Impact of birth HIV PCR testing on uptake of follow-up EID services for HIV-exposed infants in Cape Town

Lorna Dunning1, Max Kroon2, Lezanne Fourie2, Andrea Ciaranello3, Landon Myer1

1Division of Epidemiology & Biostatistics, School of Public Health and Family Medicine, University of Cape Town, South Africa2Department of Neonatal Medicine, University of Cape Town, South Africa3 Division of Infectious Disease, Medical Practice Evaluation Center, Massachusetts General Hospital, Boston, MA, USA

Pediatric HIV Workshop 2016

Background – Early Infant Diagnosis

• WHO recommends all HIV-exposed infants should be tested by six weeks of age – Virological tests

– Detect in utero, intra-partum and very early post-partum infections*

• 94% of HIV-exposed infants have access to a virologic test in South Africa1

* Debate around effect of infant prophylaxis on the sensitivity of the assay at 6 weeks1 World Health Organization. Progress report on the global plan towards the elimination of new HIV infections among children and keeping their mothers alive. Geneva, Switzerland2015

Background – 6 week EID Cascade

Infant born HIV-

exposed

6 weeks-Offer HIV

testing

Specimen Processing

Result Return

HIV-infected InfantLink to care - ART

HIV-uninfected Exposed Infant

Repeat testing after weening

Peak age of mortality 11

weeks for HIV-infected

infants

Background – Birth Testing

• Improved PMTCT coverage

– Increased proportion of in utero infections relative to intrapartum and early post partum

– More infected infants detectable at birth

• Benefits earlier treatment

– Limit disease progression

– Lower viral reservoirs

– Increase survival1

1 Francke et al. Clinical impact and cost effectiveness of early infant HIV diagnosis in South Africa: Test timing and frequency. Pre-publication JID

Background – Birth Testing Cascade

Birth-HIV PCR testing

Specimen Processing

Result Return

Birth Testing

HIV-exposed

Infant

Infant born HIV-exposed

6 week-HIV PCR testing

Specimen Processing

Result Return

HIV-infected InfantLink to care - ART

HIV-uninfected Exposed Infant

Repeat testing after weening

HIV-infected –Linked to Care

LTFU?

Returning a negative result to women immediately after delivery could cause confusion around the need for testing at 6 weeks• Recent observational study – only 49% of infants returned for follow up testing after

an initial test at birth1

1 J Matritz et al. Low Uptake of Routine Infant Diagnostic Testing Following HIV PCR Testing at Birth. Poster Presentation. CROI 2016

Objectives

1. To compare the attendance at follow-up EID testing by receipt of an HIV PCR test at birth- Does receipt of a negative birth test impact uptake of 6 weeks testing?

2. Examine yield of targeted birth testing in this setting

Site Characteristics• Secondary level

- Obstetric Hospital• Women referred from

primary care • 13% of newborn HIV-exposed

Cohort Characteristics• July 2013 – August 2015• All mothers eligible for Option B+ • HIV PCR test at birth – Infants high risk for perinatal transmission• Birth tests occurred within 48hrs of delivery: result return prior to discharge• All HIV exposed infants tested at 6 weeks, or screened if presenting to hospital

with OI

Methods – Site and Cohort Characteristics

Methods – Study Sample

HIV-exposed High Risk (576)

HIV PCR test at Birth

Equivocal

(3)HIV+

(22)HIV-exposed

(551)HIV-exposed

(551)

HIV-exposed Low risk

Mode of Delivery & Date of Birth

HIV PCR test 6 weeks +

??

1. Results of HIV PCR tests were determined from NHLS database– Non-attendance at follow up testing:

no subsequent HIV PCR test recorded since birth test or no recorded HIV PCR test

– Linkage to care: Infect infant specimen sent to laboratory for processing

2. Participant demographic characteristics – Mean/median, frequency

3. Logistic regression models – OR with 95% confidence intervals

Analysis

6%

38%

15%

15%

14%

6%

5%

4%4%

3%

1% 1% 2%

Results – HIV-exposed infants receiving an HIV PCR test at Birth

6%

38%

15%15%

14%

6%

5%

4%

4% 3%

1% 1% 2%

6%

38%

15%15%

14%

6%

5%

4%4%

3%

1% 1% 2%

6%

38%

15%15%

14%

6%

5%

4%4%

3%

1%1%

2%

6%

38%

15%15%

14%

6%

5%

4%4%

3% 1% 1%2%

No Reason

ART <12 weeks

Low Birthweight / Preterm

High Viral Load

Default from Treatment

Unbooked

No ART during Pregnancy

Seroconverted during Pregnancy

Maternal Infection

2nd Line Treatment

Substance Abuse

Other

Results – HIV PCR test at birthHIV-exposed High Risk (576)

HIV PCR test at Birth

Equivoca

l (3)HIV+

(22)HIV-exposed

(551)

• 3.8% of birth tests positive (n=22)• 95% confirmatory test• 1 false positive result • 71% linked to care w/in 3mths• 86% linked to care w/in 6mths

HIV PCR Result at Birth

Negative 551

Positive 22

Confirmatory test 21

False Positive 1

Linked to care w/in 3mths 15

Linked to care w/in 6mths 3

HIV-exposed received birth

test (551)

HIV-exposed No Birth Test

(551)

HIV PCR test 6 weeks +

Results – HIV PCR test at 6 weeks

Received birth test No birth test

HIV PCR test result(6wks)

