South African HIV Guidelines — Questions

The first thing to recognise is that the patient is not failing treatment. An “LDL” viral load means the assay cannot detect viral RNA above its limit (typically below 50 copies/mL): the patient is virologically suppressed. The CD4 of 203 is a separate question about immune reconstitution, not virological success.

Possible explanations for a discordant CD4 to VL picture in a suppressed patient:

Late start with a low CD4 nadir. Antiretroviral therapy (ART) was started at a low nadir CD4, perhaps below 200, and immune recovery has been slow. The lower the starting point, the slower and more limited the recovery: after two years of suppression a CD4 of 203 from a baseline of, say, 80 represents real progress, and recovery often takes 3 to 5 years to plateau.

Persistent immune activation. Even on suppressive ART, residual low-level inflammation (driven by microbial translocation from the gut, ongoing low-level viral replication, and other factors) limits peripheral CD4 reconstitution. Co-infections such as hepatitis C virus (HCV), hepatitis B and recurrent tuberculosis (TB) exacerbate this.

Spurious result. A single CD4 result is biologically variable: diurnal variation, recent vaccination, intercurrent infection (even minor viral illness), or laboratory variability can produce a result around the 200 threshold. A repeat sample after correcting any modifiable factor is sensible before drawing conclusions.

Management. Reassure the GP that virological suppression is the correct indicator of treatment success, and do not change the ART regimen on the basis of the CD4 count; the sensible steps are to screen for and treat co-morbidities (especially TB and HCV) and repeat the CD4 in 6 months.

An indeterminate qualitative HIV PCR is a low-level positive that does not meet the reporting threshold for a definitive positive (commonly defined by the laboratory as a cycle-threshold (CT) value above a set cut-off, e.g. CT above 36). It may represent: very early infection with low viral copy number; contamination during sample handling; technical issue with the assay; or, in an infant, passive transfer of maternal HIV nucleic acids in transplacentally acquired plasma.

The general approach

The first principle: do not ignore the result, and do not initiate antiretroviral therapy (ART) on a single indeterminate PCR. The next step depends on the infant’s testing history.

Step 1: review the infant’s prior HIV-PCR or viral-load results.

• Prior PCR positive, or prior VL detectable, or prior PCR indeterminate: the indeterminate result is corroborated by earlier evidence of infection. Treat as HIV-infected: initiate ART immediately and confirm on a fresh sample after starting therapy.
• Prior PCR negative, or no prior PCR: the indeterminate result stands alone. Repeat both the HIV PCR and a viral load urgently on a fresh sample.
  • Repeat PCR positive or indeterminate → treat as HIV-infected, initiate ART.
  • Repeat PCR negative and repeat VL undetectable → routine HIV-exposed follow-up; continue the standard EID schedule with the next PCR at 10 weeks.

Worked example: indeterminate PCR at birth, CT 38.1

A birth PCR with CT 38.1 is just above the threshold for a positive result. The clinical context matters:

• Maternal viral load at delivery. If the mother was virally unsuppressed (VL at or above 50 copies/mL on the most recent measurement), the indeterminate result is much more likely to represent true infection.
• Was prevention-of-mother-to-child-transmission (PMTCT) prophylaxis given? Dual nevirapine (NVP) plus zidovudine (AZT) prophylaxis at birth has been received by all HIV-exposed infants since the 2026 National Consolidated Guidelines (NCG) change. Continue the prophylaxis pending clarification.
• Was there any maternal antiretroviral exposure that might lower the infant’s HIV viral copy number at birth? Maternal ART (and infant prophylaxis itself) suppresses early infant viraemia and can produce a low-positive or indeterminate PCR pattern.

Immediate management:

• Repeat the HIV PCR and a viral load urgently, same week if possible.
• Continue the infant prophylaxis (NVP daily ± AZT twice daily) until the result is clarified.
• If maternal VL at delivery is at or above 50 copies/mL and the result is highly suspicious, consider starting infant ART pending confirmation rather than waiting.
• Counsel the mother: discuss the possible meanings honestly without making a premature diagnosis.

Subsequent action depends on the repeat:

• Repeat positive or indeterminate → start infant ART immediately; confirm with viral load; continue the EID follow-up schedule.
• Repeat negative, VL undetectable → return to routine HIV-exposed follow-up (PCR at 10 weeks, 6 months, etc.). Document the indeterminate result and the subsequent negative for clarity.

Indeterminate PCRs are not rare in early infant diagnosis: they represent a confluence of the test’s sensitivity at low copy numbers, the effect of PMTCT prophylaxis on early infant viraemia, and the inherent uncertainty of sub-threshold signals. A clear local algorithm prevents both over-treatment (initiating ART on a falsely positive result) and under-treatment (missing a true infection because the result was dismissed).

The South African EID programme exists to identify HIV infection in infants exposed to HIV, as early as possible, so that antiretroviral therapy (ART) can be initiated before the rapid disease progression characteristic of paediatric HIV (untreated infant HIV mortality is up to 50% by two years of age).

Current EID schedule (National Consolidated Guidelines, NCG 2026)

• Birth PCR. Performed on all HIV-exposed infants within 48 hours of delivery.
• 10-week PCR. Aligned with the Expanded Programme on Immunisation (EPI) 10-week vaccination visit.
• 6-month PCR. Aligned with the maternal 6-month viral-load check.
• 18-month rapid antibody test. Universal for all children, regardless of HIV exposure status.
• Age-appropriate test 6 weeks after breastfeeding cessation.
• Any time the infant is unwell with features suggestive of HIV.

