Correspondence  |   February 2002
Prediction of Propofol Induction Dose Using Multiple Regression Analysis
Author Notes
  • Department of Anesthesiology and Intensive Care, Hamamatsu University School of Medicine, Hamamatsu, Japan.
Article Information
Correspondence   |   February 2002
Prediction of Propofol Induction Dose Using Multiple Regression Analysis
Anesthesiology 2 2002, Vol.96, 518-519. doi:
Anesthesiology 2 2002, Vol.96, 518-519. doi:
In Reply:—
Dr. Adachi did not make the important distinction between administration rates of 15 and 40 mg·kg−1·h−1. However, propofol administration rate has a critical impact on induction dose. 1 The pharmacokinetics are not stationary. Distribution volumes and clearances vary over time. Schnider et al.  2 administered propofol at the very slow infusion rate of 15 mg·kg−1·h−1. Although this slow propofol administration rate is useful for studying pharmacokinetic parameters, it is not suitable for investigation of clinical induction doses and times. As I previously reported, induction doses are highly variable at administration rates of less than 20 mg·kg−1·h−1. 1 At these low rates, propofol metabolized during infusion may have important effects. As a consequence, the selected parameters in our study were age, lean body mass (LBM), central blood volume (CBV), and hepatic blood flow (HBF) at the rate of 40 mg·kg−1·h−1. 3 
The propofol administration rate of 15 mg·kg−1·h−1that Dr. Adachi used is not clinically acceptable because it takes over 5 min for loss of consciousness. Moreover, I was very surprised to see that cardiac output (CO) was selected as a parameter of induction dose at this slow administration rate. Although CO was also correlated with induction dose, it could not be identified as a significant variable in the multiple linear regression model in our study. 3 I could not explain the difference clearly.
The total R2of hypnotic dose by Dr. Adachi was 0.468. This means that only 46.8% of induction dose can be explained with their selected parameters of age, body weight, CO, and k (the plasma disappearance rate of indocyanine green). Our total R2for induction dose was 0.85 (at 40 mg·kg−1·h−1). 3 Dr. Adachi reanalyzed his data using the parameters identified in our study. However, the R2was still less than 0.5. The small total R2of Dr. Adachi might be a result of the large variability in the induction dose and the slow administration rates.
There are no complete independent variables among age, sex, body weight, height, LBM, hemoglobin, CO, BV, CBV, and HBF. It is important to select parameters that make R2close to 1.0 in multiple linear regressions. Multicolinearity among variables makes regression difficult to interpret. We used forward and backward selection to identify the most useful variables for predicting the induction dose. The criterion for adding and deleting variables was a minimum of 4.0 for the F ratio, which was the square of the value obtained from a t  test with the hypothesis that the coefficient of the variable in question was equal to zero.
Kazama T, Ikeda K, Morita K, Kikura M, Ikeda T, Kurita T, Sato S: Investigation of effective induction doses using a wide range of infusion rates with undiluted and diluted propofol. A nesthesiology 2000; 92: 1017–28Kazama, T Ikeda, K Morita, K Kikura, M Ikeda, T Kurita, T Sato, S
Schnider TW, Minto CF, Gambus PL, Andresen C, Goodale DB, Shafer SL, Youngs EJ: The influence of method of administration and covariates on the pharmacokinetics of propofol in adult volunteers. A nesthesiology 1998; 88: 1170–82Schnider, TW Minto, CF Gambus, PL Andresen, C Goodale, DB Shafer, SL Youngs, EJ
Kazama T, Ikeda K, Morita K, Ikeda T, Kikura M, Sanjo Y, Sato S: The relation between initial blood distribution volume and propofol induction dose requirement. A nesthesiology 2001; 94: 205–10Kazama, T Ikeda, K Morita, K Ikeda, T Kikura, M Sanjo, Y Sato, S