A framework enabling LLMs into regulatory environment for transparency and trustworthiness and its application to drug labeling document.

Regulatory agencies consistently deal with extensive document reviews, ranging from product submissions to both internal and external communications. Large Language Models (LLMs) like ChatGPT can be invaluable tools for these tasks, however present several challenges, particularly the proprietary information, combining customized function with specific review needs, and transparency and explainability of the model's output. Hence, a localized and customized solution is imperative. To tackle these challenges, we formulated a framework named askFDALabel on FDA drug labeling documents that is a crucial resource in the FDA drug review process. AskFDALabel operates within a secure IT environment and comprises two key modules: a semantic search and a Q&A/text-generation module. The Module S built on word embeddings to enable comprehensive semantic queries within labeling documents. The Module T utilizes a tuned LLM to generate responses based on references from Module S. As the result, our framework enabled small LLMs to perform comparably to ChatGPT with as a computationally inexpensive solution for regulatory application. To conclude, through AskFDALabel, we have showcased a pathway that harnesses LLMs to support agency operations within a secure environment, offering tailored functions for the needs of regulatory research.

Healthcare providers' use of a concise summary to prescribe for lactating patients.

BACKGROUND: Most breastfeeding individuals take at least one prescription drug, yet limited data from lactation studies are available to inform the safety of these drugs during breastfeeding. As a result, healthcare providers (HCPs) rely on available information about safety of drugs used during pregnancy or on personal experiences to inform prescribing/counseling decisions for breastfeeding individuals. To improve risk communication regarding drugs used during lactation, the U.S. Food and Drug Administration published the Pregnancy and Lactation Labeling Rule (PLLR) in 2015, which added a narrative summary of available risk information to the lactation section of Prescribing Information (PI). Prior studies on labeling in PLLR format revealed that although HCPs found these details valuable, they regarded the narrative as too long to support decision-making during patient encounters. OBJECTIVE: This qualitative study's objective was to assess the utility of adding a concise summary to the Lactation subsection of PI to complement the narrative and succinctly communicate to busy HCPs a drug's risks when used during lactation. The concise summary consisted of a bolded headline, bulleted descriptions of available study findings and potential adverse reactions, and recommendations for risk mitigation. METHODS: Twenty-five online focus groups were conducted with five segments of HCPs to obtain their feedback on the concise summary and discuss their prescribing/counseling decisions for four fictitious prescription drugs including one vaccine. RESULTS: HCPs utilized the concise summary to make initial prescribing/counseling decisions. Many also used the labeling narrative for a comprehensive benefit-risk assessment. CONCLUSION: The findings indicate a need to continue to improve communication about safety of drugs used during lactation, and that the concise summary may help facilitate this communication. The study also highlights the need to educate HCPs about PI limitations when clinical data are lacking and the need to encourage clinical studies to be conducted to support actionable recommendations about use of prescription drugs during lactation.

Revealing the reporting disparity: VigiBase highlights underreporting of clozapine in other Western European countries compared to the UK.

BACKGROUND: Pharmacovigilance studies indicate clozapine history is marked by adverse drug reactions (ADRs). OBJECTIVE: In a 2021 article, the United Kingdom (UK) had >90 % of European clozapine-related fatal outcomes in VigiBase, the World Health Organization's pharmacovigilance database. Two possibly opposing hypotheses could explain this disparity: 1) fewer reported fatal outcomes in other Western European countries mainly reflect underreporting to VigiBase, and 2) the higher number of UK reports reflects higher real relative mortality. METHODS: VigiBase reports from clozapine's introduction to December 31, 2022, were studied for ADRs and the top 10 causes of fatal outcomes. The UK was compared with 11 other top reporting Western countries (Germany, Denmark, France, Finland, Ireland, Italy, Netherlands, Norway, Spain, Sweden and Switzerland). Nine countries (except Ireland and Switzerland) were compared after controlling for population and clozapine prescriptions. RESULTS: The UK accounted for 29 % of worldwide clozapine-related fatal outcomes, Germany 2 % and <1 % in each of the other countries. The nonspecific label "death" was the top cause in the world (46 %) and in the UK (33 %). "Pneumonia" was second in the world (8 %), the UK (12 %), Ireland (8 %) and Finland (14 %). Assuming that our corrections for population and clozapine use are correct, other countries underreported only 1-10 % of the UK clozapine fatal outcome number. CONCLUSIONS: Different Western European countries consistently underreport to VigiBase compared to the UK, but have different reporting/publishing styles for clozapine-related ADRs/fatal outcomes. Three Scandinavian registries suggest lives are saved as clozapine use increases, but this cannot be studied in pharmacovigilance databases.

