跳到正文
制药经理人· GxP知识岛·· 18 天前AI 评分68

ISPE刊文讨论GxP项目AI系统第一个小时如何界定验证力度

AI系统,如何验证?

AI 导读

ISPE《制药工程》iSpeak栏目刊出Sachin Bhandari的文章,主张GxP项目AI系统的验证力度应随患者风险而非算法复杂度确定。文章以两家企业验证同类色谱图标记系统的六个月与八周投入作对照,提出四个角色用一小时做初始评估、对假阳性与假阴性各问一遍并把最坏情形推理写入验证计划的做法。

正文

ISPE官网《制药工程》(Pharmaceutical Engineering)栏目10月30日发了一篇ispeak文章,《The First Hour of an AI Project in GxP》,作者Sachin Bhandari。

文章开头是一个对照案例:两家公司同一个月里评估几乎相同的AI系统(给色谱图打标记、提示检验员复核),一家按高风险做了六个月验证,一家八周完成。作者的观点是:差距不在系统,在项目第一个小时怎么谈——验证力度跟着患者风险走,不跟算法复杂度走。四个角色、一个问题问两遍、把推理写进验证计划,这是作者给的方法。全文对照如下,英文原文在上,中文译文在下。

The First Hour of an AI Project in GxP

GxP项目里AI系统的第一个小时

Sachin Bhandari | Pharmaceutical En6gineering | 30 October 2026

Sachin Bhandari|ISPE《制药工程》|2026年10月30日

Two firms assessed much the same system in the same month. Both had bought a model that flags chromatograms for analyst review. The first firm treated it as high risk, and it is understandable why: AI, laboratory data, and batch decisions somewhere downstream. They budgeted six months of validation, wrote scripted tests against every function, and built a monitoring program fit for a system that acts on its own. The second firm started with the workflow. A qualified analyst reviews every flag before anything touches a batch record, the analyst can override any flag, and the override is logged. They validated in eight weeks.

同一个月里,两家公司评估了几乎相同的系统:都采购了一个给色谱图打标记、提示检验员复核的模型。第一家把它当成高风险——这可以理解:AI、实验室数据、下游某处的批决策。他们按六个月验证做预算,针对每个功能写脚本化测试,还建了一套适合"自主运行系统"的监控方案。第二家从工作流入手:任何东西进入批记录之前,先由一名有资质的检验员复核每一个标记,检验员可以否决任何标记,否决动作留有记录。他们八周完成了验证。

It’s easy to read that as a story about human review making a system safe. The more useful lesson sits one step further in. The analyst sees the flags. The analyst doesn’t see the chromatograms the model passed over. A false positive costs a few minutes, because a qualified person looks at it and sets it aside. A false negative is a peak the model should have flagged and didn’t, and if nothing else looks at that result, the error can travel toward a batch decision and, in the end, a patient. Review of the flags protects against one kind of error and can be blind to the other.

这很容易被读成一个"人工复核让系统变安全"的故事。更有用的教训在再往里一步。检验员看得见标记,看不见模型放过的那些色谱图。假阳性只花几分钟——有资质的人看一眼就放下了。假阴性则是模型本该标记却没有标记的峰,如果没有其他环节再看那个结果,错误就会一路走向批决策,最终走向患者。复核标记防住了一类错误,却可能对另一类视而不见。

So the question that decides whether eight weeks is enough is what happens to the results the model doesn’t flag. If they still go through the lab’s normal second-person review, a miss falls back on a control that already exists. If the model’s silence means no one looks again, the model is choosing which results get scrutiny, and the risk to product quality is a different size. At the second firm, every unflagged chromatogram still went through the lab’s normal second-person review, so that is where a miss would be caught. The time the second firm saved came from asking questions like that and putting the testing effort where a miss could do harm: scripted tests on the miss rate against acceptance criteria agreed up front, and unscripted testing around the edge cases. ISPE GAMP 5 Guide (Second Edition) calls this critical thinking (Appendix M12), and the ISPE GAMP Guide: Artificial Intelligence applies it to AI in Appendix M4.

