{"id":1244,"date":"2026-10-31T09:00:00","date_gmt":"2026-10-31T13:00:00","guid":{"rendered":"https:\/\/drugchatter.com\/insights\/?p=1244"},"modified":"2026-09-26T13:56:03","modified_gmt":"2026-09-26T17:56:03","slug":"the-untracked-layer-of-pharma-competitive-intelligence-what-people-actually-ask-ai","status":"publish","type":"post","link":"https:\/\/drugchatter.com\/insights\/the-untracked-layer-of-pharma-competitive-intelligence-what-people-actually-ask-ai\/","title":{"rendered":"The Untracked Layer of Pharma Competitive Intelligence: What People Actually Ask AI"},"content":{"rendered":"\n
\"\"<\/figure>\n\n\n\n

The short answer:<\/strong> Most pharma competitive intelligence still runs on search rankings, social listening, and claims data. None of that captures the fastest-growing channel where patients, caregivers, and clinicians now form opinions about a drug: a private conversation with ChatGPT, Gemini, Claude, or Perplexity. Nearly a third of U.S. adults now use AI tools for health information at least monthly, and independent studies keep finding that AI chatbots answer drug-specific questions incorrectly or incompletely a large share of the time. DrugChatter’s monitoring work starts from a different data source than most AI-visibility tools: the actual prompts people type, not a simulated list of “likely” questions.<\/p>\n\n\n\n

The Five Findings That Matter<\/strong><\/h2>\n\n\n\n
    \n
  • Monthly AI use for health information among U.S. adults rose to 29% in a June 2026 KFF poll, up from 17% in June 2024 [1].<\/li>\n\n\n\n
  • In a Pew Research Center survey fielded in June 2026, 20% of U.S. adults said they use chatbots to get medical advice, rising to 25% among adults aged 18 to 49 [2].<\/li>\n\n\n\n
  • In a study presented at the ASHP Midyear Clinical Meeting, free ChatGPT gave a satisfactory answer to only 10 of 39 real drug-information questions pulled from a university drug information service, and every reference it cited for the questions it did answer turned out to be fabricated [6][7].<\/li>\n\n\n\n
  • In a 2024 study of 70 real-world drug inquiries, GPT-4 produced a completely accurate response 64.3% of the time, an improvement over earlier GPT versions [9].<\/li>\n\n\n\n
  • Most adults, including half of those who use AI regularly, say they are not confident they can tell true health information from false information when it comes from an AI chatbot [5].<\/li>\n<\/ul>\n\n\n\n

    What This Means, in One Sentence<\/strong><\/h2>\n\n\n\n

    People are already asking AI models drug-specific questions at scale, the models are demonstrably capable of answering those questions wrong, and almost no pharma competitive intelligence function has systematically collected the actual questions being asked in order to know where the gaps are.<\/p>\n\n\n\n

    How Fast AI Became a Drug-Information Channel<\/strong><\/h2>\n\n\n\n

    The shift happened over roughly two years. A KFF Health Misinformation Tracking Poll fielded in June 2024 found that 17% of adults used AI chatbots at least monthly for health information, with confidence in the accuracy of that information already low [5]. By December 2025, a Rock Health survey of 8,000 U.S. adults put chatbot-for-health-information use at 32%, up from 16% the year before, and found that 64% of AI users engage with it for health questions weekly or more often. ChatGPT and Gemini were the most-used platforms, at 23% and 15% of respondents respectively [3]. A KFF poll fielded roughly six months later measured monthly AI use for health information at 29%, alongside 31% of adults who say they use social media for the same purpose [1].<\/p>\n\n\n\n

    The demographic pattern matters for brand and medical affairs teams building content strategy. Pew Research data from June 2026 shows chatbot use for medical advice concentrated among younger adults and varying sharply by race and ethnicity: 39% of Asian adults reported using chatbots for medical advice, compared with 23% of Latino adults, 21% of Black adults, and 18% of white adults [2]. Data on what happens after an AI health query shows the conversation rarely ends there: 42% of users said they searched for more information afterward, 40% consulted a provider, and 18% said the AI conversation led them to adjust a medication [4]. That last figure is the one medical affairs and pharmacovigilance teams should sit with.<\/p>\n\n\n\n