Negative 399 479

Positive 2 2

Confirmatory test 2 2

False Positive 0 0

Linked to care w/in 3mths 1 2

Linked to care w/in 6mths 1 -

• 0.4% of 6wk PCR tests positive (n=4)• 100% confirmatory test• 75% linked to care w/in 3mths• 100% linked to care w/in 6mths

Results – Follow up EID test results

Result Birth Test

(N=551) %

No Birth Test

(N=551) %

Total

(N=1102) %

Attendance of follow-up HIV PCR test:

Tested

Not Tested

(401) 73%

(150) 27%

(470) 85%

(81) 15%

(878) 79%

(231) 21%

Age if returned for follow-up HIV PCR test:

Mean (SD) 59.6 (41.8) 49.5 (22.8) 54.1 (33.3)

Age if returned for 6wk HIV PRC

Less than 28 days

4 – 10 weeks

10 weeks – 6 months

6 – 9 months

9 – 12 months

(4) 1%

(337) 84%

(50) 13%

(5) 1%

(5) 1%

(10) 2%

(436) 91%

(32) 7%

(2) 0.4%

(1) 0.2%

(14) 1%

(773) 70%

(82) 7%

(7) 0.6%

(6) 0.5%

Variable Birth Test

% (N=551)

No Birth Test

% N=551)

Total

% (N=1102)

Sex:

Female (279) 51% (276) 50% (555) 50%

Birthweight (g):

Low (<2500) (184) 34% (45) 7% (229) 21%

Mode of Delivery

C/S (291) 53% (291) 53% (582) 53%

Gestation (wks)

Preterm (34-37) (148) 27% (50) 9% (198) 17%

Infant Feeding

Breastfeeding (438) 80% (460) 83% (898) 82%

Maternal Age (years)

Adolescent Mothers (<24) (157) 29% (102) 19% (259) 24%

Maternal Population Group

Black African (433) 79% (454) 82% (887) 81%

Results – Demographics

Results – Regression AnalysisUnivariate Model

Risk Factors Categories Crude OR 95% CI

Receipt of Birth Test No Birth Test

Received Birth Test

-

2.17

-

(1.61-2.93)

Sex Male

Female

-

0.92

-

(0.69-1.23)

Birthweight Weight (g)

Normal Birthweight

Low Birthweight (<2500g)

1.00

-

0.79

(0.99-1.00)

-

(0.56-1.12)

Maternal Age Age (years)

Mature Mother (≥24)

Adolescent Mother (<24)

1.01

-

0.86

(0.98-1.03)

-

(0.62-1.21)

Gestation Gestation (wks)

Full Gestation

Preterm (<37)

1.00

-

0.93

(0.95-1.07)

-

(0.64-1.35)

Maternal Population

Group

Black African

Coloured

Foreign

-

0.69

0.37

-

(0.43-1.09)

(0.06-2.25)

PMTCT Coverage Received ART 12+weeks

Received ART <12weeks

No ART

-

2.03

3.24

-

(1.44-2.86)

(1.85-5.66)

Default on Treatment No recorded default

Default recorded

-

0.94

-

(0.52-1.69)

Viral Load Vl ≥1000

VL >1000

-

1.42

-

(0.85-2.37)

Multivariate Model

Adjusted OR 95% CI

-

1.68 (1.16-2.43)

(not included)

-

(0.69-1.23)

-

1.02

-

(0.70-1.48)

-

1.03

-

(0.73-1.44)

(not included)

0.95-1.07)

-

(0.64-1.35)

(not included)

-

(0.43-1.09)

(0.06-2.25)

-

1.43

2.33

-

(0.94-2.15)

(1.28-4.26)

(not included)

-

(0.52-1.69)

(not included)

-

(0.85-2.37)

Results – Subgroup analyses

Adjusted OddsRatio

Restricted analyses showing effect size in multiple subpopulations: Association between non-attendance at follow-up EID testing for infants who received a birth test compared to those that did not

All

HIV

exp

osed In

fants

Pre

term

Full Ges

tatio

n

Low B

irth

wei

ght

Norm

al B

irthw

eight

Adole

scen

t Moth

er

Mat

ure M

other

Popula

tion G

roup-B

lack

Afric

an

Popula

tion G

roup-O

ther

0.1

1

10

AdOR for non-reportingat follow up EID testing forno HIV PCR test receivedat birth

All LBW Norm BW

Pre-term

Full term

Mother <24yrs

Mother >24yrs

Black African

Other

All

HIV

exp

osed In

fants

Pre

term

Full Ges

tatio

n

Low B

irth

wei

ght

Norm

al B

irthw

eight

Adole

scen

t Moth

er

Mat

ure M

other

Popula

tion G

roup-B

lack

Afric

an

Popula

tion G

roup-O

ther

0.1

1

10

AdOR for non-reportingat follow up EID testing forno HIV PCR test receivedat birth

Limitations

• Results reflect a single urban setting within South Africa. All participants delivered at an obstetric hospital

– Generalizations should be made with caution

• Social and demographic variables would provide greater insight into predictors for non-attendance

– Risk factors used to identify infants as high risk at delivery could also be predictors for non-attendance

Conclusions

1. Targeted birth testing successfully identified mothers at high risk for transmission to their infants

2. Neonates undergoing HIV testing at birth appear less likely to receive subsequent EID testing compared to infants who did not receive a birth test. – More emphasis on negative results

– How does universal birth testing affect these results?

Acknowledgements

Research supported by the Elizabeth Glaser Pediatric AIDS Foundation

Leigh Anne van Balla & Alex Paone for their contribution during data collection, along with all the

staff at MMH.