In Reply:

We thank Dr. Hahn for his excellent comments1  and thorough interest in our study.2  In this study, we tried to identify important covariates that could be used for designing future volume kinetic studies.

We acknowledge our study limitations. Namely, in the present study, imprecise urine data were obtained via bladder ultrasound resulting in an uncertainty in urinary output with regard to timing and volume. Thus, to not introduce bias, urine measurements were not incorporated in the modeling. We appreciate that this drawback potentially resulted in a less stable model with higher interindividual variability. In addition, the design of the study, the study population, and the low number of subjects3  and observations could also result in higher estimated interindividual variability of the elimination rate constant (ke) compared with previous studies.

In our study, mean arterial pressure was strongly associated with the subject’s state (e.g., being either anesthetized or awake). Including highly associated covariates in a stepwise covariate model building procedure will result in high imprecision and instability in the covariate analysis.4,5  Therefore, we chose to only include the subject’s state as covariate. Additionally, because the subject’s state would likely be known before any intervention, this covariate could be easier to apply when designing future studies. Consequently, because the correlation between mean arterial pressure and the subject’s state is high, we do not believe that including anesthetized-induced hypotension would improve the model fit but instead would likely dilute the impact of the subject’s state covariate.

We have thoroughly reexamined table 1 in the original manuscript2  and confirm that the estimates from the model building are correct. Thus, the model (represented by table 1)2  could be used for extrapolation and design of future studies. There is, however, a typographical error in the simulations, switching the central-to-peripheral transfer rate constant to the peripheral-to-central transfer rate constant when simulating the subject’s state effect. This error would impact the simulations for the subject’s state (anesthetized or awake) in the opposite direction (i.e., resulting in a slightly lower area under the curve and maximum plasma dilution with anesthetized subjects compared with awake subjects). We thank Dr. Hahn for detecting this error.

Competing Interests

Dr. Svensen has financial support from Masimo Inc. (Irvine, California), Braun (Sweden) and Fresenius (Sweden) not related to this project. Dr. Kinsky has financial support from U.S. Department of Defense (Arlington, Virginia) not related to this project. Dr. Nyberg declares no competing interests.

References

References
1.
Hahn
RG
:
Anesthesiology
2020
;
132
:
•••
•••
2.
Nyberg
J
,
Li
H
,
Wessmark
P
,
Winther
V
,
Prough
DS
,
Kinsky
MP
,
Svensén
CH
:
Population kinetics of 0.9% saline distribution in hemorrhaged awake and isoflurane-anesthetized volunteers.
Anesthesiology
2019
;
131
:
501
11
3.
Pradhan
S
,
Song
B
,
Lee
J
,
Chae
JW
,
Kim
KI
,
Back
HM
,
Han
N
,
Kwon
KI
,
Yun
HY
:
Performance comparison of first-order conditional estimation with interaction and Bayesian estimation methods for estimating the population parameters and its distribution from data sets with a low number of subjects.
BMC Med Res Methodol
2017
;
17
:
154
4.
Ahamadi
M
,
Largajolli
A
,
Diderichsen
PM
,
de Greef
R
,
Kerbusch
T
,
Witjes
H
,
Chawla
A
,
Davis
CB
,
Gheyas
F
:
Operating characteristics of stepwise covariate selection in pharmacometric modeling.
J Pharmacokinet Pharmacodyn
2019
;
46
:
273
85
5.
Ribbing
J
,
Jonsson
EN
:
Power, selection bias and predictive performance of the population pharmacokinetic covariate model.
J Pharmacokinet Pharmacodyn
2004
;
31
:
109
34