|
Bobby P Smyth, Eamon Keenan, John
J O`Connor
| Table 1
Period of commencement of injecting, duration injecting,
age, sex, partner's injecting status, principal drug injected,
and employment status in relation to risk for hepatitis
C among injecting drug users with injecting
histories of less than 25 months; univariate and multivariate
analyses |
|
|
|
|
|
|
|
|
|
Univariate Analysis |
Multivariate
Analysis |
|
|
|
|
|
|
|
| |
Number |
Prevalence
of anti HGV % |
Number |
p value
† |
Odds
ratio (95% confidence intervals) |
p value
|
|
|
|
|
|
|
|
| Period when commenced injecting
|
|
|
|
|
|
|
|
| Before January 1994 (pre-94 group) |
172 |
64.5 |
1.0 |
|
1.0 |
|
| After January 1994 (post-93 group) |
181 |
40.3 |
0.37
(0.24, 0.57) |
<
0.001 |
0.43
(0.27, 0.67) |
<
0.001 |
|
|
|
|
|
|
|
| Duration since commenced injecting (months)
|
|
|
|
|
|
|
|
| 1 to 12 |
216 |
44.4 |
1.0 |
|
1.0 |
|
| 13 to 24 |
137 |
64.2 |
2.25
(1.45, 3.48) |
<
0.001 |
1.76
(1.10, 2.80) |
0.017 |
|
|
|
|
|
|
|
| Age (years) |
|
|
|
|
|
|
| Under 21 |
168 |
47.0 |
1.0 |
|
1.0 |
|
| 21 and over |
185 |
57.7 |
1.54
(1.01, 2.34) |
0.044 |
1.51
(0.98, 2.34) |
0.061 |
|
|
|
|
|
|
|
| Sex |
|
|
|
|
|
|
| Male |
241 |
51.0 |
1.0 |
|
|
|
| Female |
112 |
54.5 |
1.15
(0.73, 1.8) |
0.55 |
|
|
|
|
|
|
|
|
|
| Sexual Partner's injecting status
|
|
|
|
|
|
|
|
| Partner injecting |
109 |
53.2 |
1.0 |
|
|
|
| No Partner injecting |
236 |
52.1 |
0.96
(0.61, 1.51) |
0.85 |
|
|
|
|
|
|
|
|
|
| Primary drug injected
|
|
|
|
|
|
|
|
| Heroin |
274 |
52.2 |
1.0 |
|
|
|
| Other |
79 |
51.9 |
0.99
(0.60, 1.63) |
0.96 |
|
|
|
|
|
|
|
|
|
| Employment status
|
|
|
|
|
|
|
|
| Employed |
43 |
41.9 |
1.0 |
|
|
|
| Unemployed |
318 |
53.5 |
1.60
(0.84, 3.06) |
0.15 |
|
|
|
|
|
|
|
|
|
*N=345, status of partner unknown for
eight patients. Pearson X2 test
Injecting drug users represent a high risk
group for hepatitis C virus (HCV) infection and, in many locations,
the majority will test positive for antibody to HCV (anti-HCV)
within two years of starting to inject.1 Although
there is evidence of a reduction in rates of unsafe injecting,
2 3 there is little published research demonstrating
that programmes that facilitate safe injecting have reduced
the occurrence of HCV.4 Consequently, some commentators
are not optimistic that we will see a decrease in HCV prevalence
among injecting drug users.5
Harm reduction programmes include methadone treatment, education
regarding safer injecting and the provision of syringe exchange.
These services vastly expanded in Dublin over the period 1991
to late 1993. The number of community outreach workers and
addiction counselors increased by 74%. We sought to test the
hypothesis that, among injecting drug users with short injecting
histories, the prevalence of HCV would be lower in those who
started injecting during the period after this expansion in
services.
The setting for this study was McCarthy Centre, which is the
largest and longest established addiction treatment center
in Dublin. Services provided include counselling, methadone
maintenance and detoxification and assessment regarding medical
problems associated with drug use such as HCV.
Data have been recorded on an ongoing basis on the results
of all HCV tests on injecting drug users attending Trinity
Court since 1992. In this study, consecutive new attenders,
resident in Dublin, with a reported injecting history less
than 25 months, tested for anti-HCV between July 1993 and
December 1996 were included. We used a third generation enzyme
linked immunosorbent assay for anti-HCV (Ortho Clinical Diagnostics,
Amersham, England). Positive results were confirmed with a
further third generation test.
In all 353 injecting drug users were tested. The primary drug
of choice was heroin for 78%, morphine sulphate for 21%, and
benzodiazepines for 1%.
Those with injecting histories of less than 13 months were
over-represented in the group that started injecting in the
period after January 1994 (75.1% v 46.5%, x2 = 30.4, p<0.001).
