Correction to: Model-Informed Precision Dosing Software Tools for Dosage Regimen Individualization: A Scoping Review (Pharmaceutics, (2023), 15, 7, (1859), 10.3390/pharmaceutics15071859)

Abstract

There was an error in the original publication [1]; there is a mistake in the number of software. A correction has been made in the Abstract: “Background: Pharmacokinetic nomograms, equations, and software are considered the main tools available for Therapeutic Drug Monitoring (TDM). Model-informed precision dosing (MIPD) is an advanced discipline of TDM that allows dose individualization, and requires a software for knowledge integration and statistical calculations. Due to its precision and extensive applicability, the use of these software is widespread in clinical practice. However, the currently available evidence on these tools remains scarce. Objectives: To review and summarize the available evidence on MIPD software tools to facilitate its identification, evaluation, and selection by users. Methods: An electronic literature search was conducted in MEDLINE, EMBASE, OpenAIRE, and BASE before July 2022. The PRISMA-ScR was applied. The main inclusion criteria were studies focused on developing software for use in clinical practice, research, or modelling. Results: Twenty-eight software were classified as MIPD software. Nine are currently unavailable. The remaining 19 software were described in depth. It is noteworthy that all MIPD software used Bayesian statistical methods to estimate drug exposure and all provided a population model by default, except NONMEN. Conclusions: Pharmacokinetic software have become relevant tools for TDM. MIPD software have been compared, facilitating its selection for use in clinical practice. However, it would be interesting to standardize the quality and validate the software tools.” The corrections have been made to Section 3.2. Characteristics of the Articles and Software in order to include ID-ODS software, which was originally mistaken as an old version of RxStudio: “Among the 25 articles included (Supplementary Table S2), 37 software tools were identified. The 37 software tools identified were classified: 8 population modeling software, 1 research software, and 28 MIPD software. Additionally, four software were located by manual search: ADAPT, iDose [11], RxStudio [12], and CAPCIL. The last program mentioned is the new version of SIMKIN. Supplementary Table S3 shows the software identified and designed for modeling or research. During the analysis, a total of 28 MIPD software were thoroughly examined. However, nine of these software were excluded for various reasons. Five of them were no longer available or marketed (PHAR-MONITOR, PKRD, OPT, DataKineticsTM, and RADKi-netics). Three software were previous versions of other included software, and the newer versions were considered instead (SIMKIN, USC*PACK, and T.D.M.S. 2000). TDM for R was dismissed as it is a JPKD plugin. A full description of the main characteristics of the 19 MIPD software selected are detailed in Table 1. These software were developed between the years 1979 and 2020. Ten of them were developed by commercial companies, while the remaining were created by non-company providers. Eight software have been updated within the last year. Twelve software offer the option of a free trial, except for NextDose, which does not provide this option. Information regarding free access was not available for the remaining six software. Regarding free access, it is available for four software (JPKD, TCIWorks, TUCUXI, and TDMx). RxStudio, ID-ODS, and NextDose offer different types of licenses, including paid and free subscriptions with varying functionalities. For instance, the free access to RxStudio only includes simulations of empiric treatments. The access method for CAPCIL is unknown. Out of the 19 selected MIPD software, 13 of them are web-based. Among these, four software (Autokinetics, MwPharm, PrecisePK, and RxStudio) also offer a desktop version in addition to the web-based platform. Additionally, DoseMeRx, MwPharm, and RxStudio can also be accessed through mobile applications. The remaining software are solely available in desktop versions. All of the MIPD software analyzed in the study employ Bayesian statistical methods to estimate drug exposure and generate dosing recommendations. From a statistical perspective, population PK modeling approaches can be categorized as either parametric or nonparametric [13]. In the parametric Bayesian approach, the population parameters are treated as random variables with known prior distributions. Estimating the conditional distribution of the population parameters can be challenging in this case. On the other hand, the nonparametric approach assumes that the population distribution is completely unknown and random, making it a more flexible Bayesian approach [13]. It is worth noting that Autokinetics, CAPCIL, and PKS additionally offer the capability to perform linear and nonlinear regression. By default, all the MIPD software provide a population model, except for NONMEN, within which models might be defined by users. Fourteen software include models developed for both adult and pediatric populations. Ten software also include models specific to the neonatal population. Moreover, several software tools offer models for special populations: dialysis patients (DoseMeRx, MwPharm, PrecisePK, ID-ODS, and RxStudio), obese patients (ID-ODS, DoseMeRx, and PrecisePK), and critically ill patients (MwPharm, ID-ODS, PrecisePK, and PKS). Information regarding this aspect was not available for Autokinetics, BestDose, JPKD, and Kinetidex. Among the MIPD software analyzed, there are options for including new population PK models and drugs based on user requests or developer support. Eight software (DoseMeRx, ID-ODS, iDose, InsightRX Nova, MwPharm, PrecisePK, RxKinetics, and RxStudio) allow the inclusion of new population PK models and drugs upon user request, while four software (JPKD, MwPharm, NONMEM, and TCIWorks) enable users to add their own models. TUCUXI allows users to include new population PK models and drugs supported by the developer. TDMx offers the possibility of user-defined new population models but only for the drugs already included by default in the software. InsightRX Nova provides automated population model selection based on patient input data. ID-ODS includes free-style simulation routines that allow the incorporation of one-, two-, and three-compartment models for the drugs available in the system. Sixteen MIPD software include the option to generate reports, and eleven of them support graphical representation. Integration with EHR is feasible with seven software (Autokinetics, DoseMeRx, InsightRX Nova, MwPharm, ID-ODS, PrecisePK, RxStudio, and TUCUXI). Regarding the preset drugs accessible in each software, antibiotics are the most commonly implemented drug class. Immunosuppressants are included in eleven software, with specific drugs used in transplant patients, such as mTOR inhibitors, being included in DoseMeRx, JPKD, NextDose, PrecisePK, RxStudio, TCIWorks, and TUCUXI. Biologic immunomodulator agents are included in iDose, MwPharm, and TDMx. InsightRX Nova covers both transplant drugs and biologics. Antiepileptics are incorporated in four software (JPKD, Kinetidex, MwPharm, and PrecisePK), while kinase inhibitors are only included in TUCUXI. InsightRX Nova covers the most extensive range of drug classes, while Autokinetics and iDose focus on antibiotics and biological agents, respectively. ID-ODS focuses on antibiotics and antifungal agents. NONMEN does not include any preset drugs, and all drugs must be added by the user. Ease of use was assessed based on factors such as simple access, a user-friendly interface, and a smooth flow through the necessary steps to perform a simulation. Among the 19 included software, eight were considered intuitive and user-friendly (Tucuxi, PrecisePK, ID-ODS, RxStudio, MwPharm, iDose, DoseMe, and InsightRX Nova). All of these software were web-based, and all, except for ID-ODS, had the capability of EHR integration, which enhanced the user experience. They also shared a modern interface where the path to run a simulation was easily identifiable. Four other software (NextDose, BestDose, TDMx, and RxKinetics) were also considered relatively easy to use, although their interfaces were less user-friendly and the workflow was less assisted compared to the previous group. PKS and JPKD were not considered user-friendly due to their outdated interfaces and lack of smooth flow through the screens. NONMEN deserves special mention as its use requires a high level of expertise not only in pharmacokinetics and pharmacometrics but also in programming, making it less accessible in clinical settings. Lastly, Kinetidex, Autokinetics, CAPCIL, and TCIworks were not accessible to the investigators, so their ease of use could not be evaluated.” The corrections have been made to Section 4 (Paragraph 8), which are a consequence of the inclusion of ID-ODS as a separate software in the review: “Indeed, the quality of the software, user support services, supported drugs and population models, as well as output generation, are crucial considerations for selecting a MIPD software. The study by Kantasiripitak et al. [1] positioned DoseMeRx, InsightRX Nova, and MwPharm as excellent options for dose optimization in clinical practice. Similarly, the review by Fuchs et al. [7] identified MwPharm, MM-USC*PACK (currently BestDose), and TCIWorks as the most complete programs. In our opinion, software that includes population PK models for special populations, such as obese patients, critically ill patients, neonates, or patients on dialysis, provides added benefits for professionals who frequently treat patients with more complex requirements. PrecisePK and ID-ODS stand out as the most complete tool in this regard, as it includes models for each of the mentioned target populations. However, ID-ODS is only focused on antimicrobial and antifungal agents, whereas PrecisePK includes a wide range of drug types. Additionally, DoseMeRx, MwPharm, and RxStudio offer the ability to predict optimal dosing regimens for various special populations. Apart from special populations, DoseMeRx, InsightRX Nova, and MwPharm provide the largest collection of population PK models by default. These aspects facilitate the implementation of such software in clinical practice.” The corrections were made in Table 1, which include the information related to ID-ODS software. The corrected Table 1 appears below. The authors state that the scientific conclusions are unaffected. This correction was approved by the Academic Editor. The original publication has also been updated.