Cost-benefit analysis of Canada's Prison Needle Exchange Program for the prevention of hepatitis C and injection-related infections

Abstract

Background: Needle exchange programs are effective public health interventions that reduce blood-borne infections, including hepatitis C, and injection-related infections. We sought to assess the return on investment of existing Prison Needle Exchange Programs (PNEPs) in Canadian federal prisons and their expansion to all 43 institutions.

Methods: We developed a stochastic compartmental model that estimated hepatitis C and injection-related infections under different PNEP scenarios in Canadian federal prisons. Scenarios projected for 2018-2030 were no PNEP, status quo (actual PNEP implementation 2018-2022, with coverage maintained to 2030), and PNEP scale-up (coverage among people who inject drugs in prison increasing over 2025-2030 to reach 50% by 2030). We calculated the benefit-cost ratio as benefits from health care savings, divided by PNEP costs.

Results: By 2019, PNEPs were implemented in 9 of 43 federal prisons, with uptake reaching 10% of people who injected drugs in prison in 2022. Compared with no PNEP, this was estimated to cost Can$0.45 (uncertainty interval [UI] $0.32 to $0.98) million and avert 37 (UI 25 to 52) hepatitis C and 8 (UI -1 to 16) injection-related infections over 2018-2030, with a benefit-cost ratio of 1.9 (UI 0.56-3.0). Compared with the status quo, the PNEP scale-up scenario cost an additional $2.7 (UI $1.8 to $7.0) million and prevented 224 (UI 218 to 231) hepatitis C and 77 (UI 74 to 80) injection-related infections, with a benefit-cost ratio of 2.0 (UI 0.57 to 3.3).

Interpretation: Every dollar invested in the current PNEP or its expansion is estimated to save $2 in hepatitis C and injection-related infection treatment costs. This return on investment strongly supports ongoing maintenance and scale-up of the PNEP in Canada from an economic perspective.

Figure 1:

Schematic of the Prison Needle Exchange Program (PNEP) enrolment and participation process in Canadian federal prisons. A detailed description of the service delivery model can be found in Appendix 1, Section A, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.240648/tab-related-content.

Figure 2:

Schematic representation of the compartmental model for a Prison Needle Exchange Program (PNEP). The yearly number of people sentenced to full-time custody in Canadian federal prisons determines the number of prison entrants into the model, and individuals are instantly classified according to their injecting drug use behaviour in prison (injecting v. noninjecting) and hepatitis C RNA status (positive v. negative and susceptible) (blue arrows). Transitions are described as probabilities of becoming infected with hepatitis C (black downward arrow), receiving treatment for hepatitis C (black upward arrows), or experiencing an injection-related infection (black curved arrows). People exit the model according to a calibrated average sentence length (dashed arrows). The PNEP intervention acts by reducing the probabilities of experiencing an injection-related infection or acquiring hepatitis C among people who inject drugs in prison (red arrows). A detailed description of the model can be found in Appendix 1, Section B, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.240648/tab-related-content.

Figure 3:

Scenario outcomes. (A, B): Status quo (blue) and counterfactual (orange). (C, D): Status quo (blue) and scale-up (red). The lines are annual medians from 500 sampled model trajectories, and the shaded areas represent the range of values from the runs. (A) and (C): Incidence of injection-related infections per 100 people who inject drugs in prison per year. (B) and (D): Hepatitis C incidence per 100 people who inject drugs in prison per year.

Figure 4:

Incremental costs and associated health care savings (2023 Canadian dollars): status quo versus counterfactual and Prison Needle Exchange Program scale-up versus status quo, 2018–2030. The left-hand bar shows the cost breakdown, including one-off, per unit, annual start-up and operational costs. The right-hand bar shows the health care savings associated with reduced hepatitis C transmission and incidence of injection-related infection. Note that the scale-up scenario and status quo are the same up to the year 2024, so the graph represents differences across the years 2025–2030. Thin vertical bars represent 95% uncertainty intervals for total costs and benefits.

Figure 5:

Tornado chart illustrating the parameters that were most influential on the benefit–cost ratio. The point estimate refers to the benefit–cost ratio of the scale-up scenario compared with the status quo. The blue and red bars indicate the lower and upper bounds of the parameters, respectively. The $21 163 baseline hepatitis C treatment cost is the weighted average of the $20 000 cost in prison and $60 000 in community (Table 3), as some prison-acquired infections are treated after release (Appendix 1, Section B, available at www.cmaj.ca/lookup/doi/10.1503/cmaj.240648/tab-related-content). Additional sensitivity analyses are provided in Appendix 1, Section D. Note: IDU = injecting drug use.