    What the Evidence Shows About AI Drug-Information Accuracy<\/strong><\/h2>\n\n\n\n

    The Long Island University \/ ASHP Finding<\/strong><\/h3>\n\n\n\n

    The most cited real-world test of AI drug-information accuracy came from pharmacy researchers Sara Grossman, Joseph Nathan, and Tina Zerilli at Long Island University, presented at the American Society of Health-System Pharmacists’ 2023 Midyear Clinical Meeting. The team pulled 45 questions directly from LIU’s College of Pharmacy drug information service, spanning a 16-month period in 2022 and 2023, and had pharmacists research and validate answers first [6]. Six questions were dropped for lack of a literature-based standard, leaving 39 for ChatGPT. The free version of ChatGPT produced a satisfactory answer to only 10 of those 39. Of the 29 unsatisfactory responses, 11 failed to address the question at all, 10 were inaccurate, and 12 were incomplete, with some answers falling into more than one category [6].<\/p>\n\n\n\n

    The citation problem was arguably worse than the accuracy problem. ChatGPT supplied references in only 8 of the 28 questions it directly answered, and investigators found that every single cited reference did not exist [7]. One example illustrates the clinical stakes: asked whether Paxlovid and verapamil interact, ChatGPT said the combination would produce no adverse effects, when in reality the combination carries a real risk of significant hypotension [8].<\/p>\n\n\n\n

    \n

    The study authors concluded that ChatGPT was unable to provide an accurate and complete response to most questions presented to a drug information service, and recommended that healthcare professionals and consumers verify any drug-related response from the chatbot against trusted sources [7].<\/p>\n<\/blockquote>\n\n\n\n

    Real-World Inquiries Against a Drug Information Database<\/strong><\/h3>\n\n\n\n

    A separate 2024 study published in Digital Health took a different approach: instead of hypothetical questions, researchers pulled 70 real inquiries logged in a drug information inquiry database, most of them submitted by physicians (41%) and pharmacists (44%), covering dosage and administration (34.2%), drug interactions (12.8%), and pregnancy or lactation questions (15.7%). Two independent pharmacists scored GPT-3, GPT-3.5, and GPT-4 against accuracy, detail, and risk-of-harm criteria. GPT-4 reached 64.3% completely accurate responses, a result the authors described as comparable to licensed pharmacists on this sample [9].<\/p>\n\n\n\n

    Model Versions Are Improving, Unevenly<\/strong><\/h3>\n\n\n\n

    Later model generations do perform better on structured knowledge tests. A 2025 JMIR Medical Education study testing GPT-3.5 and GPT-4 against Taiwan’s National Pharmacist Licensing Examination found GPT-4 reached 72.9% accuracy versus 59.1% for GPT-3.5 [10]. A separate evaluation of five LLMs (GPT-3.5, GPT-4, Claude 2, and two Llama 2 variants) on 219 critical care pharmacotherapy questions found GPT-4 led with 71.6% accuracy, and that every model tested performed better on knowledge-recall questions than on knowledge-application questions: 87% versus 67% for GPT-4 specifically [11]. That recall-versus-application gap is a useful distinction for medical affairs teams: a model can “know” a fact about a drug and still fail to apply it correctly to a specific patient scenario.<\/p>\n\n\n\n

    A 2025 pharmacology-focused survey in the journal Drugs found a more specific weak point: general-purpose models handle broad pharmacology concepts reasonably well but struggle specifically with drug interaction retrieval and drug identification queries, exactly the categories most likely to come up in a real patient or caregiver question [12]. Licensing-exam accuracy and real-world query accuracy are not the same measurement, and the gap between them is where brand-specific risk tends to live.<\/p>\n\n\n\n