Do pediatric oral suspension acetaminophen and ibuprofen product labeling and online resources facilitate intended use?

Background: Caregivers often have difficulty administering pediatric medications which frequently results in increased dosing error risk. Objective: We examined health literacy characteristics of pediatric over-the-counter (OTC) oral suspension acetaminophen and ibuprofen instructional materials and dosing instruments. Methods: We conducted a descriptive analysis of dosing instructions, measuring syringe characteristics, and internet-based resources among a sample of OTC pediatric oral suspension acetaminophen and ibuprofen products (n = 14). Results: All products included Drug Facts Panels, employed consistent abbreviation use, and stated measuring dosage with syringe provided. However, oral syringe dosing increment markings did not match box or bottle dosing charts. Most products had supplemental English-language internet-based content resources available. Conclusions: While OTC pediatric oral suspension acetaminophen and ibuprofen products labeling included key drug fact elements, there were inconsistencies between medication dosing chart labeling guidelines and oral syringe dosing increments/markings. It is vital that oral dosing syringes are clearly marked to match product dosing chart labeling s as a means of potentially reducing caregiver dosing errors.

Real-World Amount of Clotting Factor Concentrates Dispensed and Annual Medical Expenditures for Japanese Patients with Hemophilia B.

Introduction: Until extended half-life (EHL) factor IX (FIX) concentrates became available in Japan in 2010, patients with hemophilia B received intravenous FIX replacement therapy with standard half-life (SHL) FIX concentrates. Purpose: To investigate the amount of factor dispensed and the associated medical expenditures for the treatment of hemophilia B in the real-world clinical setting in Japan. Methods: This retrospective study comprised patients with hemophilia B (N=197) who had filled prescriptions for FIX concentrates reported in Japan's Medical Data Vision database from 2015 to 2019. Patients were included if they had 2 or more prescriptions for the same FIX concentrates within the first 6 months of the study period and the interval between prescriptions was at least 2 weeks. Results: Since 2015, there was a decrease in the proportion of patients using SHL FIX concentrates and a corresponding increase in international units of dispensed EHL FIX concentrates. Median annualized dispensed dosages (IU/kg body weight) of EHL FIX concentrates were lower than for SHL concentrates for outpatient use only. Annual total health care expenditures per patient and annual expenditures for prescribed FIX concentrates increased each year during the study period. Following a switch from an SHL to an EHL concentrate, the median amount of prescribed FIX concentrate decreased slightly, although median total health care expenditures and FIX concentrate expenditures increased. Conclusion: In the real-world setting in Japan, medical expenditures and the proportion of patients prescribed EHL FIX concentrates for the treatment of hemophilia B have increased.

Guía internacional para una dosificación más segura de la clozapina en adultos mediante el uso de 6 titulaciones personalizadas de dosis basados en la etnicidad, la proteína C reactiva y los niveles de clozapina / An international adult guideline for making clozapine titration safer by using six ancestry-based personalized dosing titrations, CRP, and clozapine levels

Esta guía internacional propone mejorar los prospectos de la clozapina en todo el mundo mediante la inclusion de información sobre la titulación del fármaco en función de la ascendencia del paciente. Las bases de datos de reacciones adversas a medicamentos (RAM) sugieren que la clozapina es el tercer fármaco más tóxico en los Estados Unidos de América (EE. UU.) y que produce una mortalidad por neumonía en todo el mundo 4 veces mayor que la correspondiente a la agranulocitosis o la miocarditis. El rango terapéutico de referencia para las concentraciones séricas estables de clozapina es estrecho, de 350 a 600 ng/ml, con potencial de toxicidad y reacciones adversas más fecuentes a medida que aumentan las concentraciones. La clozapina se metaboliza principalmente por CYP1A2 (las mujeres no fumadoras requieren la dosis más baja y los hombres fumadores la dosis más alta). A través de la conversión fenotípica, la prescripción conjunta de inhibidores del metabolismo de la clozapina (incluidos los anticonceptivos orales y el valproato), la obesidad o la inflamación con elevaciones de la proteína C reactiva (PCR), pueden convertir al paciente en un metabolizador lento/pobre (MP). Las personas de ascendencia asiática (de Pakistán a Japón) o los habitantes originarios de las Américas tienen menor actividad de CYP1A2 y requieren dosis más bajas de clozapina para alcanzar concentraciones de 350 ng/ml. En los EE. UU. se recomiendan dosis diarias de 300-600 mg/día. La dosificación personalizada lenta puede prevenir RAM tempranas (incluidos el síncope, la miocarditis y la neumonía). La esencia de esta guía se fundamenta en 6 esquemas de titulaciones personalizadas para pacientes hospitalizados...(AU)

