Analysis of Real-World Data and update of a Clinical Decision Support System (CDSS) to investigate and support quality and sustainability of pharmacological therapies

This research project was inspired by the promotion of industrial PhDs focused on innovative and green topics, as well as by the need to identify tools and methodologies to improve the management of polypharmacy, with the final aim to improve quality and sustainability of healthcare. Indeed, in recent years medications have been recognized as emerging pollutants, with potential negative effects on both the ecosystem and human health. Pharmaceutical residues can enter the environment through various mechanisms, first among them human excretion and improper drug disposal. In addition, the increasing number of patients with chronic conditions, largely due to an aging population and increasingly complex clinical profiles, contributes to a broader context of medication-related problems (MRPs) and increased medication waste, often linked to the prescription of potentially inappropriate medications (PIMs) and excessive polypharmacy. This research project was first aimed at updating a Clinical Decision Support System (CDSS) developed by the Italian IT company Infologic s.r.l. (Padua, Italy) to promote technological transition in healthcare and support prescriptive appropriateness and optimization of pharmacological therapies. The secondary objective was about drug utilisation research (DUR) to support healthcare providers in the identification of areas for intervention to improve the use of medications and implement strategies to enhance the quality of care. Specifically, the CDSS developed by Infologic is a digital platform designed to support medication review and deprescription and may represent an effective strategy to reduce MRPs and PIMs. The dissertation presents in detail the CDSS developed by the company, as well as the methodologies adopted to update and extend its knowledge-base, which represents the system information database and includes evidence-based information derived from the revision of scientific literature. A version of the CDSS with the updated knowledge-base was published in 2023 and tested in an observational pilot study on a small sample of older patients admitted to the internal medicine department of a Piedmont hospital. The experience of the pilot study and feedback from CDSS users enabled the inclusion in the CDSS of new features to optimize the medication-review process and improve user experience, reducing both alert fatigue and work time. The CDSS database is fed by a large amount of information that can be organized in 33 tables containing alerts on potential MRPs and recommendations for the proper use of medications. New information added to the system includes, for example, potentially nephrotoxic drugs, guidelines for deprescribing specific drug classes and the environmental impact of medications. To complete DUR activities, research agreements were made with healthcare facilities in Piedmont and the Valle d’Aosta Region to exchange anonymized electronic health data, including drug dispensing data and hospital discharge records. Since DUR yielded numerous results that led to at least eight peer-reviewed scientific publications, the dissertation focuses in depth on three large studies aimed at investigating adherence, persistence and intensifications of antidiabetic drug treatment in naïve patients, at evaluating the impact of the introduction of new regulations for the prescription of antidiabetic drugs on the prevalence of drug use and at describing the use of lipid-lowering drugs in patients hospitalized for major cardio and/or cerebrovascular events. All analyses were performed using the R programming language. Activities carried out during the PhD should continue in the coming years, both in terms of collaboration with Infologic to keep the CDSS updated and improved, and with healthcare facilities to investigate the use of drugs in the real world. Both activities are essential to ensure optimal care and improve therapy management.