HIV Sentinel Surveillance 2006HIV Sentinel Surveillance 2006

Surveillance UnitNational Center for HIV/AIDS Dermatology and STD

15 August 2008

1

Outline

ObjectivesMethodologyMethodologyResults

lConclusionRecommendation

2

ObjectivesTo estimate the prevalence of HIV infection in two key female sentinel groups in 2006To estimate HIV incidence in these groupsTo estimate HIV prevalence in the general populationpopulation To obtain data for monitoring trends in HIV prevalenceprevalenceTo provide information on HIV prevalence for prevention planning and care efforts

3

MethodsStudy design: Cross sectional study (October 2006-February 2007)Sentinel groups:Sentinel groups:

Brothel based female sex workers (FSW)Pregnant women attending antenatal clinics (ANC)

Sentinel sites: 22 cities/provincesSentinel sites: 22 cities/provincesSample size:

FSW: 150 per provinceANC: 600 per province (300 from provincial capital/urban area and 300 from remaining districts/rural area)

Specimen collected: 5 ml of whole blood

4

Methods: 22 sentinel provinces

BMC

OMC

PVHSRP

RTKSTGS

BTBPLN KTM KRT MDK

PSTKCN

KCM

KHK KSP

TAK

PVG

SVRKAN

PHPIncluded

Not included

Tonle Sap Lake

5

KPTSHV

TAK

KEP

Methods: Sampling• FSW

• In provinces with less than the required sample size of 150, sampling was “take-all”

• In provinces with at least 150 FSWs, sampling was “take-all” from randomly selected brothels

• ANC• Separate samples of 300 women were selected from provincial

capitals (PC) and 300 from remaining districts (RD)

• Pregnant women were selected consecutively from designated ANC clinics at health centers until the required sample size was reachedclinics at health centers until the required sample size was reached

• Duration of data collection was limited to three months

6

Methods: sample size by survey year

2000 2002 2003 2006(Provinces) (21) (20) (22) (22)( ) ( ) ( ) ( ) ( )

DFSW 2,180 2,110 2,411 2,266

IDFSW 1 799 1 232 1 633IDFSW 1,799 1,232 1,633 --

Police 4,711 4,379 5,796 --

ANC 6,562 9,168 10,867 12,464

Total 17 991 19 247 20 707 14 730

7

Total 17,991 19,247 20,707 14,730

Methods: HIV testingT f d h i i l l lTests performed at the provincial level

Two rapid tests were used (Determine & Stat-Pak)A serial two-test algorithm was used for FSW and a parallel two-test algorithm was used for ANCgDried blood spot (DBS) specimens were prepared for quality control

Quality control testing performed by NIPH lab FSWFSW:

Phase 1: 10% of all specimens were testedPhase 2: All reactive specimens not tested in phase 1 were tested

ANC: All reactive specimen and 10% of all non-reactive specimens were tested

Two enzyme immunoassays were used for quality control testing (Vironostika & Murex)

8

Methods: HIV prevalence testing algorithm

Se=100 0%Determine

Female sex workers ANC women

DetermineSe 100.0%Sp=99.8%

Negative

Determine

Positive Negative

Determine

Positive

Se=99.5%Sp=100.0%Stat-Pak Stat-Pak Stat-Pak

Positive Negative Positive Negative Positive? NegativeR t b th t t

9Sensitivity (Se) and specificity (Sp) data from phase 1 validation of 5 rapid HIV tests conducted in Cambodia in 2004. The combined algorithm was found to be 99.5% sensitive and 100% specific.

Repeat both tests

Methods: Quality control testing algorithm

DBS specimens selected for QC testing

Vironostika

Murex

Positive

NegativePositive Negative

10

NegativePositive Negative

Methods: HIV incidence estimationT ti f t i f tiTesting for recent infection

All DBS specimens identified as HIV positive were tested for recent infectionC l t HIV 1 I id BED EIA dCalypte HIV-1 Incidence BED EIA was usedSpecimens with normalized optic density (OD_n) values ≤0.8 on confirmatory testing were considered to be from persons with recent HIV infectionpersons with recent HIV infection

Calculating HIV incidenceOverall window period used was 155 daysHIV i id li d i th f lHIV incidence was annualized using the formula

I =({(365/155)Ninc}/{Nneg+[(365/155)Ninc]/2})*100where: N : number of recent infections

11

where: Ninc: number of recent infectionsNneg: number of HIV seronegatives

Results: HIV prevalence* in 2006, by sentinel group

14.014.0

16.0

8.0

10.0

12.0

erce

nt

0.92.0

4.0

6.0Pe

0.0

FSW ANC**

12

*Adjusted for results of quality control**Weighted for provincial population size

