IJE Advance Access originally published online on January 11, 2007
International Journal of Epidemiology 2007 36(2):396-405; doi:10.1093/ije/dyl276
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Modelling the impact on Hepatitis C transmission of reducing syringe sharing: London case study
1 HIVTools Research Group, London School of Hygiene and Tropical Medicine, London, UK.
2 Centre for Research on Drugs and Health Behaviour, London School of Hygiene and Tropical Medicine, London, UK
3 Social Medicine, University of Bristol, Bristol, UK
4 Medical Research Council, London, UK
* Corresponding author: Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK. E-mail: peter.vickerman{at}lshtm.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
Background Hepatitis C virus (HCV) prevalence and incidence among injecting drug users (IDUs) has increased in London and rest of UK. To inform public health action, mathematical modelling is used to explore the possible impact of strategies to decrease syringe sharing.
Methods A mathematical model was developed to simulate HCV transmission amongst IDUs in London. Because of parameter uncertainty, numerical search algorithms were used to obtain different model fits to HCV seroprevalence data from London for 2002–03. These simulations were used to explore the likely impact of HCV prevention activities that reduce syringe sharing amongst all IDUs, IDUs that have injected for greater than one year, or IDUs with lower or higher frequencies of syringe sharing.
Results Key differences between model fits centred on how they simulated the high HCV incidence amongst new injectors, either through assuming increased HCV infectivity during acute infection, a large sub-group of high frequency syringe sharers, or increased sharing among new IDUs. Despite parameter uncertainty, the model projections suggest that modest reductions in syringe sharing frequency (<25%) will reduce the HCV seroprevalence in newly initiated IDUs (injecting less than four years) but much larger and sustained reductions (>50%) are required to reduce the HCV seroprevalence in long-term IDUs (injecting more than 8 years). Critically the model also suggested that large reductions in HCV seroprevalence will be achieved only if interventions target all IDUs and reach IDUs within 12 months of injecting.
Discussion Public health interventions must reduce syringe sharing amongst all IDUs, including newly initiated IDUs, and be sustained for many years to reduce HCV infection. More accurate data on key behavioural (sharing frequency) and biological (percentage of infected IDUs that clear infection) parameters is required to improve model projections.
Keywords Hepatitis C, modelling, injecting drug use, UK
Accepted 10 November 2006
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
World wide, 170 million people are estimated to be infected with Hepatitis C (HCV) virus, while 9 million are thought to be infected in Europe.1,2 HCV can be easily transmitted through blood products and infected syringes,3–6 and infection rates are typically high amongst IDUs.7–9 HCV infection is an important public health concern because the majority of infections do not resolve but lead to chronic infection.10–17 After 30 years, approximately 30% of the people chronically infected with HCV develop cirrhosis of the liver and once cirrhosis has developed 1–4% of individuals per year progress to liver cancer (hepatocellular carcinoma) which often leads to death.18
In the UK, and other countries that have safe blood supplies and low iatrogenic risk, the prevention of HCV depends largely on reducing transmission among injecting drug users (IDUs). Indeed, the population burden of HCV is closely related to the number of people with an injecting history. In the UK, there are an estimated 200 000–500 000 people infected with HCV, over 90% of diagnoses are attributable to IDU,19 and over 40% of the current IDU population are HCV antibody-positive.19,20 The annual cost of treatment or care for individuals with chronic HCV infection are currently estimated to be 
750 million for the European Union and 100 million for the UK.21 This highlights the importance of prevention interventions, especially considering the number of hepatocellular and carcinoma deaths that are projected to increase.22,23
Syringe distribution and other interventions seeking to reduce syringe sharing are the main strategies for preventing the transmission of blood-borne viruses amongst IDUs.24 However, although syringe exchange interventions and the provision of methadone are associated with reduced HIV transmission,25–27 the evidence of their impact on HCV transmission is modest.28–31 Indeed, evidence suggests that current prevention strategies are insufficient in London and elsewhere in the UK because HCV and HIV transmission among IDUs are increasing.19,32–34
Clearly, the size of the problem and the serious long-term consequences make HCV an important public health concern.35–37 In this analysis, we develop a mathematical model for simulating the transmission of HCV. The model is fitted to HCV antibody prevalence (sero-prevalence) data from London, UK, for 2002–03 in order to explore the nature of HCV transmission in this setting, and to determine the possible impact of prevention activities that reduce the frequency of syringe sharing amongst IDUs. The model is built on previous studies that have developed mathematical models of the epidemiology of HCV,38,39 or have estimated the impact of harm reduction interventions on HCV transmission,38,40 in order to inform policy-makers and public health action.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
The analysis involved several steps. First, a HCV model was developed based on data available in the literature. Then, because of uncertainty in key biological and behavioural parameters, the model was fit in a staged process to HCV sero-prevalence data from London (2002–03). The different model fits were then used as a baseline comparison that could be used to estimate the potential impact of reductions in sharing behaviour on HCV transmission.