A positive HIV-PCR result in a child under 2 years must be confirmed on a fresh sample (or a baseline viral load) before reporting as definitive.

Evolution of the South African EID programme

• Pre-2008: a single PCR at 6 weeks for HIV-exposed infants. This identified most intrapartum infections but missed in-utero infections, and the 6-week wait delayed ART start.
• From 2008: birth PCR introduced initially for high-risk infants, progressively universalised. The motivation was to detect in-utero infection at the earliest possible point. Birth-positive infants, by definition infected before delivery, represent a small but high-risk group (intrauterine acquisition is associated with rapid disease progression).
• 2015: HIV PCR at birth made universal for all HIV-exposed infants.
• 2019 (vertical transmission prevention update): the 6-week PCR moved to 10 weeks to align with the EPI immunisation visit, reducing the number of clinic touch-points and improving programme efficiency. The 10-week timepoint also gives a wider detection window for intrapartum infections (which typically become PCR-positive after the first 4 to 6 weeks of life under nevirapine prophylaxis).
• 6-month PCR added alongside the maternal 6-month viral load (VL) check, to catch HIV infections acquired during breastfeeding while the maternal VL is being monitored anyway.
• 18-month rapid antibody test retained as the definitive end-of-EID test: the residual maternal immunoglobulin G (IgG) that confounds infant antibody testing has waned by 18 months, so a positive rapid test at this age is a true infant antibody, not a maternal one.
• Post-cessation test: any breastfeeding-acquired infection that emerged after the 6-month PCR is captured by a test 6 weeks after weaning.
• NCG 2026 retains this five-point schedule (birth, 10 wk, 6 mo, 18 mo, post-cessation), reinforces the 10-week/EPI alignment, and standardises indeterminate-PCR management as a national algorithm.

Rationale for each timepoint

• Birth PCR: detects in-utero infection (the highest-mortality group, infants infected before labour). Early diagnosis means ART by the first month.
• 10-week PCR: detects intrapartum infections (those acquired during labour), now visible above the suppressive effect of infant prophylaxis. EPI alignment reduces visits.
• 6-month PCR: detects breastfeeding-acquired infection during the period of exclusive breastfeeding; aligned with maternal VL monitoring.
• 18-month rapid antibody: the definitive diagnostic endpoint after maternal antibody wanes; also universal-population HIV screen.
• Post-cessation test: catches infections acquired in the late breastfeeding period, after the 6-month PCR.

Taken together, the schedule reflects three biological realities: maternal antibody persists until about 18 months (so antibody tests in infants are uninterpretable before then); prevention-of-mother-to-child-transmission (PMTCT) drug prophylaxis suppresses early infant viraemia (so the PCR window opens after about 4 to 6 weeks); and breastfeeding remains a transmission route until weaning (so a single early test cannot exclude infection in a breastfed infant). Stringing the timepoints together gives the programme a near-complete diagnostic safety net for HIV-exposed infants.

The drugs used to prevent vertical HIV transmission are highly effective, and they make early infant diagnosis harder. The same antiretroviral exposure that prevents infection also suppresses the viral signal in an exposed-but-infected infant, blunting and delaying the diagnostic markers EID relies on. The problems sit on three axes: maternal antiretroviral therapy (ART), infant prophylaxis, and the timing of each test.

Maternal ART during pregnancy and breastfeeding

A mother on suppressive tenofovir/lamivudine/dolutegravir (TLD) throughout pregnancy and breastfeeding:

• Reduces but does not abolish vertical transmission. Residual transmission risk remains around 1%.
• Lowers infant viraemia in any infected infant, slowing the emergence of detectable HIV RNA/DNA. PCR results may take longer to become reliably positive.
• Selects (rarely) for resistance in the infant, complicating any future paediatric ART decisions.

Infant prophylaxis

The current South African regimen (dual nevirapine (NVP) plus zidovudine (AZT) for all HIV-exposed infants at birth, then stratified by maternal delivery viral load (VL)) is:

• Highly effective at preventing infection.
• A confounder for EID. NVP and AZT directly suppress whichever infant viraemia exists, blunting the PCR signal at and around the next testing timepoints.
• A possible cause of indeterminate results. A low-positive PCR in a prophylaxed infant may be the diagnostic signature of an infection that has been partially but not fully suppressed by the prophylaxis.

Timing problems

• Birth PCR. Detects in-utero infection, but in an infant on prophylaxis the viral signal may be borderline, producing indeterminate results that need confirmation.
• 10-week PCR. Aligned with the Expanded Programme on Immunisation (EPI), designed to detect intrapartum infection, but in an infant on extended NVP (higher-risk infant prophylaxis is continued through and beyond this timepoint), even an intrapartum infection may show below-threshold viraemia at 10 weeks.
• 6-month PCR. Detects breastfeeding-acquired infection during the exclusive breastfeeding period, but again, NVP prophylaxis continuing through breastfeeding suppresses any emerging infant viraemia.
• Post-cessation test. Captures infection that emerged in late breastfeeding: once prophylaxis is stopped, the diagnostic window for any remaining infection re-opens.
• 18-month antibody. Now reliable as the definitive endpoint because maternal antibody has waned.

