Analysis of four syringe dispensing machine point-of-access data 2017-2020 in Melbourne, Australia: machine...

Australian needle and syringe distribution occurs via a mix of modalities, including syringe dispensing machines (SDMs). SDMs are electronic vending machines providing (often) 24-h access to needles/syringes and may attract greater numbers of people who are younger, female, and/or have limited connection to health care services compared to individuals accessing fixed-site needle and syringe programs (NSPs). However, validating the demographic characteristics of SDM clients has proven difficult in previous research.

In this paper, we analyse SDM order and client demographic data from four SDMs located in South-East Melbourne, Australia, and compare this against the managing fixed-site NSP between May 2017 and December 2020. SDM data were collected via a novel 0-9 numeric keypad input tool. Via the tool, SDM clients were requested to input their categorised age, gender and postcode. Given the novelty of the tool, we evaluate the feasibility of the data collection method. We analysed data according to: (1) total SDM orders made, (2) estimated 'unique SDM presentations' and (3) describing the demographics of unique SDM clients. Importantly, we noted substantial invalid demographic data, and consequently, severely restricted data for analysis.

There were 180,989 SDM orders made across the four SDMs to an estimated 90,488 unique SDM presentations. There was little variation in unique presentations across days of the week, but 69% occurred out of NSP operating hours. Across the study period, the SDMs distributed 66% of the number of syringes distributed by the fixed-site NSP. Due to invalid demographic data, our restriction method provided only 10,914 (6% of all data) unique presentations for analysis. There were some demographic differences between SDM and NSP client, but these should be treated with caution.

The data collection tool provides a novel means of comparing SDM and fixed-site presentations, demonstrating the substantial expansion of service via the SDMs. However, the validity of the demographic data was highly questionable and requires significant data coding, meaning it is not feasible for community NSPs. While we recommend the inclusion of automatically collected SDM order data, the use of a 0-9 numeric keypad to collect demographic data-while an innovation-requires alteration to support NSP data.