Electronic Health Records for Drug Repurposing: Current Status, Challenges, and Future Directions

Optimising electronic documentation of medication in Hungary: Itemised, complete, historical, and standardised event recording

Hospital care is a highly complex process, requiring comprehensive documentation of all aspects of the patient journey in electronic health records. A critical component of this care is the accurate tracking of patient medications. International standards are not consistently incorporated into the electronic medication systems currently in use worldwide, and their interoperability remains an unresolved issue. We recognised the need to develop a set of standardised data elements that ensure consistent and accurate documentation. Although the medication systems studied exhibit various strengths and weaknesses and can satisfactorily document certain aspects of the medication process, none achieve the necessary level of optimal documentation. Our paper presents a new perspective on medication recording by identifying the electronic data requirements for all events in an itemized, complete, historical, and standardized manner. To address this gap, we collected, defined, and introduced the essential data elements required for the comprehensive documentation of medication sub-processes for the first time in our study. The Fast Health Interoperability Resources (FHIR) data exchange standard was employed for designing these data requirements. Our research identified and categorised 138 data elements essential for describing the complete medication process, including medication description, requests, dispensation, and administration. These data elements were divided into fundamental and supplementary categories. We developed a survey form to assess medication systems. In a pilot study, we tested the quality of 5 medication systems, currently in operation in Hungary. Our analysis assessed the accuracy of the electronic recording of medication and the correspondence of the recorded data elements with international standards. None of the systems demonstrated the ability to document medication accurately or capture all fundamental data elements. The best-performing system managed to record 63 % of all fundamental data elements, while the worst-performing system managed only to document 30 %. The names and the values of data elements in these systems did not comply with international standards either. The primary clinical pharmaceutical usefulness of this study was to enhance the digital documentation of medication in hospitals to meet comprehensive data recording requirements, ensure greater compliance, and improve their suitability for enriching clinical health data files, enabling real-world studies, pharmacovigilance analyses, and the identification of drug repositioning opportunities.

Combination Therapies in Drug Repurposing: Personalized Approaches to Combatting Leukaemia and Multiple Myeloma

Despite advances in cancer research, treating malignancies remains challenging due to issues like drug resistance, disease heterogeneity, and the limited efficacy of current therapies, particularly in relapsed or refractory cases. In recent years, several drugs originally approved for non-cancer indications have shown potential in cancer treatment, demonstrating anti-proliferative, anti-metastatic, and immunomodulatory effects. Drug repurposing has shown immense promise due to well-established safety profiles and mechanisms of action of the compounds. However, the implementation is fraught with clinical, logistical, regulatory, and ethical challenges, especially in diseases such as leukaemia and multiple myeloma. This chapter examines the treatment challenges in leukaemia and multiple myeloma, focusing on the role of drug repurposing in addressing therapeutic resistance and disease variability. It highlights the potential of personalized, tailored combination therapies, using repurposed drug components, to offer more effective, targeted, and cost-efficient treatment strategies, overcoming resistance and improving patient outcomes.

Linking Electronic Health Record Prescribing Data and Pharmacy Dispensing Records to Identify Patient-Level Factors Associated With Psychotropic Medication Receipt: Retrospective Study

Background

Pharmacoepidemiology studies using electronic health record (EHR) data typically rely on medication prescriptions to determine which patients have received a medication. However, such data do not affirmatively indicate whether these prescriptions have been filled. External dispensing databases can bridge this information gap; however, few established methods exist for linking EHR data and pharmacy dispensing records.

Objective

We described a process for linking EHR prescribing data with pharmacy dispensing records from Surescripts. As a use case, we considered the prescriptions and resulting fills for psychotropic medications among pediatric patients. We evaluated how dispensing information affects identifying patients receiving prescribed medications and assessing the association between filling prescriptions and subsequent health behaviors.

Methods

This retrospective study identified all new psychotropic prescriptions to patients younger than 18 years of age at Duke University Health System in 2021. We linked dispensing to prescribing data using proximate dates and matching codes between RxNorm concept unique identifiers and National Drug Codes. We described demographic, clinical, and service use characteristics to assess differences between patients who did versus did not fill prescriptions. We fit a least absolute shrinkage and selection operator (LASSO) regression model to evaluate the predictability of a fill. We then fit time-to-event models to assess the association between whether a patient filled a prescription and a future provider visit.