Results: HIV l f l kResults: HIV prevalence among female sex workers, by survey year

44.742.6

39.936.840

45

50

42 3 42 5 45.8

33.630.2

26.823.4

20.317 4

25

30

35

42.3 42.5 45.841.3 41.9

25.721.4

14

17.414.7

5

10

15

20

0

5

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

QC adjusted QC adjusted EPP smoothed

13

QC-adjusted QC-adjusted, EPP smoothed

Results: HIV l * f l kResults: HIV prevalence* among female sex workers, by age group

43.540.4

34.2

41.340

45

50

34.2

29.9

24.8

32.2

24.1

20

25

30

35

Per

cent

14.419.7

7.85

10

15

20P

2.90

5

1998 1999 2000 2001 2002 2003 2004 2005 2006

20 ld d ld <20 ld

14

20 years old and older <20 years old

*Adjusted for results of quality control

Results: HIV l * f l kResults: HIV prevalence* among female sex workers, by survey year and duration of sex work

30.8

37.3

34.635

40

26.9

22.1

34.6

22.920

25

30

rcen

t

18.6

9.2 10.310

15

Per

0

5

< 1 Year 1-2 Years > 2 Years

2002 2003 2006

15

2002 2003 2006

*Adjusted for results of quality control

Results: HIV l * f l kResults: HIV prevalence* among female sex workers, by age and duration of sex work

16.0

14

16

18

10.8

8

10

12

Per

cent

2.8 2.9

2

4

6

P

06 months or less More than 6 months

Duration of work

16

Aged <20 yrs old Aged 20 yrs or more

*Adjusted for results of quality control

Results: HIV prevalence* among FSWs byResults: HIV prevalence among FSWs, by province

12 6

20.6

25.8

30.726.7

20.7

20.4Odd M h

Siem ReapBattambang

Koh KongKampong Speu

SihanoukvilleBanteay Meanchey

10 5

11.1

11.5

12.612.5

11.3

11.1Pailin

Preah VihearKampong Cham

Phnom PenhStung Treng

Kampong ChhnangOddar Meanchey

rovi

nce

5.3

9.1

10.1

10.510.4

10.0

5.6Rattanakiri

KampotKratie

Svay RiengKampong Thom

PursatPailinPr

4.2

5.34.4

3.1

0 5 10 15 20 25 30 35

KandalTakeo

Prey Veng

S

17

HIV Seroprevalence Percent

*Adjusted for results of quality control

Results: HIV prevalence among ANC, by survey year

3.0

3.5

1.92.1 2.1 2.1 2.0 1.8 1.6

1.5 1.3 1 1

1.51 5

2.0

2.5

1.82.5

2.9

1.92.4 2.2

0.9

1.3 1.1

0.5

1.0

1.5

0.01996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

QC adjusted weighted QC adjusted weighted and EPP smoothed

18

QC-adjusted, weighted QC-adjusted, weighted, and EPP smoothed

Results: T d i ti t d HIV l *Results: Trends in estimated HIV prevalence* among ANC women, by ANC location

2.62.8 2.8 2.7

2.52.32.5

3.0

1.5

2.1 2.11.9

1.71.4

1.82.0 2.0 2.0 1.9

1 71.5

1.92.1 2.1 2.1

2.0

1.51 3

1.61.8

1 5

2.0

2.5

cent

1.4

1.8 1.71.6

1.41.2

1.11.0

1.0 1.11.3

0 5

1.0

1.5

Perc

0.0

0.5

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

19

Urban Rural Total

* QC-adjusted, weighted, and smoothed with EPP

Result: Estimated HIV prevalence* amongResult: Estimated HIV prevalence* among general population aged 15-49 years old3 0

2.32.5

2.62.4

2.12.02 0

2.5

3.0

1.8 1.81.6

1.41.2

1.11.51.7

1.91.7

1.5 1 5 1.11.2

1.41.51.6

1.92.0

1.91.7

1.21.5

2.0

Perc

ent

1.1

1.51.3

1.21.1

0.90.8

0.91.0

0.5

1.0

P

0.01995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006

20

Urban Rural Total

*Qc adjusted, weighted and EPP smoothed

Results: HIV prevalence* among ANC byResults: HIV prevalence among ANC, by province

