293 Full
293 Full
293 Full
Methods
Study recruitment was conducted over a 6-month period from March to August 1998 through 4 locations in Greater
Vancouver: 2 “street outreach clinics” operating in the Vancouver downtown area, one of which mainly serves a street youth
population, whereas the other clinic, which is located near to the Vancouver needle exchange facility, was frequented
predominantly by injec- tion drug users; a needle exchange facility serving injection drug users in Surrey; and an STD clinic
located in Vancouver.
All those who attended the outreach clinics and the needle ex- change facility were invited to participate in this study, whereas
at the STD clinic the invitation was extended only to males. This re- striction was dictated by the feasibility of securing a
sufficient sample of MSM and heterosexual men who did not use drugs and were of similar social status. Excluded from the
study were those individuals who said that they had already been immunized against hepatitis A. At each location, recruitment
was conducted until about 120 individuals were enrolled. The rate of refusal to participate in the study was not recorded.
Saliva collection was facilitated by the use of Salivette (Sarstedt, Rommelsdorf, Germany) with a neutral cotton insert. Study
par- ticipants were asked to place the insert into the mouth, chewing occasionally until the insert was wet. Within 24 hours of
collec- tion, the saliva was recovered by centrifugation, divided into aliquots and frozen at –70° C until assayed. The presence of
HAV- specific IgG was evaluated using an ultrasensitive capture enzyme immunoassay–based method, which demonstrates 99%
sensitivity and specificity when compared with serum-based tests.10
Information about demographics, sexual orientation and prac- tices, previous injection drug use and hours per day spent on the
streets was collected using a 1-page, self-administered, structured questionnaire. The questionnaire was based on a similar instru-
ment used successfully to survey Vancouver MSM at the onset of
294 JAMC • 7 AOÛT 2001; 165 (3)
a hepatitis A outbreak. The study design, including the question- naire, was reviewed and approved by the Clinical Research
Ethics Board at the University of British Columbia
Individuals who reported having injected drugs in the past were identified as injection drug users (IDUs), whereas males who
reported a homosexual or bisexual orientation were identified as men who have sex with men (MSM). To qualify as street youth,
respondents had to be less than 25 years of age and had to have reported spending at least 8 hours daily on the streets.
The association between potential risk factors and groups and past infection with HAV was evaluated using stepdown logistic
re- gression. Goodness-of-fit statistics were examined to determine the appropriateness of the final model.
Besides the 3 groups (MSM, IDUs and street youth) who were originally thought to be at increased risk for HAV infec- tion,
the initial model also included sex, age group and birth in a country where there was an increased risk of HAV infection, the last
2 being well-recognized factors associated with past HAV infection. The 2-way interactions between age and hypothesized risk
groups were examined, as were the 2- and 3-way interac- tions among the risk groups themselves. The site of recruitment was not
considered to be a separate risk factor, because it was self-evident that there was a substantial overlap of the risk groups under
investigation due to the specialized nature of these clinics.
Results
A total of 494 individuals were recruited into the study (Table 1). The majority (77.5%) of the 111 study partici-
pants who met street youth criteria were enrolled through a street youth–oriented, downtown outreach clinic,
whereas most of the 235 subjects who reported past injection drug use were recruited through the needle exchange
facility (41.7%) and through the second downtown outreach clinic (42.1%). An MSM orientation was reported by 51
males,
Table 1: Prevalence of anti-HAV antibodies among Vancouver street youth, injection drug users and men
who have sex with men
Study subjects
Risk group
Total no.
% positive for anti-HAV
Mean age , yr (and SD)
IDUs (all) 235 42.6 34.6 (10.5) MSM (all) 51 25.5 33.7 (12.2) IDUs who are MSM 13 23.1 34.5 (13.0) Street youth
(all) 111 6.3 19.6 (2.7) Street youth who are IDUs 42 9.5 20.4 (2.6) Street youth who are not MSM or IDU 64 3.1 19.0
(2.5) Born in countries where HAV is prevalent (all) 32 40.6 34.6 (11.8)
Total 494 27.9 31.9 (11.2)
Note: HAV = hepatitis A virus, SD = standard deviation, IDU = injection drug user, MSM = men who have sex with
men.
most of whom were recruited at the STD clinic (54.9%) and at the youth-oriented outreach clinic (25.5%). Most
(71%) of the 157 participants who did not fall into any of the 3 risk groups were re- cruited at the STD clinic.