Model structure
The initial form of the model is an adaptation of a model developed by Kretzschmar and Wiessing,39 modified to allow for two types of acute infection, one leading to chronic infection and the other leading to resolved infection (allows for differences in viraemia during acute phase). Appendix 1 shows the flow diagram for the model.
The model simulates the transmission of HCV in a cohort of IDUs that start injecting at the same point in time. It assumes all new IDUs are susceptible and simulates the dynamics of infection over their duration of injecting ‘a’. Susceptible IDUs (x) are infected at a per capita rate ‘
’ through the use of infected syringes.3 Some epidemiological studies report an association between HCV infection and sharing of paraphernalia.4–6 However, this was not explicitly included in the model due to a lack of data on both the transmission probability and frequency of paraphernalia sharing. Sexual risk is assumed to be low.41–43 All newly infected IDUs become HCV RNA-positive (active infection) and progress to a phase of acute infection (duration 1/
) which lasts for 6–24 weeks,16,44–46 during which individuals usually develop an antibody response (anti-HCV-positive)11,13,15–17,44 and may have elevated viraemia16,17,44 and so could be more infectious. Acute infecteds are divided into those that resolve their infection (proportion ‘
’) after the acute phase (h2)10,13,17 or progress to chronic infection (h1). Chronic infecteds (y) are assumed to remain infected (HCV RNA-positive) and anti-HCV-positive until death. IDUs that resolve their infection are assumed to become HCV RNA-negative, immune for life46–52 and are initially anti-HCV-positive (z1), but may lose their antibody response (serorevert z2) after an average duration 1/
.51,53–55
The corresponding differential equations for the initial version of the model are as follows:
|
| (1) |
i’ is the same as the average probability per unit time that a susceptible IDU in syringe-sharing group ‘i’ becomes infected with HCV, and is dependent on the number of IDUs in the acute and chronic phase of HCV infection, the probability of HCV transmission per syringe-sharing incident for each phase of infection, and the syringe-sharing behaviour of the IDU. The parameter ‘i’ is assumed to be zero for IDUs that do not syringe share (i = 0).
If an IDU has ‘mi’ syringe-sharing partners per unit time then ‘i’ is the sum of the probabilities that the IDU will get infected from any of these partners. For each of these syringe-sharing partners, there is a certain probability that they will be a low or high syringe-sharer (
ij, where i and j denote the syringe-sharing sub-group of the IDU and their partner, respectively), which then determines the probability that their partner will be in the acute or chronic phase of infection, and the probability of being infected by this IDU per unit time (Bk). This gives the following formulation for i (for i = 1 or 2):
|
| (2) |
ij that a particular syringe-sharing partner is a low- (j = 1) or high- (j = 2) frequency syringe sharer is calculated using a standard method,56 by estimating the proportion of all syringe-sharing partners provided by that syringe-sharing sub-group and weighing it by a parameter ‘
’ that denotes the degree of assortative mixing ( = 0 for random mixing and 1 for fully assortative):
|
| (3) |
ij is the dirac-delta function, and equals one if i = j and zero otherwise. Lastly, the model allowed recently initiated ‘new’ injectors (IDUs injecting for less than one year) to have a higher frequency of syringe sharing (mi is scaled up by a factor 
) and to share partly with older injectors (with a greater HCV seroprevalence). This heterogeneity was incorporated as a possible reason for the high HCV incidence observed amongst ‘new’ IDUs in London, and because other studies have reported similar behaviours amongst recently initiated IDUs.57,58
Parameterizing the model
The model was parameterized using data from a number of sources, and these are shown in Table 1 with their uncertainty bounds. HCV seroprevalence data (based on oral fluid tests which have over 90% sensitivity and 99% specificity for HCV antibody59) against duration of injecting was obtained from a prospective cohort study and routine surveillance data from London for 2002–03.19,32,60 The cohort study recruited over 400 IDUs from community settings and networks; and routine surveillance data were available on over 1000 IDUs recruited from specialist drug treatment and syringe exchange agencies. All IDUs reported injecting in the last 4 weeks before recruitment, and completed an injecting risk questionnaire. These studies are the key source of data on HCV prevalence in London, and contribute to surveillance in the UK.
|
Unfortunately, no direct data were available on the frequency with which IDUs share, and so a proxy estimate was generated based on the estimated shortfall between the injection frequency, the number of syringes distributed and average syringe re-use.61,62 This estimate was used as the overall average syringe- sharing rate amongst all syringe sharers, and was kept constant in the fitting process. There was also no data on the number of people IDUs share with per month and so the ‘Bk’ terms in equation (2) were expanded to obtain the product of ‘n’ and ‘mi’ (the total syringe- sharing frequency in different syringe- sharing groups) in the formulation of ‘
i’. Data on other important aspects of IDU risk behaviour were also lacking, and so were given large uncertainty bounds, such as the degree of mixing between IDUs with different sharing rates (); the proportion of IDUs with different sharing frequencies; the percentage of new IDUs that share with older IDUs; and the degree to which new IDUs share with a higher frequency than older IDUs (). The rate of leaving the population (µ) was assumed to be 10% per year.63 Estimates for the HCV biological parameters were obtained from the scientific literature. However, several biological parameters were uncertain, such as the proportion of infected IDUs that resolve their infection.10–17 In addition, the HCV transmission probability per needle- sharing act (1–10%) was estimated by multiplying the HIV transmission probability per IDU syringe-sharing act (0.63–2.4%)64,65 by the factor difference between the HCV transmission probability following needle-stick injury (0.31–9%)3,66 and the corresponding HIV needle- stick transmission probability (0.25%, 95% CI 0.01–0.49%).67 Because of the large degree of uncertainty in this parameter, any uncertainty in the frequency of syringe sharing was also assumed to be accounted for in this parameter.
