這篇文章，是那個決定之後的故事。
某一天，我盯著自己的個人網站，突然覺得哪裡不對。
網站確實存在——有網址、有頁面、有按鈕。但點開來，它只是一串連結：Medium、Substack、Facebook、Twitter……一個個箭頭，把訪客送離這裡，卻沒有一句話說清楚「我是誰」、「我在想什麼」、「我為什麼值得你認識」。
那個頁面，像一張沒有照片的名片。
我決定，這不夠。
在開始動工之前，我和 AI 做了一件事：把我在網路上所有的身份找出來。
這個過程，意外地像是一場考古。
我以為自己有七八個帳號，翻著翻著，最後數出來還記得或曾經使用的是 17 個。有些是幾年前熱情滿滿開設的寫作平台，有些是當時覺得很重要的社群帳號，有些則是學術圈的識別身份。它們散落在網路各個角落，各自沉默著。
最讓我心頭一緊的發現，是一個叫 Matters 的平台即將關閉部分平台功能。
我在那裡寫過兩個帳號，累積了將近 176 篇文章。那些文字，是 2019 到 2021 年間的我——剛開始認真寫作的那段時光，留下的痕跡。
幸好，Matters 在關閉部分功能前推出了備份工具。我搶在最後，把所有文章完整儲存下來。176 篇，一篇都沒有少。
那個下午，我鬆了一口氣，也悄悄地感謝了過去那個持續寫作的自己。
備份完成之後，真正的工程才開始。
建一個新的個人網站，第一個問題不是「要放什麼」，而是「這個網站要說一個什麼樣的故事」。
我現在是醫療科技評估（HTA）的研究者，這是我希望建立且深化的現階段專業。但我同時也對 AI 與科技著迷、關心歷史與人文、思考科技和社會之間的關係。我不是一個容易被單一標籤概括的人。
我們花了一些時間，找到了這樣一句話：
「從科學到社會，好奇心是我探索世界的羅盤。」
這句話，後來成為整個網站的靈魂。每一個設計決定、每一篇文章、每一個頁面，都從這裡長出來。
建網站的時候，有一個問題我思考了很久：要不要放本人照片？
最後，我選擇了另一種方式。
我用了一張自己喜歡的插畫作為頭像——藍色的頭髮、側臉的構圖、帶著一種若有所思的神情。它不是我，但某種程度上它又很像我——或者說，像我希望自己呈現出來的樣子：有溫度、有深度，不急著把自己解釋清楚。
那抹藍色，和整個網站的色調幾乎一模一樣。像是早就說好的事。
最終完成的網站，有六個頁面，每一頁都同時以中文和英文呈現：
每一頁都可以切換深色與淺色模式，讓不同習慣的訪客都能舒適閱讀。
這篇文章，就是這段旅程最新的一章。
現在，當我有想法想分享，我坐下來寫，寫完之後，一個動作，它就出現在網站上。電子報會自動送到訂閱者的信箱，部分內容也會同步出現在我的社群媒體上。
我不需要在各個平台間奔波，不需要反覆複製貼上，不需要記住一堆流程。
寫，就夠了。
這是我們花了好幾天，一起把它建起來的。
這個網站還在成長。
176 篇塵封多年的文章，還在等待以更好的方式重新出現。我的知識庫和網站之間的深度連結，還在慢慢摸索。也許還有更多我現在還沒想到的可能。
但最重要的是——
這裡，現在有了自己的聲音。
從科學到社會，好奇心是我探索世界的羅盤。
— Michael Wei-Chih Liu 劉威志
English version below
Sometimes, one decision changes a lot of things.
This article is the story of what happened after that decision.
One day, I stared at my personal website and felt something was off.
The website existed — it had a URL, a page, buttons. But clicking through, it was just a list of links: Medium, Substack, Facebook, Twitter… arrows pointing visitors away from here, without a single sentence explaining who I am, what I think about, or why I might be worth knowing.
That page was like a business card with no photo.
I decided: this isn’t enough.
Before building anything new, I did something with AI: I tracked down every version of myself on the internet.
It turned out to be a kind of archaeology.
I thought I had seven or eight accounts. When I finished counting, the number was 17. Some were writing platforms I’d set up years ago with great enthusiasm. Some were social accounts that felt important at the time. Some were academic identity profiles. They were scattered across the internet, each quietly sitting in silence.
The discovery that unsettled me most: a platform called Matters was shutting down.
I had written across two accounts there, accumulating nearly 176 articles — traces of who I was between 2019 and 2021, during the years I first started writing seriously.
Fortunately, Matters launched a backup tool before closing. I managed to save everything just in time. All 176 articles, not one missing.
That afternoon, I breathed a sigh of relief — and quietly thanked the version of myself who had kept writing.
Once the backup was done, the real work began.
Building a new personal website, the first question wasn’t “what should I put here?” It was “what story should this website tell?”
I’m a Health Technology Assessment (HTA) researcher — that’s my profession. But I’m also fascinated by AI and technology, drawn to history and the humanities, and deeply interested in the relationship between science and society. I’m not someone easily captured by a single label.
After some thought, we arrived at this:
“From science to society — curiosity is my compass.”
That sentence became the soul of the entire website. Every design decision, every article, every page grows from it.
When building the site, I spent a long time on one question: Should I use a real photo?
In the end, I chose a different path.
I used an illustration I love — blue hair, a side profile, an expression of quiet contemplation. It isn’t me, but in some way it feels like me — or like the version of myself I want to present: warm, layered, not in a hurry to explain everything.
That shade of blue matched the site’s color palette almost exactly. As if it had always been meant to be.
The finished website has six pages, each presented in both Chinese and English:
Every page supports both dark and light mode, so every visitor can read comfortably.
This article is the newest chapter of that journey.
Now, when I have something to say, I sit down and write. When I’m done, one action later, it appears on the website. A newsletter goes out automatically to subscribers. Some of it syncs to my social media.
I don’t need to rush between platforms. I don’t need to copy and paste over and over. I don’t need to remember a complicated process.
Writing is enough.
That’s what we spent several days building together.
This website is still growing.
176 long-archived articles are waiting to resurface in a better form. The deeper connection between my knowledge base and my website is still being figured out. There are probably possibilities I haven’t even thought of yet.
But most importantly —
This place now has its own voice.
From science to society — curiosity is my compass.