所以,决定八周够不够的问题,是模型没有标记的结果去了哪里。如果它们仍走实验室常规的第二人复核,漏报就落回到一个已经存在的控制上。如果模型的沉默意味着没人再看,那就是模型在替你挑哪些结果值得细看,对产品质量的风险就是另一个量级。第二家公司里,每一个未标记的色谱图仍然走实验室常规的第二人复核,漏报会在那里被拦下。第二家省下的时间,来自把这类问题问清楚,并把测试力气花在漏报可能造成伤害的地方:对漏报率做脚本化测试、验收标准事先商定,对边界情形做非脚本化测试。《ISPE GAMP 5指南(第二版)》把这叫批判性思维(附录M12),《ISPE GAMP指南:人工智能》在附录M4把它用到AI上。

Both approaches could be defended. One of them spent four extra months and a good part of its validation budget on effort the risk didn’t call for, and that’s budget and expert time taken from systems where a patient could be hurt. I’ve watched the first hour of an AI project shape what follows more often than any protocol or vendor claim. When the conversation goes well, the later decisions line up behind it: how deep the testing goes, how closely the system is monitored, how changes are controlled, and who signs. When it goes badly, a team builds a six-month file on the wrong foundation, or worse, an eight-week file on a system that needed the six months.

两种做法都说得过去。但其中一家多花了四个月,把验证预算的一大块用在了风险并不要求的力气上——这些预算和专家时间,本可以用在那些真正可能伤到患者的系统上。我见过太多次:一个AI项目的第一个小时,比任何方案或供应商承诺都更能决定后面的一切。谈话谈得好,后面的决策会跟着排好队:测试做多深、系统盯多紧、变更怎么控、谁来签字。谈得不好,团队要么在错误的地基上写出一份六个月的文件,要么更糟——在一个本需要六个月的系统上,写出一份八周的文件。

Start from What Happens to the Patient When the Model Is Wrong

从"模型错了,患者会怎样"出发

When a team meets its first AI system, the validation effort usually gets scoped from the technology. The model is complex, so the file must be heavy. It sounds prudent and it’s backwards, and everyone raised on GAMP already knows why: assurance effort follows risk to patient safety, product quality and data integrity, judged in the system’s actual context of use. ICH Q9(R1) says the same in its own words, linking the evaluation of risk to the protection of the patient and matching the level of effort to the level of risk. The complexity of an algorithm tells a team how hard the system is to build. It says very little about how much assurance the system needs.

团队第一次遇到AI系统时,验证力度通常按技术来定:模型复杂,文件就得厚。听着谨慎,其实方向反了——在GAMP体系下长大的人都知道为什么:保证的力度跟着对患者安全、产品质量和数据完整性的风险走,并且要在系统实际的使用场景里判断。ICH Q9(R1)用它的语言说了同一件事:把风险评估与患者保护挂钩,让努力程度与风险等级匹配。算法的复杂度告诉团队系统有多难造,却几乎说明不了系统需要多少保证。

Take a pair of examples. A neural network that summarizes deviation trends for a monthly leadership review is technically sophisticated and almost harmless in operation: if it’s wrong, a chart misleads a meeting until someone checks the numbers. A plain decision rule that routes out-of-specification (OOS) investigations is technically trivial and serious in operation: if it’s wrong, investigations go down the wrong path, the records they produce inherit the error, and a batch that should have been held can move on. Scope by complexity and the first system gets heavy validation while the second is waved through. Scope by consequence and both get what they need.

举一对例子。一个为月度管理层评审汇总偏差趋势的神经网络,技术上很复杂,运行中几乎无害:它错了,无非一张图误导一场会,直到有人去核对数字。一个给OOS(超标)调查分派路径的简单决策规则,技术上平平无奇,运行中却很要紧:它错了,调查就会走错方向,产生的记录继承这个错误,本该被扣住的批可能继续往前走。按复杂度定范围,第一个系统会被重验,第二个会被挥手放行。按后果定范围,两个系统都各得其所。

Who Needs to Be in the Room for That First Hour

第一个小时,谁该在房间里

If the first hour shapes so much, it deserves real design, and the most common failure is letting one function decide alone. A data scientist working alone reads the model and under-weights the workflow. A quality assurance (QA) reviewer working alone reads the worst case everywhere. A business owner working alone reads the benefit case. The initial assessment needs four people in one room for one structured hour: the process owner who knows what the system does in the workflow, the data scientist who knows what the model can and can’t do, the QA reviewer who knows what an error would mean for product and patient, and the validation lead who carries the lifecycle consequences of what the room decides.