Specific diagnostic pitfalls

• Increased indeterminate PCR rate. Particularly at birth, and in higher-risk infants on extended dual prophylaxis. Algorithmic management is essential.
• Delayed seroconversion in an infected infant. Maternal antibody persists for up to 18 months, masking infant antibody until then; infant prophylaxis may further blunt the infant’s own antibody development.
• False reassurance from interim negatives. A negative PCR at 10 weeks does not exclude intrapartum infection if prophylaxis is still actively suppressing viraemia. Hence the layered schedule.
• Transmission late in breastfeeding may not be detected until the post-cessation test (potentially many months after acquisition).

Programme-level implications

The South African EID schedule is constructed precisely to see around these prophylaxis-induced diagnostic gaps. Multiple testing timepoints (birth, 10 weeks, 6 months, post-cessation, 18 months) cover the periods during which prophylaxis is active and the periods during which it is not. The post-breastfeeding-cessation test is critical: it is the window after which prophylaxis no longer affects detection. Indeterminate-PCR algorithms standardise management of borderline results, and the 18-month rapid antibody as the universal endpoint catches anything missed.

Mothers need to understand that prophylaxis works, but the diagnostic timeline is longer than it would be without prophylaxis, and the EID schedule is not over until the 18-month test or 6 weeks after weaning is complete. Premature reassurance (“the baby is HIV-negative”) at an interim timepoint is the commonest counselling error; defer the definitive statement until the end of the schedule.

A dried blood spot (DBS) is a small volume of whole blood (typically obtained by capillary fingerprick or heel-prick) collected onto filter paper, dried, and shipped at ambient temperature. DBS extends molecular diagnostic capacity to settings where plasma logistics (refrigerated transport, rapid processing, large sample volumes) are impractical.

Applications

• HIV viral-load monitoring in remote or hard-to-reach sites.
• Drug-resistance surveillance (pre-treatment and acquired drug resistance studies, particularly the WHO HIVResNet programme).
• Early infant diagnosis in some settings (although South Africa primarily uses plasma EDTA for EID, DBS is acceptable).
• National seroprevalence and HIV-incidence surveys.
• Birth cohort and longitudinal study sampling where repeated venous draws are impractical.

Advantages

• Simple collection: no venepuncture skills required at the collection site.
• Stable at ambient temperature: can be transported by ordinary post or courier.
• Small volume: suitable for paediatric and capillary sampling.
• Cost-effective at scale: much cheaper logistically than plasma.

Challenges for viral-load testing

• Lower limit of detection. DBS VL assays typically require a viral load above 1,000 copies/mL to reliably amplify; the South African plasma assay threshold for suppression (below 50 copies/mL) cannot be reliably read off DBS. DBS cannot distinguish between truly suppressed and persistent low-level viraemia.
• Proviral DNA contamination. Whole-blood DBS contains both cell-free RNA and cell-associated proviral DNA. Most DBS VL assays detect both, leading to overestimation of viral load in suppressed patients, a critical limitation. Specialised RNA-only extraction protocols can mitigate this.
• Sample quality issues: under-spotting, blood mixed with sweat or alcohol, prolonged ambient exposure, humidity damage during transport.
• Inter-assay variability is higher than plasma.
• Subtype-specific sensitivity varies across DBS platforms.

Challenges for drug-resistance testing

• Higher VL threshold for sequencing success: typically VL above 1,000 copies/mL is needed for reliable Sanger amplification from DBS, similar to plasma.
• Sub-optimal amplification rates: DBS DRT amplification rates are typically lower than plasma (around 70 to 80% vs 90 to 95% for plasma).
• Proviral DNA contamination can confound RT/protease sequencing, picking up archived resistance mutations that are not present in the current plasma virus, relevant for surveillance interpretation.
• Subtype-specific primers may need optimisation.
• DBS is acceptable for surveillance (where archived mutations are arguably informative) but is not preferred for clinical case management.

South African programmatic context

The South African programme has used DBS primarily for surveillance and operational research rather than for routine clinical viral-load monitoring or resistance testing. Plasma EDTA remains the standard for clinical care. DBS is useful in:

• Hard-to-reach populations: mobile clinics, prison health services, remote rural sites.
• WHO HIVResNet surveillance: South Africa contributes pre-treatment and acquired drug-resistance data via DBS-based methodologies.
• Cohort studies: birth cohorts, key-population research.

DBS is a powerful tool for extending diagnostic and surveillance reach, but it is not a substitute for plasma in clinical case management. The South African virology laboratory should maintain DBS capability as a parallel competency for the surveillance and operational-research roles it serves, while continuing to drive clinical viral-load and resistance testing on plasma, particularly given the suppression threshold (below 50 copies/mL) that current South African practice demands.

HIV-exposed but uninfected (HEU) infants consistently show higher rates of infectious-disease morbidity, hospitalisation and mortality than HIV-unexposed infants in cohort studies, even in the post antiretroviral therapy (ART) era. The phenomenon is well described in southern African populations and has multiple causes.

Underlying reasons

1. Reduced placental transfer of maternal antibody. HIV-positive mothers transfer lower titres of protective immunoglobulin G (IgG) to their infants, for measles, pertussis, Streptococcus pneumoniae, and other pathogens, because of disrupted placental Fc-receptor function. HEU infants are functionally less protected during the first 6 months of life.