The presence of HAV-specific IgG was de- tected in saliva specimens from 138 individuals, resulting in an
overall prevalence rate of 27.9% (Table 1). Evidence of past HAV infection was detected in 7 of 111 participants
who met the street youth criteria, indicating a prevalence of 6.3% (95% confidence interval [CI] 2.6–12.6). Street
youth who were IDUs had higher rates of past infection with hepatitis A (4/42, 9.5%, 95% CI 2.7%–22.6%)
compared with street youth who denied a history of injection drug use (3/69, 4.3%, 95% CI 0.9%–12.2%), although
these dif- ferences were not statistically significant. The presence of anti-HAV was almost 3 times more common
among all IDUs (100/235, 42.6%, 95% CI 36.2%–48.9%) than among subjects who de- nied injection drug use
(38/258, 14.7%, 95% CI 10.4%–19.1%). The differences in rates of previ- ous HAV infection among IDUs and
non-IDUs were most pronounced in those aged 25–34 and 35–44 years and not present among those aged 45 years
and above, as illustrated in Fig. 1. Anti- HAV prevalence among MSM (13/51, 25.5%) was almost identical to that
of heterosexual men (83/315, 26.3%). After the exclusion of IDUs, however, it was more than twice as high among
MSM (10/38, 26.3%) than among heterosexuals (21/175, 12%).
Logistic regression identified 2 factors to be significantly associated with increased anti-HAV prevalence (Table
2): most obviously with greater age but also with birth in a country where there was increased risk for HAV
infection. IDUs and MSM were also at higher risk for past HAV in- fection, although these associations were nomi-
nally significant.
The regression analysis also identified 2 inter- actions that were significantly associated with in- creased
anti-HAV prevalence. First, an interac- tion between IDUs and age (p = 0.05) indicates that injection drug use
especially among young adults (25–34 years old) is a significant risk factor for a positive anti-HAV test (p = 0.009).
Second, a significant interaction between IDUs and MSM (p = 0.03) indicates that among MSM injection drug use
is not associated with the increased presence of anti-HAV. Among those who are not MSM, however, the anti-HAV
rate was 3.9 times higher in IDUs than in those who denied injec- tion drug use (95/218, 43.6%) versus (24/213,
11.3%).
Past infection with hepatitis A virus
60
50
40
30
20
10
0
< 25 25–34 35–44 > 44 Age, yr
Injection drug users Participants who denied injection drug use
Fig. 1: Prevalence of antibodies against hepatitis A virus (HAV) according to age and self-reported injection
drug use.
Table 2: Factors and groups associated with anti-HAV antibodies
Variable
CMAJ • AUG. 7, 2001; 165 (3) 295
Unadjusted OR* (95% CI)
Adjusted OR*
(95% CI) p value
Sex Male 0.72 (0.46–1.13) 0.75 (0.43–1.30) 0.30 Female† 1.00 – 1.00 – – Age group, yr < 0.001 < 25 0.09
(0.04–0.20) 0.10 (0.06–0.31) < 0.001 25–34 0.40 (0.22–0.73) 0.14 (0.05–0.36) < 0.001 35–44 1.17 (0.65–2.11) 0.84
(0.29–2.42) 0.75 > 44† 1.00 – 1.00 – – Risk group‡ IDUs 4.70 (3.01–7.34) 2.48 (0.88–6.99) 0.09 MSM 0.90
(0.46–1.74) 2.39 (0.93–6.14) 0.07 Street youth 0.13 (0.06–0.29) 0.69 (0.18–2.60) 0.59 Born in countries where HAV
is prevalent 1.85 (0.86–3.94) 2.93 (1.13–7.58) 0.03 Interaction IDUs x MSM – 0.15 (0.03–0.80) 0.03 Age group, yr ×
IDUs – 0.05 < 25 × IDUs – 1.26 (0.26–6.18) 0.78 25–34 × IDUs – 6.45 (1.58–26.26) 0.009 35–44 × IDUs – 1.59
(0.43–5.84) 0.48
Note: OR = odds ratio, CI = confidence interval. *Ten individuals were not included in the crude and adjusted models
because of gaps in the completeness of their questionnaire. Two individuals identified as transsexuals were also
excluded. †Reference category. ‡Study subjects who did not belong to a given risk group constituted the reference
category.
Ochnio et al
Interpretation
Our findings stress once again that injection drug use is not only associated with an unfavourable course of HAV
infection due to the likelihood of underlying chronic liver diseases,11,12 but also with an increased risk for HAV
infection itself, particularly among young adults. The stable, as opposed to increasing, prevalence of anti-HAV in
the older IDUs could reflect a past pat- tern of risk reduction behaviour and the cyclic nature of HAV transmission.