如果第一个小时能定这么多事,它就值得认真设计。最常见的失败,是让一个职能单独拍板。数据科学家独自看,看的是模型,低估了工作流。QA审评人独自看,处处都是最坏情形。业务负责人独自看,看的是收益。初始评估需要四个人在同一间房里度过一个结构化的小时:知道系统在工作流里干什么的过程负责人,知道模型能与不能什么的数据科学家,知道出错对产品和患者意味着什么的QA审评人,以及要为房间里所有决定承担生命周期后果的验证负责人。

They need an agreed process scope and objective before they sit down, so they’re all assessing the same thing. And they need to reach a shared view of the worst case, because that view informs a long list of decisions later in the project and in operation.

坐下之前,他们需要先就过程范围和目标达成一致,这样四个人评估的才是同一件事。他们还需要就最坏情形形成共同看法——这个看法会决定项目后续和运行期的一长串决策。

A Question That Gets the Conversation Going

一个能打开话匣子的问题

When the room stalls, and it will, one question helps: what happens to patients and product if this system is wrong on 10 percent of records for six months, without anyone noticing? The numbers are deliberately arbitrary. Ten percent is big enough to matter and small enough that no single-record alarm will catch it, and six months is long enough for an error to spread into decisions. For some uses, a team should change them, because a model that screens for rare events can do real harm at a far lower error rate. The question is there to start a discussion, and what comes out of it is a list of what the team needs to do.

房间冷场时——一定会冷场——有一个问题好用:如果这个系统在10%的记录上出错、持续六个月、而没有任何人察觉,患者和产品会怎样?两个数字是故意随便定的。10%大到足以造成影响,又小到任何单条记录的报警都抓不住它;六个月长得足以让一个错误扩散进决策。有些用途应该换数字——筛罕见事件的模型,错误率低得多也能造成实际伤害。这个问题的作用是启动讨论,讨论的产出是一份"团队需要做什么"的清单。

Ask it twice, once for each kind of error. Too many false alarms and reviewers start to tire, which matters because a tired reviewer is the control the system relies on. Too many misses and the question becomes who, if anyone, looks at what the model didn’t flag. Then follow the answers to the controls.

这个问题要问两遍,两类错误各问一遍。假警报太多,复核人会开始疲劳——这很要紧,因为疲劳的复核人正是系统赖以运转的那个控制。漏报太多,问题就变成:还有没有人看模型没标记的东西?然后,顺着答案去找控制措施。

A misleading trend line in a monthly meeting points to a periodic plausibility check on a sample of outputs. A stream of misclassified deviations points to the reviewer: if the reviewer caught them, the override log should show it, and if the reviewer approved them by reflex, the human oversight was never there. Months of misrouted investigations quietly corrupting corrective and preventive action records across sites point to acceptance criteria broken down by site and product family, and to monitoring that starts on the day of go-live. A wrong decision that could release product to market raises a harder question, which is whether AI belongs in that decision at all, and with what human oversight. Draft EU GMP Annex 22 is cautious on exactly this point for critical applications.

月度会上一条误导性的趋势线,指向对输出抽样做定期合理性核查。一批被错误分类的偏差,指向复核人:复核人若真拦过,否决日志里应当有记录;复核人若是条件反射地点了同意,那个人监督从来就不存在。数月间被分错路径的调查悄悄污染了跨场站的CAPA记录,指向按场站和产品族拆分的验收标准,以及上线当天就启动的监控。一个可能把产品放进市场的错误决策,引出一个更难的问题:那个决策到底该不该交给AI,以及要配什么样的人工监督。欧盟GMP附录22草案对关键应用的这一点,态度正是谨慎。

Keep the Harm and the Controls Apart on the Record

在记录里,把危害和控制分开写

It’s tempting to let the controls answer the risk question. "A qualified person reviews every recommendation" describes the control design. The harm a wrong output could do to a patient, to product quality or to data integrity stays the same whatever review sits around it. What the controls change is how likely that harm is to get through, and how quickly someone would notice.

用控制措施回答风险问题,是很有诱惑力的。"有资质人员复核每一条建议"描述的是控制设计。错误输出对患者、产品质量或数据完整性可能造成的危害,不因围绕它的复核而改变。控制改变的是危害穿透的可能性有多大、多快会被人发现。

Writing the two down separately matters because controls weaken in operation in ways the harm doesn’t. When the override rate drifts toward zero, look closer before celebrating, because it can mean the reviewers have stopped reviewing. A simple test of any human review claim is to ask for the last time a reviewer disagreed with the system, and to check the records can show it. The exception path deserves the same test. If overriding the system’s routing takes four screens and a supervisor password while accepting it takes one click, the exception path exists on paper and the oversight is weaker than the file says.

