Robert S. Remis, Carol Major, Carol Swantee, Margaret Fearon, Evelyn Wallace, Elaine Whittingham
Post on 13-Jan-2016
41 Views
Preview:
DESCRIPTION
Transcript
Using the detuned assay to Using the detuned assay to determine HIV incidence determine HIV incidence in Ontario: Results and in Ontario: Results and methodologic perspectivesmethodologic perspectives
Robert S. Remis, Carol Major, Carol Swantee, Robert S. Remis, Carol Major, Carol Swantee, Margaret Fearon, Evelyn Wallace, Elaine Margaret Fearon, Evelyn Wallace, Elaine WhittinghamWhittinghamDepartment of Public Health Sciences, University of TorontoDepartment of Public Health Sciences, University of Toronto
HIV Laboratory, Laboratory Services, HIV Laboratory, Laboratory Services, Ontario Ministry of Health and Long-Term CareOntario Ministry of Health and Long-Term Care
Public Health Branch, Public Health Branch, Ontario Ministry of Health and Long-Term CareOntario Ministry of Health and Long-Term Care
STAHRS WorkshopSTAHRS WorkshopCenters for Disease ControlCenters for Disease Control
Albany, New York, USA, November Albany, New York, USA, November 14-16, 200114-16, 2001
MOHLTC, Laboratories Branch, IMC – 2001
AcknowledgementsAcknowledgements• At the HIV LaboratoryAt the HIV Laboratory
• Lisa Santangelo and Cindi Farina, data collectionLisa Santangelo and Cindi Farina, data collection• Lynda Healey, detuned assayLynda Healey, detuned assay• Elaine McFarlane, data entry screens Elaine McFarlane, data entry screens • Len Neglia, mailout of questionnairesLen Neglia, mailout of questionnaires
• Regional PHLs, mailout of questionnairesRegional PHLs, mailout of questionnaires• Physicians who prescribe HIV testing, supplementary Physicians who prescribe HIV testing, supplementary
datadata• Frank McGee, AIDS Bureau for base fundingFrank McGee, AIDS Bureau for base funding• Ontario HIV Treatment Network for initial project Ontario HIV Treatment Network for initial project
fundingfunding• CIDPC, Health Canada for continued project fundingCIDPC, Health Canada for continued project funding
MOHLTC, Laboratories Branch, IMC – 2001
IntroductionIntroduction• Serodiagnostic data useful for surveillanceSerodiagnostic data useful for surveillance• However,However,
• persons who test may not be representativepersons who test may not be representative• data quality often poordata quality often poor
• Unbiased estimates of HIV incidence and Unbiased estimates of HIV incidence and prevalence prevalence cannotcannot be derived directly be derived directly
MOHLTC, Laboratories Branch, IMC – 2001
IntroductionIntroduction• Testing of HIV-positive specimens using Testing of HIV-positive specimens using
less sensitive (“detuned”) assay permits less sensitive (“detuned”) assay permits the identification of persons who recently the identification of persons who recently seroconverted;seroconverted;
• Can calculate HIV incidence density, a critical Can calculate HIV incidence density, a critical indicator otherwise difficult to measureindicator otherwise difficult to measure
MOHLTC, Laboratories Branch, IMC – 2001
Study objectivesStudy objectives• Determine the number of persons newly Determine the number of persons newly
testing positive for HIV testing positive for HIV • Determine the distribution of exposure Determine the distribution of exposure
category among newly diagnosed HIV-category among newly diagnosed HIV-infected personsinfected persons
• Estimate HIV incidence density among Estimate HIV incidence density among persons undergoing HIV testing persons undergoing HIV testing
MOHLTC, Laboratories Branch, IMC – 2001
Data collection and managementData collection and management• Questionnaire sent with HIV-positive results Questionnaire sent with HIV-positive results
and 1:200 sample of HIV-negative resultsand 1:200 sample of HIV-negative results• Coolect data on risk factors for HIV infection Coolect data on risk factors for HIV infection
and HIV test history and HIV test history• Questionnaire returned by mail, fax or Questionnaire returned by mail, fax or
