In more than seventeen years of practice advising hospitals, research institutions and contract research organisations across Sichuan, I have watched China's clinical trial landscape transform almost beyond recognition. Trials that once moved forward on paper records and institutional trust are now conducted under a dense framework of statutes, regulations and judicial interpretations built for a single purpose: to ensure that the data submitted to regulators is true. That framework acquired sharp teeth in 2017, when the Supreme People's Court and the Supreme People's Procuratorate issued the Judicial Interpretation on Several Issues Concerning the Application of Law in Handling Criminal Cases Involving the Falsification of Materials in Drug and Medical Device Registration Applications. Since that interpretation took effect, fabricating clinical trial data has ceased to be a regulatory embarrassment that can be settled with a fine or a warning. It is a criminal offence, punishable by imprisonment of up to five years together with a fine.
In more than seventeen years of practice advising hospitals, research institutions and contract research organisations across Sichuan, I have watched China's clinical trial landscape transform almost beyond recognition. Trials that once moved forward on paper records and institutional trust are now conducted under a dense framework of statutes, regulations and judicial interpretations built for a single purpose: to ensure that the data submitted to regulators is true. That framework acquired sharp teeth in 2017, when the Supreme People's Court and the Supreme People's Procuratorate issued the Judicial Interpretation on Several Issues Concerning the Application of Law in Handling Criminal Cases Involving the Falsification of Materials in Drug and Medical Device Registration Applications. Since that interpretation took effect, fabricating clinical trial data has ceased to be a regulatory embarrassment that can be settled with a fine or a warning. It is a criminal offence, punishable by imprisonment of up to five years together with a fine.
Why Data Falsification Became a Crime
For years, a loophole troubled regulators and honest industry participants alike. The deliberate fabrication of non-clinical research data or clinical trial data — inventing animal study records, inventing subject information, rewriting records of the main trial process, adjusting test data to make results look cleaner — could at worst attract administrative sanctions. The 2017 judicial interpretation closed that loophole. It provides that where such fabrication affects the evaluation of a drug's safety and efficacy, the conduct is to be treated as the offence of intentionally providing false certification documents under Article 229 of the Criminal Law, which carries a maximum penalty of five years' imprisonment or criminal detention, together with a fine.
Background & legal framework
The interpretation is deliberately broad in its reach. It expressly applies to staff of drug non-clinical research institutions, drug clinical trial institutions and contract research organisations — in other words, the clinical research associates, data managers, investigators and site coordinators whose signatures and keystrokes stand behind every page of a trial record. It then enumerates six circumstances that constitute the offence:
- knowingly using counterfeit investigational drugs in the course of non-clinical research or a clinical trial;
- concealing serious adverse events associated with the investigational drug;
- intentionally destroying original non-clinical research data or clinical trial data;
- fabricating information on test animals or trial subjects, records of the main trial process, research data or test data in a manner that affects the evaluation of the drug's safety and efficacy;
- having previously been criminally punished for providing false certification materials in a drug or medical device registration application, or having received administrative punishment for such conduct within the preceding two years, and providing false materials again; and
- other circumstances of a serious nature.
In my compliance practice, the fifth circumstance deserves special emphasis for sponsors and sites. A repeat offence triggers criminal liability even where the first violation was resolved only administratively; under this framework there is no third chance. I therefore advise clinical research organisations to treat a first inspection finding, regulatory query or data discrepancy as a serious event requiring immediate investigation, root-cause analysis and remediation — not as an acceptable cost of doing business.
Data Integrity Under GCP: The ALCOA Standard
The criminal law sits on top of a regulatory framework that defines, in practical terms, what "true data" means. The current Good Clinical Practice standard in China — the Drug Clinical Trial Quality Management Standard issued by the National Medical Products Administration in 2020 and effective on 1 July 2020 — requires that trial records be accurate, complete, legible and timely, and that every piece of source data be attributable to the person who created it. Inspectors, both Chinese and international, describe this requirement through the ALCOA acronym: Attributable, Legible, Contemporaneous, Original and Accurate, extended in modern practice by completeness, consistency, endurance and availability.
What ALCOA means on the ground is often less glamorous than the acronym suggests. It means that a research nurse's entry is made on the day the observation is made, not reconstructed from memory a week later. It means that a correction to a case report form carries a date, a signature and a reason, and that the original entry remains visible. It means that electronic data capture systems maintain audit trails in which every change is logged with user identification and a timestamp, and that access to those systems is controlled by role. The 2020 GCP devotes real attention to electronic systems precisely because digital records, once considered more trustworthy than paper, are in fact easy to alter invisibly unless the system is designed and validated to prevent it.