Methods for modelling the transmission of HCV in London
The model simulates the transmission of HCV against duration of injecting and was fitted to cross-sectional epidemiological data on HCV sero-prevalence from London for 2002–03.19,60 No single ‘best fitting’ model simulation could be determined because of the large degree of uncertainty present in the model parameters. Therefore, a numerical search algorithm was used to find different possible model fits from different starting points in the parameter uncertainty space.68 The model fitting process had a number of stages.
First, random sampling was used to obtain 1000 model parameter sets from the uncertainty ranges for each parameter.69 Second, using quarter monthly time steps the model was run with each parameter set, and the chi-squared error between each simulation and the epidemiological data was calculated and ranked.70 Third, the 20 best fitting parameter sets were used as starting points for using a numerical optimization algorithm (Newton's method used in the Solver tool of Excel) to find local optima that minimize the chi-square error. Fourth, the process was replicated for 10 additional model parameter sets sampled randomly from the remaining 980 simulations of the uncertainty analysis, thus generating 30 best-fits of the model to the cross-sectional HCV seroprevalence data. Fifth, the best-fits from this analysis were grouped together into classes with similar characteristics as defined by their parameter values. Finally, the model simulation with the smallest chi-squared error from each of these model types was selected for the following analyses.
Impact of decreasing syringe sharing on transmission of HCV
The model simulations were used to explore the likely impact of specific decreases in syringe sharing on HCV seroprevalence. The model was used to explore the effect of reduction in syringe-sharing amongst: all IDUs, IDUs injecting for greater than one year and IDUs with lower- or higher-frequencies of syringe-sharing.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
Figure 1 shows that the best-fit model simulation and the lower and upper bounds of the other 30 best-fits accurately project the observed HCV sero-prevalence in London for 2002–03. Critically, HCV sero-prevalence increases rapidly among recent IDUs with over 20% infected within 2 years of injecting.
|
Four classes of model type (A–D) were identified with the following common attributes: no change in the transmission probability during acute infection amongst those IDUs that resolve infection; a very low proportion (<4%) of new IDUs sharing with older IDUs; 60% or more of IDUs sharing syringes; IDUs mainly sharing syringes with IDUs with similar sharing rates; and IDU sub-groups with low- and high-frequencies of syringe sharing. The key differences between the model types was in how they simulated the high incidence of HCV infection observed amongst new injectors. One of the following combinations was required: a large sub-group of high-frequency syringe sharers either with a low (class C) or high proportion that resolve infection (class A); or a smaller sub-group of high-frequency syringe sharers and either new IDUs have a higher sharing rate (class B) or acute infected individuals have an increased HCV transmission probability before they progress to chronic infection (class D). Although they all closely fitted the observed data, projecting the same HCV sero-prevalence for IDUs injecting for less than eight years (
44%), they give different projections for the proportion of IDUs that are actively infected (RNA-positive) or have resolved their infection (RNA-negative and immune), ranging from 25% to 35% and 7–25%, respectively. Table 2 shows the precise differences in the parameter values and summary projections for the different model types.
|
For the following analyses, we have focused on the best-fits for class A, B and C, as D gave similar projections to A.
Impact of decreasing syringe sharing in all IDUs
Figure 2 shows the impact of decreasing syringe-sharing amongst all IDUs on HCV seroprevalence for models A and B (model C projections are similar to A). First, the figure shows that for model A, a modest decrease in syringe sharing will result in a notable decrease in HCV seroprevalence amongst IDUs that have been injecting for 1 year. In contrast, for model B, which assumes a small sub-group of higher frequency syringe sharers, a much greater decrease in syringe sharing is required to decrease the HCV seroprevalence. Amongst IDUs who have been injecting for 4 years, relatively modest decreases in syringe sharing result in reductions in HCV seroprevalence in both models. However, amongst IDUs that have been injecting for longer, more substantial decreases in syringe sharing, up to 25–50%, are required to achieve notable decreases in HCV sero-prevalence. For example, amongst IDUs that have been injecting for 8 years, syringe sharing has to decrease by at least 25%, from 16 to 12 receptive syringe-shares per month, for any decrease in HCV sero-prevalence to occur. Lastly, to reduce HCV seroprevalence to <10% amongst all IDUs that have been injecting for 
8 years, the rate of syringe sharing has to decrease to about 1–2 times a month (see Figure 3 for these projections). Furthermore, any projected reductions in HCV seroprevalence assume that the reduction in syringe sharing is sustained over an IDU's injecting life course, i.e. for at least 8 years for these projections.