This is the Spanish translation of an international guideline which proposes improving clozapine package inserts worldwide by using ancestry-based: 1) dosing and 2) titration. Adverse drug reaction (ADR) databases suggest clozapine: 1) is the third most toxic drug in the United States (US), and 2) produces worldwide pneumonia mortality four times greater than that of agranulocytosis or myocarditis. For trough steady-state clozapine serum concentrations, the therapeutic reference range is narrow, from 350 to 600 ng/mL with the potential for toxicity and ADRs as concentrations increase. Clozapine is mainly metabolized by CYP1A2 (female non-smokers require the lowest dose and male smokers the highest dose). Poor metabolizer (PM) status through phenotypic conversion is associated with co-prescription of inhibitors (including oral contraceptives and valproate), obesity or inflammation with C-reactive protein (CRP) elevations. People with ancestry from Asia (Pakistan to Japan) or the Americas’ original inhabitants have lower CYP1A2 activity and require lower clozapine doses to reach concentrations of 350 ng/ml. Daily doses of 300-600 mg/day are recommended in the US. Slow personalized titration may prevent early ADRs (including syncope, myocarditis and pneumonia). The core of this guideline consists of six personalized titration schedules for inpatients...(AU)

Leveraging GPT-4 for food effect summarization to enhance product-specific guidance development via iterative prompting.

Food effect summarization from New Drug Application (NDA) is an essential component of product-specific guidance (PSG) development and assessment, which provides the basis of recommendations for fasting and fed bioequivalence studies to guide the pharmaceutical industry for developing generic drug products. However, manual summarization of food effect from extensive drug application review documents is time-consuming. Therefore, there is a need to develop automated methods to generate food effect summary. Recent advances in natural language processing (NLP), particularly large language models (LLMs) such as ChatGPT and GPT-4, have demonstrated great potential in improving the effectiveness of automated text summarization, but its ability with regard to the accuracy in summarizing food effect for PSG assessment remains unclear. In this study, we introduce a simple yet effective approach,iterative prompting, which allows one to interact with ChatGPT or GPT-4 more effectively and efficiently through multi-turn interaction. Specifically, we propose a three-turn iterative prompting approach to food effect summarization in which the keyword-focused and length-controlled prompts are respectively provided in consecutive turns to refine the quality of the generated summary. We conduct a series of extensive evaluations, ranging from automated metrics to FDA professionals and even evaluation by GPT-4, on 100 NDA review documents selected over the past five years. We observe that the summary quality is progressively improved throughout the iterative prompting process. Moreover, we find that GPT-4 performs better than ChatGPT, as evaluated by FDA professionals (43% vs. 12%) and GPT-4 (64% vs. 35%). Importantly, all the FDA professionals unanimously rated that 85% of the summaries generated by GPT-4 are factually consistent with the golden reference summary, a finding further supported by GPT-4 rating of 72% consistency. Taken together, these results strongly suggest a great potential for GPT-4 to draft food effect summaries that could be reviewed by FDA professionals, thereby improving the efficiency of the PSG assessment cycle and promoting generic drug product development.

An expert review of clozapine in Eastern European countries: Use, regulations and pharmacovigilance.