— Michael Wei-Chih Liu 劉威志
]]>昨晚的HTAi線上研討會帶來了滿滿的共鳴。身為醫療科技評估研究員，我在準備這場會議的過程中，不斷反覆探究國際指引與殘酷的全球現實。我甚至拿出了我們內部療效心法文件，與國際標準逐一對照。
為什麼這麼做？因為這代表著我對自身專業的焦慮與期許。我想知道的是：「我們目前的做法，距離完美的國際標準，到底還有多少差距？」
昨晚各國專家的分享給了我深刻的解答。
追求最完美、最標準的完整醫療科技評估報告固然是所有人的理想。但各國的現實處境——無論是醫療環境、預算限制，或是醫療科技評估投入的人力與資源匱乏——造就了大家實務上「長得都不一樣」的評估產出。
標準一直都在那裡，但面對現實的「彈性」必須跟著變化。
然而，在這些妥協與彈性中，唯一不能退步與退讓的，是我們對真實科學證據的追求。
哪怕面臨現實環境與時間的極度逼迫——我們必須在短短的42天內完成評估——身為醫療科技評估研究人員，我們盡可能地在所有有限的條件中，交出了我們所認知、甚至超越國際基本門檻的簡明式/快速審查醫療科技評估報告。
面對我們產出的報告，我一點都不感到心虛。因為我們在台灣，已經在現有資源下盡到了自己最大的努力。當然，我們深知自己還有進步的空間，我們也清楚「最高標準」在哪裡，而我們正堅定地朝著那個方向邁進。
在推進台灣醫療科技評估與健保給付決策的這條路上，或許酸言酸語可以不用。我們需要的是「一起把手弄髒」——動手把環境與處境弄好，真正地把事情做好做滿，這才是對病患與醫療體系負責任的態度。
這是我在這次HTAi研討會後最深刻的個人心得與反思。
致所有在資源限制下，仍堅持科學證據底線的醫療科技評估夥伴們，我們繼續前行！💪
有關行政法人設置條例草案，昨日07/01已於立法院黨團協商達成共識並順利通過。
May the force be with you in HTA!
English version below
Last night’s HTAi webinar brought profound resonance. As a health technology assessment researcher, I spent the preparation process repeatedly examining international guidelines and the harsh realities of global practice. I even pulled out our internal efficacy documentation and aligned it point-by-point with international standards.
Why do this? It reflects my anxiety and aspiration about my own profession. What I want to know is: “How much gap remains between our current practices and the perfect international standard?”
Last night’s presentations from international experts provided me a profound answer.
Pursuing the most perfect and standardized complete health technology assessment report is certainly everyone’s ideal. Yet each country’s real circumstances—whether healthcare environment, budget constraints, or scarce human and financial resources dedicated to HTA—result in practically “different-looking” assessment outputs everywhere.
The standards are always there, but the “flexibility” needed to face reality must adapt accordingly.
Yet amid these compromises and flexibility, there is one thing we must never compromise on: our commitment to authentic scientific evidence.
Despite facing extreme pressure from environment and time—we must complete assessments in just 42 days—as HTA researchers, we do our utmost within all limited conditions to deliver mini/rapid HTA reports that meet, and even exceed, international baseline standards.
Facing the reports we produce, I feel no shame whatsoever. Because in Taiwan, we have given our maximum effort within available resources. Of course, we know there is room for improvement, and we are well aware of where the “highest standard” lies. We are marching steadily toward that goal.
On the path of advancing Taiwan’s health technology assessment and health insurance reimbursement decision-making, perhaps pointless criticism can be spared. What we need is to “get our hands dirty together”—to roll up our sleeves, improve the environment and circumstances, and truly do things well and thoroughly. This is the responsible attitude toward patients and the healthcare system.
This is my deepest personal reflection and insight from this HTAi webinar.
To all HTA colleagues who remain committed to the scientific evidence baseline even under resource constraints, let us continue forward! 💪
Regarding the Administrative Agency Establishment Act draft, consensus was reached through caucus negotiations at the Legislative Yuan on 07/01 and successfully passed.
May the force be with you in HTA!
]]>昨晚由 HTAi 主辦的線上研討會「Beyond the ‘Full HTA Report’」圓滿落幕！醫療科技評估（HTA）機構在全球面臨共同的挑戰：如何在「時間壓力、資源限制與科學嚴謹度」之間取得平衡？這場討論延續前兩篇探討「理想定義」與「現實落差」的內容，真實呈現了全球各地的務實做法。
Tracy Merlin 帶我們回顧了 2014 年國際醫療科技評估組織網絡（INAHTA）訂定 HTA 產出標準的初衷——為了在混亂的名詞中建立「共同語言」（Shared understanding）。
Barbara Clyne 則透過 10 年的實證數據直擊痛點：現實中的「快速/迷你 HTA」充滿異質性。多數機構為了求快，妥協了方法學或省略了經濟/倫理評估，卻沒有「透明地交代理由」，這為決策帶來了潛在風險。
下半場的 Panel Discussion 更是整場會議的精華！來自三個不同新興醫療科技評估系統的專家，分享了他們如何在資源與時間的極端壓力下「務實地求生」。
菲律賓的 Anne Julienne Genuino-Marfori 分享了該國的做法——為了因應不同情況，他們建立了明確的「三軌制」：
透過清晰的分類與限縮適用族群來控管風險，這是一個系統化的應對策略。
烏拉圭的 Ana Pérez 點出了拉丁美洲特有的困境——「司法化」。面對病患透過法律途徑要求國家給付高價藥物的壓力，醫療科技評估機構必須產出特定、動態的快速報告來因應訴訟，並將其作為資源「優先排序」的談判工具。
阿布達比的 Kefah Al Alqawasmeh 面對長官要求「今天下班前就要決定」的極端時效性，以及針對人工智慧等新科技高度不確定的證據，他們的解方是「擁抱不確定性」：
這個做法既滿足了決策時程，也為證據補強預留了空間。
會議最後，兩位主講者的總結非常有力——我們需要保持彈性，但必須伴隨「標準化」與「透明度」。Tracy 更直言：「決策者必須了解並『承擔抄捷徑的風險』，而非將責任全推給產出報告的醫療科技評估機構。」
雖然礙於時間關係，會議未能進行線上 Q&A，但我在會中的一個線上提問——「目前學界仍缺乏專為快速 HTA 設計的品質評估工具」，恰好與講者們最後呼籲建立「新標準」的結論不謀而合！這無疑是未來 HTA 社群亟需共同努力的方向。
感謝講者與所有參與者帶來如此精彩的「繼往開來」對話！在追求「快與準」的決策道路上，您的機構又是如何平衡這些挑戰的呢？歡迎交流！
English version below
Last night, the online webinar ‘Beyond the Full HTA Report’ hosted by HTAi concluded successfully! Health technology assessment (HTA) bodies worldwide face a common challenge: how to balance ‘time pressure, resource constraints, and scientific rigor’? This discussion extends our previous exploration of ‘ideal definitions’ and ‘reality gaps,’ showcasing pragmatic approaches from different regions globally.
Tracy Merlin walked us through the rationale behind the health technology assessment standards established by the International Network of Agencies for Health Technology Assessment (INAHTA) in 2014—to create a ‘shared understanding’ amid the chaos of terminology.