|
|
The importance of reaching all injectors
Figure 3 compares the relative impact on HCV infection of reducing syringe sharing amongst all IDUs or only amongst IDUs who have been injecting for >1 year. The model projections suggest that HCV seroprevalence can be reduced even with moderate reductions in sharing. However, as syringe sharing is reduced further all the models predict a substantial reduction in impact if the intervention fails to reduce sharing amongst new injectors. For example, if the frequency of syringe sharing is reduced by 75%, to an estimated four sharing events per month, the HCV sero-prevalence would reduce to 13–22% (amongst IDUs injecting for
8 years) if all IDUs were reached, whereas it would only reduce to 25–29% if recent IDUs were missed. Model B projects a much greater decrease in HCV sero-prevalence when all IDUs are reached because the model's increased frequency of syringe sharing amongst new injectors initiates the HCV epidemic amongst the lower frequency syringe sharing IDUs, and so reducing syringe sharing in this period has a larger effect on the projected epidemic than for the other models.
Impact of decreasing syringe sharing in low- or high-frequency syringe-sharing IDUs
Figure 4 compares the impact of targeting higher frequency or lower frequency syringe sharers and shows that the impact of moderate decreases in syringe sharing is much greater if it is achieved in the low-frequency syringe sharing IDUs than if it was achieved in the higher frequency syringe sharing IDUs. Indeed, unless substantial decreases in sharing can be achieved in the higher frequency syringe sharers (>60%) little impact on HCV seroprevalence will be achieved unless the rest of the IDUs are also targeted. This is because of the high level of transmission occurring in the higher frequency syringe sharing IDUs.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
We developed a model of HCV transmission that incorporated key biological features of HCV infection and summary behavioural data from IDUs in London. The model accurately fitted the observed HCV seroprevalence in London (2002–03) by injecting duration. However, because of parameter uncertainty several different models, with contrasting biological and behavioural assumptions, gave equally good fits to the observed data. The key differences centred on how they simulated the rapid spread of HCV infection amongst new injectors, either assuming a large sub-group of high-frequency syringe sharing IDUs, increased syringe sharing among new injectors, or higher transmission during the acute infection phase. These models gave different projections for the proportion of IDUs with active HCV infection or that have resolved their infection (assumed to be immune), and the reduction in HCV seroprevalence in response to reductions in sharing.
Currently, there are insufficient data to determine which model type is the better description of HCV transmission in London, and probably also for other sites. Better information on several of the key biological or behavioural parameters could reduce the overall amount of parameter uncertainty and so improve model selection and the accuracy of projections. For example, if new experimental data strongly suggested that there was no increase in transmission risk during the acute phase of infection, then this could be ruled out as a possible reason for the rapid initial spread of HCV and one of our models could be rejected. Conversely, more accurate estimates for the proportion that resolve their infection could set stricter limits on estimates of syringe sharing.
Despite uncertainty in some important model parameters, the models provide a number of insights, highlighting the critical importance of reaching new injectors, and the potential difficulty in effectively controlling HCV transmission amongst IDUs, as highlighted by other modelling studies.38–40
Evidence suggests that the current level of intervention activity has been insufficient in preventing increases in HCV transmission in London.19,33,34 Because of this, the model was used to explore the possible impact of further reducing syringe sharing. The models project that if all IDUs reduce their syringe sharing the HCV seroprevalence in shorter term IDUs (injecting for less than four years) will be reduced, but that greater reductions are required to produce similar reductions amongst longer term IDUs (>eight years of injecting). In addition, and in agreement with a previous modelling study,28 the frequency of syringe sharing has to decrease to very low levels (one to two times per month) for the HCV seroprevalence to be reduced to 10%. Moreover, if interventions fail to reach recent injectors then the models project considerably smaller decreases in HCV seroprevalence. Targeting recent injectors may require the design of new interventions or innovative ways of delivery.
Lastly, as for sexually transmitted infection (STI) prevention interventions, one possible strategy for safer injecting and syringe distribution interventions could be to target higher risk IDUs. However, our model projections suggest that the most effective prevention strategy is to target all IDUs, rather than just higher frequency syringe sharers. Indeed, our model projections suggest that targeting higher risk IDUs could have less effect on the overall HCV seroprevalence than targeting lower frequency syringe sharers.