    Where AI and the Evidence Diverge: A Comparison<\/strong><\/h2>\n\n\n\n
    Study<\/th>Question source<\/th>Model(s)<\/th>Key accuracy finding<\/th><\/tr><\/thead>
    LIU \/ ASHP, presented Dec 2023 [6][7]<\/td>39 real drug information service questions<\/td>Free ChatGPT (GPT-3.5)<\/td>10\/39 satisfactory; fabricated references in every cited response<\/td><\/tr>
    Digital Health, May 2024 [9]<\/td>70 real physician\/pharmacist inquiries<\/td>GPT-3, GPT-3.5, GPT-4<\/td>GPT-4: 64.3% completely accurate<\/td><\/tr>
    JMIR Med Educ, 2025 [10]<\/td>Taiwan pharmacist licensing exam<\/td>GPT-3.5, GPT-4<\/td>72.9% vs. 59.1% (exam accuracy, not real-world query accuracy)<\/td><\/tr>
    Frontiers in AI, 2024\/2025 [11]<\/td>219 critical care pharmacotherapy questions<\/td>GPT-3.5, GPT-4, Claude 2, Llama2-7b\/13b<\/td>GPT-4 highest at 71.6%; recall 87% vs. application 67%<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

    Read across these four studies and a pattern holds regardless of which model or dataset is used: general drug knowledge has improved materially since 2023, but application to a specific, real question, especially one involving an interaction, a dose, or a special population, remains the point where models are most likely to get it wrong.<\/p>\n\n\n\n

    Why Conventional Competitive Intelligence Misses This Layer<\/strong><\/h2>\n\n\n\n

    Established pharma competitive intelligence tooling was not built to see this. Platforms like IQVIA’s Social Media Intelligence give brand teams a dashboard of digital performance KPIs benchmarked against competitors, drawing on social listening and web signals [13]. Patient-sentiment platforms such as Talking Medicines’ PatientMetRx, built by a Glasgow-based team that has raised $3.4 million to date, turn patient social commentary into a confidence score for a brand [14]. A newer category of AI-visibility or answer-engine-optimization (AEO) tools has emerged specifically to track how AI systems describe drug brands, indications, and safety profiles across model families, positioning itself as the pharma-specific analogue to SEO tracking.<\/p>\n\n\n\n

    All of these approaches share a structural limitation: they measure what AI systems say back, or what patients say about a brand elsewhere online. Almost none of them start from the actual prompt a person typed into the AI system in the first place. That is a different question. Knowing that a chatbot describes a drug’s contraindications inaccurately is useful. Knowing which specific question, worded which specific way, triggered that inaccurate answer, and how often real users ask it that way, is a different and more actionable input for medical affairs, brand, and regulatory teams deciding where to spend limited content-remediation resources.<\/p>\n\n\n\n

    What a Question-First Approach Changes<\/strong><\/h2>\n\n\n\n

    DrugChatter’s monitoring work starts from a database of the actual questions people have put to AI systems about specific drugs, rather than a list of questions an analyst guesses might be asked. The distinction sounds small and is not. A simulated prompt list reflects what a marketing or medical affairs team assumes patients and clinicians care about. A real prompt log reflects what they are actually typing, phrased the way they actually phrase it, including the colloquial, symptom-first, or comparison-shopping language that a brand’s own approved content rarely uses.<\/p>\n\n\n\n

    That distinction changes what gets prioritized. If the real-question data shows that a drug interaction question is being asked in a particular phrasing far more often than a company’s content team assumed, that phrasing, not a generic “drug interactions” FAQ page, is the content gap worth closing first. If a competitor’s molecule keeps surfacing in answers to a question a brand’s own team never anticipated, that is a competitive-displacement signal conventional social listening will not catch, because it never appears in public posts, reviews, or forums. It only exists inside the AI conversation itself.<\/p>\n\n\n\n

