Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Imputation of concentration==0 when no concentration record appears at TIME==0/time of dose for Model 1 Single Dose #284

Open
tensfeldt opened this issue Mar 5, 2021 · 0 comments
Labels
3.0 defect Defect identified during code development for Computation Engine v3.0

Comments

@tensfeldt
Copy link
Owner

tensfeldt commented Mar 5, 2021

This is illustrated with testcase tc1822_M1SD and illustrated with commit c3d3f48:

In the following output, profiles SDEID==1 and 2 are identical except that the first profile has a zero concentration record at time==0 (ATIME==0), the second profile (SDEID==2) doesn't have a record with ATIME==0 (time of dose for single dose).

Note that AUCALL and AUCLAST differ considerably in their values but would have been expected to be identical if CE 3.0 commit c3d3f48 was following the requirements of Section 13.3.1 "Missing Data at time of Dose" of the Computation Engine Specifications document that requires that for the single dose model if there is no time or concentration at time of dose, the concentration at time of dosing will be imputed as zero (0.0).

Note also a requirement of the CE specification is that CTOLDesti value be generated that for single dose models will record the imputed value of concentration at time of dosing even for single dose models (see the entry for CTOLDesti for section 10.2 "Matrix Dependent Parameters"). Not that CTOLDest1 was generated and indicates that for the second profile (SDEID==2) that the concentration at time of dose WAS indeed imputed as zero (0). However this doesn't appear to impact the values of the AUCs generated.

A cursory review of the run_M1_SD_computation.R code indicates that there is no code incorporated that would update the profile records to impute concentration of zero at time zero. There is code incorporated into auc_tau.R and aumc_tau.R to incorporate ctoldest into the single dose models. There is a get_told_concentration.R routine that updates dataset for steady state models but this isn't incorporated into the single dose models.

 r %>% select(SDEID, CTOLDest1, AUCALL, AUCLAST, matches("^(CONC|CONCTIME|AUC|AUCINT)[0-9]+?$"))
  SDEID CTOLDest1   AUCALL  AUCLAST     AUC1     AUC2     AUC3     AUC4     AUC5     AUC6     AUC7     AUC8    AUC9 AUCINT1     AUCINT2
1     1        NA 996.2250 996.2250 11.32625 39.95335 102.7670 176.5979 231.9766 331.7574 422.9514 466.5495 996.225 0_0.533         0_1
2     2         0 984.8987 984.8987 28.62710 91.44070 165.2717 220.6504 320.4311 411.6251 455.2232 984.8987      NA 0.533_1 0.533_1.833
  AUCINT3     AUCINT4     AUCINT5 AUCINT6 AUCINT7     AUCINT8 AUCINT9 CONC1 CONC2 CONC3 CONC4 CONC5 CONC6 CONC7 CONC8 CONC9 CONC10
1 0_1.833         0_3     0_4.033 0_6.033     0_8         0_9 0_23.65   0.0  42.5  80.1  70.9  56.2  51.1  48.7  44.1  43.1     30
2 0.533_3 0.533_4.033 0.533_6.033 0.533_8 0.533_9 0.533_23.65    <NA>  42.5  80.1  70.9  56.2  51.1  48.7  44.1  43.1  30.0     NA
  CONCTIME1 CONCTIME2 CONCTIME3 CONCTIME4 CONCTIME5 CONCTIME6 CONCTIME7 CONCTIME8 CONCTIME9 CONCTIME10
1     0.000     0.533     1.000     1.833     3.000     4.033     6.033         8      9.00      23.65
2     0.533     1.000     1.833     3.000     4.033     6.033     8.000         9     23.65         NA
@tensfeldt tensfeldt added the 3.0 defect Defect identified during code development for Computation Engine v3.0 label Mar 5, 2021
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
3.0 defect Defect identified during code development for Computation Engine v3.0
Projects
None yet
Development

No branches or pull requests

1 participant