2. Antenatal co-morbidity and immune activation. HEU infants are more likely to have been born to mothers carrying HIV-driven immune activation and chronic inflammation during pregnancy, which affects placental function and transmits a pro-inflammatory milieu to the fetus. Co-morbidities cluster with maternal HIV: tuberculosis (TB), syphilis, anaemia and malnutrition.

3. Higher prematurity and low-birthweight rates. HIV-positive mothers have higher rates of preterm delivery and low birth weight, both established drivers of infant morbidity. Tenofovir use in pregnancy has been associated with marginally lower birth weight in some cohorts.

4. Socioeconomic exposure. HEU infants face greater infectious exposure through poverty, food insecurity, household HIV burden (siblings and parents on ART, with intermittent illness and high TB exposure), maternal hospitalisation disrupting infant care, and reduced access to support (stigma, female-headed households without paternal involvement).

Measures to improve outcomes

• Maternal viral suppression throughout pregnancy and breastfeeding. The single most important determinant of both transmission risk and HEU outcomes; keeping the mother well is the single most effective intervention for the infant.
• EPI immunisations strictly on schedule to compensate for reduced placental antibody transfer, with pneumococcal and Haemophilus influenzae vaccines particularly important given reduced maternal antibody.
• Exclusive breastfeeding for the first 6 months, then continued breastfeeding with appropriate complementary feeding. Exclusive breastfeeding is now strongly recommended over formula feeding for HEU infants: the immune and microbiological protection outweighs the small residual transmission risk on suppressive maternal ART.
• Cotrimoxazole preventive therapy where indicated. The 2026 National Consolidated Guidelines (NCG) withdrew routine cotrimoxazole from HEU infants, driven by reduced transmission rates and concerns about antimicrobial pressure; it remains indicated for higher-risk infants and where the infant’s HIV status is not yet excluded, and the withdrawal will require monitoring of post-implementation HEU morbidity to confirm safety.

Both qualitative and quantitative nucleic-acid tests amplify HIV nucleic acid by reverse-transcription PCR (or transcription-mediated amplification, TMA), but they answer different clinical questions.

Comparison

Feature	Qualitative HIV PCR	Quantitative HIV viral load
Clinical question answered	Is the virus present (yes / no)?	How much virus is present (copies/mL)?
Output	Positive / negative / indeterminate	A numerical concentration (e.g. “1,200 copies/mL”) plus a lower limit of detection
Primary clinical application	Early infant diagnosis in children under 18 months (where maternal antibody confounds serology)	Antiretroviral therapy (ART) monitoring: measuring response, detecting failure
Target sequence	Conserved region of pol or gag for high subtype-inclusive detection	Conserved RT/gag region for accurate quantification
Sample type	EDTA whole blood or plasma; dried blood spot (DBS) acceptable	EDTA plasma (preferred); DBS in some surveillance contexts
Detection limit	Approximately 50 to 100 copies/mL on standard platforms	Lower limit typically 20 to 50 copies/mL; upper linear range to 10⁷
Quantification accuracy	Not relevant (qualitative output only)	Critical: used to discriminate suppression, low-level viraemia, and failure
Cost	Lower per test	Higher per test
Turnaround	Generally faster	Comparable or slightly slower
Subtype performance	Optimised for inclusive detection across subtypes	Validated across major subtypes including C
Confirmation	Result confirmed on a fresh sample for any child under 2 with a positive screen	A single number; trend over serial samples drives clinical decisions
Indeterminate results	Possible; requires algorithmic management	Less common; assay platform reports a numerical value or “below limit of detection”

Similarities

Both assays:

• Use reverse-transcription PCR or TMA to detect HIV nucleic acid.
• Are subject to PCR inhibition (heparin, haemoglobin, urine, faecal contamination, etc.); internal controls flag failed runs.
• Are vulnerable to subtype-specific primer mismatches if not properly validated for the local subtype (subtype C in South Africa).
• Detect virus earlier than serology, closing the diagnostic window.
• Require careful specimen handling to preserve RNA integrity (timely separation of plasma; appropriate shipping).

Key differences

• Purpose. Qualitative PCR answers “infected or not”; quantitative viral load (VL) answers “how well is therapy working”.
• Quantification. Qualitative outputs no number; quantitative outputs a precise viral concentration that drives clinical action (repeat-and-persistent-check eligibility at below 50 copies/mL; low-level viraemia between 50 and 999 copies/mL; failure assessment at 1,000 copies/mL or above).
• Indeterminate handling. Qualitative PCR has a defined indeterminate category and algorithm; quantitative VL produces a clean numerical result.
• Confirmation requirement. A positive qualitative result in a child under 2 always requires confirmation; a quantitative VL is a clinical trajectory, not a single-result diagnosis.

Clinical implications

• Early infant diagnosis uses qualitative PCR because the question is binary (infected or not) and the test must operate near the diagnostic threshold.
• ART monitoring uses quantitative VL because the clinically meaningful information is the numerical trajectory (rising, falling, stable around suppression).
• DBS is acceptable for both purposes in surveillance contexts but has greater quantitative inaccuracy.
• A qualitative result is not a substitute for a viral load in an established patient on ART, and vice versa for an infant diagnosis.