Barbara Clyne, drawing on 10 years of empirical data, highlighted the critical pain point: rapid and mini HTA approaches in practice are highly heterogeneous. Many agencies compromise on methodology or omit economic and ethics assessments in pursuit of speed, yet fail to ‘transparently account for their reasons.’ This creates potential risks for decision-making.
The Panel Discussion in the second half was truly the highlight of the conference! Experts from three different emerging health technology assessment systems shared how they ‘pragmatically survive’ under extreme resource and time pressures.
Anne Julienne Genuino-Marfori from the Philippines shared their approach—to address different situations, they established a clear ‘three-track system’:
Through clear classification and limited scope definition, they manage risk systematically.
Ana Pérez from Uruguay highlighted a challenge unique to Latin America—’judicialization.’ Facing patients using legal pathways to demand government coverage of expensive medicines, health technology assessment bodies must produce specific, dynamic rapid reports to address litigation and use them as negotiation tools for resource ‘prioritization.’
Kefah Al Alqawasmeh from Abu Dhabi faces extreme time constraints with leadership demanding decisions ‘before end of business today,’ plus highly uncertain evidence for emerging technologies like artificial intelligence. Their solution is to ‘embrace uncertainty’:
This approach satisfies decision timelines while reserving space for evidence strengthening.
At the end, both presenters’ conclusions were powerful—we need flexibility, but it must be accompanied by ‘standardization’ and ‘transparency.’ Tracy was particularly direct: ‘Decision makers must understand and accept the risks of taking shortcuts, rather than pushing all responsibility onto the HTA agencies producing reports.’
Although time constraints prevented a live Q&A session, my online question during the webinar—’Currently, academia still lacks quality assessment tools specifically designed for rapid HTA’—remarkably aligns with speakers’ final call for establishing ‘new standards.’ This is undoubtedly a direction where the HTA community urgently needs to collaborate.
Thank you to the speakers and all participants for this inspiring dialogue of ‘inheriting and innovating’! In the pursuit of balancing ‘speed and accuracy’ in decision-making, how does your organization navigate these challenges? Feedback welcome!
]]>在上一篇貼文中，我們回顧了 Merlin et al. (2014) 為醫療科技評估報告所訂定的「黃金標準」。按照定義，一份合格的「迷你醫療科技評估」必須包含成本與財務衝擊資訊，且必須具備嚴謹的系統性文獻回顧。但十年過去了，我們真的有照著做嗎？為了準備 HTAi 主辦的線上研討會，我深入研讀了即將在會中被探討的最新重磅實證研究——McLaughlin et al. (2026)。
McLaughlin et al. (2026) 針對過去十年間全球 203 份快速與迷你醫療科技評估報告進行了大規模的「現實體檢」，結果揭露了巨大的理論與現實落差。
最令人震驚的發現是：
這代表絕大多數機構在面臨時間壓力時，根本沒有向決策者透明說明自己抄了哪些捷徑。
近 38% 的報告承認僅依賴「一般文獻檢索」，而非定義中要求的系統性文獻回顧。這是最令人擔憂的發現之一，因為方法學的嚴謹性直接影響報告的可信度與通用性。
幾乎所有報告都優先確保了臨床有效性（99%）與安全性（82.3%），但倫理、社會與法律層面的評估大多被犧牲。這反映出在緊急評估情境下，醫療科技評估機構傾向於「聚焦於最直接的臨床證據」。
在醫療資源有限且決策時間緊迫的當下，醫療科技評估機構為了「求快」而抄捷徑是無可厚非的妥協。但如果我們沒有在報告中「透明地交代這些限制」，決策者極有可能將這些縮水版的報告過度解讀，甚至誤當成「完整醫療科技評估」來做出長期的健保給付承諾。這對醫療衛生系統的永續性將帶來潛在風險。
還是應該建立更嚴格的「透明度與報告標準」，強迫機構交代抄了哪些捷徑？這正是本次 HTAi 線上研討會（主題：『Beyond the “Full HTA Report”』）要直擊的核心與痛點。
這篇文獻的作者 Barbara Clyne 應該也會分享她的第一手觀點。強烈建議所有關注醫療政策與醫療科技評估發展的夥伴們一起上線參與這場「繼往開來」的精彩對話。
您的機構在面臨緊急評估壓力時，最常「妥協」的是哪一個環節？是經濟評估的深度？是文獻回顧的系統性？還是倫理與社會影響的考量？歡迎在底下留言交流！
English version below
In a previous post, we reviewed the ‘gold standard’ set by Merlin et al. (2014) for health technology assessment reports. According to the definition, a qualified ‘mini health technology assessment’ must include cost and financial impact information and be supported by rigorous systematic reviews. But has the industry actually followed these standards over the past decade? To prepare for an online seminar hosted by HTAi, I delved into the latest evidence-based research that will be explored during the event—McLaughlin et al. (2026).
McLaughlin et al. (2026) conducted a large-scale ‘reality check’ of 203 rapid and mini health technology assessment reports published over the past decade, revealing a significant gap between theory and practice.
The most striking findings include:
This suggests that most institutions, facing time pressure, fail to transparently communicate which corners they cut to their decision-makers.
Nearly 38% of reports relied solely on ‘generic literature searches’ rather than the systematic reviews required by definition. This is one of the most concerning findings, as methodological rigor directly impacts the credibility and generalisability of the report.
Almost all reports prioritized clinical efficacy (99%) and safety (82.3%), but ethical, social, and legal assessments were largely sacrificed. This reflects a tendency in urgent evaluation contexts for health technology assessment institutions to ‘focus on the most direct clinical evidence.’
Given limited healthcare resources and tight decision-making timelines, it is understandable that health technology assessment institutions compromise to ‘speed up’ the process. However, if these limitations are not ‘transparently disclosed in the report,’ decision-makers may easily overinterpret these abbreviated versions, potentially treating them as ‘complete health technology assessments’ and making long-term reimbursement commitments. This poses a potential risk to the sustainability of health systems.
Or should we establish stricter ‘transparency and reporting standards’ that force institutions to disclose which corners they cut? This is precisely what the HTAi online seminar (topic: ‘Beyond the “Full HTA Report”’) aims to address.
Barbara Clyne, one of the authors of this research, is expected to share her firsthand insights. We strongly encourage all stakeholders interested in health policy and health technology assessment development to join this forward-looking dialogue.
When facing emergency evaluation pressure, which aspect does your institution most often compromise on? The depth of economic evaluation? The systematic nature of literature review? Or consideration of ethical and social impacts? Feel free to share your thoughts in the comments below!
