Evaluation of machine learning methods for covariate data imputation in pharmacometrics
Abstract Missing data create challenges in clinical research because they lead to loss of statistical power and potentially to biased results. Missing covariate data must be handled with suitable approaches to prepare datasets for pharmacometric analyses, such as population pharmacokinetic and pharm...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-12-01
|
| Series: | CPT: Pharmacometrics & Systems Pharmacology |
| Online Access: | https://doi.org/10.1002/psp4.12874 |
| _version_ | 1811198867603980288 |
|---|---|
| author | Dominic Stefan Bräm Uri Nahum Andrew Atkinson Gilbert Koch Marc Pfister |
| author_facet | Dominic Stefan Bräm Uri Nahum Andrew Atkinson Gilbert Koch Marc Pfister |
| author_sort | Dominic Stefan Bräm |
| collection | DOAJ |
| description | Abstract Missing data create challenges in clinical research because they lead to loss of statistical power and potentially to biased results. Missing covariate data must be handled with suitable approaches to prepare datasets for pharmacometric analyses, such as population pharmacokinetic and pharmacodynamic analyses. To this end, various statistical methods have been widely adopted. Here, we introduce two machine‐learning (ML) methods capable of imputing missing covariate data in a pharmacometric setting. Based on a previously published pharmacometric analysis, we simulated multiple missing data scenarios. We compared the performance of four established statistical methods, listwise deletion, mean imputation, standard multiple imputation (hereafter “Norm”), and predictive mean matching (PMM) and two ML based methods, random forest (RF) and artificial neural networks (ANNs), to handle missing covariate data in a statistically plausible manner. The investigated ML‐based methods can be used to impute missing covariate data in a pharmacometric setting. Both traditional imputation approaches and ML‐based methods perform well in the scenarios studied, with some restrictions for individual methods. The three methods exhibiting the best performance in terms of least bias for the investigated scenarios are the statistical method PMM and the two ML‐based methods RF and ANN. ML‐based approaches had comparable good results to the best performing established method PMM. Furthermore, ML methods provide added flexibility when encountering more complex nonlinear relationships, especially when associated parameters are suitably tuned to enhance predictive performance. |
| first_indexed | 2024-04-12T01:38:26Z |
| format | Article |
| id | doaj.art-41d2fb2990be42f0ad388eb4544e9aab |
| institution | Directory Open Access Journal |
| issn | 2163-8306 |
| language | English |
| last_indexed | 2024-04-12T01:38:26Z |
| publishDate | 2022-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | CPT: Pharmacometrics & Systems Pharmacology |
| spelling | doaj.art-41d2fb2990be42f0ad388eb4544e9aab2022-12-22T03:53:16ZengWileyCPT: Pharmacometrics & Systems Pharmacology2163-83062022-12-0111121638164810.1002/psp4.12874Evaluation of machine learning methods for covariate data imputation in pharmacometricsDominic Stefan Bräm0Uri Nahum1Andrew Atkinson2Gilbert Koch3Marc Pfister4Pediatric Pharmacology and Pharmacometrics University Children's Hospital Basel (UKBB), University of Basel Basel SwitzerlandPediatric Pharmacology and Pharmacometrics University Children's Hospital Basel (UKBB), University of Basel Basel SwitzerlandPediatric Pharmacology and Pharmacometrics University Children's Hospital Basel (UKBB), University of Basel Basel SwitzerlandPediatric Pharmacology and Pharmacometrics University Children's Hospital Basel (UKBB), University of Basel Basel SwitzerlandPediatric Pharmacology and Pharmacometrics University Children's Hospital Basel (UKBB), University of Basel Basel SwitzerlandAbstract Missing data create challenges in clinical research because they lead to loss of statistical power and potentially to biased results. Missing covariate data must be handled with suitable approaches to prepare datasets for pharmacometric analyses, such as population pharmacokinetic and pharmacodynamic analyses. To this end, various statistical methods have been widely adopted. Here, we introduce two machine‐learning (ML) methods capable of imputing missing covariate data in a pharmacometric setting. Based on a previously published pharmacometric analysis, we simulated multiple missing data scenarios. We compared the performance of four established statistical methods, listwise deletion, mean imputation, standard multiple imputation (hereafter “Norm”), and predictive mean matching (PMM) and two ML based methods, random forest (RF) and artificial neural networks (ANNs), to handle missing covariate data in a statistically plausible manner. The investigated ML‐based methods can be used to impute missing covariate data in a pharmacometric setting. Both traditional imputation approaches and ML‐based methods perform well in the scenarios studied, with some restrictions for individual methods. The three methods exhibiting the best performance in terms of least bias for the investigated scenarios are the statistical method PMM and the two ML‐based methods RF and ANN. ML‐based approaches had comparable good results to the best performing established method PMM. Furthermore, ML methods provide added flexibility when encountering more complex nonlinear relationships, especially when associated parameters are suitably tuned to enhance predictive performance.https://doi.org/10.1002/psp4.12874 |
| spellingShingle | Dominic Stefan Bräm Uri Nahum Andrew Atkinson Gilbert Koch Marc Pfister Evaluation of machine learning methods for covariate data imputation in pharmacometrics CPT: Pharmacometrics & Systems Pharmacology |
| title | Evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| title_full | Evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| title_fullStr | Evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| title_full_unstemmed | Evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| title_short | Evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| title_sort | evaluation of machine learning methods for covariate data imputation in pharmacometrics |
| url | https://doi.org/10.1002/psp4.12874 |
| work_keys_str_mv | AT dominicstefanbram evaluationofmachinelearningmethodsforcovariatedataimputationinpharmacometrics AT urinahum evaluationofmachinelearningmethodsforcovariatedataimputationinpharmacometrics AT andrewatkinson evaluationofmachinelearningmethodsforcovariatedataimputationinpharmacometrics AT gilbertkoch evaluationofmachinelearningmethodsforcovariatedataimputationinpharmacometrics AT marcpfister evaluationofmachinelearningmethodsforcovariatedataimputationinpharmacometrics |