Most deaths in the CHER study1 were early, sudden and without symptoms. Testing uptake may improve if results are given quickly.
Risks or undesired effectsMaintaining confidentiality
May decrease rates of offering testing if health-care workers are aware of the need to give results quickly and see this as increasing their work
Values and preferencesStrongMay improve health systems' ability to ensure children get into care Mothers and families likely to prefer quicker results
Costs and feasibilityStrongCosts might be reduced as morbidity-related costs are avoided.
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 8
It is strongly recommended that all infants with unknown or uncertain HIV exposure being seen in healthcare facilities at or around birth or at the first postnatal visit (usually 4–6 weeks) or other child health visit have their HIV exposure ascertained.
Population:
Infants with unknown or uncertain HIV exposure being seen in health-care facilities at the first postnatal or other child health visit
Intervention:
  1. Determine HIV status of the mother in this pregnancy through review of records, maternal or caregiver questioning (STRONG recommendation).
  2. If maternal HIV testing has not been done or the HIV status of the mother remains unclear for the duration of the pregnancy, an HIV serological test in the mother should be performed after obtaining informed consent (STRONG recommendation).
  3. If the mother is unavailable or does not consent to testing, a single HIV serological test in the infant should be performed to detect HIV exposure (STRONG recommendation).
Action:
  • If the infant is seen <72 hours after delivery and exposure documented, post-exposure prophylaxis (PEP) should be given, and mothers should be counselled on safe infant-feeding practices according to the national/local recommendations.
  • For infants first seen at 4–6 weeks or the earliest thereafter and in whom HIV exposure is documented, HIV virological testing should be performed and the mother should receive safe infant-feeding counselling.
FactorDecisionExplanation
Quality of evidenceStrongEffective interventions to prevent new infection in the mother and infant are available.
Interventions save lives and reduce mortality and hospitalization.
A negative test excludes HIV exposure.
The mother needs care and treatment.
Benefits or desired effectsStrong (benefits outweigh risks)Prefer to confirm HIV exposure by assessment of the mother or testing
In HIV exposed, cotrimoxazole saves lives.
An appropriate infant-feeding choice can be offered.
Benefits, especially in high-prevalence settings
Facilitates a higher detection of exposed infants
Opportunity for diagnosis of HIV infection and care for the mothers
Risks or undesired effectsCaution if older infant and breastfeeding or other exposure not known or recent
May miss incident HIV infection (window period)
May need to retest mother in high-incidence areas (due to infection acquired late in pregnancy)
A positive HIV serological test may be misinterpreted by the mother as the child being HIV infected
In low-prevalence settings there is a higher likelihood that a positive result is an FP result.
Values and preferencesWeakPotential harms – FP results and associated anxiety
Prevalence in children will determine how acceptable testing is to children or their parents.
Health-care workers might be reluctant to test because of increased workload.
Mothers may be more willing to accept infant testing than maternal testing.
Costs and feasibilityWeakIn low-prevalence areas the value of detecting rare conditions should be balanced against the costs.
Opportunity costs are higher in low-prevalence settings.
Retesting cost needs to be considered (tested early in pregnancy).
Overall ranking of recommendationStrength Of recommendation
Strong
Recommendation 9
It is strongly recommended that HIV-exposed infants who are well have HIV serological testing at around 9 months of age (or at the time of the last immunization visit). Infants who have reactive serological assays at 9 months should have a viral test to identify infected infants who need ART.
Population:
HIV-exposed children who are well and aged 9–18 months
Intervention:
HIV serological testing at 9 months
Action:
If non-reactive and no exposure via breastfeeding within the past six weeks, infant uninfected; discontinue co-trimoxazole and discharge from programme
If reactive and infant is well, repeat at 18 months of age. If infant is sick, perform virological test if available.
FactorDecisionExplanation
Quality of evidenceQuality is low but felt could support a strong recommendationAt 9 months of age, maternal HIV antibodies are no longer detectable in between 40% and 50% of HIV-exposed children.
Benefits or desired effectsBenefits outweigh risksLeads to early exclusion of HIV where possible
Follow up for HIV and cotrimoxazole can be stopped.
Virological testing for seropositive infants will help identify infected infants early.
Risks or undesired effectsIf very recent exposure to HIV cannot be ruled out, a negative serological test does not exclude HIV infection (window period).
Values and preferencesStrongReduces the mother's anxiety regarding the child being infected
Costs and feasibilityStrongAvoids HIV care-related costs for HIV-exposed but –uninfected child.
Inexpensive intervention
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 10
It is strongly recommended that infants with signs or symptoms suggestive of HIV infection should have HIV serological testing and, if positive (reactive), virological testing.
Population:
Sick infants and children (4 weeks–9 months of age)
Intervention:
Serological testing followed by virological testing when positive
FactorDecisionExplanation
Quality of evidenceWeakThere are concerns as not all rapid tests can be used in this case.
Serological assays used must be highly sensitive, e.g. current Determine test. Further research is necessary.
Impaired immune system of the baby could reduce HIV antibody detection in the baby.
Persisting maternal antibodies could affect the serological test result.
The performance of the clinical algorithm depends on the healthcare worker.
Benefits or desired effectsWeakIn sick children who are HIV infected, the earliest possible start of HIV treatment and care reduces mortality.
Risks or undesired effectsPossible FN results (see above)
