Aims To illustrate the impact of different proportions of injecting drug users (IDUs) sharing high dead-space syringes (HDSS) or low dead-space syringes (LDSS) on the probability of human immunodeficiency virus (HIV) transmission; and thus the impact on injection-related HIV prevalence and incidence.
Design A stochastic mathematical model was used to evaluate the impact of HDSS use in high- and low-risk IDU populations. Model parameters were obtained from peer-reviewed publications. Analytical solutions of a simplified deterministic model were obtained to explain the effect of HDSS on HIV endemic states.
Findings Simulation analysis shows that the HIV epidemic could be sustained even when a small percentage of sharing (10%) involved HDSS. The effect is much stronger in high-risk compared with low-risk populations. Steady state HIV prevalence increases with the proportion of HDSS, and for high- and low-risk populations reaches around 80% and 20%, respectively. For low-risk populations, the use of LDSS could result in the virtual elimination of HIV. These results are dependent upon an evidence-supported assumption of a significant difference in HIV transmission risk associated with HDSS versus LDSS.
Conclusions Our models suggest that injection-related HIV epidemics may not occur when most (e.g. 95% or more) IDUs use LDSS. While these results are based on indirect risk measures and a number of simplifying assumptions, the effect of blood retained in high dead-space syringes on HIV prevalence seems to be very strong, even using relatively conservative assumptions. The findings have potential implications for needle exchange programs and the types of syringes produced and distributed world-wide.