Periodic Reporting for period 1 - SafePolyMed (Improve Safety in Polymedication by Managing Drug-Drug-Gene Interactions)
Okres sprawozdawczy: 2022-06-01 do 2023-11-30
Adverse drug reaction (ADRs), a major cause of hospitalization, rank among the leading causes of death in developed countries. SafePolyMed addresses significant public health issues by focusing on drug-drug and drug-gene interactions (DGIs and DDIs) as major causes of ADRs. These can occur, when genetic polymorphisms (DGIs) or drug co-administrations (DDIs), alter a drug’s behavior, and are strongly interconnected. However, studying complex real-world drug-drug gene interactions (DDGIs) scenarios in clinical trials is not feasible due to combinatorial explosion, high costs and ethical concerns, leading to significant knowledge gaps. Furthermore, insufficient patient education on drug intake, polypharmacy and associated risks along with low levels of patient participation in their healthcare management, poses challenges to patient safety and adherence, especially among elderly, and often comorbid individuals under polymedication. SafePolyMed will develop a novel and innovative framework to define, assess, and manage DDGIs resulting in education and empowerment of citizens as well as in reduced healthcare costs by improving patient safety. SafePolyMed aims: (1) to develop an evidence-based risk scoring system using machine learning on real-world datasets (RWD) to identify patients at risk of developing ADRs; (2) to develop patient-reported outcome measures (PROMs) for patient safety; (3) to empower citizens through a digital Medication Management Center (MMC) allowing patients to manage their therapies, learn about interactions, and collect patient-reported outcomes; (4) to establish individualized model-based dose adaptations of clinically relevant compounds; (5) to validate the safety tools through a proof-of-principle study involving diverse patient cohorts from European clinical sites.
To predict the individual risk of a patient to develop ADRs, statistical and AI models are used that will be trained using RWD. Substantial progress has been made in data processing and model development. As RWD contain sensitive information on disease state, medication intake and genetics, a SafePolyMed analysis platform incorporating three different federated learning (FL) approaches is applied, that enables data analyses without sharing sensitive data.
To analyse free text in the RWD, natural language processing was applied for data analysis, and dictionaries, query lists as well as fact extraction pipelines were established to extract information on treatment outcome or ADR from the free text.
To increase overall citizen participation in health care, SafePolyMed aims to develop PROMs regarding patient safety as well as patient training in cooperation with patient organisations. Therefore, a patient engagement hub (PEH) was established, involving patient organisations with different fields of expertise. The PEH will be consulted throughout the project to ensure that patients are actively engaged and that the developed tools effectively consider patients’ needs.
For the development of safety PROMs, an initial list of side effect symptoms was curated based on symptoms reported in clinical studies conducted by the polypharmacy outpatient clinic of partner UKA and in the real-world databases available to partner UTARTU. Within a currently ongoing DELPHI process, a final set of relevant PROMs will be selected from the list of symptoms by patients and healthcare professionals.
In collaboration with the PEH, the content for a patient training curriculum has been defined that has the overall aim of educating patients on actively managing their polymedication. The training addresses D(D)GIs, side effects, regulatory aspects, GDPR and patient-led evidence generation.
The foundation for the development of the MMC was laid by specifying the architecture of the system and its core components and main use scenarios. Here, the patient's perspective and needs were taken into consideration, using a user-centered approach. Based on these specifications, a vertical prototype of the MMC as well as a user interface were established.
Compounds for mathematical modelling were selected based on information derived from real-world data and literature analyses. PBPK models of mirtazapine, atorvastatin, methotrexate as well as amitriptyline and nortriptyline are currently under development. PBPK models of carbamazepine, metformin, tacrolimus, simvastatin and rosuvastatin are readily available in the PBPK model library of USAAR and will be refined to describe SafePolyMed-relevant DD(G)I scenarios.
To analyse free text in the RWD, natural language processing was applied for data analysis, and dictionaries, query lists as well as fact extraction pipelines were established to extract information on treatment outcome or ADR from the free text.
To increase overall citizen participation in health care, SafePolyMed aims to develop PROMs regarding patient safety as well as patient training in cooperation with patient organisations. Therefore, a patient engagement hub (PEH) was established, involving patient organisations with different fields of expertise. The PEH will be consulted throughout the project to ensure that patients are actively engaged and that the developed tools effectively consider patients’ needs.
For the development of safety PROMs, an initial list of side effect symptoms was curated based on symptoms reported in clinical studies conducted by the polypharmacy outpatient clinic of partner UKA and in the real-world databases available to partner UTARTU. Within a currently ongoing DELPHI process, a final set of relevant PROMs will be selected from the list of symptoms by patients and healthcare professionals.
In collaboration with the PEH, the content for a patient training curriculum has been defined that has the overall aim of educating patients on actively managing their polymedication. The training addresses D(D)GIs, side effects, regulatory aspects, GDPR and patient-led evidence generation.
The foundation for the development of the MMC was laid by specifying the architecture of the system and its core components and main use scenarios. Here, the patient's perspective and needs were taken into consideration, using a user-centered approach. Based on these specifications, a vertical prototype of the MMC as well as a user interface were established.
Compounds for mathematical modelling were selected based on information derived from real-world data and literature analyses. PBPK models of mirtazapine, atorvastatin, methotrexate as well as amitriptyline and nortriptyline are currently under development. PBPK models of carbamazepine, metformin, tacrolimus, simvastatin and rosuvastatin are readily available in the PBPK model library of USAAR and will be refined to describe SafePolyMed-relevant DD(G)I scenarios.
SafePolyMed employs ML and AI techniques to analyze real-world datasets, including genetic information, demographic data, health conditions, and medication history to extract patterns related to ADRs. This approach goes beyond traditional statistical analysis, aiming to provide a more comprehensive understanding of individual risk factors. Furthermore, the project will empower citizens by establishing a training curriculum to improve overall knowledge on polymedication and active patient engagement. Safety-related PROMs will be developed to enable individual standardized safety reporting. This will contribute to a more comprehensive understanding of drug safety. Additionally, the project aims to develop a prototype of an easy-to-use clinical decision support system based on comprehensive PBPK models. Hereby, SafePolyMed aims to promote the adoption of model-based dose recommendations among physicians, enhancing the precision of patient-tailored drug therapy. The MMC will be developed as a harmonised European solution for medication management. It will integrate the different tools developed within SafePolyMed to increase access to health-relevant information and allow citizens to manage their drug therapy.
Overall, SafePolyMed aims to develop a framework to empower citizens, focusing on improving citizen education and their understanding of complex drug therapies, as well as to improve exchange between citizens and healthcare professionals.
The development of respective novel tools is currently ongoing. To ensure future uptake and success of the developed tools, considering the patient’s perspective and needs is substantial. Therefore, a patient engagement hub has been established which will be consulted throughout the project to ensure that the innovative tools consider relevant aspects and are easy to use.
Overall, SafePolyMed aims to develop a framework to empower citizens, focusing on improving citizen education and their understanding of complex drug therapies, as well as to improve exchange between citizens and healthcare professionals.
The development of respective novel tools is currently ongoing. To ensure future uptake and success of the developed tools, considering the patient’s perspective and needs is substantial. Therefore, a patient engagement hub has been established which will be consulted throughout the project to ensure that the innovative tools consider relevant aspects and are easy to use.