]]>為了準備由 INAHTA 主辦的線上研討會（主題：超越「完整醫療科技評估報告」），重新複習了醫療科技評估領域中非常經典的一篇文獻：Merlin et al. (2014) 發表的《What’s in a Name?》。這篇文獻對於理解全球醫療科技評估報告的標準化至關重要。
在十多年前，全球醫療科技評估機構在面臨決策時間與資源壓力時，發展出了各種精簡版的評估報告。但當時最大的痛點是「定義極度混亂」：你說的「快速審查」跟我以為的「迷你醫療科技評估」可能是兩回事！再者，如果無法反映報告的真實嚴謹度，導致決策者容易產生誤解。
為此，INAHTA 凝聚了全球共識，明確界定了在「速度」與「科學嚴謹度」取捨下的三大產出標準。
最完整的黃金標準。必須包含全面的系統性文獻回顧、嚴格的證據品質評估、經濟模型與廣泛的組織及社會考量。
範圍縮小，但「嚴謹度不減」。必須包含系統性文獻回顧、證據品質把關，並且必須提供成本與預算衝擊資訊。
為了搶時間，允許研究方法學的「捷徑」。通常縮減文獻檢索範圍，且證據品質評估與成本資訊皆為「非必要」。
這套定義與隨後推出的 INAHTA 產品類型標章（IPT Marks），為全球醫療科技評估報告建立了一套「防誤用」的視覺防護網，讓決策者一眼就能看懂他們手中的證據效力。
但十餘年過去了，我們真的照著做嗎？根據近期的實證分析（如 McLaughlin 等人 2026 年的研究），當今許多號稱「迷你醫療科技評估」或「快速審查」的報告，高達近四成僅使用一般文獻檢索，且近九成省略了預算評估卻未給出理由。這意味著「理論定義」與「現實妥協」之間出現了差距。
我們是否該重新定義迷你醫療科技評估？還是該嚴格要求機構在報告中透明地交代「抄了哪些捷徑」以保護決策品質？
這會是今晚 INAHTA 線上研討會要探討的核心議題！非常期待與來自愛爾蘭、澳洲、菲律賓、烏拉圭等地的專家們深入交流。
您所屬的機構在處理急迫性評估時，又是如何拿捏這條界線的呢？歡迎在底下留言交流！
相關標籤： #醫療科技評估 #INAHTA #實證醫學 #迷你醫療科技評估 #快速審查 #醫療政策 #決策科學
參考文獻：
Merlin, T., et al. (2014). What’s in a Name? Developing Definitions for Common Health Technology Assessment Product Types of the International Network of Agencies for Health Technology Assessment (INAHTA). International Journal of Technology Assessment in Health Care. https://www.cambridge.org/core/journals/international-journal-of-technology-assessment-in-health-care/article/whats-in-a-name-developing-definitions-for-common-health-technology-assessment-product-types-of-the-international-network-of-agencies-for-health-technology-assessment-inahta/9525F3145C0A60F897BA50BAD47E3389#metrics
English version below
In preparation for an online seminar hosted by INAHTA (topic: Beyond the ‘Full HTA Report’), I revisited a classic piece of literature in the field of health technology assessment: ‘What’s in a Name?’ published by Merlin et al. (2014). This paper is essential for understanding the standardization of HTA reports globally.
More than a decade ago, as global HTA agencies faced pressure on decision-making timelines and resources, they developed various abbreviated assessment reports. However, the biggest pain point at that time was ‘extreme definitional confusion’: the ‘Rapid Review’ you mentioned might mean something completely different from what I understand as ‘Mini-HTA’! Furthermore, when reports failed to reflect their true rigor level, decision makers were easily misled.
To address this, INAHTA brought global consensus together and clearly defined three major output standards under the tradeoff between ‘speed’ and ‘scientific rigor’.
The most comprehensive gold standard. Must include comprehensive systematic review, strict evidence quality assessment, economic modeling, and broad organizational and societal considerations.
Narrower in scope, but ‘rigor unchanged’. Must include systematic review, evidence quality assurance, and must provide cost and budget impact information.
To save time, methodological ‘shortcuts’ are permitted. Usually reduces the scope of literature search, and evidence quality assessment and cost information are both ‘optional’.
These definitions, along with the subsequently launched INAHTA Product Type Marks (IPT Marks), established a ‘safeguard against misuse’ visual protection network for global HTA reports, allowing decision makers to understand the strength of evidence in their hands at a glance.
But more than a decade later, are we actually following these standards? According to recent empirical analysis (such as the 2026 research by McLaughlin et al.), today many reports claiming to be ‘Mini-HTA’ or ‘Rapid Review’ use only general literature searches in nearly 40% of cases, and nearly 90% omit budget assessments without providing reasons. This means a gap has emerged between ‘theoretical definition’ and ‘practical compromise’.
Should we redefine Mini-HTA? Or should we strictly require institutions to transparently disclose which ‘shortcuts’ they took in their reports to protect decision quality?
This will be a core topic at tonight’s INAHTA webinar! I’m looking forward to in-depth exchanges with experts from Ireland, Australia, the Philippines, Uruguay, and other regions.
How does your institution balance this line when handling urgent assessments? Please share your thoughts in the comments!
Related Tags: #HealthTechnologyAssessment #INAHTA #EvidenceBasedMedicine #MiniHTA #RapidReview #HealthPolicy #DecisionMaking
Reference:
Merlin, T., et al. (2014). What’s in a Name? Developing Definitions for Common Health Technology Assessment Product Types of the International Network of Agencies for Health Technology Assessment (INAHTA). International Journal of Technology Assessment in Health Care. https://www.cambridge.org/core/journals/international-journal-of-technology-assessment-in-health-care/article/whats-in-a-name-developing-definitions-for-common-health-technology-assessment-product-types-of-the-international-network-of-agencies-for-health-technology-assessment-inahta/9525F3145C0A60F897BA50BAD47E3389#metrics
]]>由Adelaide Health Technology Assessment（hOurHTA）主辦的線上會議「GLP-1療法真的與眾不同嗎？臨床、經濟與政策觀點」，邀集了三位專家從不同維度深入剖析GLP-1藥物（如Wegovy、Mounjaro）帶來的革命性影響與挑戰。這不僅是一場醫療科技評估的討論，更涉及未來全球醫療體系與健保政策的重新定義。
肥胖管理專科醫師Dr. Remon Eskandar強調，肥胖是一種慢性且漸進的生物學疾病，而非單純的生活方式失敗。
臨床數據顯示，GLP-1藥物不僅能帶來16%～22%的減重效果，更重要的是，即使在沒有減輕體重的情況下，它也能降低20%的心血管疾病（如心臟病、中風）風險。
它的本質是針對心臟、腎臟和血管的抗發炎藥物，未來的目標是預防洗腎、心臟衰竭等高成本的末期併發症。
應用健康經濟學家Chris Schilling點出了現實的殘酷：在澳洲，目前已有超過25萬人自費使用GLP-1。若將此推廣至100萬名符合條件的病人，每年可能需要耗費50億澳幣，這相當於澳洲藥品給付計畫（PBS）每年總預算的25%。
儘管預防末期併合症能省錢，但在藥物高昂的單價與需要終身服用的前提下，目前尚需發展「疾病專屬決策模型」，來驗證未來10到20年的長期成本效益。這正是醫療科技評估在新興藥物評估上面臨的核心課題。
健康經濟與政策專家Prof. Andrew Mitchell提出了政策面的巨大挑戰：GLP-1的潛在適應症跨度極大，部分甚至遊走在醫療與美容的灰色地帶。
澳洲政府不太可能為「美容減重」買單。面對龐大的潛在使用人口，健保體系必須嚴格把關，設立明確的客觀指標與持續用藥標準，將資源精準投資在能獲得最大經健康生活品質校正生命年（QALY）的高風險族群上（例如：已有心血管病史的次級預防）。
會中的交流探討了一個關鍵痛點：雖然臨床試驗的停藥率僅4～5%，但真實世界中卻有高達15～50%的病人在一年內停藥並面臨復胖。這背後反映的可能是「藥物短缺」與「長期自費負擔過高」的雙重困境。
GLP-1絕對是極具潛力的突破性療法，但要將其納入常規的健保或公共醫療給付，我們仍需要：
面對這個可能改變全球醫療體系的藥物，我們如何在「病人健康、醫療創新與健保財政」之間取得平衡？這不僅是醫療科技評估的議題，更是21世紀公共衛生政策的關鍵課題。
English version below
An online conference organized by Adelaide Health Technology Assessment (hOurHTA), titled ‘Is GLP-1 Therapy Really Different? Clinical, Economic, and Policy Perspectives,’ brought together three experts to examine from different angles the revolutionary impact and challenges brought by GLP-1 drugs (such as Wegovy and Mounjaro). This is not merely a discussion of health technology assessment, but touches on the future redefinition of global healthcare systems and reimbursement policies.