OBJECTIVES: To compare the prevalence, regulations, and pharmacovigilance practices of clozapine use in Eastern European countries (except Russia). METHODS: Questionnaires and data from administrative databases (2016 and 2021), package inserts and national guidelines were collected from 21 co-authors from 21 countries. Reports of clozapine adverse drug reactions (ADRs) sent to the global pharmacovigilance database (VigiBase™) were analyzed from introduction to December 31, 2022. RESULTS: Clozapine prescription among antipsychotics in 2021 varied six-fold across countries, from 2.8 % in the Czech Republic to 15.8 % in Montenegro. The utilization of antipsychotics in both 2016 and 2021 was highest in Croatia, and lowest in Serbia in 2016, and Montenegro in 2021, which had half the defined daily dose (DDD)/1000/day compared to the Croatian data. From 2016 to 2021, the prevalence of antipsychotic use increased in almost all countries; the proportion of clozapine use mainly remained unchanged. Differences were detected in hematological monitoring requirements and clozapine approved indications. Only a few national schizophrenia guidelines mention clozapine-induced myocarditis or individual titration schemes. The VigiBase search indicated major underreporting regarding clozapine and its fatal outcomes. By comparison, the United Kingdom had less than half the population of these Eastern European countries but reported to VigiBase more clozapine ADRs by 89-fold and clozapine fatal outcomes by almost 300-fold. CONCLUSION: Clozapine is under-utilized in Eastern European countries. Introducing individualized clozapine treatment schedules may help to maximize clozapine benefits and safety. Major improvement is needed in reporting clozapine ADRs and fatal outcomes in Eastern European countries.

Comprehensibility of South African pictograms by sub-Saharan African prospective-users.

OBJECTIVES: This work aimed to explore the validity of South African pictograms to be comprehended by sub-Saharan migrant populations in Spain. METHODS: A quantitative cross-sectional study was carried out. A voluntary survey was conducted among a population ≥ 18 years of age in a Temporary Stay Centre for Immigrants located in Spain, in June 2022. The form included a culturally and linguistically validated version of the Health Literacy questionnaire (HLS-EU-Q16), a brief questionnaire of sociodemographic variables and 10 South African pictograms selected to assess the comprehension. According to the International Organization for Standardization, a pictogram must receive 66.70% correct answers to be acceptable. Mann-Whitney U or Kruskal-Wallis tests were used to assess differences between group variables. Statistical significance was set at P < 0.05. RESULTS: 55 sub-Saharan migrants who newly arrived in Europe voluntarily agreed to participate. None of the pictograms evaluated met the validity criterion. For the selected pictograms, the range of correct answers was between 52.50% and 20.00%. Neither sociodemographic variables nor Health Literacy level were significantly associated with pictogram comprehensibility. CONCLUSION: New culturally sensitive pictograms that are easily comprehended by sub-Saharan African population are needed. PRACTICE IMPLICATIONS: These results may serve as a starting point for future studies that aim to create and evaluate new pictograms or create new culturally adapted versions of existing pictograms involving the target population.

Evaluation of off-label medication use and drug safety in a pediatric intensive care unit.

Safety and efficacy are essential in the process of disease treatment. However, off-label medication use is inevitable because various medications do not contain regulatory labels for pediatric use. We aimed to examine off-label medication use and analyze the risk factors correlated with adverse drug reactions (ADRs). This study was performed retrospectively using electronic medical data from a pediatric intensive care unit (PICU) of a tertiary hospital in Korea from July 2019 to June 2020. A total 6,183 prescribed medications from 502 PICU patients were examined in the present study. A total of 80% were infants or children, and 96.0% of them were treated with off-label medications. It was discovered that 4,778 off-label cases (77.2%) of the top 100 drugs had prescriptions with dosage (67.8%). Drugs prescribed to patients admitted to the cardiothoracic department (odds ratio [OR], 3.248; p = 0.019), total number of medications (OR, 1.116; p = 0.001), and length of PICU stay of ≥ 7 days (OR, 4.981; p = 0.008) were significantly associated with ADRs. ADRs were noted to be more severe in off-label use (p = 0.0426). For appropriate medication use, evidence regarding the safety of off-label medications is required and ultimately reflected in the official regulation.

Comparison of pharmacogenomic information for drug approvals provided by the national regulatory agencies in Korea, Europe, Japan, and the United States.