Results

We identified 1254 pediatric patients with a new psychotropic prescription. In total, 976 (77.8%) patients filled their prescriptions within 30 days of their prescribing encounters. Thus, we set 30 days as a cut point for defining a valid prescription fill. Patients who filled prescriptions differed from those who did not in several key factors. Those who did not fill had slightly higher BMIs, lived in more disadvantaged neighborhoods, were more likely to have public insurance or self-pay, and included a higher proportion of male patients. Patients with prior well-child visits or prescriptions from primary care providers were more likely to fill. Additionally, patients with anxiety diagnoses and those prescribed selective serotonin reuptake inhibitors were more likely to fill prescriptions. The LASSO model achieved an area under the receiver operator characteristic curve of 0.816. The time to the follow-up visit with the same provider was censored at 90 days after the initial encounter. Patients who filled prescriptions showed higher levels of follow-up visits. The marginal hazard ratio of a follow-up visit with the same provider was 1.673 (95% CI 1.463-1.913) for patients who filled their prescriptions. Using the LASSO model as a propensity-based weight, we calculated the weighted hazard ratio as 1.447 (95% CI 1.257-1.665).

Conclusions

Systematic differences existed between patients who did versus did not fill prescriptions. Incorporating external dispensing databases into EHR-based studies informs medication receipt and associated health outcomes.

Natural language processing of electronic medical records identifies cardioprotective agents for anthracycline induced cardiotoxicity

In this retrospective observational study, we aimed to investigate the potential of natural language processing (NLP) for drug repositioning by analyzing the preventive effects of cardioprotective drugs against anthracycline-induced cardiotoxicity (AIC) using electronic medical records. We evaluated the effects of angiotensin II receptor blockers/angiotensin-converting enzyme inhibitors (ARB/ACEIs), beta-blockers (BBs), statins, and calcium channel blockers (CCBs) on AIC using signals extracted from clinical texts via NLP. The study included 2935 patients prescribed anthracyclines at a single hospital, with concomitant prescriptions of ARB/ACEIs, BBs, statins, and CCBs. Upon propensity score matching, groups with and without these medications were compared, and expressions suggestive of cardiotoxicity, extracted via NLP, were considered as the outcome. The hazard ratios for ARB/ACEIs, BBs, statins, and CCBs were 0.58 [95% CI: 0.38-0.88], 0.71 [95% CI: 0.35-1.44], 0.60 [95% CI 0.38-0.95], and 0.63 [95% CI: 0.45-0.88], respectively. ARB/ACEIs, statins, and CCBs significantly suppressed AIC, whereas BBs did not demonstrate statistical significance, possibly due to limited statistical power. NLP-extracted signals from clinical texts reflected the known effects of these medications, demonstrating the feasibility of NLP-based drug repositioning. Further investigation is needed to determine if similar results can be replicated using electronic medical records from other institutions.

Using administrative healthcare data to evaluate drug repurposing opportunities for cancer: the possibility of using beta-blockers to treat breast cancer

Introduction: Cancer registries and hospital electronic medical records are commonly used to investigate drug repurposing candidates for cancer. However, administrative data are often more accessible than data from cancer registries and medical records. Therefore, we evaluated if administrative data could be used to evaluate drug repurposing for cancer by conducting an example study on the association between beta-blocker use and breast cancer mortality. Methods: A retrospective cohort study of women aged ≥50 years with incident breast cancer was conducted using a linked dataset with statewide hospital admission data and nationwide medication claims data. Women receiving beta blockers and first-line anti-hypertensives prior to and at diagnosis were compared. Breast cancer molecular subtypes and metastasis status were inferred by algorithms from commonly prescribed breast cancer antineoplastics and hospitalization diagnosis codes, respectively. Subdistribution hazard ratios (sHR) and corresponding 95% confidence intervals (CIs) for breast cancer mortality were estimated using Fine and Gray's competing risk models adjusted for age, Charlson comorbidity index, congestive heart failure, myocardial infraction, molecular subtype, presence of metastasis at diagnosis, and breast cancer surgery. Results: 2,758 women were hospitalized for incident breast cancer. 604 received beta-blockers and 1,387 received first-line antihypertensives. In total, 154 breast cancer deaths were identified over a median follow-up time of 2.7 years. We found no significant association between use of any beta-blocker and breast-cancer mortality (sHR 0.86, 95%CI 0.58-1.28), or when stratified by beta-blocker type (non-selective, sHR 0.42, 95%CI 0.14-1.25; selective, sHR 0.95, 95%CI 0.63-1.43). Results were not significant when stratified by molecular subtypes (e.g., triple negative breast cancer (TNBC), any beta blocker, sHR 0.16, 95%CI 0.02-1.51). Discussion: It is possible to use administrative data to explore drug repurposing opportunities. Although non-significant, an indication of an association was found for the TNBC subtype, which aligns with previous studies using registry data. Future studies with larger sample size, longer follow-up are required to confirm the association, and linkage to clinical data sources are required to validate our methodologies.

