Investigation of the utility of vitreous humour drug concentrations when compared to peripheral blood using the National Programme on Substance Use Mortality database.

Peripheral blood (PB) remains the standard matrix for postmortem toxicological analysis, yet its interpretive reliability is often compromised by postmortem redistribution (PMR), trauma, or decomposition. Vitreous humour (VH), due to its anatomical isolation and reduced susceptibility to PMR, has emerged as a promising alternative matrix. This study investigates the correlation between VH and PB drug concentrations using data from the National Programme on Substance Use Mortality (NPSUM), aiming to enhance interpretive accuracy and forensic utility of VH. A total of 110 cases with paired VH and PB drug concentrations were identified. Quantitative toxicology data were extracted from coronial submissions across the UK, with analysing laboratories accredited by UKAS or an international equivalent. Statistical comparisons were performed using Wilcoxon signed-rank tests and coefficients of determination (R²) to assess correlation strength. Drug concentrations were consistently lower in VH than PB, yet all drugs detected in PB were also present in VH. Strong correlations were observed for pregabalin (R²=0.86), methadone (R²=0.81), gabapentin (R²=0.76), and codeine (R²=0.70), while cocaine (R²=0.56) showed moderate correlation. Weak correlations were noted for BZE (R²=0.44), EDDP (R²=0.20), and morphine (R²=0.19). These findings suggest that VH concentrations reflect systemic levels for certain drugs, particularly those with limited PMR and favourable physicochemical properties. The study highlights VH's value as a confirmatory matrix, especially when PB is unavailable. However, the variability in VH-PB correlations underscores the need for matrix-specific interpretive frameworks. Factors such as lipophilicity, protein binding, and molecular size appear to influence VH penetration more than PMR alone. This work represents one of the most comprehensive evaluations of VH-PB drug correlations to date and supports the development of a systematic reference database. Future research should expand case numbers and incorporate additional covariates to refine VH-based interpretive models. VH offers a viable alternative for postmortem toxicology, but its evidential weight depends on drug-specific characteristics and robust reference data.