Modelling the hepatitis C virus disease burden among injecting drug users in Scotland
A forward projection model was used to estimate the numbers of, both current and former, IDUs who acquired HCV infection and progressed to mild, moderate and severe HCV disease in Glasgow and Scotland between 1960 and 2030. The model was developed initially for Glasgow because more epidemiological information exists for this region, than elsewhere in Scotland, to calibrate model outcomes with local data relating to HCV and its consequence. Insights gained from the model fitting process in Glasgow were then used to extend the model to the rest of Scotland. First, the incidence and cessation of injecting drug use in Glasgow during 1960-2000 were derived through the use of a modified Delphi approach. Instead of the usual iterative process to refine experts’ estimates, the elicitation of IDU incidence and cessation provided an opportunity to combine these data and examine coherence with capture-recapture IDU prevalence estimates. Coherent estimates indicated that incidence (median: 28 to 49) and cessation (1 to 24%) remained low and stable during 1960-1975, rose steeply between 1975-1985 (incidence from 49 to 1,335; cessation from 2% to 6%), and by 2000 there had been a decline in incidence (1,195) but a further rise in cessation (15%). Secondly, stochastic simulation was used to model the transmission of HCV among current IDUs in Glasgow, according to their injecting risk behaviours, and estimate the past incidence of HCV infection. The model that considered higher infectivity during acute viraemia following infection produced seroprevalences (median: 62-72%) and incidences (18-30 per 100 susceptible injector-years) consistent with observed data during the 1990s. The annual number of new HCV infections among current IDUs in Glasgow was estimated to be low during 1960-1976 (median: 10-60), rise steeply during 1960-1976 (median: 10-60), rise steeply during the early 1980s to peak in 1985 (1,120), stabilise during 1991-1997 (510-610) and rise again during 1998-2000 (710-780). Scenario analyses indicated that potentially as many as 4,500 HCV infections (10th and 90th percentiles: 2,400-7,700) had been prevented in Glasgow during 1988-2000 as a result of harm-reduction measures. Scenario analyses also permitted the gauging of changes in risk behaviours required to effect appreciable reductions in the incidence of HCV infection. Incidence can be successfully reduced if IDUs who, unavoidably, share needles/syringes confine their borrowing to one person; with this strategy alone, an estimated 5,300 HCV infections (10th and 90th percentiles: 4,100-6,700) could have been averted in Glasgow during 1988-2000. Such insights will inform those responsible for developing new ways to prevent HCV transmission among IDU populations. Thirdly, linkage of laboratory data on diagnosed HCV antibody positive persons in Scotland to clinical data from hospital and death records provided a unique national epidemiological dataset to estimate the number who had progressed to severe HCV disease.