Human immunodeficiency vi rus as a sexually transmitted disease : Manitoba ' s HIV unlinked seroprevalence study

OBJECTIVE
To determine the prevalence of human immunodeficiency virus (HIV)-1 infections.


DESIGN
An unlinked seroprevalence study on all 'left over' blood submitted for syphilis testing during three consecutive six month periods (April 1, 1990 to September 30, 1991). After the capture of demographic data, sera to be tested for antibodies were stripped of all personal identifiers and assigned a random number. The study fulfilled Canadian ethical guidelines.


PARTICIPANTS
Persons with a valid Manitoba Health Services Commission number were included only once in a study year. Persons were enrolled into one of five categories: infants, injection drug users, pregnant women, persons investigated for sexually transmitted diseases (STD) and routine testing.


RESULTS
Of the 50,929 persons tested. 133 (0.26%) were positive, a low prevalence of HIV infections. A significant increase in positivity was noticed in the second study period. While 77.7% of those tested were women, only 10 (0.025%) were positive for HIV. The rate of positivity for pregnant women was 0.72 per 10,000 tested. Of the 11,314 males tested. 123 (1.08%) had antibodies. Most of the positives were males in the STD category residing in Winnipeg; for them, the rate of positivity increased by middle age, while the rate of testing decreased with each decade of age (from age 15 to 44 years). DISCUSSION/RECOMMENDATION: This is the first report from a sentinel laboratory in Canada of an unlinked seroprevalence study in persons with STD. Results suggest an urgent need to target strategies for the control of HIV for persons with STD, particularly males in urban centres, by routine HIV testing for STD patients.

T HE PREVALE C E OF HUMA ' IMMUNODEFIC ISNCY V IRUS (H!V) -1 infection in Canada is unknown : yet U1is information is essential to target healUl resources.To obtain accurate data.Ule Royal Society of Canada (l) recommended anonymous unlinked eroprevalence studies.and Ule Federal Centre for Acquired Immun e Deficiency Syndrome (AJDS) set up <Ju id elines for such tudies to meet legal.eU1ical and scientific requirements (2).The National H a lUl Re earch and Development Program provided funds for proposals subm itted from various laboratories across Canada.Results of U1e 18-monUl Manitoba unlinked HlV seroprevalen cc study are presented.This report.which focuses on persons tested for a sexually transmitted disease (STD).represents the first such report in Canada.

METHODOLOGY
The study design has been d etailed previou ly (3).BrieDy, all 'left over• blood submilted to Ule Cadh am Provincial Laboratory for syphilis testin<J were reviewed for inclus ion in the HlV seroprevalence study (this labo ratory is Ule sole provider of ti1is diagnostic testing in Manitoba).Persons wiUl a valid Manitoba HealU1 Services Comm ission number were included only once during each study year.A mainframe computer progr am verified and sto red relevant demographic data (age.sex, region of residence and patient enrollm ent category).and a randomly generated uniqu e number was assigned to each serum t steel for a ntibody to HlV.All personal identifiers were destroyed before testing.
Based on informat ion obtained from specimen requis ition.patient sera were divided into five categori es: infants (serodiagnostic testing of sick babies).injection drug users.pregnant women (all are routinely tested for syp hilis ).persons with STD or ilieir contacts .or persons tested for routin e reasons such as visa.admission to hospita ls , etc.The STD categ01y included persons wiU1 STD symptoms as well as persons (age 15 to 30 years) s en in clinics identified as caring for a 'high risk STD clientele .A<Je. sex and regional data for the composition of Manitoba's population were obtained from Manitoba Healti1 Services Commission.
From April l. 1990 to September 30.1991.50.929 persons were ntered into ti1e study.and U1eir sera were tested for antibody to HIV (3).A total of 2789 per ons tested during the first year were retested during ilie last six monilis of U1e study (2207 of U1ese were women).The total number of persons enrolled in the study by age group .sex. region of residence and patient catego ry were entered in a personal comp uter database.along wiUl data on Manitoba's population.A tabulation by sex.age, region and patient catego ry.maintained at Cadh am Provincial Laborat01y for all positive specimens.was m erged with Ule database to permit calculation of rates of positivity a nd testing.Statistical analysis software was used in ti1e statistical analys is.Rates were calcu lated per 10.000 population (or 10.000 per ons te ted).The 95% confidence interva ls for rates of positivity were calcu lated using tile Poisson dislribution.In instances where less U1an 100 samples were tested.conference intervals wen• not reported.