把两者分开写下来很重要,因为控制会在运行中弱化,危害不会。当否决率漂向零,先别急着庆祝,仔细看看——那可能意味着复核人已经不复核了。检验任何"人工复核"声明的一个简单办法:要来最近一次复核人不认同系统的时间,并核实记录真的能证明。例外路径值得做同样的测试。如果否决系统的分派要过四个界面外加主管密码,而接受只要点一下,例外路径就只存在于纸面上,监督比文件写的弱。

The GAMP AI Guide Already Gives the Room a Common Language

GAMP AI指南已经给了房间里一种共同语言

Section 21.2 of the ISPE GAMP Guide: Artificial Intelligence sets out a classification of AI sub-system autonomy, and it’s the right shared language for how far a system acts on its own in the workflow. The first hour builds on it with what only the people in the room know: the specific harm in this process, how it could reach a patient, and which controls the team is relying on to stop it.

《ISPE GAMP指南:人工智能》第21.2节给出了AI子系统自主性的分类,这是描述"系统在工作流里能自己走多远"的合适共同语言。第一个小时在它之上,加上只有房间里的人才知道的东西:这个过程里具体的危害是什么、它可能怎样到达患者、团队指望哪些控制去拦住它。

The same Guide asks a second question alongside autonomy, which is whose data established the model. A model trained entirely on a firm’s own data sits under the firm’s own change control. A vendor’s pre-trained model, which the firm can’t retrain or inspect, changes when the vendor decides. A system with modest consequences built on a vendor model still carries real change-control and supplier-oversight obligations, because the thing that changes its behavior lives outside the firm’s quality system. That’s the half of the assessment leaders miss most often, and it’s the half the supplier audit depends on.

同一份指南在自主性之外问了第二个问题:模型是用谁的数据建立的。完全用自己数据训练的模型,归自己公司的变更控制管。供应商的预训练模型,公司既不能重新训练也无法检查,它什么时候变由供应商决定。一个后果有限的系统,如果建在供应商模型上,仍然背着实打实的变更控制和供应商监督义务——因为改变它行为的东西,在公司质量体系之外。这是评估中最常被领导者漏掉的一半,也是供应商审计赖以成立的一半。

Write the Decision into the Validation Plan

把决定写进验证计划

A decision that lives only in meeting minutes is hard for anyone to follow later. The validation plan is the right home for it, with the reasoning alongside. A short section is enough: the system and its intended use stated in workflow terms, the worst case in each direction and how it could reach a patient, the controls the team is relying on, the names of the four people, and the triggers that would reopen the discussion. When someone later asks how the AI was validated, the answer starts there, and it’s a strong start because it shows the thinking as well as the conclusion.

只活在会议纪要里的决定,事后没人跟得下去。验证计划才是它的家,推理过程要跟在旁边。一小节就够:用工作流语言陈述的系统及其预期用途、两个方向各自的最坏情形及其到达患者的路径、团队依赖的控制、四个人的名字,以及会重新打开讨论的触发条件。以后有人问"这个AI是怎么验证的",答案就从这里开始——这是个有力的开头,因为它同时展示了思考和结论。

图1 初始评估的一小时:四个角色,两类错误各问一遍的那个问题,以及写进验证计划的决定。/Figure 1. The initial assessment hour: four roles, one question asked for each kind of error, and the decision recorded in the validation plan.

Where the First Hour Usually Goes Wrong

第一个小时通常在哪里出错

The assessment looks simple on a slide. In practice a handful of patterns cause most of the trouble, and each one has a working countermeasure.

这套评估在幻灯片上看着简单。实践中,少数几种模式惹出了大部分麻烦,每一种都有能用的对策。

表1 初始评估的常见陷阱与规避方法。/Table 1. Common pitfalls in the initial assessment and the ways to avoid them.