Pharmacogenomics, which is defined as the study of changes in the properties of DNA and RNA associated with drug response, enables the prediction of the efficacy and adverse effects of drugs based on patients' specific genetic mutations. For the safe and effective use of drugs, it is important that pharmacogenomic information is easily accessible to clinical experts and patients. Therefore, we examined the pharmacogenomic information provided on drug labels in Korea, Europe, Japan, and the United States (US). The selection of drugs that include pharmacogenomic information was based on the drug list that includes genetic information from the Korea Ministry of Food and Drug Safety (MFDS) and US Food and Drug Administration (FDA) websites. Drug labels were retrieved from the sites of MFDS, FDA, European Medicines Agency, and Japanese Pharmaceuticals and Medical Devices Agency. Drugs were classified as per the Anatomical Therapeutic Chemical code, and the biomarkers, labeling sections, and necessity of genetic tests were determined. In total, 348 drugs were selected from 380 drugs with available pharmacogenomic information in Korea and the US after applying the inclusion and exclusion criteria. Of these drugs, 137, 324, 169, and 126 were with pharmacogenomics information in Korea, the US, Europe, and Japan, respectively. The most commonly represented drug class was antineoplastic and immunomodulating agents. Regarding the classification as per the mentioned biomarkers, the cytochrome P450 enzyme was the most frequently mentioned information, and the targeted anticancer drugs most commonly required genetic biomarker testing. The reasons for differences in drug labeling information based on country include differences in mutant alleles according to ethnicity, frequencies at which drug lists are updated, and pharmacogenomics-related guidelines. Clinical experts must continuously strive to identify and report mutations that can explain drug efficacy or side effects for safe drug use.

Utility of patient information leaflet and perceived impact of its use on medication adherence.

BACKGROUND: Although patients frequently use patient information leaflets (PILs) to obtain information about medicine, their confidence in using it may be diminished after reading it. This study aimed to assess the public perception of PIL's quality and the perceived impact of its use on medication adherence. METHODS: A community-based cross-sectional study of 1,138 adult individuals in Saudi Arabia, April-May 2020, was conducted via Survey Monkey using an anonymous validated e-questionnaire. Data were collected on personal characteristics, PIL readership and preferences, perception towards PIL quality and impact of its use on taking medication, and reasons for not reading PIL. In addition, logistic regression analysis was performed to identify the significant predictors of reading PIL. Significance was considered at p < 0.05. RESULTS: Nearly all participants (91.1%) reported reading PIL. The more read PIL's sections were directions of use (52.7%) and side effects (30.3%). Female gender (OR = 5.64, 95%CI: 3.53,9.02), age over 40 years (OR = 2.80, 95%CI: 1.69,4.64), and secondary education or more (OR = 1.74, 95%CI: 1.06,2.85) were the significant predictors of reading PIL. The majority of PIL readers reported their preference for verbal information (65.8%), hard copy presentation (77%), adding graphics (71.1%), and concise content of PIL (68.8%). In addition, most participants reported PIL always/usually adds to their knowledge of medicines (70.6%) and said that PIL reading positively impacted their medication adherence (64.9%). For only 8.8%, PIL reading negatively impacted their adherence, primarily because of reading information on medicine's side effects and complications (74.4%). More than one-half of participants perceived the PIL quality as good/excellent in terms of; font size (51.3%), language comprehensiveness (64.9%), paper quality (68.0%), and general appearance (64.9%). Getting sufficient information from doctors and pharmacists was the main reason for not reading the PIL (59.2%). Most participants (92.5%) agreed on standardizing how information is displayed in the PIL among all PILs of all companies. CONCLUSION: PIL is read by nearly all the study sample, especially females, older, and educated subjects. It was perceived as beneficial in upgrading medication adherence. Effective designing of PILs should focus on patients' literacy level and age. Standardization of the PIL structure in all pharmaceutical companies is recommended.

Neural text generation in regulatory medical writing.

Background: A steep increase in new drug applications has increased the overhead of writing technical documents such as medication guides. Natural language processing can contribute to reducing this burden. Objective: To generate medication guides from texts that relate to prescription drug labeling information. Materials and Methods: We collected official drug label information from the DailyMed website. We focused on drug labels containing medication guide sections to train and test our model. To construct our training dataset, we aligned "source" text from the document with similar "target" text from the medication guide using three families of alignment techniques: global, manual, and heuristic alignment. The resulting source-target pairs were provided as input to a Pointer Generator Network, an abstractive text summarization model. Results: Global alignment produced the lowest ROUGE scores and relatively poor qualitative results, as running the model frequently resulted in mode collapse. Manual alignment also resulted in mode collapse, albeit higher ROUGE scores than global alignment. Within the family of heuristic alignment approaches, we compared different methods and found BM25-based alignments to produce significantly better summaries (at least 6.8 ROUGE points above the other techniques). This alignment surpassed both the global and manual alignments in terms of ROUGE and qualitative scoring. Conclusion: The results of this study indicate that a heuristic approach to generating inputs for an abstractive summarization model increased ROUGE scores, compared to a global or manual approach when automatically generating biomedical text. Such methods hold the potential to significantly reduce the manual labor burden in medical writing and related disciplines.