Dr. Remon Eskandar, an obesity management specialist, emphasizes that obesity is a chronic and progressive biological disease, not merely a lifestyle failure.
Clinical data shows that GLP-1 drugs not only deliver 16%–22% weight reduction, but more importantly, they can reduce the risk of cardiovascular diseases (such as heart disease and stroke) by 20% even without weight loss. The essential mechanism is an anti-inflammatory drug targeting the heart, kidneys, and blood vessels, with the future goal of preventing costly end-stage complications such as renal failure and heart failure.
Chris Schilling, an applied health economist, points out a harsh reality: in Australia, over 250,000 people currently self-pay for GLP-1. If expanded to 1 million eligible patients, the annual cost could reach 5 billion AUD, equivalent to 25% of the annual budget of the Pharmaceutical Benefits Scheme (PBS).
Although preventing end-stage complications can save money, given the high drug cost and the need for lifelong treatment, we currently need to develop ‘disease-specific decision models’ to verify long-term cost-effectiveness over the next 10 to 20 years. This is a core challenge in health technology assessment when evaluating emerging drugs.
Prof. Andrew Mitchell, a health economics and policy expert, points out a major policy challenge: GLP-1 has a vast potential scope of indications, some of which even blur the line between medical and cosmetic purposes.
The Australian government is unlikely to fund ‘cosmetic weight reduction.’ Facing a massive potential user population, the healthcare system must strictly control access by establishing clear objective criteria and continuing treatment standards, directing resources precisely toward high-risk populations that can achieve the maximum quality-adjusted life years (QALY)—for example, secondary prevention in patients with established cardiovascular disease history.
Discussions at the conference explored a critical pain point: although clinical trial discontinuation rates are only 4–5%, real-world data shows 15–50% of patients discontinue within one year and face weight regain. This may reflect a dual challenge of ‘drug shortages’ and ‘high long-term out-of-pocket costs.’
GLP-1 is undoubtedly a promising breakthrough therapy, but to integrate it into routine public healthcare or insurance reimbursement, we still need:
Facing a drug that could reshape the global healthcare system, how do we balance ‘patient health, medical innovation, and healthcare finances’? This is not only a health technology assessment issue but a key question in 21st-century public health policy.
]]>我實在很不擅長身處在這喧鬧吵雜的世界。特別是當外界充滿歡愉，就像這整台車的酸臭雄性，參加完 #ITZY 兩天週末演唱會的歡喜回程，但我卻充滿著悲傷到無意自拔的憤恨。
我就像夾在塞入垃圾車的一包又乾又濕的垃圾，而我也被一而再再而三地被 #借錢。
一開始是 3000，這一次 5000，後來變 7000。每一次都有理由——
前面是印尼同鄉得癌回國，她要幫忙。這一次是她門齒假牙換了，要換新的。但她不要看醫生用健保，卻要自己找自己同鄉認識的人搞，還不讓我跟我媽講。
通常健康相關的事情，我媽會同意，更會主動地負擔。但她還早就已經跟我媽借了六萬，因為她在印尼的女兒眼睛長瘤要開刀。
我已經說過了，這件事情私底下我媽是會直接轟人出門的，叫她滾的。我原則先說了：借錢可以，先還了前一筆，再借不難。在加上她金額越來越高，所以我加了第二原則——我不會借超過一萬。因為我不想要讓事情越來越複雜。
我自以為她下個月和我媽借的六萬就會還款還完了，所以我自以為她到時候可能就會比較有餘裕了。所以這一次我才會答應借給她。
但沒關係，我以為我借了，我可以換來更好的她對我父母的服務……沒有。她請假，臨時請假。看準我在家裡，讓她可以去忙她要做的假牙，甚至可以不用跟我說，然後是跟我媽。
我真的有夠不爽，但我還能夠如何？
她就算了，我爸更不給任何人面子。他失智好了，但他還能吃，但他不能喝。他失智算了，但他還能走，但他隨時眼睛閉著。我不知道他要怎樣走進廁所。
今天下午我叫他起床，他坐起。我要他在床邊上廁所尿尿，他起立了，雙手扶著桌子。我抽出已經濕透的護墊，我拿了水桶要督過去，但它鳥不抓，任尿隨便流，隨便滴。整個外面的尿褲都要被他滴濕透了。
我滿手都是尿。我不停說：「手抓起來，你的鳥，扶好尿。」但他只顧著扶桌子，根本不理我。我真的非常抓狂與憤怒。
我只能夠自己心想，至少他不是拉屎排便。
之後坐在床上喝水。一口水，不到 10cc，他吞不下去，含不下去。我已經喊了，叫了，超過半小時，他一口都吞不下去。最後我叫他吐在尿桶裡面。
我超級憤怒，我超級生氣。我就想要去死。
而這時候生命的荒唐就在——外籍看護回來了。
這兩面夾擊還好嗎？沒有，還有一面我媽的巨牆。
她可以對我爸又打又罵，讓我以為她就是在洩憤，但不准任何人代勞。更同樣視子女如糞土。
兩週前我回高雄，已經跟她大吵一架。她除了不停地哭、情勒，就好像全世界她最委屈，更又要遷怒。當然我爸是主要的源頭。又要把我爸給趕出去，然後同樣的戲碼繼續上演。
「你給我滾出去，你不准幫你爸，你出去給人幹，你給我當狗做！」
我真的沒有要什麼。我只想要死。
因為在你們之間，我應該連狗都不如。只能夠被你們一個接著一個，把我當垃圾般，推進那不停滾動的垃圾車裡面。
死可能更簡單。這一切都沒有任何意義。
English version below
I have never been comfortable existing in this loud and chaotic world. Especially when the outside is filled with joy—like this entire car of people returning happily from the two-day #ITZY concert weekend—yet I am filled with sadness and helpless rage.