    What This Means for Medical Affairs<\/strong><\/h2>\n\n\n\n

    Medical affairs teams already field unsolicited medical information requests through structured channels. An AI-question dataset extends visibility into a parallel, much larger channel that operates entirely outside those structured intake forms. The LIU finding that fabricated citations accompanied every answered question in its sample is directly relevant here: a clinician or caregiver who asks a chatbot a dosing question and receives a fluent, confidently cited, but fictitious answer has no built-in signal that anything is wrong [7]. Medical information teams that know which specific questions are recurring can pre-position accurate, source-linked answers in the channels AI systems actually draw from, rather than reacting after a discrepancy is reported.<\/p>\n\n\n\n

    What This Means for Pharmacovigilance<\/strong><\/h2>\n\n\n\n

    The finding that 18% of AI health-query users reported adjusting a medication afterward is a pharmacovigilance-relevant statistic, not just a marketing one [4]. If AI systems are influencing real medication-taking behavior, and independent studies show those systems getting interaction and dosing questions wrong at meaningful rates [6], then knowing what questions are actually being asked about a specific product is relevant to understanding where AI-influenced adverse outcomes could originate, even though it does not substitute for a formal signal-detection process.<\/p>\n\n\n\n

    What This Means for Brand and Marketing Teams<\/strong><\/h2>\n\n\n\n

    Brand teams have historically optimized for search intent. Query-fan-out and long-tail keyword strategy assumed a searcher would land on a page and read it. An AI answer compresses that same intent into a single generated response the user may never click through from. A brand’s approved content can be technically excellent and still lose the moment if the underlying question, in its real phrasing, is not the one the content was built to answer. Real-question data closes that gap directly rather than through inference from keyword-volume tools built for a different kind of search behavior.<\/p>\n\n\n\n

    Methodology Notes for Anyone Building This Kind of Analysis<\/strong><\/h2>\n\n\n\n

    Studies like the ones cited above are reproducible only when they specify their inputs clearly. At minimum, that means naming the model and version tested, the exact date range of testing, whether the questions came from real users or were constructed by researchers, the number of queries, how a “satisfactory” or “accurate” response was defined, and how discrepancies against a reference standard, such as an FDA label or peer-reviewed literature, were classified. The LIU study is a useful model of this transparency: it specifies the 16-month collection window, the 45-to-39 question filtering process, the pharmacist-validation step, and the specific criteria used to sort unsatisfactory answers into “non-responsive,” “inaccurate,” and “incomplete” categories [6]. Any AI-monitoring claim that does not disclose an equivalent level of methodological detail should be treated as directional rather than as a citable statistic.<\/p>\n\n\n\n

    Key Takeaways<\/strong><\/h2>\n\n\n\n
      \n
    • Monthly AI use for health information among U.S. adults reached 29% in mid-2026, roughly double the 2024 rate [1].<\/li>\n\n\n\n
    • Independent, peer-reviewed and conference-presented studies from 2023 through 2025 consistently find real accuracy gaps in AI-generated drug information, concentrated in interaction, dosing, and application questions rather than general pharmacology knowledge [6][11][12].<\/li>\n\n\n\n
    • Model accuracy is improving version over version, but the improvement is uneven and concentrated in recall rather than clinical application [11].<\/li>\n\n\n\n
    • Existing pharma competitive intelligence tools largely measure AI output or public sentiment, not the underlying real-world question volume and phrasing driving that output.<\/li>\n\n\n\n
    • A meaningful share of AI health-information users report acting on what they were told, including adjusting medications, which connects this data directly to medical affairs and pharmacovigilance functions, not only to marketing [4].<\/li>\n<\/ul>\n\n\n\n

      FAQ<\/strong><\/h2>\n\n\n\n

      How many people actually use AI chatbots for drug or health questions?<\/strong><\/h3>\n\n\n\n

      Estimates vary by survey and definition, but multiple independent 2025-2026 surveys converge on roughly 29% to 32% of U.S. adults using AI tools for health information at least monthly or within the past year [1][3].<\/p>\n\n\n\n

