Individual and population pharmacokinetic compartment analysis: a graphic procedure for quantification of predictive performance

Abstract

ObjectivesPharmacokinetic studies are important for optimizing of drug dosing, but requires proper validation of the used pharmacokinetic procedures. However, simple and reliable statistical methods suitable for evaluation of the predictive performance of pharmacokinetic analysis are essentially lacking. The aim of the present study was to construct and evaluate a graphic procedure for quantification of predictive performance of individual and population pharmacokinetic compartment analysis.MethodsOriginal data from previously published pharmacokinetic compartment analyses after intravenous, oral, and epidural administration, and digitized data, obtained from published scatter plots of observed vs predicted drug concentrations from population pharmacokinetic studies using the NPEM algorithm and NONMEM computer program and Bayesian forecasting procedures, were used for estimating the predictive performance according to the proposed graphical method and by the method of Sheiner and Beal.ResultsThe graphical plot proposed in the present paper proved to be a useful tool for evaluation of predictive performance of both individual and population compartment pharmacokinetic analysis.ConclusionThe proposed method is simple to use and gives valuable information concerning time- and concentration-dependent inaccuracies that might occur in individual and population pharmacokinetic compartment analysis. Predictive performance can be quantified by the fraction of concentration ratios within arbitrarily specified ranges, e.g. within the range 0.8–1.2.

Challenges associated with test dose pharmacokinetic predictions of high dose melphalan exposure in patients with multiple myeloma

“…Predictive performance was evaluated as previously described [ 27 , 28 ]. Prediction error (PE) was first calculated using the equation PE = observed ( obs )–predicted ( pred ) observation, then expressed as a percentage of the observed value; PE% = ( obs–pred )/ obs *100).…”

“…Mean absolute prediction error (MAPE) was the mean of │PE│. The differences between observed and predicted observations were also evaluated by generation of a scatterplot of observed versus the ratio of predicted/observed observation as previously described [ 27 , 28 ], and the fraction of ratios falling within a range of 0.8 to 1.2 was also calculated. All manipulation of data and statistical analyses were performed using SPSS version 25 or Microsoft Excel (Microsoft office 2010).…”

Dosing clonidine for sedation in intensive care

“…In this issue of Pediatric Anesthesia, Hayden and colleagues 3 present a cleverly designed study that uses published data to provide further guidance to drug dosing There are limitations to population pharmacokinetics. 5 The PK estimates must be represented by the samples collected; study design has impact. Dosing history and sampling times must be pre- We need more good PKPD based dosing guidelines to help us deliver the ideal drug dose regime to our pediatric patients.…”

“…There are limitations to population pharmacokinetics . The PK estimates must be represented by the samples collected; study design has impact.…”

vaCATE: A Platform for Automating Data Output from Compartmental Analysis by Tracer Efflux

“…Compartmental analysis is a methodology used to investigate the fluxes and accumulation of matter and energy in compartmented systems [1]. It is used to model phenomena from a diverse breadth of areas, including pharmacokinetics [2,3], geography [4], ecology [5,6], oncology [7,8], physiology [9,10,11,12], and even building design [13]. This broad utility is facilitated by the uniform theory compartmental analysis is built upon, and the systematic modelling methods it provides [1].…”