Evaluating the Impact of Black Box Warning Updates on the Reporting of Drug-Related Adverse Events: a Cross Sectional Study of the FAERS Database.

BACKGROUND: The Food and Drug Administration (FDA)'s Adverse Event Reporting System (FAERS) is a post-marketing surveillance system which relies on spontaneous reports of adverse drug reactions (ADRs). Our objective was to evaluate how black box warning (BBW) updates impact ADR reporting rates. RESEARCH DESIGN AND METHODS: We searched MEDWATCH for all BBW updates issued between January 2014 and December 2016 and categorized them as new, major, and minor. Rates of relevant ADR reports from the FAERS database in the 4 years preceding and following a BBW update were assessed among the different BBW categories. RESULTS: Forty BBW updates were included (16 major, 3 new, and 21 minor). A meaningful increase in the proportion of relevant ADRs of all ADRs reported following BBW updates was documented for 53% of new or major updates and 24% of minor updates (p = 0.06). The median percentage of reported relevant ADRs increased by 5% following new and major BBW updates and decreased by 30% following minor BBW updates (p = 0.3). CONCLUSIONS: Reporting of adverse events to the FAERS database is affected by the severity and timing of related BBW updates, highlighting the drawbacks of spontaneous reporting systems. Regulators should promote proactive pharmacovigilance strategies to cope with these limitations.

Fine-tuning BERT for automatic ADME semantic labeling in FDA drug labeling to enhance product-specific guidance assessment.

Product-specific guidances (PSGs) recommended by the United States Food and Drug Administration (FDA) are instrumental to promote and guide generic drug product development. To assess a PSG, the FDA assessor needs to take extensive time and effort to manually retrieve supportive drug information of absorption, distribution, metabolism, and excretion (ADME) from the reference listed drug labeling. In this work, we leveraged the state-of-the-art pre-trained language models to automatically label the ADME paragraphs in the pharmacokinetics section from the FDA-approved drug labeling to facilitate PSG assessment. We applied a transfer learning approach by fine-tuning the pre-trained Bidirectional Encoder Representations from Transformers (BERT) model to develop a novel application of ADME semantic labeling, which can automatically retrieve ADME paragraphs from drug labeling instead of manual work. We demonstrate that fine-tuning the pre-trained BERT model can outperform conventional machine learning techniques, achieving up to 12.5% absolute F1 improvement. To our knowledge, we were the first to successfully apply BERT to solve the ADME semantic labeling task. We further assessed the relative contribution of pre-training and fine-tuning to the overall performance of the BERT model in the ADME semantic labeling task using a series of analysis methods, such as attention similarity and layer-based ablations. Our analysis revealed that the information learned via fine-tuning is focused on task-specific knowledge in the top layers of the BERT, whereas the benefit from the pre-trained BERT model is from the bottom layers.

RxBERT: Enhancing drug labeling text mining and analysis with AI language modeling.

The US drug labeling document contains essential information on drug efficacy and safety, making it a crucial regulatory resource for Food and Drug Administration (FDA) drug reviewers. Due to its extensive volume and the presence of free-text, conventional text mining analysis have encountered challenges in processing these data. Recent advances in artificial intelligence (AI) for natural language processing (NLP) have provided an unprecedented opportunity to identify key information from drug labeling, thereby enhancing safety reviews and support for regulatory decisions. We developed RxBERT, a Bidirectional Encoder Representations from Transformers (BERT) model pretrained on FDA human prescription drug labeling documents for an enhanced application of drug labeling documents in both research and drug review. RxBERT was derived from BioBERT with further training on human prescription drug labeling documents. RxBERT was demonstrated in several tasks using regulatory datasets, including those involved in the National Institutes of Technology Text Analysis Challenge Dataset (NIST TAC dataset), the FDA Adverse Drug Event Evaluation Dataset (ADE Eval dataset), and the classification of texts from submission packages into labeling sections (US Drug Labeling dataset). For all these tasks, RxBERT reached 86.5 F1-scores in both TAC and ADE Eval classification, respectively, and prediction accuracy of 87% for the US Drug Labeling dataset. Overall, RxBERT was shown to be as competitive or have better performance compared to other NLP approaches such as BERT, BioBERT, etc. In summary, we developed RxBERT, a transformer-based model specific for drug labeling that outperformed the original BERT model. RxBERT has the potential to be used to assist research scientists and FDA reviewers to better process and utilize drug labeling information toward the advancement of drug effectiveness and safety for public health. This proof-of-concept study also demonstrated a potential pathway to customized large language models (LLMs) tailored to the sensitive regulatory documents for internal application.