RESULTS
The numbe r of males and females tested during each monlh of U1e 18-monUl study for U1e m a in enrollment catego ri e is presented in Figu re 1. Th e numbe r of persons tested each monlh remained relatively conslant U1rough out U1e study a nd only a small fraction of U1 es were in the STD category.
Table 1 provides an overa ll study ummary.a nd indicates U1e number of sera tested and U1e results obtained in each patient ca tegory.One hundred and U1irty-U1ree persons were found lo have anUbod ies lo HN (0.26%).The number of HN-pos iUve m a les a nd females by e nrollment category for each of U1e sludy"s 18 months is illustrated in Fig ure 2 .which clearly shows Ulat almost aU lhe posilives were in Ule STD category and U1at more posi lives we re detected in 199 1 U1 an in Lhe corresponding nine-month period in 1990.Females represented 77.7% of U1e persons tested.but on ly 7.5% of the posilives.with an overa ll HIV prevalence of 0.025% .Only two of Ule 27.627 pregnant women tested had HIV a nlibod.ies.fo r a rate of 0. 72 per 10.000 (confidence interval: 0.1 lo 2 .6).
In Ule rouline L est a nd STD categories.an a lm ost equal number of males and females we re te led.The re were s ignificanUy more positive males U1 an females in bolli these categories (P=0.003 and P<O.OO l. respectively).Overall.123 of U1e 11 .314 m a les tes ted (1.08%) had HIV anlibod.ies and 111 of U1e 123 posilive males (90.2%) were in lh e STD cateo•ory.
The lola! number of persons tested was a lmost identical in each of Ule Ulree conseculive six-monlli periods studied.yet a ignificant increase in Ule number of po itives was noUced in th e second six-month period when compared \villi U1 e first six monU1s (P<O .OO 1) (Figure 3).Since no positives were detected in lhe 137 infants and 83 inj ection drug users tested .U1ese categories a re not shOW11 in Figure 3 .
An a n a ly is of the age group.sex.region or residen ce and enrollm ent catego ry of lhe 54 positives de tec ted from April l to September 30.1991.suggests Ulat a maximum of 13 persons may have tested positive during llie preceding 12 -monlli period.The study design ens ured Ulat Ule 79 person s found posilivc during U1e first 12 monU1s of U1 e study had been tes ted on ly on ce durin <1 Ulat period .Similarly .individuals in lhe last six monU1s of the study were tes ted only once.
Approximately 56% of Ma nitoba's population o f 1.130.845reside in Winnipeg: of U1e 133 positives.111 (83.4%) reside in Winnipeg.While Ule ra te of tesling decreases wiU1 age fo r Winnipeg ma les in U1e STD category.the rate of HIV positivity increases for each decade of age (from 15 to 44) (Figure 4 ).A similar pattern (data not shown) was noticed in Ul e 16 HIVpo itive males in U1e STD cate<1ory who resided outs ide of Winnipeg.Th is obse rvation may pe rh aps be explained by th e long incuba tion period of the HIV viru and the fact lh at a ntibodies usu a lly rem a in detec table HIV as STD DISCUSSION This is Ul e first report in Canada from a public health la bora tory of an unl inked HTV seroprevalence study in pe rson s investigated for an STD using blood drawn fo r syph ilis serology .Ma nitoba's health <1uidelines recommend tes ung for syphilis for all pregnant women a nd a ll 1991 10 Seplember 30.199 I: T Upper lim it q( 95% corlficlence patients with STD.In Manitoba.a lm ost a ll pregnant women and all women sc heduled for abortion a re tested for syphili s .but the compliance for s u ch testing for STD patients is unlmown.A questionna ire sent to Ma nitoba physicians a fte r the co mpl etion of this seroprevale nce study indicated that only 50% of responding practitioners comp ly wiU1 the syphili s testing guidelin es (6) .
All laboratory-based s urveys .such as this study.have oU1e r inherent limitation becau se laboratories have liltle co n trol over the information acco mpa nying the specime ns to be tested.partic ula rly with rega rd to U1 e exact clini cal data .Despite these limita tions.Manitoba• unlinked seraprevale n ce study h as s hown U1at the overall I-IN positivity is low in Ma nitoba (0.26%).The prevalen ce of I-IN in pregnant women is mu c h lower U1a n the 2.7 pe r 10.000 pregnant wome n reported in British Co lum bia (4) or U1e 6.1 per 10.000 live births in Quebec (5). Thi study s ugges ts Lhat strategies fo r the co ntrol of HIV s h o uld be focu sed on males a nd on persons with a histo ry of STD.part ic ularly thosc residing in la rge

Figure 1 )
Figure 1) The number of males andfemales testedfor each of the main enrollment categories for each of the 18 months of study.-Male sexually transmitted diseases: D Male routine: [J §[] Pemale sexually transmitted diseases:

Figure 2 )Figure 3 )
Figure 2) The nwnbe,. of males andfemctles found positive for human immunodeficiency virus (HJV] antibodies for each enroll• ment category in which positives were detected andjor each of the 18 monl.hs of study.-Male sexually transmitted diseases: D Male routine: B Pemale sexuall.ytransmitted diseases: ~ Pemale routine

Figure 4 )
Figure 4) Rate of testing contrasts with the rate q{ human immw10df!ficiencyvirus (HTV] pos itivity for jive 10-year age groups oJmales in sexua lly transmilied diseases category res iding in Winnipeg .T Upper limit of 95% conjidence

TABLE 1
Manitoba unlinked human immunodeficiency se roprevalence study: April 1, 1990 to Se ptember30, 1991 Cl Co nfidenc e interval: IOU Inj ec tio n drug user: STD Sexually transmitted diseases lor life.possibly signifying U1a t U1 e HIV infection s detected we re old HTV infec tions .