Graphical Learning and Causal Inference for Drug Repurposing

Gene expression profiles that connect drug perturbations, disease gene expression signatures, and clinical data are important for discovering potential drug repurposing indications. However, the current approach to gene expression reversal has several limitations. First, most methods focus on validating the reversal expression of individual genes. Second, there is a lack of causal approaches for identifying drug repurposing candidates. Third, few methods for passing and summarizing information on a graph have been used for drug repurposing analysis, with classical network propagation and gene set enrichment analysis being the most common. Fourth, there is a lack of graph-valued association analysis, with current approaches using real-valued association analysis one gene at a time to reverse abnormal gene expressions to normal gene expressions. To overcome these limitations, we propose a novel causal inference and graph neural network (GNN)-based framework for identifying drug repurposing candidates. We formulated a causal network as a continuous constrained optimization problem and developed a new algorithm for reconstructing large-scale causal networks of up to 1,000 nodes. We conducted large-scale simulations that demonstrated good false positive and false negative rates. To aggregate and summarize information on both nodes and structure from the spatial domain of the causal network, we used directed acyclic graph neural networks (DAGNN). We also developed a new method for graph regression in which both dependent and independent variables are graphs. We used graph regression to measure the degree to which drugs reverse altered gene expressions of disease to normal levels and to select potential drug repurposing candidates. To illustrate the application of our proposed methods for drug repurposing, we applied them to phase I and II L1000 connectivity map perturbational profiles from the Broad Institute LINCS, which consist of gene-expression profiles for thousands of perturbagens at a variety of time points, doses, and cell lines, as well as disease gene expression data under-expressed and over-expressed in response to SARS-CoV-2.

Drug repurposing using real-world data

The use of real-world data in drug repurposing has emerged due to well-established advantages of drug repurposing in supplementing de novo drug discovery and incentives in incorporating real-world evidence in regulatory approvals. We conducted a scoping review to characterize repurposing studies using real-world data and discuss their potential challenges and solutions. A total of 250 studies met the inclusion criteria, of which 36 were original studies on hypothesis generation, 101 on hypothesis validation, and seven on safety assessment. Key challenges that should be addressed for future progress in using real-world data for repurposing include isolated data sources with poor clinical granularity, false-positive signals from data mining, the sensitivity of hypothesis validation to bias and confounding, and the lack of clear regulatory guidance.

Computational drug repurposing based on electronic health records: a scoping review

Computational drug repurposing methods adapt Artificial intelligence (AI) algorithms for the discovery of new applications of approved or investigational drugs. Among the heterogeneous datasets, electronic health records (EHRs) datasets provide rich longitudinal and pathophysiological data that facilitate the generation and validation of drug repurposing. Here, we present an appraisal of recently published research on computational drug repurposing utilizing the EHR. Thirty-three research articles, retrieved from Embase, Medline, Scopus, and Web of Science between January 2000 and January 2022, were included in the final review. Four themes, (1) publication venue, (2) data types and sources, (3) method for data processing and prediction, and (4) targeted disease, validation, and released tools were presented. The review summarized the contribution of EHR used in drug repurposing as well as revealed that the utilization is hindered by the validation, accessibility, and understanding of EHRs. These findings can support researchers in the utilization of medical data resources and the development of computational methods for drug repurposing.