Pharmacists' Knowledge and Perceptions of FDA Approval Standards and the Breakthrough Therapy Designation.

The "breakthrough therapy" designation (BTD) is a recent mechanism implemented by the United States Food and Drug Administration (FDA) to expedite access to drugs that address unmet needs. The purpose of this study is to describe pharmacists' knowledge of FDA drug-approval standards and knowledge and perceptions of the BTD. Pharmacists engaged in advanced clinical practice were identified through membership profiles of a professional pharmacy organization. Eligible participants were then sent a questionnaire to assess knowledge of FDA approval standards and the BTD. A total of 226 pharmacists responded. The majority of respondents were women (70.2%) and had completed post-graduate training (85.8%). Over half correctly answered at least two of three questions on FDA approval standards (58.1%) and the BTD (78.1%). Only 24.1% of respondents identified as being familiar with the BTD. The majority of pharmacists (62.8%) were certain that FDA-approved "breakthrough" drugs represented a major advance over currently approved therapies and most (88.5%) preferred the drug designated as "breakthrough" in a hypothetical scenario. In conclusion, pharmacists were able to correctly answer questions about FDA approval standards and the BTD. However, they were unfamiliar with the implications of a BTD and may overestimate the benefit demonstrated by these drugs. Future research should identify knowledge gaps in pharmacist understanding of regulatory mechanisms designed to expedite drug approval.

A Review of Patient-Reported Outcome Labeling of FDA-Approved New Drugs (2016-2020): Counts, Categories, and Comprehensibility.

OBJECTIVES: A review of new drug approvals (NDAs) by the US Food and Drug Administration (FDA) for 2006 to 2015 showed that approximately 20% of new drugs had labeling based on patient-reported outcomes (PROs). The purpose of this study was to review labeling text based on PRO endpoints for NDAs from 2016 to 2020, with a special focus on the comprehensibility of such statements when included. METHODS: We reviewed drug approval reports on the Drugs@FDA web page of the FDA website to determine the number of NDAs from 2016 to 2020. For all identified NDAs, drug approval package and product labels were reviewed. NDAs from 2016 to 2020 were grouped by disease category as per International Classification of Diseases 10th Revision. Data were summarized for diseases that traditionally rely on PROs for evaluating treatment benefit (PRO dependent) and for diseases that traditionally do not rely on PROs (non-PRO dependent). Results were compared with NDAs from 2006 to 2010. RESULTS: NDAs amounting to 228 were identified from 2016 to 2020, 26.3% of which had labeling statements based on PRO endpoints. From 2006 to 2015 and from 2016 to 2020, PRO labeling statements were included in 46.5% (46 of 99) and 50.0% (47 of 94), respectively, of NDAs for PRO-dependent new molecular entities and in 6.0% (12 of 199) and 9.7% (13 of 199), respectively, of NDAs for non-PRO-dependent new molecular entities. Comprehensibility of labeling statements based on PRO endpoints was judged to be complex in 56.7% of product labels. CONCLUSIONS: The increase in labeling text based on PRO endpoints in product labels is encouraging. However, there is room for improvement on the comprehensibility of labeling statements based on PRO endpoints.

Regulation of Drug Prescribing Information in Latin America and the Caribbean.