In drug-sensitive extra-neurological tuberculosis (TB) with CD4 below 50 cells/µL, ART starts within 2 weeks of TB treatment, reflecting the high short-term mortality of untreated HIV at low CD4.

At CD4 at or above 50 the start is deferred to 8 weeks, which does not apply here. Same-day initiation is not the rule with TB co-treatment; deferring until TB treatment ends abandons a severely immunocompromised patient; and immune reconstitution inflammatory syndrome (IRIS) is a risk of early ART, not a trigger for it. The earlier start here is offset by prophylactic prednisone (40 mg for 2 weeks then 20 mg for 2 weeks) when CD4 is at or below 100 and ART begins within 30 days of TB treatment.

AHD is any one of CD4 at or below 200 cells/µL, World Health Organization (WHO) clinical stage 3 or 4, or age under 5 years (automatic unless stable on antiretroviral therapy for at least 12 months).

The threshold is 200, not 50, and stage 3 counts as well as stage 4, so the single-criterion options are too narrow; viral load and hospitalisation alone do not define AHD. The classification triggers the nine-step package: reflex cryptococcal antigen (CrAg), urine tuberculosis lipoarabinomannan (TB-LAM), cotrimoxazole prophylaxis, tuberculosis preventive therapy, pre-emptive fluconazole, rapid ART and intensified follow-up. About one in five people living with HIV present with AHD in South Africa.

Any positive HIV PCR in a child under 2 years must be confirmed on a fresh sample before a definitive report.

A single positive neither confirms nor excludes infection, so reporting on it, discharging, or deferring to a 6-month or 18-month test are all wrong. South African practice substitutes a viral load (VL) for the confirmatory PCR, giving confirmation and a baseline VL at once. Antiretroviral therapy (ART) starts immediately without waiting for the confirmatory result, but the report rests on the confirmed sample.

A maternal delivery VL below 50 copies/mL is low risk, so the infant receives nevirapine (NVP) daily for 6 weeks.

Prophylaxis is not omitted, 12 weeks of NVP or the zidovudine (AZT)-plus-NVP regimen is the higher-risk option (maternal VL at or above 50 while breastfeeding), and full combination therapy is treatment, not prophylaxis. Every exposed infant first gets dual NVP plus AZT at birth until the delivery VL is known; here the below-50 result stops the AZT and leaves 6 weeks of NVP.

The immunity threshold is HBsAb at or above 10 mIU/mL, so a value of 4 is non-immune and needs a full three-dose HBV vaccination at 0, 1 and 6 months.

The HBsAb is not protective, a single booster is insufficient in a never-immune non-responder, the vaccine is safe rather than contraindicated in HIV, and immunoglobulin is not the action here. Recheck HBsAb 2 months after the last dose; if still below 10, repeat the three-dose course, then refer if it remains below 10.

Tenofovir suppresses both HIV and HBV, so abrupt cessation of tenofovir disoproxil fumarate (TDF), lamivudine (3TC) or emtricitabine (FTC) can precipitate a severe, sometimes fatal hepatitis B flare in a co-infected patient.

HIV rebound, resistance from the drug tail, renal injury and prodrug withdrawal are not the driving concern. Check hepatitis B surface antigen (HBsAg) before stopping any of these agents; if TDF must be replaced (for example eGFR 30 to 50), the substitute is tenofovir alafenamide (TAF) plus FTC plus dolutegravir, which preserves HBV cover.

Cryptococcal meningitis defers ART 4 to 6 weeks after antifungal therapy starts, reducing the risk of paradoxical cryptococcal immune reconstitution inflammatory syndrome (IRIS) in the central nervous system.

Immediate or 2-week starts are too early for meningitis, waiting for antifungal completion is too long, and ART is not contraindicated. Tuberculous meningitis has a comparable 4-to-8-week window, whereas asymptomatic antigenaemia with a negative lumbar puncture starts ART immediately alongside pre-emptive fluconazole (1,200 mg daily for 14 days).

The ABCDE differential runs Adherence, Bugs, Correct dose for weight, Drug interactions, then Resistance, so resistance testing is the last step, reached only once the others are excluded.

The distractors substitute lifestyle or generic factors (body mass index, diet, exercise, counselling, disclosure) that do not belong in the viraemia work-up. In detail: adherence is the dominant cause of detectable viraemia on tenofovir/lamivudine/dolutegravir (TLD); bugs means concurrent infection (tuberculosis, sexually transmitted infections, recent vaccination); correct dose matters especially in paediatrics; drug interactions include rifampicin, anticonvulsants and polyvalent cations. This complements the reflex dolutegravir (DTG) drug-level workflow.

Prophylactic prednisone is reserved for the highest-risk patients: CD4 at or below 100 cells/µL with ART started within 30 days of TB treatment, given as 40 mg daily for 2 weeks then 20 mg daily for 2 weeks.

The other options are too broad: prophylaxis is not for every HIV-TB patient, nor defined by drug-resistant or extrapulmonary TB alone, and CD4 above 200 with delayed ART is low-risk. Established non-neurological IRIS is instead treated reactively (prednisone 1.5 mg/kg for 2 weeks then 0.75 mg/kg for 2 weeks); IRIS is a diagnosis of exclusion, ruling out multidrug-resistant or extensively drug-resistant TB and non-adherence.