One potential limitation was that the model did not incorporate the possibility of HCV transmission through sharing other injecting equipment,4–6 primarily because of a lack of data. Equally the model did not explicitly model periods of imprisonment when IDUs may have increased risk. The implication of these limitations is that the model may over-estimate the impact of reducing syringe sharing on HCV prevalence. Critically, the analysis also lacked reliable data on the frequency and nature of syringe sharing, though not uncommon in behavioural surveys.71 We therefore combined information on injection frequency, syringe re-use and syringe coverage to provide an estimate of 16 sharing-events per month. The uncertainty in this parameter estimate was accounted for in the large uncertainty range given to the HCV transmission probability per syringe sharing act. Further, the model allowed for some heterogeneities in both syringe sharing (among recent or higher frequency syringe sharers) and the degree of mixing between different groups, but still assumed sharing was a random event. Clearly this is a simplification because syringe sharing, like drug taking, usually occurs within social groups72 and networks.73 We hypothesize that in the same way as concurrent sexual partnerships are important for STI spread,74 concurrent sharing partnerships formed in drug sharing networks may be important in determining the spread of HCV. However, such network effects cannot be incorporated and explored until better and more accurate data are collected on these risk behaviours.75
|
Appendix |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
|
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
The work was partly funded by DTI Foresight Programme and we are very grateful for their support and interest. M.H. is funded by an NHS Career Scientist grant and P.V. also receives funding from the DFID funded AIDS Knowledge Programme. The views expressed are those of the authors and cannot be taken to reflect the official opinion of the London School of Hygiene and Tropical Medicine, Bristol University and Medical Research Council or Foresight.
Conflict of interest: None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAppendixAcknowledgementsReferences |
|---|
1 Wong J, Bennett W, Raymond S, Pauker S. Pretreatment evaluation of chronic hepatitis C. JAMA (1998) 280:2088–93.
2 WHO. HepatitisCfactsheet. (2000) Geneva. (http://www.who.int/mediacentre/factsheets/fs164/en/).
3 De Carli G, Puro V, Ippolito G. Risk of hepatitis C virus transmission following percutaneous exposure in healthcare workers. Infection (2003) 31(Suppl 2):22–27.[Web of Science][Medline]
4 Thorpe LE, Ouellet LJ, Hershow R, et al. Risk of hepatitis C virus infection among young adult injection drug users who share injection equipment. Am J Epidemiol (2002) 155:645–53.
5 Zhou F, Ma ZE, Hu W, et al. Study on the relationship between hepatitis C virus infection and sharing injection equipment, sexual behavior among injecting drug users. Zhonghua Liu Xing Bing Xue Za Zhi (2004) 25:329–32.[Medline]
6 Vidal Trecan GM, Varescon Pousson I, Gagniere B, Tcherny Lessenot S, Madariaga E, Boissonnas A. Association between first injection risk behaviors and hepatitis C seropositivity among injecting drug users. Ann Med Interne (Paris) (2002) 153:219–25.[Medline]
7 Villano SA, Vlahov D, Nelson KE, Lyles CM, Cohn S, Thomas DL. Incidence and risk factors for hepatitis C among injection drug users in Baltimore, Maryland. J Clin Microbiol (1997) 35:3274–77.
8 Lucidarme D, Bruandet A, Ilef D, et al. Incidence and risk factors of HCV and HIV infections in a cohort of intravenous drug users in the North and East of France. Epidemiol Infect (2004) 132:699–708.[CrossRef][Medline]
9 Van Ameijden EJ, Van den Hoek JA, Mientjes GH, Coutinho RA. A longitudinal study on the incidence and transmission patterns of HIV, HBV and HCV infection among drug users in Amsterdam. Eur J Epidemiol (1993) 9:255–62.[CrossRef][Web of Science][Medline]
10 Kenny-Walsh E. Clinical outcomes after Hepatitis C infection from contaminated anti-D immune globulin. New Engl Jf Med (1999) 340:1228–34.[CrossRef]
11 Hwang SJ, Lee SD, Lu RH, et al. Hepatitis C viral genotype influences the clinical outcome of patients with acute posttransfusion hepatitis C. J Med Virol (2001) 65:505–09.[CrossRef][Web of Science][Medline]
12 Spada E, Mele A, Berton A, et al. Multispecific T cell response and negative HCV RNA tests during acute HCV infection are early prognostic factors of spontaneous clearance. Gut (2004) 53:1673–81.
13 Hofer H, Watkins Riedel T, Janata O, et al. Spontaneous viral clearance in patients with acute hepatitis C can be predicted by repeated measurements of serum viral load. Hepatology (2003) 37:60–64.[CrossRef][Web of Science][Medline]
14 Santantonio T, Sinisi E, Guastadisegni A, et al. Natural course of acute hepatitis C: a long-term prospective study. Dig Liver Dis (2003) 35:104–13.[CrossRef][Web of Science][Medline]
15 Gerlach JT, Diepolder HM, Zachoval R, et al. Acute hepatitis C: high rate of both spontaneous and treatment-induced viral clearance. Gastroenterology (2003) 125:80–88.[CrossRef][Web of Science][Medline]
16 Kamal SM, Rasenack JW, Bianchi L, et al. Acute hepatitis C without and with schistosomiasis: correlation with hepatitis C-specific CD4(+) T-cell and cytokine response. Gastroenterology (2001) 121:646–56.[CrossRef][Web of Science][Medline]
17 Villano SA, Vlahov D, Nelson KE, Cohn S, Thomas DL. Persistence of viremia and the importance of long-term follow-up after acute hepatitis C infection. Hepatology (1999) 29:908–14.[CrossRef][Web of Science][Medline]
18 Limburg W. Natural history, treatment and prevention of hepatitis C in injecting drug users: an overview. In: Hepatitis C and Injecting Drug Use: Impact, Costs and Policy Options—Jager JC, Limburg W, Kretzschmar M, Postma MJ, Wiessing L, eds. (2004) Lisbon: European Monitoring Centre For Drugs And Drug Addiction. 21–40.