FP results, especially in settings where the result cannot be further confirmed.
Values and preferencesStrongIf clinical suspicion is still high, may need to treat while seeking further testing
HIV infection is an unlikely cause of signs and symptoms such as cough fever, diarrhoea, pneumonia, tuberculosis in low-prevalence settings. However, in settings where the prevalence is >5%, HIV infection is a more likely cause of such signs and symptoms.
Costs and feasibilityStrongDepends on the HIV prevalence
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 11
In breastfeeding infants or children, it is strongly recommended that breastfeeding not be discontinued before performing any kind of diagnostic HIV test.
Population:
HIV-exposed infants who are breastfeeding.
Intervention:
Do not discontinue breastfeeding at the time of virological or HIV serological testing.
Action:
Breastfeeding should be continued for HIV-infected infants and ART started.
In infants who have negative results, safer infant-feeding options will need to be considered.
FactorDecisionExplanation
Quality of evidenceStrongTransmission rates through breastfeeding are established:
Mixed feeding 0–6 months: 1.5%/month
Exclusive breastfeeding 0–6 months: 0.75%/month
Breastfeeding 7–36 months: 0.75%/month
Breastfeeding mother on ART: 0.3%/month
Mixed feeding and abrupt cessation of breastfeeding are recognized to increase the risks of transmission.
Benefits or desired effects versus risks or undesired effectsBenefits greater than risksThe risk of transmission due to breastfeeding is known.
The risk of death or morbidity for the nonbreastfed, HIV-infected or HIV-exposed infant is increased.
Very abrupt cessation of breastfeeding leads to risk for the HIV-infected or -uninfected infant and is therefore undesirable.
Infant-feeding decisions are more easily made if based on the results of testing, and not prior to receiving the results. May increase the rates of returning for results
Values and preferencesStrongMay be more acceptable to mothers
Some health-care workers feel breastfeeding at any time is unacceptable for HIV-positive women.
Costs and feasibilityStrongMay reduce health systems costs due to infant morbidity, and reduce mortality in infected infants
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 12
It is strongly recommended that children (18 months or older) with suspected HIV infection or HIV exposure have HIV serological testing performed according to the standard diagnostic HIV serological algorithm used in adults.
Population:
All children 18 months or older suspected to be HIV infected or recommended to have HIV testing based on other clinical or exposure criteria
Intervention:
HIV serological testing according to the diagnostic algorithm
Action:
If reactive, confirms HIV infection; provide HIV care and treatment.
If non-reactive, excludes HIV; discharge from programme.
FactorDecisionExplanation
Quality of evidenceStrongThere is no evidence to suggest that HIV serological testing performs differently in children.
Concerns about time to seroreversion in breastfeeding children Concerns about fourth-generation assays
Benefits or desired effectsBenefits outweigh risksAn HIV-positive result is most likely to indicate infection; young children need access to the full package of care as soon as possible.
Risks or undesired effectsFP results more likely if fourth-generation assays used as very sensitive
Values and preferencesStrongAcceptable to patients
Health-care providers at all levels have been shown to be able to use simple rapid tests.
Biggest fear is taking blood and giving results to parents
Costs and feasibilityStrongSimple inexpensive intervention
Overall ranking of recommendationStrength of recommendation
Strong
Recommendation 13
In sick infants in whom HIV infection is being considered as an underlying cause of symptoms and signs and virological testing is not available, perform HIV serological testing and use the clinical algorithm for presumptive clinical diagnosis of HIV infection.
Population:
Sick infants and children (18 months of age) in settings where virological testing not available
Intervention:
Perform HIV serological testing and follow the clinical algorithm for presumptive diagnosis of HIV infection.
FactorDecision
Quality of evidenceWeakConcerns as not all rapid tests can be used in this case
Serological assays used must be highly sensitive, e.g. current Determine. Further research necessary
Impaired immune system of the baby could reduce HIV antibody detection in the baby.
Persisting maternal antibodies could affect serological test result.
The performance of the clinical algorithm depends on the health-care worker.
Benefits or desired effectsWeakIn sick children who are HIV infected, start of HIV treatment and care at the earliest possible reduces mortality.
Risks or undesired effectsPossible FN results (see above)
FP results especially in settings where the result cannot be further confirmed
Values and preferencesStrong/weakSaves lives
Do not stop child from getting ART because not able to perform virological testing
If clinical suspicion is still high may need to treat while seeking further testing
HIV infection is an unlikely cause of signs and symptoms in low-prevalence settings. However, in settings where the prevalence is >5%, HIV infection is a more likely cause of signs and symptoms.
Costs and feasibilityStrong/weakDepends on the prevalence
Overall ranking of recommendationStrength of recommendation
Strong for high-prevalence settings
Weak for low-prevalence settings (<1% maternal ANC seroprevalence)

References

1.
Violari A, et al. Early antiretroviral therapy and mortality among HIV-infected infants. New England Journal of Medicine. 2008;359:2233–2244. [PMC free article: PMC2950021] [PubMed: 19020325]
Copyright © 2010, World Health Organization.

All rights reserved. Publications of the World Health Organization can be obtained from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: tni.ohw@sredrokoob). Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution – should be addressed to WHO Press, at the above address (fax: +41 22 791 4806; e-mail: tni.ohw@snoissimrep).

Bookshelf ID: NBK138557
PubReader format: click here to try

Views

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to pubmed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...