Absorption profiles of rectally administered midazolam estimated by deconvolution

An algebraic deconvolution procedure adopted from the literature has been used to estimate the absorption profiles of rectally administered midazolam (0.3 mg/kg) in 8 healthy subjects. The extent of absorption estimated by the conventional AUC approach has previously been published and was compared to the deconvolution results. In the original study the sampling period was 840 min. When a deconvolution approach is used there may be no need to sample after the absorption is completed. Thus in this communication only samples drawn within the first 120 min and 300 min were used. The bioavailability estimated by the AUC ratio (mean 0.52 +/- SDrel 16%) is compared to the deconvolution results obtained using only data points within 120 (0.46 +/- 22%) (p = 0.02) and 300 min (0.51 +/- 20%) (p = 0.6) after drug administration. The absorption is almost complete after 120 min. Regarding the absorption rates 40% (range 27.55%) of the dose is absorbed after 60 min. The relative (normalised) absorption profiles showed that 79% (range 63-90%) of the maximum absorbed amount is absorbed within 60 min.

Information regarding the rectal absorption of midazolam can be obtained from a sampling period of approximately 5 hours using the deconvolution approach. The algorithm used is relative easy to implement and is very easy to use with polyexponential parameters obtained by curve fitting as input.