telephone interviewtelephone interview• Data entered in Microsoft AccessData entered in Microsoft Access
MOHLTC, Laboratories Branch, IMC – 2001
Laboratory methodsLaboratory methods• Modified Abbott 3A11 EIA kit (Oct 1999-Oct Modified Abbott 3A11 EIA kit (Oct 1999-Oct
2000)2000)• Serum diluted to 1:20,000Serum diluted to 1:20,000• Incubation period reduced to 30 minutesIncubation period reduced to 30 minutes• Cut-off value increasedCut-off value increased
• Organon-Teknika (Oct 2000-Jul 2001)Organon-Teknika (Oct 2000-Jul 2001)• Similar principle to Abbott EIASimilar principle to Abbott EIA• Allows use of variable cut-off value Allows use of variable cut-off value
reflecting varying “window period” reflecting varying “window period”
MOHLTC, Laboratories Branch, IMC – 2001
Data analysisData analysis• NumeratorNumerator
• Non-reactive (discordant specimens) Non-reactive (discordant specimens) without without risk factors imputed to NIR risk factors imputed to NIR specimens based on reclassification from specimens based on reclassification from LES LES
• Initially, imputed as proportion of those with Initially, imputed as proportion of those with risk factor informationrisk factor information
• Denominators (testers) handled similarlyDenominators (testers) handled similarly• Incidence density =Incidence density =
• NR NR * 100* 100 Testers * (t / 365) Testers * (t / 365)
MOHLTC, Laboratories Branch, IMC – 2001
Study questionnaires mailed and Study questionnaires mailed and returned, Oct 1999 - Dec 2000returned, Oct 1999 - Dec 2000
QuestionnairesQuestionnairesKaplan - Meier Kaplan - Meier
returned byreturned by
MailedMailed ReturnedReturned ProportionProportion 4 mon.4 mon. 8 mon.8 mon.
HIV-positiveHIV-positive
HIV-negativeHIV-negative
TotalTotal
1,071 1,071
1,392 1,392
2,463 2,463
733 733
1,001 1,001
1,734 1,734
68%68%
72%72%
70%70%
72%72%
73%73%
79%79%
80%80%
MOHLTC, Laboratories Branch, IMC – 2001
Exposure category classified according to Exposure category classified according to HIV test requisition, returned questionnaires HIV test requisition, returned questionnaires and modeled distribution, HIV-positives and modeled distribution, HIV-positives
HIV test HIV test requisitionrequisition
Returned Returned questionnaires questionnaires
of NIRof NIRProjected finalProjected final
distributiondistribution
MSMMSM
MSM-IDUMSM-IDU
IDUIDU
EndemicEndemic
HR heteroHR hetero
LR heteroLR hetero
OtherOther
NIRNIR
307 307
15 15
68 68
33 33
30 30
7979
11 11
528 528
57%57%
3%3%
13%13%
6%6%
6%6%
15%15%
2%2%
115115
66
3030
6565
2121
6666
44
221221
37%37%
2%2%
10%10%
21%21%
7%7%
21%21%
1%1%
505 505
25 25
120 120
145 145
66 66
193193
18 18
47%47%
2%2%
11%11%
14%14%
6%6%
18%18%
2%2%
TotalTotal 1,0711,071 528528 100%100% 1,0711,071 100%100%
%NIR%NIR 49%49% 42%42%
MOHLTC, Laboratories Branch, IMC – 2001
Incidence (per 100 py) by Incidence (per 100 py) by exposure categoryexposure categoryOntario, Oct 1999 - Jul 2001Ontario, Oct 1999 - Jul 2001
2.34 2.34
1.82 1.82
0.44 0.44
MSMMSM
MSM-IDUMSM-IDU
IDUIDU
0.22 0.22
0.04 0.04
HR heteroHR hetero
LR heteroLR hetero
Incidence Incidence per 100 p-yper 100 p-y
Exposure Exposure categorycategory
MOHLTC, Laboratories Branch, IMC – 2001
Incidence (per 100 py) amomg Incidence (per 100 py) amomg MSM, MSM-IDU and IDU by health MSM, MSM-IDU and IDU by health region,region,Ontario, Oct 1999 Ontario, Oct 1999 –– Jul 2001 Jul 2001
0.44 0.44 1.82 1.82 2.34 2.34 OntarioOntario
0.51 0.51
1.36 1.36
0.24 0.24
6.32 6.32
18.118.1
0.24 0.24
3.38 3.38
0.82 0.82
1.38 1.38
TorontoToronto
OttawaOttawa
OtherOther
IDUIDUMSMMSM MSM-IDUMSM-IDU
MOHLTC, Laboratories Branch, IMC – 2001
Incidence (per 100 py) among LR Incidence (per 100 py) among LR and and HR heterosexual by health regionHR heterosexual by health regionOntario, Oct 1999 - Jul 2001Ontario, Oct 1999 - Jul 2001
0.040.040.220.22OntarioOntario
0.04 0.04
0.07 0.07
0.02 0.02
0.26 0.26
0.17 0.17
0.21 0.21
TorontoToronto
OttawaOttawa
OtherOther