19 Health Protection Agency. Shooting up. Infections among injecting drug users in the United Kingdom 2003. (2004) London.
20 Roderick P, Parkes J, Rosenberg W, et al. The epidemiology and health care burden of chronic liver disease. (2004) Ringwood: British Liver Trust and Foundation for Liver Research. Available at (http://www.britishlivertrust.org.uk/content/research/current_content_associated_docs/impact%20and%20burden%20final%20report.doc).
21 Postma MJ, Wiessing L, Jager JC. Updated healthcare cost estimates for drug-related hepatitis C infections in the European Union. In: Hepatitis C and Injecting Drug use: Impact, Costs and Policy Options—Jager JC, Limburg W, Kretzschmar M, Postma MJ, Wiessing L, eds. (2004) Lisbon: European Monitoring Centre For Drugs And Drug Addiction. 203–16.
22 Sweeting MJ, De Angelis D, Brant L, Harris HE, Mann AG, Ramsay ME. The burden of hepatitis C in England. In preparation.
23 Sypsa V, Touloumi G, Papatheodoridis GV, et al. Future trends of HCV-related cirrhosis and hepatocellular carcinoma under the currently available treatments. J Viral Hepat (2005) 12:543–50.[CrossRef][Web of Science][Medline]
24 World Health Organisation. Provision of sterile injecting equipment to reduce HIV transmission. In: Evidence for action on HIV/AIDS and injecting drug use (2004) Geneva. (http://www.who.int/hiv/pub/idu/idupolicybriefs/en/).
25 MacDonald M, Law MG, Kaldor JM, Hales J, Dore GJ. Effectiveness of needle and syringe programmes for preventing HIV transmission. Int J Drug Policy (2003) 14:353–57.[CrossRef]
26 Gibson DR, Flynn N, Perales D. Effectiveness of syringe exchange programs in reducing HIV risk behavior and HIV seroconversion among injecting drug users. AIDS (2001) 15:1329–341.[CrossRef][Web of Science][Medline]
27 Hurley SF, Jolley DJ, Kaldor JM. Effectiveness of needle-exchange programmes for prevention of HIV infection. Lancet (1997) 349:1797–800.[CrossRef][Web of Science][Medline]
28 Murray JM, Law MG, Gao Z, Kaldor JM. The impact of behavioural changes on the prevalence of HIV and hepatitus C among injecting drug users. Int J Epidemiol (2003) 32:708–14.
29 Goldberg D, Burns S, Taylor A, Cameron S, Hargreaves D, Hutchinson S. Trends in HCV prevalence among injecting drug users in Glasgow and Edinburgh during the era of needle/syringe exchange. Scand J Infect Dis (2001) 33:457–61.[CrossRef][Web of Science][Medline]
30 Hutchinson SJ, McIntyre PG, Molyneaux P, et al. Prevalence of hepatitis C among injectors in Scotland 1989–2000: declining trends among young injectors halt in the late 1990s. Epidemiol Infect (2002) 128:473–477.[CrossRef][Medline]
31 van de Laar TJ, Langendam MW, Bruisten SM, et al. Changes in risk behavior and dynamics of hepatitis C virus infections among young drug users in Amsterdam, the Netherlands. J Med Virol (2005) 77:509–18.[CrossRef][Web of Science][Medline]
32 Judd A, Hickman M, Jones S, et al. Incidence of hepatitis C virus and HIV among new injecting drug users in London: prospective cohort study. Br Med J (2005) 330:24–25.
33 Hope VD, Judd A, Hickman M, et al. HIV prevalence among injecting drug users in England and Wales 1990 to 2003: evidence for increased transmission in recent years. AIDS (2005) 19:1207–14.[Web of Science][Medline]
34 Sutton AJ, Gay NJ, Edmunds WJ, Hope V, Gill ON, Hickman M, eds. Modelling the Force of Infection for Hepatitis B and Hepatitis C in injecting drug users in England and Wales. (In review).
35 The Global Burden of Hepatitis C Working Group. Global burden of disease (GBD) for Hepatitis C. J Clin Pharmacol (2004) 44:20–29.
36 World Health Organisation. Hepatitis C. In: Fact sheets (2000) Geneva. (http://www.who.int/mediacentre/factsheets/fs164/en/).