Pitfall 陷阱

How it shows up 典型表现

How to avoid it 规避方法

Over-caution
过度谨慎

A dashboard built for awareness gets the full treatment because caution goes unchallenged. A six-month program runs against a six-week system, and the next proposal arrives arguing for less effort on cost grounds.
一个只为知情而建的数据看板被全套对待,因为没人质疑那份谨慎。六个月的验证方案跑在一个六周就能建好的系统上,下一个提案就来了,开口就是"为了省钱,验证能不能少做点"。

State in the procedure that effort follows risk in both directions. Taking effort away from a low-consequence system is as much a quality decision as adding it to a high-consequence one.
在程序文件里写明:力度跟着风险走,双向都算。从一个低后果系统上撤下力气,和往一个高后果系统上加力气,同样是质量决策。

Deadline pressure
截止日期压力

A triage model is treated as informational because go-live is near and the workflow integration is misread. It ships without subgroup acceptance criteria or monitoring, and the gaps surface 18 months later.
一个分诊模型被当成"仅供参考",因为上线临近、工作流集成被误读。它没有分组验收标准、没有监控就上了线,缺口18个月后才浮出来。

Hold the first-hour discussion before the go-live date enters the conversation, and record the worst-case reasoning. Remediation always costs more than the assurance would have.
在上线日期进入讨论之前,就把第一个小时的谈话开完,并把最坏情形的推理记下来。补救永远比当初做保证更贵。

One function decides alone
一个职能单独拍板

The outcome reflects whoever ran the assessment: the model view, the worst-case view or the benefit view.
评估结果就是主持评估那个人观点的翻版:模型视角、最坏情形视角,或者收益视角。

Make the four-role discussion part of the procedure, with all four people named in the validation plan.
把四角色讨论写进程序文件,四个人的名字都出现在验证计划里。

The demo becomes the workflow
演示变成了工作流

The system is assessed against how it was presented, and then usage quietly grows. The dashboard built for awareness starts feeding batch record commentary.
系统按它被演示的样子接受评估,然后用途悄悄长大。那个为知情而建的数据看板,开始给批记录评语供货。

Name the triggers that reopen the assessment: a new use, a new user group, a new data source, a workflow change. Check them at every periodic review.
把重新触发评估的条件写出来:新用途、新用户群、新数据源、工作流变更。每次定期审查都过一遍。

The pilot that never ended
永远没结束的试点

A proof of concept performs well, people start relying on it, and it drifts into production without an acceptance decision.
概念验证表现不错,大家开始依赖它,它就在没有任何验收决定的情况下漂进了生产。

Gate every proof of concept with evaluation criteria agreed up front and an explicit acceptance decision before its output touches anything GxP.
给每个概念验证设闸:事先商定评价标准,在它的输出碰到任何GxP的东西之前,先做一个明确的验收决定。

Change source ignored
变更来源被忽视

A vendor’s pre-trained model gets a modest risk rating and everyone relaxes, while retraining happens outside the firm on data it will never see.
供应商的预训练模型拿了个中等风险评级,大家都松了口气,而再训练在公司之外、用它永远看不到的数据进行。

Assess where change comes from as well as what an error would do, and let that drive supplier oversight and change control.
既评估错误会造成什么,也评估变更从哪里来,让后者来驱动供应商监督和变更控制。

What to Do with This on Monday

周一就能动手的动作清单

For a quality leader, three moves stand out. Commission an inventory of the AI already inside the operation, and make sure it includes the assistants, dashboards and vendor features that were never labelled AI, because the real inventory is usually bigger than the one on paper. Put the four-role initial assessment into a procedure, so the next AI system meets a process and a shared vocabulary. And when a validation plan comes up for approval, read the worst-case reasoning first, because that reasoning is what protects the patient, and a conclusion without it is only an opinion.

对质量负责人,三件事最突出。第一,清点运营中已有的AI,并确保把那些从没被叫作AI的助手、看板和供应商功能也数进去——真实的清单通常比纸面上的大。第二,把四角色初始评估写进程序文件,让下一个AI系统撞上的是一个流程和一套共同词汇。第三,验证计划送来批准时,先读最坏情形的推理——保护患者的是那个推理,没有推理的结论只是一种意见。

For the people who run validation, the moves are just as concrete. Run the initial assessment on current systems, including the ones already live, and write the result into each validation plan. Check every system where a person reviews AI output for override capture, and pull the actual override data; if the rate has sat at zero for a quarter, treat it as a signal worth following up. Find out what happens to the results the model doesn’t flag. Walk the exception paths and time them. Gate the pilots. And revisit the assessment at every periodic review against the triggers the team named, because the workflow it describes won’t stand still.