How the dispute was handled
In my experience, most serious inspection findings do not arise from sophisticated fraud. They arise from ordinary documentation habits: undated entries, backfilled data, shared passwords, spreadsheets maintained outside the validated system, source documents completed after the visit. The nationwide campaign in 2015 in which the drug regulator required sponsors to self-inspect their clinical trial data — and in which a large number of registration applications were withdrawn rather than face verification — remains a vivid warning to the industry. A trial can be scientifically excellent and still fail inspection if its records cannot be trusted.
Protecting the People Behind the Data
Data integrity is not an end in itself. Regulators treat falsification with such severity because corrupted data distorts the safety and efficacy assessment on which approval decisions rest, and because it betrays the individuals who volunteered their bodies and their medical information to the research enterprise. The legal framework protecting trial subjects in China rests on three pillars: ethics review, informed consent and the truthful reporting of adverse events.
Under the Drug Administration Law as revised in 2019, a clinical trial may not begin until an ethics committee has reviewed and approved the protocol, and the trial must be conducted in accordance with Good Clinical Practice. Articles 19 through 22 of the Law establish the core safeguards: ethics review as a precondition for the trial, the obligation of the trial institution to protect the rights and interests of subjects, informed consent obtained from each subject before enrolment, and the subject's right to withdraw from the trial at any time without prejudice, together with the right to compensation for trial-related injury. Under the GCP, informed consent must be a process, not a signature: the subject must be told, in language they can understand, the purpose of the trial, the procedures involved, the foreseeable risks and discomforts, the alternatives to participation, and the fact that withdrawal is possible at any time without penalty.
The judicial interpretation makes clear that these duties are not aspirational. Its second aggravating circumstance — concealing serious adverse events associated with the investigational drug — criminalises precisely the failure that most endangers subjects. When a participant suffers a serious adverse event and the event is buried, delayed or recharacterised in the data, that subject is denied the protection the trial system promises, and every subsequent participant is put at risk. Sponsors and sites should therefore treat serious adverse event reporting not as a regulatory chore but as the moral core of the enterprise. An escalation path that is fast, documented and free of blame culture is a compliance asset, not a burden.
Practical implications
Health Data and Human Genetic Resources: The Compliance Frontier
For sponsors and research organisations operating in China, one of the most complex layers of trial compliance concerns the data itself. Clinical trial data derived from Chinese subjects contains health information, which the Personal Information Protection Law classifies as sensitive personal information. Processing sensitive personal information requires a separate informed consent, a specific purpose and a clear necessity, and is subject to stricter security obligations than ordinary personal information. The same data may also implicate human genetic resources, which are regulated separately under the Regulations on the Administration of Human Genetic Resources. Under those Regulations, foreign organisations and individuals may not collect or preserve Chinese human genetic resources on their own, and international cooperative research using such resources must generally be carried out with a Chinese partner institution, subject to approval or record-filing with the authorities as applicable.
These two regimes interact in ways that catch organisations by surprise. A biobank holding residual trial samples, a real-world evidence study reusing historical medical records, or a data transfer to an overseas affiliate for central statistical analysis can each trigger obligations that the clinical team never anticipated. Cross-border transfer of personal information is itself subject to restrictions under the Personal Information Protection Law, and human genetic resource data may not be provided abroad except in narrowly defined circumstances. In my work with health data processors in Chengdu, I have seen governance gaps in each of these areas — and I have seen regulators and courts begin to treat data governance failures with the same seriousness as data fabrication. The reputational and legal cost of a violation can far exceed anything saved by the shortcut.
Conclusion: A Compliance Framework That Protects Everyone
The 2017 judicial interpretation, the 2019 Drug Administration Law, the 2020 GCP and the human genetic resources and data protection regimes send a coherent message: in China today, the integrity of clinical trial data is a legal duty with criminal consequences, and the protection of trial subjects is the reason that duty exists. The clients who sleep well at night are those who treat data integrity as a system rather than a slogan. That system includes written standard operating procedures and training; source data verification and independent monitoring; validated electronic systems with genuine audit trails; a serious adverse event escalation path that works after business hours; ethics committee processes that are documented rather than pro forma; and a data governance framework covering retention, access control, secure disposal and cross-border transfer in line with the Personal Information Protection Law and the human genetic resources rules.
For hospitals, research institutions and sponsors alike, the cost of building such a system is modest compared with the alternative: a criminal investigation, a failed registration, a trial stopped mid-course or, worst of all, a subject harmed by data that was never true. Data, in the end, is the only bridge between a promising molecule and the patient who needs it. When that bridge is built honestly, everyone benefits — the subject, the institution, the sponsor and the public that ultimately relies on every approved drug. That is the principle I bring to every clinical research compliance matter, and the standard I urge every organisation in the field to adopt.