37 Alter MJ. Prevention of spread of hepatitis C. Hepatology (2002) 36:93–8.[CrossRef]
38 Hutchinson S, Bird S, Taylor A, Goldberg D. Modelling the spread of hepatitis C virus infection among injecting drug users in Glasgow: implications for prevention. Int J Drug Policy (2006) 17:211–21.[CrossRef]
39 Kretzschmar M, Wiessing L. Modelling the transmission of hepatitis C in injecting drug users. In: Hepatitis C and Injecting Drug Use: Impact, Costs and Policy Options—Jager JC, Limburg W, Kretzschmar M, Postma MJ, Wiessing L, eds. (2004) Lisbon: European Monitoring Centre For Drugs And Drug Addiction.
40 Pollack HA. Cost-effectiveness of harm reduction in preventing hepatitis C among injection drug users. Med Decis Making (2001) 21:357–67.
41 Sladden TJ, Hickey AR, Dunn TM, Beard JR. Hepatitis C transmission on the north coast of New South Wales: explaining the unexplained. Med J Aust (1997) 166:290–93.[Web of Science][Medline]
42 Roy KM, Goldberg DJ, Hutchinson S, Cameron SO, Wilson K, MacDonald L. Hepatitis C virus among self declared non-injecting sexual partners of injecting drug users. J Med Virol (2004) 74:62–66.[CrossRef][Web of Science][Medline]
43 Tibbs CJ. Methods of transmission of hepatitis C. J Viral Hepat (1995) 2:113–19.[Web of Science][Medline]
44 Yamaguchi K, Tanaka E, Higashi K, et al. Adaptation of hepatitis C virus for persistent infection in patients with acute hepatitis. Gastroenterology (1994) 106:1344–348.[Web of Science][Medline]
45 Post JJ, Pan Y, Freeman AJ, et al. Clearance of hepatitis C viremia associated with cellular immunity in the absence of seroconversion in the hepatitis C incidence and transmission in prisons study cohort. J Infect Dis (2004) 189:1846–855.[CrossRef][Web of Science][Medline]
46 Lanford RE, Guerra B, Chavez D, et al. Cross-genotype immunity to hepatitis C virus. J Virol (2004) 78:1575–581.
47 Mehta SH, Cox A, Hoover DR, et al. Protection against persistence of hepatitis C. Lancet (2002) 359:1478–483.[CrossRef][Web of Science][Medline]
48 Lauer GM, Barnes E, Lucas M, et al. High resolution analysis of cellular immune responses in resolved and persistent hepatitis C virus infection. Gastroenterology (2004) 127:924–36.[CrossRef][Web of Science][Medline]
49 Shoukry NH, Grakoui A, Houghton M, et al. Memory CD8+ T cells are required for protection from persistent hepatitis C virus infection. J Exp Med (2003) 197:1645–655.
50 Shoukry NH, Cawthon AG, Walker CM. Cell-mediated immunity and the outcome of hepatitis C virus infection. Annu Rev Microbiol (2004) 58:391–24.[CrossRef][Web of Science][Medline]
51 Takaki A, Wiese M, Maertens G, et al. Cellular immune responses persist and humoral responses decrease two decades after recovery from a single-source outbreak of hepatitis C. Nat Med (2000) 6:578–82.[CrossRef][Web of Science][Medline]
52 Aitken CK, Bowden S, Hellard M, Crofts N. Indications of immune protection from hepatitis C infection. J Urban Health (2004) 81:58–60.[Web of Science][Medline]
53 Lefrere JJ, Girot R, Lefrere F, et al. Complete or partial seroreversion in immunocompetent individuals after self-limited HCV infection: consequences for transfusion. Transfusion (2004) 44:343–48.[CrossRef][Web of Science][Medline]
54 Lefrere JJ, Guiramand S, Lefrere F, et al. Full or partial seroreversion in patients infected by hepatitis C virus. J Infect Dis (1997) 175:316–22.[Web of Science][Medline]
55 Kondili LA, Chionne P, Costantino A, et al. Infection rate and spontaneous seroreversion of anti-hepatitis C virus during the natural course of hepatitis C virus infection in the general population. Gut (2002) 50:693–96.
56 Garnett GP, Anderson RM. Balancing sexual partnerships in an age and activity stratified model of HIV transmission in heterosexual populations. IMA J Math Appl Med Biol (1994) 11:161–92.
57 Becker Buxton M, Vlahov D, Strathdee SA, et al. Association between injection practices and duration of injection among recently initiated injection drug users. Drug Alcohol Depend (2004) 75:177–83.[CrossRef][Web of Science][Medline]
58 Guydish JR, Abramowitz A, Woods W, Black DM, Sorensen JL. Changes in needle sharing behavior among intravenous drug users: San Francisco, 1986–88. Am J Public Health (1990) 80:995–97.