I am like a bundle of wet and dry garbage squeezed into a garbage truck, and I am being asked to lend money repeatedly.
First it was 3,000, then 5,000, then 7,000. Each time there is a reason.
Before, an Indonesian acquaintance got cancer and returned home, and she needed to help. This time, her front tooth dentures needed replacing. But she doesn’t want to see a doctor using health insurance. Instead, she wants to find someone she knows from her hometown to handle it, and she won’t let me tell my mother.
Usually, when it comes to health-related matters, my mother would agree and even take it upon herself to pay. Yet she has already borrowed 60,000 from my mother because her daughter in Indonesia has a tumor in her eye that requires surgery.
I already said: borrowing is fine, but pay back the previous amount first, then it’s not difficult to borrow again. Because the amounts keep increasing, I added a second principle—I won’t lend more than 10,000. Because I don’t want things to become increasingly complicated.
I thought that next month when the 60,000 she borrowed from my mother would be fully repaid, she might have more room. So that’s why I agreed to lend to her this time.
But never mind. I thought that by lending, I could receive better service from her toward my parents. Nothing. She took time off—sudden leave. She saw that I was at home, so she could take care of her dentures, and she didn’t even have to tell me, let alone my mother.
I’m really upset, but what else can I do?
As for her, forget it. My father is even more disrespectful to anyone. He has dementia, yes, but he can eat. However, he cannot drink. He has dementia, yes, but he can walk. Yet his eyes are closed all the time. I don’t know how he is supposed to walk into the bathroom.
This afternoon I told him to get up. He sat up. I asked him to use the toilet on the edge of the bed to urinate. He stood up, holding onto the table with both hands. I pulled out the soaking wet pad. I grabbed a bucket to push over, but he doesn’t grip it properly. His urine just flows freely, dripping everywhere. His entire wet pants will be soaked through.
My hands are covered in urine. I kept saying, “Hold it with your hands, hold your penis, hold it so it doesn’t drip.” But he just keeps holding onto the table, not listening to me at all. I’m really going crazy and furious.
I can only think to myself, at least he’s not defecating.
Later, he sat on the bed drinking water. One sip, less than 10cc, and he can’t swallow, can’t hold it in his mouth. I have already shouted, called out, for over half an hour, and he can’t swallow even one sip. Finally, I told him to spit it into the urinal.
I’m super angry. I’m super furious. I just want to die.
And at that very moment, the absurdity of life unfolds—the foreign caregiver comes home.
Are these two sides pressing enough? No, there is also one side—my mother’s giant wall.
She can hit and scold my father, making me think she is just venting, but she won’t allow anyone else to help. She treats her children the same way—like dirt.
Two weeks ago I went back to Kaohsiung and had a big fight with her. All she does is cry and use emotional manipulation, as if she is the most wronged person in the world. She even redirects her anger. Of course, my father is the main source. She wants to kick my father out, and then the same drama repeats.
“Get out of my sight! You are not allowed to help your father! Go out and let people use you! Be my dog!”
I really don’t want anything. I just want to die.
Because between you all, I should be worth less than a dog. I can only be pushed by you, one after another, into that ever-rolling garbage truck like trash.
Death might be simpler. None of this has any meaning.
]]>在這個特殊的日子裡，我用一整天的行程來標記我的50歲生日。這一天對我來說，不僅是年齡的刻度，更是一場深刻的生命反思之旅——從早上七點半開始，我按照自己的意思安排每一個時刻，而這些選擇最終導向了一個關於生命本身的結論。
這一天從 M One Cafe 大安開始。在那裡，我親眼目睹了一段安靜的時光——一對年輕男女相隔將近一小時同桌，然後只剩半小時吃東西，沒有交談。這樣的片段觀察，成了整個生日的起點。
接著我前往 Muvie 松仁 Mucrown 看電影 ‘超少女’。這部作品講述了一個星球爆炸後，少女需要傳承活下來、前往地球成為英雄的故事。在虛幻的世界裡，我看見了一種被迫而來的責任。
在北師美術館，我參觀了 ‘遙視’ 展覽。這個展覽讓我感到莫名其妙，不知所云，跨越二樓、三樓與一樓的空間，卻始終無法完全理解。
師大美術館的 ‘共生之華’ 展覽則帶給我不同的體驗。這個展覽讓我看到感動，看到日治或日據前後台日藝術的師承脈絡，是少數如此密集又具深度的優秀展覽。
午後回到汀州稍做梳洗後，我去 Pluus三越前讓敬愛的 Babie 老師修毛。
在 ‘屯屯’ 這間書店兼咖啡店，我點的小食讓我足足等了半小時以上。這家店以水餃聞名，但水餃都破了。
這一天最關鍵的時刻，是參加報導者的新書發表會，主題是 ‘生命訂製時代’。
發表會涵蓋了 ‘人工’、’代孕’、’第三生殖’ 等核心議題。但我必須坦誠地說，除非與會者具有部分相關知識或了解，否則直接從 Dcard 切入的敘事，確實讓人缺乏框架以完整理解這些議題。
然而，最觸發我深刻思考的，是現場四位報導者作者們的一個問答：在整個產業鏈或生命週期中，最被忽略或忽視權利的是誰？
他們的答案都指向同一個群體：根本沒得決定的『孩子』。
聽到這個答案的瞬間，我感到非常不爽。特別是在這樣的背景和脈絡之下，如果各位都認為被生下來的孩子是最被忽略的，那就代表著一個根本的事實：
根本沒有小孩應該被生下來。無論你是生命訂製或不訂製的，當你無法徵求那要被生下者的權益時，你根本沒有資格決定你要不要生。
我在06/24我生日的這一天，得出了這樣的結論。我根本就不想被生下來。從來沒有人問過我的意見，就好像生下來，其他的就是我自己的責任似的。但我的認知很簡單：我沒有要，那就連從頭都不要開始。
我不能夠說，這一天對所有人的意義是如何，但至少今天我照著我自己的意思，請了特休，感覺一點都不浪費。儘管如此，還是有人不停的聯絡、分案和分配工作。
沒關係。這一切，我認為都不需要存在。因為只要我消失了，我就不用有任何不適當的想法了。
謝謝大家。祝我自己生日快樂，希望大家都能夠繼續快樂下去。而我會在有限的50結束。讓我們繼續看下去……就當作是喜事般的慶祝吧！
English version below
On this special day, I structured my entire day as a marker of my 50th birthday. This date represents more than just another year of age—it became a profound journey of life reflection. Starting at 7:30 in the morning, I arranged each moment according to my own will, and these choices ultimately led me to a fundamental conclusion about life itself.