The development and implementa- tion of a routine vaccination program for IDUs appears to be the only rational
approach to breaking the cycle of hepatitis A outbreaks and, in the long run, saving costs and health care resources.
Similarly, in the case of MSM, our limited data tend to support calls for a routine 2-dose vaccination program to
provide long-term protection, as opposed to sporadic 1-dose vaccine interventions as an outbreak control measure
requiring sudden bursts of public health activity.
The high, age-independent HAV infection rate ob- served among IDUs in Vancouver is in accordance with
previous reports associating injection drug use with in- creased risk for this infection.13–15 The difference in anti-
HAV rates between young adult IDUs and those who de- nied prior injection drug use should be interpreted with
some caution. The majority of participants who denied in- jection drug use were enrolled into the study through clin-
ics other than the needle exchange program, which also in- cluded an STD clinic located in a more affluent
Vancouver neighbourhood. It is possible that the observed low rate in this group reflects not only lack of
involvement in injection drug use but also different social background. Socioeco- nomic status is a significant factor
associated with the prevalence of HAV infection,16 and it could not be con- trolled for in our study.
Discrepancies in social background, however, are un- likely to play any significant role in the difference in anti-
HAV prevalence observed between MSM and heterosexual men, because the proportion of individuals enrolled
through the STD clinic into each group was almost identi- cal. The similar rates of anti-HAV prevalence among
MSM and heterosexual men are probably the result of there being a higher proportion of IDUs among the
heterosexual men (44%) than among MSM (25%) and the lack of any addi- tional effect of injection drug use on
anti-HAV rates in MSM. The absence of additional risk may be because the harm reduction message that has been
delivered to MSM has spilled over into their injection drug–using habits. Al- though the difference in anti-HAV rate
between MSM and heterosexual men was more than 2-fold when individuals who denied injection drug use were
considered, being MSM on its own did not emerge as a significant risk factor in logistic regression. This result may
have been a conse- quence of the relatively small number of MSM studied (51); they constituted only about 10% of
study participants,
296 JAMC • 7 AOÛT 2001; 165 (3)
Competing interests: None declared.
Contributors: Drs. Ochnio, Patrick and Dobson contributed to the design and exe- cution of this study. Ms. Ho was responsible
for the accuracy of laboratory results and for data analysis. Mr. Talling designed the logistic regression model used for evaluation
of risk factors. All the authors took part in the drafting and revising of the article.
Acknowledgements: We are indebted to the nursing staff of Vancouver Outreach and STD clinics and the Surrey Needle
Exchange for their work in study enrol- ment, supervision of specimen collection and data gathering. We also thank Dr. Robert
Strang, Dr. Danuta Skowronski and Mr. Jim Bennett for their continued support throughout the study in allowing access to
clinics.
This work was supported by a British Columbia Health Research Foundation grant.
and of these only 33 denied injection drug use and were not street youth.
Evidence of past HAV infection was present in only 6.3% of Vancouver street youth. This closely resembles the
anti-HAV prevalence of 7.1% (95% CI 4.1%–11.3%) detected 2 years earlier among grade six students in Van-
couver.17 More than 25% of the sixth graders, however, were born outside Canada, and this subgroup contained the
majority of anti-HAV-positive children, whereas the sample of street youth included only 5 individuals born
elsewhere. Although the anti-HAV positivity rate among Canadian-born street youth (6.6%) was twice as high as the
rate observed among Canadian-born grade six students (3%), it still should be considered to be relatively low. Such
a low rate means that almost all of Vancouver street youth are likely to be susceptible to HAV. When this vul-
nerability is considered in addition to the substandard liv- ing conditions of a high proportion of this population and
their exposure to other risk factors associated with HAV infection, such as injection drug use (often resulting in
needle sharing6) and involvement in the sex trade (“survival sex,” multiple sexual partners), it would seem that an
out- break of hepatitis A is just waiting to happen. The design and implementation of a routine hepatitis A
vaccination program for street youth is challenging, because it is diffi- cult to reach all of this elusive and dynamic
population, but street youth do tend to concentrate in relatively small areas in certain parts of the city. However, they
represent a heterogeneous group whose time spent on the streets may vary considerably.