59 Judd A, Parry J, Hickman M, et al. Evaluation of a modified commercial assay in detecting antibody to hepatitis C virus in oral fluids and dried blood spots. J Med Virol (2003) 71:49–55.[CrossRef][Web of Science][Medline]
60 Judd A, Hutchinson S, Wadd S, Hickman M. Prevalence of and risk factors for, hepatitus C virus infection among recent initiates to injecting in London and Glasgow - Cross sectional analysis. J Viral Hepat (2005) 12:655–62.[CrossRef][Web of Science][Medline]
61 Vickerman P, Hickman M, Rhodes T, Watts CH. Model projections on the required coverage of syringe distribution to prevent HIV epidemics among injecting drug users. J Acquir Immune Defic Syndr (2006) 42:355–61.[CrossRef][Web of Science][Medline]
62 Hickman M, Higgins V, Hope VD, et al. Injecting drug use in Brighton, Liverpool and London: best estimates of prevalence and coverage of public health indicators. J Epidemiol Commun Heal (2004) 58:766–71.
63 De Angelis D, Hickman M, Yang S. Estimating long-term trends in the incidence and prevalence of opiate/injecting drug use and the number of ex-users: the use of back-calculation methods and opiate overdose deaths. Amer J Epidemiol (2004) 160:994–1004.
64 Hudgens MG, Longini IM Jr, Vanichseni S, et al. Subtype-specific transmission probabilities for Human Immunodeficiency Virus among injecting drug users in Bangkok, Thailand. Amer J Epidemiol (2002) 155:159–68.
65 Kaplan EH, Heimer R. A model-based estimate of HIV infectivity via needle sharing. J Acq Imm Deficien Synd (1992) 5:1116–118.
66 MacDonald M, Crofts N, Kaldor J. Transmission of Hepatitis C virus: rates, routes and cofactors. Epidemiol Rev (1996) 18:137–49.
67 Baggaley RF, Boily MC, White RG, Alary M. Risk of HIV-1 transmission for parenteral exposure and blood transfusion: a systematic review and meta-analysis. AIDS (2006) (in press).
68 Weisstein EW. Global optimization. In: Mathworld (2006) Wolfram Research, Inc. (http://mathworld.wolfram.com/about/mathworld.html).
69 Blower SM, Dowlatabadi H. Sensitivity and uncertainty analysis of complex models of disease transmission: an HIV Model, as an Example. Int Stat Rev (1994) 62:229–43.[CrossRef]
70 Press W, Teukolsky S, Vetterling W, Flannery B. Least squares as a maximum likelihood estimator. In: Numerical Recipes in C: The Art of Scientific Computing (1992) Cambridge: Cambridge University Press.
71 Hickman M, Hope V, Brady T, et al. Hepatitis C prevalence and injecting risk behaviour in multiple sites in England in 2004. (In review).
72 Hunter GM, Donoghoe MC, Stimson GV, Rhodes T, Chalmers CP. Changes in the injecting risk behaviour of injecting drug users in London, 1990–1993. AIDS (1995) 9:493–01.[Web of Science][Medline]
73 Friedman SR, Kottiri BJ, Neaigus A, Curtis R, Vermund SH, Des Jarlais DC. Network-related mechanisms may help explain long-term HIV-1 seroprevalence levels that remain high but do not approach population-group saturation. Amer J Epidemiol (2000) 152:913–922.
74 Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS (1997) 11:641–48.[CrossRef][Web of Science][Medline]
75 Kretzschmar M, Wiessing LG. Modelling the spread of HIV in social networks of injecting drug users. AIDS (1998) 12:801–11.[CrossRef][Web of Science][Medline]
76 Ross RS, Viazov S, Roggendorf M. Risk of hepatitis C transmission from infected medical staff to patients: model-based calculations for surgical settings. Arch Intern Med (2000) 160:2313–316.
77 Farci P, Alter HJ, Wong D, et al. A long-term study of hepatitis C virus replication in non-A, non-B hepatitis. N Engl J Med (1991) 325:98–104.[Abstract]
78 Courouce AM, Le Marrec N, Bouchardeau F, et al. Efficacy of HCV core antigen detection during the preseroconversion period. Transfusion (2000) 40:1198–202.[CrossRef][Web of Science][Medline]
79 Logvinoff C, Major ME, Oldach D, et al. Neutralizing antibody response during acute and chronic hepatitis C virus infection. Proc Natl Acad Sci U S A (2004) 101:49–54.
80 Garson JA, Tuke PW, Makris M, et al. Demonstration of viraemia patterns in haemophiliacs treated with hepatitis-C-virus-contaminated factor VIII concentrates. Lancet (1990) 336:22–25.[CrossRef]
81 Cameron SO, Wilson K, Good T, et al. Detection of antibodies against hepatitis C virus in saliva: a marker of viral replication. J Viral Hepat (1999) 6:141–44.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P Vickerman, L Platt, and S Hawkes Modelling the transmission of HIV and HCV among injecting drug users in Rawalpindi, a low HCV prevalence setting in Pakistan Sex Transm Inf, April 1, 2009; 85(Suppl_2): ii23 - ii30. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Bayoumi MD MSc and G. S. Zaric PhD The cost-effectiveness of Vancouver's supervised injection facility Can. Med. Assoc. J., November 18, 2008; 179(11): 1143 - 1151. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||