A Study on Methodologies of Drug Repositioning Using Biomedical Big Data: A Focus on Diabetes Mellitus

Drug repositioning is a strategy for identifying new applications of an existing drug that has been previously proven to be safe. Based on several examples of drug repositioning, we aimed to determine the methodologies and relevant steps associated with drug repositioning that should be pursued in the future. Reports on drug repositioning, retrieved from PubMed from January 2011 to December 2020, were classified based on an analysis of the methodology and reviewed by experts. Among various drug repositioning methods, the network-based approach was the most common (38.0%, 186/490 cases), followed by machine learning/deep learningbased (34.3%, 168/490 cases), text mining-based (7.1%, 35/490 cases), semantic-based (5.3%, 26/490 cases), and others (15.3%, 75/490 cases). Although drug repositioning offers several advantages, its implementation is curtailed by the need for prior, conclusive clinical proof. This approach requires the construction of various databases, and a deep understanding of the process underlying repositioning is quintessential. An in-depth understanding of drug repositioning could reduce the time, cost, and risks inherent to early drug development, providing reliable scientific evidence. Furthermore, regarding patient safety, drug repurposing might allow the discovery of new relationships between drugs and diseases.

Applications of artificial intelligence in drug development using real-world data

The US Food and Drug Administration (FDA) has been actively promoting the use of real-world data (RWD) in drug development. RWD can generate important real-world evidence reflecting the real-world clinical environment where the treatments are used. Meanwhile, artificial intelligence (AI), especially machine- and deep-learning (ML/DL) methods, have been increasingly used across many stages of the drug development process. Advancements in AI have also provided new strategies to analyze large, multidimensional RWD. Thus, we conducted a rapid review of articles from the past 20 years, to provide an overview of the drug development studies that use both AI and RWD. We found that the most popular applications were adverse event detection, trial recruitment, and drug repurposing. Here, we also discuss current research gaps and future opportunities.

From Real-World Patient Data to Individualized Treatment Effects Using Machine Learning: Current and Future Methods to Address Underlying Challenges

Clinical decision making needs to be supported by evidence that treatments are beneficial to individual patients. Although randomized control trials (RCTs) are the gold standard for testing and introducing new drugs, due to the focus on specific questions with respect to establishing efficacy and safety vs. standard treatment, they do not provide a full characterization of the heterogeneity in the final intended treatment population. Conversely, real-world observational data, such as electronic health records (EHRs), contain large amounts of clinical information about heterogeneous patients and their response to treatments. In this paper, we introduce the main opportunities and challenges in using observational data for training machine learning methods to estimate individualized treatment effects and make treatment recommendations. We describe the modeling choices of the state-of-the-art machine learning methods for causal inference, developed for estimating treatment effects both in the cross-section and longitudinal settings. Additionally, we highlight future research directions that could lead to achieving the full potential of leveraging EHRs and machine learning for making individualized treatment recommendations. We also discuss how experimental data from RCTs and Pharmacometric and Quantitative Systems Pharmacology approaches can be used to not only improve machine learning methods, but also provide ways for validating them. These future research directions will require us to collaborate across the scientific disciplines to incorporate models based on RCTs and known disease processes, physiology, and pharmacology into these machine learning models based on EHRs to fully optimize the opportunity these data present.

Repurposing current therapeutics for treating COVID-19: A vital role of prescription records data mining

Since its outbreak in late 2019, the SARS‐Cov‐2 pandemic already infected over 3.7 million people and claimed more than 250,000 lives globally. At least 1 year may take for an approved vaccine to be in place, and meanwhile millions more could be infected, some with fatal outcome. Over thousand clinical trials with COVID‐19 patients are already listed in ClinicalTrials.com, some of them for assessing the utility of therapeutics approved for other conditions. However, clinical trials take many months, and are typically done with small cohorts. A much faster and by far more efficient method for rapidly identifying approved therapeutics that can be repurposed for treating COVID‐19 patients is data mining their past and current electronic health and prescription records for identifying drugs that may protect infected individuals from severe COVID‐19 symptoms. Examples are discussed for applying health and prescription records for assessing the potential repurposing (repositioning) of angiotensin receptor blockers, estradiol, or antiandrogens for reducing COVID‐19 morbidity and fatalities. Data mining of prescription records of COVID‐19 patients will not cancel the need for conducting controlled clinical trials, but could substantially assist in trial design, drug choice, inclusion and exclusion criteria, and prioritization. This approach requires a strong commitment of health provides for open collaboration with the biomedical research community, as health provides are typically the sole owners of retrospective drug prescription records.