      Is ChatGPT reliable for drug interaction questions?<\/strong><\/h3>\n\n\n\n

      The available evidence says no in a meaningful share of cases. The LIU\/ASHP study found ChatGPT incorrectly stated that Paxlovid and verapamil carried no interaction risk, when the actual risk is significant hypotension [8]. A separate 2025 survey noted drug interaction retrieval as a specific weak point for general-purpose LLMs [12].<\/p>\n\n\n\n

      Do newer AI models make fewer drug-information errors than earlier ones?<\/strong><\/h3>\n\n\n\n

      Generally yes on structured knowledge tests: GPT-4 outperformed GPT-3.5 by a wide margin on a pharmacist licensing exam, 72.9% versus 59.1% [10]. The gap narrows, and matters more in risk terms, on application-style questions rather than recall questions [11].<\/p>\n\n\n\n

      Does ChatGPT make up sources when answering drug questions?<\/strong><\/h3>\n\n\n\n

      In the LIU\/ASHP study, every reference ChatGPT supplied across its answered questions was fabricated and did not correspond to a real publication [7]. This is a specific, documented instance rather than a general claim about all AI systems in all contexts.<\/p>\n\n\n\n

      What is answer engine optimization (AEO) in a pharma context?<\/strong><\/h3>\n\n\n\n

      AEO, sometimes called AI visibility or generative engine optimization, is the practice of tracking and influencing how AI systems describe a drug brand, its indications, and its safety profile across model families, positioned by vendors as the pharma-specific analogue to search engine optimization.<\/p>\n\n\n\n

      Can AI-generated drug information affect whether someone changes their medication?<\/strong><\/h3>\n\n\n\n

      Survey data suggests it can. In a late-2025 consumer survey, 18% of respondents who had used AI for health information in the past year reported that it led them to adjust a medication [4]. That is a self-reported behavioral outcome, not a clinical verification of appropriateness.<\/p>\n\n\n\n

      How is DrugChatter’s approach different from social listening or search-rank tracking?<\/strong><\/h3>\n\n\n\n

      Social listening and search-rank tools measure public commentary or page rankings. DrugChatter’s monitoring starts from an underlying database of the actual questions users have put to AI systems about specific drugs, which surfaces phrasing and intent that never appears in public posts or search queries.<\/p>\n\n\n\n

      Do most people trust AI-generated health information?<\/strong><\/h3>\n\n\n\n

      No. Across multiple polls, a majority of adults, including about half of regular AI users, say they are not confident they can distinguish accurate from inaccurate health information generated by AI chatbots [5].<\/p>\n\n\n\n

      Which AI platforms do people use most for health questions?<\/strong><\/h3>\n\n\n\n

      In the Rock Health 2025 survey, ChatGPT and Gemini were the most-used platforms for health information, at 23% and 15% of respondents respectively [3].<\/p>\n\n\n\n

      What should a brand or medical affairs team do first with this kind of data?<\/strong><\/h3>\n\n\n\n

      Prioritize closing content gaps around the specific, real-world phrasings tied to interaction, dosing, and special-population questions, since independent studies consistently identify these as the categories where AI-generated answers are most likely to diverge from the current label or guideline [9][12].<\/p>\n\n\n\n