The EID schedule places HIV PCR at birth, 10 weeks and 6 months, so the two-timepoint and 10-week-only options miss tests and monthly testing is not the schedule.

Alongside the three PCRs, a universal rapid antibody test follows at 18 months, with an age-appropriate test 6 weeks after breastfeeding stops and testing whenever the infant is unwell. The 10-week PCR was moved from the earlier 6-week timing to align with the Expanded Programme on Immunisation (EPI) vaccination visit; the 6-month PCR aligns with the maternal 6-month viral load.

The original 90-90-90 framework has been upgraded to 95-95-95 by 2030: 95% of people living with HIV diagnosed, 95% of those diagnosed on antiretroviral therapy (ART), and 95% of those on ART virally suppressed.

The 90-90-90 targets (2020) are superseded; 100-100-100 and 70-70-70 are not real targets; and the framework does set explicit numeric goals. Achieved in cascade, these targets deliver population-level suppression sufficient to end HIV as a public-health threat, and South Africa has committed to them in the 2026 National Consolidated Guidelines.

The three-test serial strategy exists to hold the positive predictive value (PPV) at 99% or higher, minimising false-positive misdiagnosis. A highly sensitive screening test first identifies all reactive individuals; two confirmatory tests of different manufacture then rule out false-reactive results. Only a sample reactive on the screening test and both confirmatory tests is reported HIV positive.

Raising sensitivity is the role of the screening test, not the third; a surveillance case count and HIV-2-specific testing are unrelated to the algorithm’s purpose. A sample reactive on the screening test and confirmatory test 1 but non-reactive on confirmatory test 2 is reported as discrepant and referred for enzyme-linked immunosorbent assay (ELISA) reflex testing. The third test’s job is therefore to secure the PPV, not merely to break a tie between the first two.

The Advanced HIV Disease (AHD) package triggers reflex serum CrAg at CD4 at or below 200 cells/µL, the same threshold that defines AHD.

The other cut-offs (50, 100, 350, 500 cells/µL) are not the trigger. Any positive serum CrAg requires a lumbar puncture (LP): a positive cerebrospinal fluid (CSF) CrAg is cryptococcal meningitis (treat and defer antiretroviral therapy 4 to 6 weeks); a negative CSF CrAg is asymptomatic antigenaemia (pre-emptive fluconazole 1,200 mg daily for 14 days and immediate ART).

TLD 1 covers the ART-naïve patient started on a dolutegravir (DTG)-based regimen and the patient switched from a non-DTG first-line (typically TEE) with a viral load (VL) below 50 copies/mL in the last 12 months, which is exactly this patient.

TLD 2 is for a switch from a non-DTG first-line with VL at or above 50, or from a protease inhibitor (PI)-based second-line (VL below 50, or VL at or above 1,000 without a genotype). TLD 3 follows a genotypic resistance test showing PI resistance, authorised by the ARV Drug Resistance Committee (ADReC). ALD 2 is the paediatric abacavir-based series, and “recycled TLD” is not a category.

The 2026 change moved the first routine VL from 6 months to 3 dispensing cycles (DCs), about 3 months on antiretroviral therapy (ART).

The 1-, 6- and 12-month options are the earlier or wrong timings, and testing is routine rather than symptom-triggered. Subsequent VLs follow at 10 DCs (about 10 to 12 months), 22 DCs (about 24 months), then every 12 DCs annually; breastfeeding women test every 6 DCs from delivery. The eGFR and creatinine schedule was pulled earlier to align with the new VL timing.

A central 2026 change dropped the higher-risk threshold from a maternal viral load (VL) at or above 1,000 copies/mL to at or above 50 copies/mL, extending dual prophylaxis to a far larger fraction of exposed infants.

The higher thresholds (200, 500, 1,000 and 10,000 copies/mL) are the old or invented cut-offs. Higher-risk infants receive nevirapine (NVP) for at least 12 weeks plus zidovudine (AZT) for 6 weeks if breastfed (NVP and AZT each for 6 weeks if formula-fed); every exposed infant gets dual NVP plus AZT at birth until the delivery VL is known.

Confirmed virological failure needs two consecutive viral loads (VLs) at or above 1,000 copies/mL, after at least 9 months on a DTG-based regimen, with at least two documented adherence interventions.

A single high VL, a threshold of 50 or 200 copies/mL, and a doubling VL all fall short of the definition. The criteria reflect that DTG resistance is rare and most detectable viraemia is non-adherence, which the reflex DTG drug-level workflow confirms or refutes before any genotype.

A 2026 change withdrew CPT from HIV-exposed uninfected infants and reserved it for confirmed HIV infection meeting the age, CD4 or stage criteria.

CPT is not for every HIV-positive person regardless of CD4 or stage, nor limited to active Pneumocystis pneumonia, nor given to exposed uninfected or breastfeeding infants. In detail: all HIV-positive infants under 1 year whatever the CD4 or stage; children 1 to 5 years with CD4 at or below 25% or World Health Organization (WHO) stage 3 or 4; those over 5 and adults with CD4 at or below 200 or WHO stage 3 or 4. Stop once CD4 rises above 200 (or above 25% in under-5s).

The default adult TPT regimen is 12H: isoniazid (INH) 300 mg daily with pyridoxine 25 mg daily, for 12 months.