The day began at M One Cafe in Da’an. There, I witnessed a quiet moment unfold: a young couple seated at the same table for nearly an hour, then spending only half an hour eating with no conversation. This quiet observation became the opening note of my entire birthday.
Next, I went to Muvie Songrenjiong Mucrown to watch the film ‘Super Girl.’ This film depicts a planet’s explosion, after which a girl must inherit and carry on life, traveling to Earth to become a hero. In this fictional world, I saw a sense of responsibility imposed by circumstance.
At the Taipei Normal University Museum of Fine Arts Beishi campus, I visited the exhibition ‘Remote Viewing.’ This exhibition left me feeling bewildered and mystified, spanning the second floor, third floor, and basement, yet I could never fully comprehend it.
The ‘Symbiosis of Blooms’ exhibition at the Taipei Normal University Museum of Fine Arts offered a different experience. This exhibition moved me deeply, revealing the lineage of Japanese and Taiwanese art before and after Japanese colonial rule. It was one of those rare exhibitions of remarkable depth and density.
After briefly returning to Tingzhou to freshen up, I visited Pluus in front of Mitsukoshi to have my hair groomed by Teacher Babie, whom I deeply respect.
At ‘Tun Tun,’ a bookstore-café hybrid, the small food I ordered required waiting over half an hour. This establishment is known for its dumplings, though they all broke.
The most critical moment of the day came when I attended The Reporter’s book launch event on the theme of ‘The Age of Designed Life.’
The event covered core topics including ‘Artificial [reproduction],’ ‘Surrogacy,’ and ‘Third-party reproduction.’ I must be candid: unless attendees possessed some prior knowledge or understanding of these matters, the narrative approach drawn directly from Dcard left people without an adequate framework for comprehensive understanding.
However, what triggered my deepest reflection was a specific Q&A from the four author panelists: In the entire industry chain or lifecycle, whose rights are most overlooked or ignored?
Their unanimous answer pointed to a single group: the ‘child’ who has no say in the matter at all.
Upon hearing this answer, I felt profound displeasure. Particularly within this context and framework, if all the panelists acknowledged that the born child is the most overlooked party, then it reveals a fundamental truth:
No child should ever be brought into existence. Whether through designed life or undesigned birth, when you cannot seek the consent of the one being born, you have no right to decide whether to procreate.
On June 24th, my birthday, I reached this conclusion. I fundamentally did not want to be born. No one ever asked my opinion—as if being born meant everything afterward was simply my own responsibility. But my logic is simple: since I never consented, the entire enterprise should never have begun.
I cannot speak to what this day means for everyone else, but today I arranged things according to my own will, taking personal leave without feeling it was wasted. Yet people continued to contact me, reassign tasks, and delegate work.
It doesn’t matter. I believe all of this need not exist. Because if I simply disappeared, I would no longer need to harbor any inappropriate thoughts.
Thank you all. Happy birthday to myself. I hope everyone can continue to find happiness. As for me, I will end at fifty. Let us continue and see what comes next…let us celebrate this as if it were a joyous occasion!
]]>最近看完了Netflix的韓劇《鐵拳教育》（韓語：참교육，英語：Teach You a Lesson），這部作品在國內引發了不少討論與批評。不過，我想先說明一點——判斷這部劇的價值，不該只看第一集就下結論，甚至更重要的是，批評者是否有把整部劇看完。唯有真正看完全劇，才能更準確地評估這些批評究竟從何而來，又想把人帶往何處。
我認為這部劇值得觀看。更重要的是，我不認為它像表面看起來那樣淺薄，單純用幾巴掌的鐵拳就能解決問題——現實絕對沒有那麼簡單。
這部劇把許多現代社會與教育現場的問題搬上了檯面。雖然採用了架空、誇張或激進的表現方式，但涵蓋的議題卻非常多元，包括：
這些議題的龐雜與深度，足以證明這部劇的企圖心並非只停留在表面懲罰。
在觀看許多劇情時，我的心情異常沸騰，特別是在家庭與家長教育的情節。這些內容觸及了我至今仍記憶且感受到的苦痛。
儘管劇情高潮迭起，但我能肯定的是：每段劇情主題的後段或延伸，都可以看到一個重要的核心——哪怕是鐵拳的痛擊，教育的本質還是被有意識地帶回舞台的中心。
這部劇中的教權局展現了一套邏輯：
我可能無法認同有人用「以牙還牙、以眼還眼」來簡化這部劇的核心理念。劇中的懲罰制度，是在嘗試建立一套有邏輯、有原則的教育介入方式，而不是單純的報復。
透過社會勞動的陪同，以及更深入地挖掘學生為何會做出這些行為的過去，劇作展現了另一層意義——真正的教育改變，需要理解根源，而非只是懲罰表象。
這部劇對我最大的敘事啟示是：無論如何，改變都需要費盡心力。
在教育環境不穩定的場域中，無論涉及哪一種年齡的孩子、哪一類的家長、怎樣的教職人員，乃至社會、政治與國家層面，教育都需要每一個人勇敢地大聲呼喊出心中的訴求與願景。
我不認為道理需要很多，但至少大家要走在正直良善的道路之上。唯有這樣，教育的改革才有可能向前推進。
這部劇或許無法直接改變現實，也可能會招致批評，但它提出的問題、展現的多元議題，以及對教育本質的執著關注，都值得我們認真對待。真正的評價，應該建立在完整觀看與深思的基礎之上。
English version below
I recently finished watching Netflix’s Korean drama “Teach You a Lesson” (Korean: 참교육, English: Teach You a Lesson). This series has sparked considerable discussion and criticism domestically. However, I want to make one point clear—judging the value of this drama should not be based on watching only the first episode. More importantly, one should consider whether critics have actually watched the entire series through. Only by completing the full drama can we accurately assess where these criticisms come from and where they aim to lead.
I think this drama is worth watching. More significantly, I don’t believe it’s as shallow as it appears on the surface, with a few iron fist blows solving all problems—reality is definitely far more complex.
This drama brings many problems from modern society and the education field onto the screen. Although it uses fictional, exaggerated, or radical narrative methods, the issues it covers are remarkably diverse, including:
The complexity and depth of these issues prove that the drama’s ambition extends far beyond surface-level punishment.
While watching many scenes, my emotions were extraordinarily turbulent, especially during episodes about family and parental education. These scenes touched upon pain I still remember and feel to this day.
Despite the drama’s climactic moments, I can affirm one certainty: in every scene’s conclusion or extension, we can see an important core—even through iron fist blows, the essence of education is consciously brought back to center stage.