      References<\/h2>\n\n\n\n
        \n
      1. Becker’s Physician Leadership. (2026, June). 29% of US adults use AI for health information monthly: KFF. https:\/\/www.beckersphysicianleadership.com\/?p=7718<\/li>\n\n\n\n
      2. Fox, S. (2026, June 24). How U.S. adults use AI for health. susannahfox.com, citing Pew Research Center, June 2026. https:\/\/susannahfox.com\/2026\/06\/24\/how-u-s-adults-use-ai-for-health\/<\/li>\n\n\n\n
      3. FierceHealthcare. (2026). AI chatbot use for health information up 16% from 2024: Rock Health survey. https:\/\/fiercehealthcare.com\/ai-and-machine-learning\/ai-chatbot-use-health-information-16-2024-rock-health-survey<\/li>\n\n\n\n
      4. Darwin Research Group. (2026, March 30). Nearly one-third of U.S. adults use AI to get health information, surveys show. https:\/\/www.darwinresearch.com\/news-and-insights\/darwins-our-take-nearly-one-third-of-u-s-adults-use-ai-to-get-health-information-surveys-show<\/li>\n\n\n\n
      5. Presiado, M., Montero, A., Lopes, L., & Hamel, L. (2024, August 15). KFF Health Misinformation Tracking Poll: Artificial intelligence and health information. KFF. https:\/\/www.kff.org\/public-opinion\/kff-health-misinformation-tracking-poll-artificial-intelligence-and-health-information\/<\/li>\n\n\n\n
      6. Becker’s Hospital Review. (2023, December). Most of ChatGPT’s answers to drug questions are wrong: Study. https:\/\/beckershospitalreview.com\/pharmacy\/most-of-chatgpts-answers-to-drug-questions-are-wrong-study.html<\/li>\n\n\n\n
      7. Drug Topics. (2023). ChatGPT provides inaccurate responses to questions about drug information. https:\/\/www.drugtopics.com\/view\/chatgpt-provides-inaccurate-responses-to-questions-about-drug-information<\/li>\n\n\n\n
      8. Epocrates. (2023, December 18). ChatGPT fails to answer drug-related questions, new study finds. https:\/\/epocrates.com\/online\/article\/chatgpt-fails-to-answer-drug-related-questions-new-study-finds<\/li>\n\n\n\n
      9. Albogami, Y., Alfakhri, A., Alaqil, A., Alkoraishi, A., Alshammari, H., Elsharawy, Y., Alhammad, A., & Alhossan, A. (2024). Safety and quality of AI chatbots for drug-related inquiries: A real-world comparison with licensed pharmacists. Digital Health, 10. https:\/\/doi.org\/10.1177\/20552076241253523<\/li>\n\n\n\n
      10. Wang et al. (2025). GPT-3.5 and GPT-4 on the Taiwan National Pharmacist Licensing Examination. JMIR Medical Education, 11, e56850. https:\/\/doi.org\/10.2196\/56850<\/li>\n\n\n\n
      11. Yang, H., Hu, M., Most, A., Hawkins, W. A., Murray, B., Smith, S. E., Li, S., & Sikora, A. (2024\/2025). Evaluating accuracy and reproducibility of large language model performance on critical care assessments in pharmacy education. Frontiers in Artificial Intelligence, 7. https:\/\/doi.org\/10.3389\/frai.2024.1514896<\/li>\n\n\n\n
      12. Aligning large language models with humans: A comprehensive survey of ChatGPT’s aptitude in pharmacology. (2025, February). Drugs, 85(2), 231-254. https:\/\/doi.org\/10.1007\/s40265-024-02124-2<\/li>\n\n\n\n
      13. IQVIA. Social media intelligence for pharma and consumer health. https:\/\/www.iqvia.com\/de-ch\/solutions\/commercialization\/brand-strategy-and-management\/primary-intelligence\/social-media-intelligence-for-pharma-and-consumer-health<\/li>\n\n\n\n
      14. Business Wire. (2021, February 4). Talking Medicines launches AI powered patient intelligence platform. https:\/\/www.businesswire.com\/news\/home\/20210204005476\/en\/<\/li>\n<\/ol>\n","protected":false},"excerpt":{"rendered":"

        The short answer: Most pharma competitive intelligence still runs on search rankings, social listening, and claims data. None of that […]<\/p>\n","protected":false},"author":1,"featured_media":1247,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_lmt_disableupdate":"","_lmt_disable":"","site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1244","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-general"],"modified_by":"DrugChatter","_links":{"self":[{"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/posts\/1244","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/comments?post=1244"}],"version-history":[{"count":1,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/posts\/1244\/revisions"}],"predecessor-version":[{"id":1248,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/posts\/1244\/revisions\/1248"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/media\/1247"}],"wp:attachment":[{"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/media?parent=1244"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/categories?post=1244"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/drugchatter.com\/insights\/wp-json\/wp\/v2\/tags?post=1244"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}