3HP (weekly INH plus rifapentine, 12 doses) is an alternative only for clients 25 kg or more already on dolutegravir (DTG) with a recent suppressed viral load, so not for new antiretroviral therapy (ART) starters. Rifampicin monotherapy and bedaquiline are not TPT regimens, and TPT does not require confirmed active-TB exposure. TPT in pregnancy is now indicated only with Advanced HIV Disease (CD4 at or below 200 or World Health Organization stage 3 or 4).

The Children’s Act permits any child aged 12 or older to self-consent to HIV testing, and a child under 12 with sufficient maturity to understand the test (the mature-minor doctrine) may also self-consent.

A child under 12 without that capacity needs consent from a parent, caregiver or designated adult, so option E is too absolute. The 16-, 18- and 7-year thresholds are not the ones set in the framework.

A negative HIV serology at 6 months after exposure in a healthcare worker who received timely and complete post-exposure prophylaxis (PEP) is the expected and reassuring result. It effectively rules out HIV transmission from the exposure.

Interpretation of the 6-month negative

• The PEP worked, or transmission never occurred. Either is reassuring. The exposure (mucocutaneous splash to the eye) carries a transmission risk of approximately 0.09% per single contact, low even without PEP, and effectively zero with timely 28-day tenofovir/lamivudine/dolutegravir (TLD).
• Window-period testing has been completed. The South African PEP follow-up testing schedule for occupational exposure is HIV testing at 6 weeks, 3 months and 6 months (the 6-month test is the standard endpoint, with 12-month testing reserved for hepatitis C virus (HCV)-coinfected sources).
• Antibody response in PEP-treated patients. PEP itself does not produce HIV antibodies. The negative serology at 6 months means the worker did not seroconvert, confirming no transmitted infection.
• Long-acting PrEP caveat. This worker received TLD for PEP (not long-acting). The atypical-seroconversion (“LEVI syndrome”) pattern that complicates testing after long-acting pre-exposure prophylaxis (PrEP) failure does not apply here; the negative test can be taken at face value.

Management going forward

The worker is HIV-negative. Specific actions:

• Document the negative result in the occupational health record and close the exposure file.
• Reassure the worker: the exposure was managed appropriately and there is no HIV.
• Brief counselling on prevention going forward: adherence to standard precautions, prompt occupational-health reporting of future exposures.
• Mental-health debriefing if not already done: a 6-month period of uncertainty following an occupational exposure is significant. Offer counselling and employee assistance programme (EAP) referral if appropriate.
• Vaccination and immunisation review: confirm the worker’s hepatitis B virus (HBV) vaccination is current; if the source was HBV-positive and the worker was non-immune, that has been managed separately at the time of exposure.

Reflection for the programme

The case is an exemplar of how the South African occupational PEP pathway should work:

• Rapid initiation within 72 hours (in this case within 4 hours).
• Complete 28-day course of TLD.
• Structured serial follow-up at 6 weeks, 3 months and 6 months.
• Documentation and closure at the 6-month timepoint.

The detail of the SA PEP regimen and follow-up schedule sits in the standalone PEP 2019 guideline to which the National Consolidated Guidelines (NCG) defer; this case illustrates that pathway in action.

What if the 6-month test had been positive?

A positive HIV result at 6 months in a PEP-treated worker would prompt:

• Repeat HIV testing on a fresh sample to confirm.
• HIV viral load and CD4 as baseline for antiretroviral therapy (ART) initiation.
• Genotypic resistance testing: possibility of transmitted resistance from the source (especially if the source was on ART) or PEP-induced resistance during sub-therapeutic exposure.
• Initiation of ART: TLD as first-line in current practice, but with regimen choice informed by the resistance result.
• Investigation of the exposure: was the PEP regimen appropriate? Was there a missed exposure? Documentation of programme-level lessons.
• Counselling and occupational support: disclosure considerations, work restrictions (none required by NCG; the Health Professions Council of South Africa (HPCSA) position should be consulted), psychological support.

A negative 6-month HIV serology after a timely 28-day course of TLD PEP for an eye-splash exposure from a known HIV-positive source is the expected and reassuring outcome. Routine management closes the file; no further HIV testing is required.

Clinical scenario. An HIV-exposed infant has the following PCR series:

• November 2017 (birth): HIV PCR indeterminate
• February 2018 (~10 weeks): HIV PCR negative
• September 2018 (~6 months): HIV PCR negative

Result-by-result interpretation

Result 1: birth PCR, indeterminate. An indeterminate result at birth is a low-level positive (commonly defined by a cycle-threshold (CT) value above the laboratory’s positive cut-off). The result requires immediate follow-up: review of the infant’s prior testing (none, in this case, as this is the birth test), urgent repeat HIV PCR and a viral load (VL) to clarify, continuation of nevirapine (NVP) plus zidovudine (AZT) prophylaxis, and clinical assessment of risk factors (maternal VL at delivery; prevention-of-mother-to-child-transmission (PMTCT) prophylaxis status; possible PMTCT failure indicators).

The possible explanations for an indeterminate birth PCR are: very early in-utero infection at low viral copy number; PMTCT prophylaxis (NVP given in late pregnancy and at birth) producing partial suppression; sample contamination; transfer of maternal nucleic acids in transplacental plasma; technical/threshold artefact.

The clinical management of the indeterminate result is what comes next.