The Educational Authority in this drama presents a logic:
I may not agree with those who oversimplify the drama’s core message as “an eye for an eye, a tooth for a tooth.” The punishment system depicted in the drama attempts to establish a logical and principled approach to educational intervention, not mere retaliation.
Through accompanying community service and deeper exploration of why students behaved this way in their past, the drama reveals another layer of meaning—true educational change requires understanding root causes, not just punishing symptoms.
The most significant narrative insight this drama offers me is: regardless of circumstances, change requires tremendous effort.
In an unstable educational environment, whether involving children of any age, parents of any kind, teachers of any caliber, or issues at societal, political, and national levels, education requires every person to bravely voice their own demands and vision for the future.
I don’t believe we need many arguments, but at least everyone must walk the path of integrity and goodness. Only through this can educational reform move forward.
This drama may not directly change reality, and it may well face criticism, but the questions it raises, the diverse issues it presents, and its steadfast focus on the essence of education are all worth our serious consideration. True appreciation should be built upon complete viewing and deep reflection.
]]>醫療科技評估人才培訓課程第二天（05/29），進一步深入臨床試驗解讀與存活曲線外推的專業領域。這些內容對於評估藥廠提交的實證資料，以及建立科學性的批判思考能力，都相當有幫助。
臺大臨床藥學研究所蕭斐元教授首先點出重點。她提醒我們，在評估新藥是否納入健保給付時，最核心的問題之一，在於「比較品」的選擇，以及臨床試驗收案族群與實際目標族群之間的差異。
許多創新療法的早期臨床試驗，缺乏與臺灣現行標準治療進行頭對頭直接比較；此外，試驗族群也可能未充分納入亞洲或臺灣病人。這些限制都會增加相對療效評估的不確定性，並進一步影響後續健保給付條件的訂定與成本效益估算。
國家衛生研究院蕭金福教授針對次族群分析，提出重要的統計學提醒。他強調，次族群分析的主要價值，在於檢視各次族群間的結果是否一致，以及產生新的研究假設，不能直接將其視為臨床試驗的最終結論。
他以著名的 ISIS-2 試驗為例。該試驗曾出現「雙子座與天秤座病人的治療效果較差」這類明顯缺乏臨床合理性的分析結果，藉此提醒大家，不能因整體臨床試驗未達預期結果，便在事後進行大量資料探勘（data dredging），再將偶然出現統計顯著差異的特定次族群，包裝成足以支持決策的試驗結論。
現場也有學員敏銳地提問，若不同治療暴露組別中的次族群人口學特徵分布不一致，應如何處理？這正呼應教授所提醒的重點：一旦原有的隨機分派平衡遭到破壞，分析結果通常只能用於觀察趨勢，難以據此作出具有充分統計依據的結論。
在評估長期成本效益時，我們經常需要將有限期間內觀察到的存活資料向後外推。針對這項議題，中國醫藥大學譚家惠助理教授提供了相當具有實務價值的存活外推評估原則。
她提醒評估研究人員，不應直接採信未經驗證的外推結果，也不能僅依賴 AIC 或 BIC 等統計指標選擇模型，還必須同時檢視視覺配適情形，並評估結果是否具有臨床合理性。
更重要的是，應透過外部驗證進行合理性檢查，將外推結果與臺灣一般人口生命表相互比對，確認模型預測的病人存活率，不會出現明顯高於一般人口的不合理情況。
此外，也不能預設新藥療效會終生維持，而應要求廠商評估療效衰退的可能性。
面對學員提問，對於宣稱「僅治療一次即可終生治癒」、但長期效果仍具有高度不確定性的創新療法，應如何評估其外推合理性？專家建議，除了可以考慮採用混合治癒模型外，也應設定不同療效維持年限與療效衰退情境，進行敏感度分析，以釐清各項不確定性對財務影響與健康效益的實際影響。
唯有具備扎實的證據評讀與品質評估能力，我們才能在創新醫療科技與有限健保資源之間，找到兼顧科學證據、臨床價值與資源配置的適當決策平衡點。
English version below
The second day of the HTA capacity building training program (May 29) provided an in-depth exploration of clinical trial interpretation and survival curve extrapolation—fields essential for evaluating pharmaceutical evidence submissions and developing critical evidence appraisal skills.
Professor Hsiao Fei-Yuan from the National Taiwan University Department of Clinical Pharmacy identified a critical starting point. She emphasized that when assessing whether new drugs should be included in national health insurance coverage, one of the core questions concerns the selection of comparators and discrepancies between trial enrollment populations and real-world target populations.
Many innovative therapies lack head-to-head direct comparisons with Taiwan’s current standard treatments in early-phase clinical trials; moreover, trial populations often insufficiently represent Asian or Taiwanese patients. These limitations increase uncertainty in relative efficacy assessment and subsequently affect reimbursement conditions and cost-effectiveness calculations.
Professor Hsiao Chin-Fu from the National Health Research Institutes highlighted critical concerns regarding subgroup analysis. He emphasized that the primary value of such analyses lies in examining consistency across subgroups and generating new research hypotheses—not in serving as definitive trial conclusions.
He illustrated this with the famous ISIS-2 trial, which reported that patients born under Gemini and Libra had poorer treatment outcomes—a finding lacking any clinical rationale. This serves as a stark reminder: when overall trial results disappoint, researchers cannot resort to extensive data dredging, then package statistically significant differences in isolated subgroups as evidence sufficient for policy decisions.
A perceptive participant asked how to handle cases where demographic distributions of subgroups differ across treatment groups. This question reinforced the professor’s warning: once the balance achieved through randomization is disrupted, analysis results can only suggest trends—they cannot support robust statistical conclusions.
When assessing long-term cost-effectiveness, we frequently must extrapolate observed survival data beyond the trial observation period. On this topic, Assistant Professor Tan Chia-Hui from China Medical University offered highly practical guidance.
She cautioned evaluators against directly accepting unvalidated extrapolation results, and against relying solely on statistical indices like AIC or BIC for model selection. Instead, extrapolation assessment must include visual fit inspection and clinical plausibility evaluation.
Crucially, external validation should be performed by comparing extrapolation results against Taiwan’s general population life tables, ensuring that predicted patient survival rates do not implausibly exceed general population figures.
Furthermore, one cannot assume new drug efficacy will persist indefinitely; manufacturers must be required to assess potential treatment effect waning.
When asked how to assess extrapolation reasonableness for innovative therapies claiming “single-administration lifelong cure” despite high long-term uncertainty, experts recommended considering cure-fraction models and conducting sensitivity analyses across scenarios with varying efficacy duration and treatment effect waning patterns. This approach helps clarify how different sources of uncertainty impact financial impact and health benefits.
Only with solid evidence appraisal and quality assessment capabilities can we navigate the space between innovative health technologies and finite health system resources, achieving decisions that respect scientific evidence, clinical value, and judicious resource stewardship.
]]>