0 7

0.5

0.5

0.20.2

0.5

0.6Si R

Preah VihearStung Treng

KratieSvay Rieng

RattanaakiriOddar Meanchey

0 8

0.7

0.7

0.70.7

0.7

0.7Kampong Cham

PursatKandalTakeo

Kampong SpeuKampong Thom

Siem Reap

rovi

nce

1.5

1

0.8

0.80.8

0.9

1.3Battambang

Bantey MeancheyKg_Chhnang

KampotPhnom Penh

Prey VengKampong ChamPr

1.71.6

2.1

0.0 0.5 1.0 1.5 2.0 2.5

PailinSihanoukville

Koh kong

HIV S l P t

21

HIV Seroprevalence Percent

*Adjusted for results of quality control

Estimating HIV incidence

Annualized period (365 days)

Assumed window period (155 days)

In HSS 2006:

I =({(365/155)Ninc}/{Nneg+[(365/155)Ninc]/2})*100

In HSS 2006: All new infection was assumed to occur within 155 day of the cohort if ll ti i tall participants were

followed up Annualized HIV incidence (HIV incidence among 100 person per year)

22Adjusted annualized incidence

Results: HIV incidence* among FSWs inResults: HIV incidence* among FSWs in 2006

2.552.5

3

1.681.5

2

rce

nt

0.801

Pe

r

0

0.5

Without

23

Ajustment methods* With 95% confidence interval

Results: HIV incidence* among FSWs byResults: HIV incidence* among FSWs, by survey year

10.2

10

12

7.4

6.0

8

ent

4

6

Per

ce

1.7

0

2

1999 2000 2001 2002 2003 2004 2005 2006

24*Uncorrected incidence

1999 2000 2001 2002 2003 2004 2005 2006

Results: HIV incidence* among ANC womenResults: HIV incidence* among ANC women in 2006, by adjustment method

0 35

0.29

0.25

0.30

0.35

0.170.15

0.20

0.25

Per

cent

0.07 0.060.06

0.12 0.11

0.05

0.10

0.02 0.020.00

No adjustment Hargrove McDougalAjustment methods

25

* With 95% confidence interval

Results: HIV incidence* among ANCResults: HIV incidence among ANC women, by survey year

0.9

0 70.8

0.9

1.0

0.6

0.7

0.5

0.6

0.7

erce

nt

0.170.2

0.3

0.4Pe

0.0

0.1

1999 2000 2001 2002 2003 2004 2005 2006

26

* Unadjusted incidence

Results: Projected new HIV infectionResults: Projected new HIV infection among adult population aged 15+

700700

800

540

550

480

410360

630

400

500

600

mbe

r

420

330

270230

360

300260

200

300

400

Num

230200

0

100

2006 2007 2008 2009 2010 2011 2012

27

Women Men

ConclusionsHIV l h d li d b h i l fHIV prevalence has declined among both sentinel groups of female sex workers and pregnant women attending antenatal clinicsData* suggest that some female sex workers may already be ata suggest t at so e e a e se o e s ay a eady beinfected with HIV when they start sex work or become infected almost immediately after startingHIV prevalence among pregnant women who attend provincial capital/urban ANCs has been consistently higher than amongcapital/urban ANCs has been consistently higher than among those who attend remaining district/rural ANCsHIV incidence has declined among both sentinel groups and it is estimated that among 100,000 pregnant women, 70 were newly infected in 2006infected in 2006.Although unadjusted HIV incidence among FSWs and ANCs is probably an over-estimate of true incidence, the declining trend and magnitude of the decline is consistent with prevalence data

28*9% HIV prevalence among FSWs who had reported working for less than one year

RecommendationsI t t t ti f t i f ti i t ll f tIntegrate testing for recent infection into all future rounds of HSSMaintain or improve adherence to quality assurance and capacity building for surveillance at provincial levelContinue to use dried blood spot specimens in future HSS rounds because of simplicity in preparing,HSS rounds because of simplicity in preparing, processing and storingMaintain current interventions among most at risk populations—despite declining prevalence and incidencepopulations—despite declining prevalence and incidence among both sentinel groups, this step will be crucial in preventing further spread of HIV and another wave of the epidemic

29

the epidemic

Acknowledgementsl f / l• National Center for HIV/AIDS, Dermatology

and STDs (NCHADS)P i i l H lth D t t d P i i l• Provincial Health Department and Provincial AIDS ProgramsNational Institute of Public Health (NIPH)• National Institute of Public Health (NIPH)

• Collaborating partners:US CDC Gl b l AIDS P• US CDC Global AIDS Program

• USAID/PRASIT• World Health Organization

30

• World Health Organization

31