Period of commencement of injecting was not significantly
associated with age, sex, employment or injecting status of
sexual partner (x2 tests).
The prevalence of anti-HCV was 52.1%. Univariate analysis
showed that those who started injecting in the period after
January 1994 (post-1993 group) and those with injecting histories
of less than 13 months demonstrated significantly reduced
risks of HCV infection (see table 1). Age over 21 years was
weakly associated with increased risk.
| Table 2
Association between period of onset of injecting drug
use and risk of hepatitis C, adjusted within strata
of duration of injecting drug use
|
|
|
|
|
|
|
|
|
| |
Commenced
injecting before Aug 93 |
Commenced injecting
between Aug 93 and July 94 |
Commenced
injecting after July 94 |
|
|
|
|
|
|
|
|
|
| |
Number |
Prevalence of anti HGV % |
Number |
Prevalence
of
anti HGV % |
Number |
Prevalence
of
anti HGV % |
p
value * |
|
|
|
|
|
|
|
|
| |
125 |
64.8 |
112 |
51.8 |
116 |
38.8 |
|
| Stratified by duration injecting (months)
|
| 1-12 |
48 |
60.4 |
73 |
46.6 |
95 |
34.7 |
0.003 |
| 13-24 |
77 |
67.5 |
39 |
61.5 |
21 |
57.1 |
0.33 |
|
|
|
|
|
|
|
|
* Mantel-Haenszel X2
test for trend
Nivariate analysis was repeated with data stratified by
length of injecting history (data not shown). This demonstrated
a significant reduction in HCV in the post-1993 group with
injecting histories of less than 13 months (odds ratio 0.36,
(95% confidence intervals 0.21, 0.64) p=0.001) but the reduction
was not significant in those with injecting histories of
13 months and over (odds ratio 0.57, (95% confidence intervals
0.28, 1.20_ p=0.20). Multivariate analysis was then performed
with the three variables that were significant on univariate
analysis being entered into a logistic regression equation.
This resulted in a weakening of the association with duration
since starting injecting (see table 1). Also the effect
of age became of borderline significance. There was no evidence
of interaction between independent variables.
To further explore for the presence of a trend of reducing
prevalence of HCV, subjects were ranked chronologically
in terms of their date of commencement of injecting and
then divided into thirds - that is, those who began injecting
before August 1993, those who started between August 1993
and July 1994 inclusive, and thirdly, those who first injected
after July 1994. Table 2 shows the highly significant downward
trend in HCV prevalence. When data are stratified by length
of injecting history, the trend remains one of reducing
HCV prevalence over time in both those with short and longer
injecting careers. However, the fall in prevalence is statistically
significant only in those with injecting histories of less
than 13 months.
practices after the service expansion.
Unfortunately, we were unable to control for other factors
that may explain this decline in HCV. Alternatively explanations
might include a possible reduction in overall injecting
frequency among the more recent injectors or continued rates
of unsafe injecting but confined within safer groups. Therefore,
while acknowledging that our detection of a declining prevalence
of HCV infection after an expansion in harm reduction services
does not conclusively prove causality, we believe that this
is an encouraging finding. However, we consider it premature
to assume that this protective effect will persist over
time, as a reduced rate of unsafe injecting by people within
this group could still lead to a very high prevalence of
HCV infection.5 It may simply take longer to do so.
We wish to thank Drs E O'Callaghan, J Barry, C Moran, and
Z Johnson for their advice and criticism in the preparation
of the manuscript. We also wish to acknowledge the staff
at the Virus Reference Laboratory, Dublin, where all blood
tests were analysed.
Funding: none.
Conflicts of interest: none.
1 Garfein RS, Vlahov D, Galia C, et al.
Viral infections in sort-term injectin drug users: The prevalence
of the hepatitis C, hepatitis B, human immunodeficiency,
and human T-lymphocyte viruses. Am J Public Health 1996;86:655-61.
2 Robertson JR, Ronald PJM, Raab GM, et al. Deaths, HIV
infection, abstinence, and other outcomes in a cohort of
injecting drug users followed up for 10 years. BMJ 1994;309:369-72.
3 Hunter GM, Donoghoe MC, Stimson GV, et al. Changes in
injecting risk befavious of injecting drug users in London
1990-1993. AIDS 1995;9:493-501.
4 Hagan H, Des Jarlais DC, Freidman SR, et al. Reduced risk
of hepatitis B and hepatitis C among infection drug users
in the Tacoma sytinge exchange program. Am J Public Health
1995;85:1531-7.
5 Wodak A, Crofts N. Once more unto the breach: controlling
hepatitis C in njecting drug users. Addiction 1996;91:181-4.
|
|
|
