Midazolam is one of top three drugs used in palliative care. Its use increases in the last days of hospice patients’ lives while safe dosage can be challenging. Equations currently used to estimate glomerular filtration rate, e.g: the Cockroft-Gault (eGFR_CR) and the Modification of Diet in Renal Disease (eGFR_MDRD) ones, do not generate precise calculations, especially in palliative patients exhibiting variations in body parameters. Our aim was to seek new relationships between mean midazolam (M_avg) and alfahydroxymidazolam (OH-M_avg) concentrations in plasma, and selected biochemical and physiological parameters of palliative patients, to enable optimal midazolam pharmacotherapy.

The pilot study included 11 Caucasians, aged 42–95, with advanced cancer disease, receiving midazolam in a hospice in-patient unit. We tested correlations among M_avg, BMI, eGFR_MDRD, midazolam clearance (CL), OH-M_avg, bilirubin (Bil) and blood creatinine concentration (Cr). F test and leave-one out (LOO) validation was applied to verify the correlations’ significance and predictive ability.

We found ten statistically significant (p < 0.05) correlations related to midazolam pharmacokinetics and physiological factors. We formulated two equations with high degree of predictive ability, based on the eGFR_MDRD→CL and the (Bil + BMI × Ln(Cr))→M_avg-(OH-M_avg) correlations.

The limitations of the study mainly revolve around its pilot nature and the need to continue testing the results on a bigger population.

No funding to disclose.

The significance of correlations corresponding to the arithmetic expressions confirms that Bil, BMI, Ln(Cr) analyzed simultaneously report a series of processes on which midazolam metabolism depends. Two of ten correlations proposed came close to meet all LOO validation criteria. Current findings can help optimize midazolam treatment in palliative therapy.