Reports from countries such as Brazil where there is an increased risk of contracting HAV infection indicate a
very high prevalence of 80%–92% in street youth.18 The preva- lence of HAV infection in street youth in other
Canadian ur- ban populations or in other low-risk countries might be very different from Vancouver’s, reflecting
local outbreak histo- ries. Such outbreaks involving street youth, however, might be difficult to recognize and
attribute to this population, be- cause it is elusive, it overlaps considerably with IDUs and it is often characterized by
limited access to health care ser- vices. The availability of a vaccine, which is easy to use and, thus, a very effective
preventive tool, calls for investigation of HAV prevalence in street youth populations particularly in low-risk
countries that have not adopted universal hepatitis A vaccination programs for children and youth.
References
1. Bell BP, Shapiro CN, Alter MJ, Moyer LA, Judson FN, Mottram K, et al. The diverse patterns of hepatitis A epidemiology in
the United States — im- plications for vaccination strategies. J Infect Dis 1998;178:1579-84. 2. Notifiable disease annual
summary 1997. Can Commun Dis Rep 1999;25(S6):
119-21. 3. BC reportable communicable diseases, annual summary 1998. Vancouver: UBC
Centre for Disease Control; 1999. p. 23-5. 4. Advisory Committee on Immunization Practices. Prevention of hepatitis A
through active or passive immunization. MMWR Morb Mortal Wkly Rep 1999;48(RR12):1-37. 5. BC reportable communicable
diseases, annual summary 1997. Vancouver: UBC
Centre for Disease Control; 1998. p. 21-3. 6. Roy E, Lemire N, Haley N, Boivin JF, Frappier JV, Classens C. Injection
drug use among street youth. Can J Public Health 1998;89(4):239-40. 7. Roy E, Haley N, Lemire N, Boivin JF, Leclerc P,
Vincelette J. Hepatitis B virus infection among street youths in Montreal. CMAJ 1999;161(6):689-93. Available:
www.cma.ca/cmaj/vol-161/issue-6/0689.htm 8. Greene JM, Ennett ST, Ringwald CL. Prevalence and correlates of survival sex
among runaway and homeless youth. Am J Public Health 1999;89(9):1406-9. 9. MacDonald NE, Fisher WA, Wells GA, Doherty
JA, Bowie WR. Canadian street youth: correlates of sexual risk-taking activity. Pediatr Infect Disease J 1994;13(8):690-7. 10.
Ochnio JJ, Scheifele DW, Ho M, Mitchell LA. New, ultrasensitive enzyme immunoassay for detection of vaccine- and
disease-induced hepatitis A virus-
specific immunoglobulin G in saliva. J Clin Microbiol 1997;35(1):98-101. 11. Vento S, Garofano T, Renzini C, Cainelli F,
Casali F, Ferraro T, et al. Ful- minant hepatitis associated with hepatitis A virus superinfection in patients with chronic hepatitis
C. N Engl J Med 1998;338(5):26-90. 12. Akriviadis EA, Redeker AG. Fulminant hepatitis A in intravenous drug users
with chronic liver disease. Ann Intern Med 1989;110(10):838-9. 13. Widell A, Hansson BG, Moestrup T, Nordenfelt E.
Increased occurrence of hepatitis A with cyclic outbreaks among drug addicts in a Swedish commu- nity. Infection
1983;11:198-200. 14. Hepatitis A among drug abusers. MMWR Morb Mortal Wkly Rep 1988;37(19):
297-300,305. 15. Harkess J, Gildon B, Istre GR. Outbreaks of hepatitis A among illicit drug
users, Oklahoma, 1984-87. Am J Public Health 1989;79(4):463-6. 16. Szmuness W, Dienstag JL, Purcell RH, Harley EJ,
Steavens SE, Wong DC. Distribution of antibody to hepatitis A antigen in urban adult populations. N Engl J Med
1976;295(14):755-9. 17. Ochnio JJ, Scheifele DW, Ho M. Hepatitis A virus infections in urban chil- dren — are preventive
opportunities being missed? J Infect Dis 1997;176: 1610-3. 18. Queiroz DA, Cardoso DD, Martelli CM, Martins RM, Porto SO,
Azevedo MS, et al. Seroepidemiology of hepatitis A virus infection in street children of Goiania-Goias. Rev Soc Bras Med Trop
1995;28(3):199-203.
Correspondence to: Dr. Jan J. Ochnio, Vaccine Evaluation Center, Rm 317A, 950 W 28th Ave., Vancouver BC V5Z
4H4; fax 604 875-2496; [email protected]