对做验证的人,动作同样具体。给现有系统补做初始评估——包括已经上线的——并把结果写进各自的验证计划。逐个检查"由人复核AI输出"的系统有没有把否决记录下来,把真实的否决数据拉出来看;如果否决率在一个季度里一直是零,把它当作值得追查的信号。弄清楚模型没标记的结果去了哪里。把例外路径走一遍并计时。给试点设闸。并且在每次定期审查时,对照团队写下的触发条件重新看一遍评估——它描述的工作流不会原地不动。

All of it runs on one disciplined hour per system and the honesty to scope effort from consequence. That’s the whole method, and it costs less than the first day of the remediation it prevents.

这一切只需要每个系统一次守纪律的一小时,以及按后果定力度的诚实。方法就这么全部,成本还比不上它所防止的那次整改的第一天。

Objects in the Mirror are Closer Than They Appear

后视镜里的物体,比看起来更近

Every passenger-side mirror carries that etched warning, and I’ve started reading it as a regulatory forecast. A couple of years ago, AI guidance in GxP felt like something on the horizon. Look at the mirror now. US Food and Drug Administration’s (US FDA) Computer Software Assurance guidance is final, and it anchors assurance effort to risk. The draft EU GMP Annex 22, the first EU GMP text written specifically for AI, has been through public consultation, and its draft wording points to human oversight proportionate to risk and to performance at least as good as the process the AI replaces. And on 2 April 2026, US FDA issued the first warning letter citing AI use in drug manufacturing (320-26-58). The firm had let an AI agent draft its specifications, procedures and master production records and took the agent’s word for what the regulations required. US FDA’s answer was that AI output must be reviewed and cleared by an authorized person in the quality unit. The direction is consistent, and the distance is shorter than it looks.

每一面副驾驶侧的后视镜上都蚀刻着那句警告,我开始把它当作监管预报来读。几年前,GxP领域的AI指南还像地平线上的东西。现在看看镜子。美国FDA的《计算机软件保证》(CSA)指南已经定稿,把保证力度锚定在风险上。欧盟GMP附录22草案——第一份专门为AI写的欧盟GMP文本——已经过公开征求意见,其草案措辞指向与风险相称的人工监督,以及"性能至少不差于被AI替代的工艺"。而2026年4月2日,美国FDA发出了第一封援引AI用于药品生产的警告信(320-26-58)。那家公司让一个AI智能体起草了质量标准、程序文件和主生产记录,并听信智能体所说的法规要求。FDA的回答是:AI输出必须由质量部门的授权人员审评放行。方向是一致的,距离比看起来近。

The encouraging part is that everything the regulators are converging on is thinking GAMP has carried for more than 30 years, applied to a new kind of system. There’s no need to wait for a final text. Inventory the AI and put 4 people in a room for an hour per system. Ask what happens to the patient when it’s wrong, in both directions, and write the reasoning into the validation plan. The regulations in the mirror are closer than they appear, and that hour is still the cheapest distance a team can buy.

令人鼓舞的是,监管机构正在汇合的那些要求,是GAMP背了三十多年的思想,只是套到了一种新系统上。没必要等定稿。清点AI,给每个系统找四个人放进房间一小时。问它出错时患者会怎样——两个方向都问——并把推理写进验证计划。镜子里的法规比看起来更近,而那一小时,仍是一支团队能买到的最便宜的距离。

iSpeak blog posts provide an opportunity for the dissemination of ideas and opinions on topics impacting the pharmaceutical industry. Ideas and opinions expressed in iSpeak blog posts are those of the author(s) and publication thereof does not imply endorsement by ISPE.

iSpeak博客文章为传播影响制药行业的观点与意见提供平台。文中表达的观点与意见属作者本人,文章的发布不构成ISPE对相关内容的认可。

原文为ISPE官网《制药工程》(Pharmaceutical Engineering)栏目ispeak文章"The First Hour of an AI Project in GxP",作者Sachin Bhandari,发表于2026年10月30日,可在ISPE官网查阅。译文仅供学习参考,如与英文原文有歧义,以原文为准。

来源:制药经理人 · mp.weixin.qq.com

所属事件

查看事件全部
ISPE 文章主张按患者风险定 GxP AI 验证深度