OBJECTIVE: To describe the status of drug regulations in Latin America and the Caribbean, in force as of May 2021, and assess through a comparative exercise the differences between the countries under scope on prescribing information of drugs for human use. MATERIALS AND METHODS: A narrative review allowed the identification of the regulations concerning the prescribing information of drugs in 25 countries in Latin America and the Caribbean for the registration of prescription medications. On this basis, terms and concepts regarding this topic, the general provisions by the regulatory authorities for these products, applications for health registration and further amendments were identified for each country. RESULTS: The Latin American and the Caribbean countries included, manage and regulate drug prescribing information differently in terms of concepts, information publishing, structure for product information, among other criteria. Few health authorities publish product information on their website. Additionally, the patient information leaflet is not requested for prescription drugs in most of the studied countries. There is no standardized structure for drug product information within the region. CONCLUSIONS: A poor level of harmonization among the regulations from these countries regarding the content and management (e.g. if physical package insert is required or not, if it is subject to notification or approval) of the prescribing information of human use drugs became evident. Also, there is a visible lack of standardization of concepts for referring to a specific document (e.g., package insert for healthcare professionals, patient information leaflet and technical information for the drug product) and in the content itself.

Prenatal exposure to teratogenic medications in the era of Risk Evaluation and Mitigation Strategies.

BACKGROUND: Prevention of prenatal exposures to teratogenic drugs is a significant clinical and public health concern. With the enactment of the US Food and Drug Administration Amendments Act in 2007, the US Food and Drug Administration has begun to require manufacturers to implement Risk Evaluation and Mitigation Strategies to prevent prenatal exposures. Among 12 risk evaluation and mitigation strategy drugs, several had predecessor risk mitigation plans (eg, isotretinoin) and some were newly required (eg, mycophenolate). Only a small proportion of teratogenic drugs are currently subject to Risk Evaluation and Mitigation Strategies, and the extent of prenatal exposure to the universe of teratogenic drugs compared with drugs subject to Risk Evaluation and Mitigation Strategies is unknown. Moreover, the effectiveness of such advanced risk mitigation programs in preventing prenatal exposure is not clear. OBJECTIVE: This study aimed to characterize the epidemiology of prenatal exposures to definite and potential teratogens during the risk evaluation and mitigation strategy era. STUDY DESIGN: We constructed a time-series of pregnancies identified from a national private insurance claims database (IBM MarketScan) to estimate prenatal exposures to teratogenic drugs (2006-2017). Pregnancy outcomes, gestational age, and the onset of pregnancy were determined with previously validated algorithms. The Teratology Information Service and Clinical Pharmacology databases were used to identify drugs with definite (n=141) or potential (n=65) teratogenic effects, and drugs with debatable risks such as benzodiazepines, statins, tetracyclines, sex hormones, infertility treatments, and gonadotropin-releasing hormone analogs were excluded. We defined prenatal exposure as ≥1 prescription fill or medical encounter involving administration of drugs with a definite teratogenic risk (including 12 for which there is a "current or discontinued" risk evaluation and mitigation strategy) or a potential teratogenic risk. We evaluated secular trends and modeled the effects of age, preconception exposure, and state healthcare quality rankings on prenatal exposure, adjusting for demographic factors and clinical conditions. RESULTS: The cohort included 3,445,612 pregnancies (2,532,444 live deliveries). Prenatal exposures to definite teratogens decreased slightly during the study years from 1.86 to 1.24 per 100 pregnancies between 2006 and 2017, whereas exposure increased for potential teratogens from 3.40% to 5.33%. Prenatal exposure prevalences were higher during the first trimester and for pregnancies that ended in nonlive outcomes. Drugs subject to Risk Evaluation and Mitigation Strategies had low background utilization and contributed to a small proportion of prenatal exposures (15.1 per 100,000 pregnancies). We also observed fewer prenatal exposures to risk evaluation and mitigation strategy drugs among women of childbearing age who used these treatments (0.14% vs 0.36% for any definite teratogen). Age extremes and low state-level healthcare quality rankings were independent predictors of prenatal exposure. CONCLUSION: Fetuses in more than 1 in 16 pregnancies continued to be exposed to teratogenic drugs during the past decade. Drugs with Risk Evaluation and Mitigation Strategies imposed a small burden of prenatal exposure because of the low background utilization rates and lower pregnancy prevalence among women of childbearing age who used these drugs. Although the declining exposure rates to teratogenic drugs with definite risk are encouraging, the rising prenatal exposure to drugs with potential risk calls for more assessments. Future research is needed to elucidate the health outcomes of fetuses exposed to potential risk drugs, understand the effectiveness of risk evaluation and mitigation strategy programs, and prioritize teratogenic drugs for advanced risk mitigation.