Result 2: 10-week PCR (February 2018), negative. The 10-week test is the Expanded Programme on Immunisation (EPI)-aligned PCR that screens primarily for intrapartum-acquired infection. A negative result at this timepoint:

• Argues strongly against an established intrapartum infection.
• Effectively resolves the November 2017 indeterminate result: the indeterminate at birth was almost certainly not true infection (or was, but at a copy number that infant prophylaxis suppressed below subsequent detection). With prior negative testing, the standard National Consolidated Guidelines (NCG) algorithm would have closed the issue here.
• Returns the infant to routine HIV-exposed follow-up.
• Continues breastfeeding prophylaxis if still indicated by maternal VL.

A useful caveat: a negative 10-week PCR does not exclude HIV acquired during the intrapartum or early postnatal period if PMTCT prophylaxis is still actively suppressing infant viraemia. This is one reason the EID schedule continues with a 6-month PCR.

Result 3: 6-month PCR (September 2018), negative. The 6-month PCR is timed to detect breastfeeding-acquired infection that emerges during the period of exclusive or mixed breastfeeding. A negative result:

• Indicates no detectable HIV at 6 months, covering breastfeeding-acquired infection up to that point.
• Aligns with the maternal 6-month VL check, allowing simultaneous reassessment of vertical transmission risk.
• Does not exclude infection acquired after 6 months, particularly if breastfeeding continues, hence the post-cessation testing requirement (and the 18-month antibody test) that complete the early infant diagnosis (EID) schedule.

Overall interpretation

The infant has had an indeterminate birth result that was not confirmed on follow-up testing. The two subsequent negative PCRs at 10 weeks and 6 months effectively rule out vertically acquired infection up to 6 months. The infant is presumed HIV-negative for the purposes of clinical decision-making, while remaining on routine HIV-exposed follow-up (including the post-breastfeeding-cessation test and the 18-month rapid antibody).

Counselling and ongoing care

• Reassurance to the mother: the most likely interpretation is no vertical transmission, with the indeterminate birth result being a sub-threshold artefact or transient maternal nucleic-acid carryover.
• Continued breastfeeding management: the infant needs ongoing prophylaxis if the mother is unsuppressed and breastfeeding; the prophylaxis is reviewed at each maternal VL check.
• Complete the schedule: testing at 18 months and 6 weeks post-cessation of breastfeeding is mandatory for definitive clearance from the EID programme.
• Maternal care: confirm the mother’s ongoing antiretroviral therapy (ART) suppression, address any adherence issues.

This case illustrates how the layered EID schedule delivers progressively stronger confidence about an infant’s HIV status: no single test result alone is conclusive, but the series builds toward a defensible conclusion.

The 95-95-95 targets (95% of people living with HIV diagnosed, 95% of those diagnosed on antiretroviral therapy (ART), and 95% of those on ART virally suppressed) define the operational programme goal for ending HIV as a public-health threat by 2030. The earlier UNAIDS 90-90-90 framework was upgraded to reflect what is now achievable with current tools. The virology laboratory underpins all three pillars.

Pillar 1: 95% diagnosed

The laboratory provides the testing infrastructure for diagnosis at scale:

• The three-test serial rapid algorithm in primary care, validated and quality-assured, backed by external quality assurance to maintain the 99% positive predictive value the algorithm depends on.
• Fourth-generation antigen/antibody assays for laboratory confirmation, and HIV PCR for early infant diagnosis (birth, 10-week, 6-month).
• Surveillance and seroprevalence studies (including dried blood spot surveys) to track progress against the first 95.

Pillar 2: 95% of diagnosed on ART

The laboratory enables rapid same-day ART initiation through:

• Baseline workup at the point of diagnosis: hepatitis B surface antigen (HBsAg), syphilis, creatinine/estimated glomerular filtration rate (eGFR), CD4, tuberculosis nucleic acid amplification test (TB-NAAT), pregnancy test.
• Reflex cryptococcal antigen (CrAg) below CD4 200 and TB lipoarabinomannan (TB-LAM) for advanced HIV disease (AHD) identification, enabling rapid ART for those eligible and appropriate deferral for those who are not.
• Pre-treatment drug-resistance surveillance (transmitted drug-resistance studies) to inform regimen choice.

Pillar 3: 95% of those on ART virally suppressed

This is the pillar where the laboratory does the largest volume of work:

• Routine viral-load (VL) monitoring at scale: South Africa runs millions of HIV VL tests per year. The National Consolidated Guidelines (NCG) 2026 schedule (VL at 3, 10 and 22 dispensing cycles, then annually; breastfeeding 6-monthly) is feasible only because of centralised high-throughput laboratory capacity.
• The 2026 reflex dolutegravir (DTG) drug-level testing workflow: the laboratory measures DTG levels on every drug-resistance sample, gate-keeping unnecessary resistance testing in non-adherent patients.
• Genotypic drug-resistance testing for confirmed virological failure on a DTG-based regimen (integrase, reverse transcriptase and protease regions), plus dried blood spot VL where plasma logistics fail.

Cross-cutting laboratory roles

Underpinning all three pillars are surveillance reporting to the National Department of Health and UNAIDS for the 95-95-95 estimates, operational research informing programme evolution, and quality management systems (SANAS accreditation, proficiency testing, external quality assurance).