HR heteroHR hetero LR heteroLR hetero
MOHLTC, Laboratories Branch, IMC – 2001
Incidence (per 100 py) among MSM Incidence (per 100 py) among MSM by health region and study period, by health region and study period, Ontario, Ontario, Oct 1999 - Jul 2001Oct 1999 - Jul 2001
0.0
1.0
2.0
3.0
4.0
5.0
6.0
1999-02 2000-01 2000-02 2001-01
Study period
HIV
inc
ide
nc
e (
pe
r 1
00
py
)
TorontoOttawaRest of Ontario
MOHLTC, Laboratories Branch, IMC – 2001
Incidence (per 100 py) among IDU Incidence (per 100 py) among IDU by health region and study period, by health region and study period,
Ontario, Ontario, Oct 1999 - Jul 2001Oct 1999 - Jul 2001
0.0
0.5
1.0
1.5
2.0
2.5
1999-02 2000-01 2000-02 2001-01
Quarter
HIV
inc
ide
nc
e (
pe
r 1
00
py
)
Toronto
Ottawa
Rest of Ontario
MOHLTC, Laboratories Branch, IMC – 2001
HIV incidence by age group, HIV incidence by age group, selected exposure categoriesselected exposure categories
0.0
1.0
2.0
3.0
4.0
5.0
6.0
15-24 25-29 30-34 35-39 40-44 45-49 50+
Age Group
HIV
inci
dence
(p
er
100 p
y)
MSMIDULR hetero
MOHLTC, Laboratories Branch, IMC – 2001
Incidence calculated for selected Incidence calculated for selected exposure categoriesexposure categories using different using different "window" "window" periods with the OT assay, Jan-Jul 2001periods with the OT assay, Jan-Jul 2001
0.0
0.5
1.0
1.5
2.0
2.5
133 170 336
Window period (days)
Inci
dence
(per
100 p
ers
on-y
ears
) MSM
MSM-IDU
IDU
HR hetero
LR hetero
MOHLTC, Laboratories Branch, IMC – 2001
InterpretationInterpretation
• Number of discordant samples and HIV tests Number of discordant samples and HIV tests by exposure category were modeled by exposure category were modeled
• Interpretation of HIV incidence must Interpretation of HIV incidence must incorporate knowledge of patterns in HIV test incorporate knowledge of patterns in HIV test seeking behavioursseeking behaviours
• Observed HIV incidence likely higher than for Observed HIV incidence likely higher than for actual population actual population
MOHLTC, Laboratories Branch, IMC – 2001
Methodologic issues 1Methodologic issues 1
• Risk factors unknown for significant proportion Risk factors unknown for significant proportion of both HIV-positive and HIV-negative testersof both HIV-positive and HIV-negative testers
• Distribution of those with unknown risk factors Distribution of those with unknown risk factors different than that among those with known different than that among those with known risk factorsrisk factors
MOHLTC, Laboratories Branch, IMC – 2001
Methodologic issues 2Methodologic issues 2
• Testers may include persons with risk Testers may include persons with risk behaviours in distant pastbehaviours in distant past
• This would tend to underestimate HIV This would tend to underestimate HIV incidenceincidence
MOHLTC, Laboratories Branch, IMC – 2001
Methodologic issues 3Methodologic issues 3
• For MSM, analysis with varying “window For MSM, analysis with varying “window period” in later samples showed period” in later samples showed substantial decrease in estimated HIV substantial decrease in estimated HIV incidence with longer interval but not incidence with longer interval but not for other groupsfor other groups
• Likely due to increased probability of Likely due to increased probability of HIV testing related to isolated high risk HIV testing related to isolated high risk exposures and seroconversion illnessexposures and seroconversion illness
MOHLTC, Laboratories Branch, IMC – 2001
ConclusionsConclusions
• Detuned assay a powerful tool to Detuned assay a powerful tool to estimate HIV incidence at low costestimate HIV incidence at low cost
• However, further work is required to However, further work is required to develop methodolgic approaches to develop methodolgic approaches to account for missing data, account for missing data, unrepresentative samples and sources unrepresentative samples and sources of bias related to HIV testing behavioursof bias related to HIV testing behaviours
top related