How Modern Clinical Research Systems Actually Work

Most people who describe clinical trials as "complex" are describing the science. The operational infrastructure beneath those trials, the stack of software systems, document repositories, randomization engines, and data platforms that have to function together without failure, rarely gets the same attention. That gap is where trials slow down, protocols drift out of sync across sites, and inspections surface documentation gaps no one planned for.

As of 2024, more than 500,000 clinical studies are registered in ClinicalTrials.gov, a milestone reached during that calendar year according to the National Library of Medicine ^[1]. Managing that volume requires purpose-built technology. Understanding what that technology actually does, and where it breaks down, is essential for anyone running, sponsoring, or supporting a modern trial.

Why the System Architecture Matters

A clinical trial is not a single workflow. It is a set of parallel, interdependent activities: site activation, patient screening, randomization, investigational product logistics, data collection, monitoring, document management, and regulatory submission preparation. Each of these runs on at least one dedicated system. When those systems do not communicate well, or when teams maintain parallel records in spreadsheets alongside official platforms, the consequences are measurable.

A 2024 peer-reviewed analysis from the Tufts Center for the Study of Drug Development, drawing on 409 clinical trial budgets, estimated that the average direct cost of conducting a Phase II or Phase III trial runs approximately $40,000 per day in 2023 dollars ^[2]. At that rate, a single month of avoidable delay costs a sponsor more than $1.2 million in direct expenditure alone, before accounting for market access impact.

An ICON survey conducted in June 2025 among over 100 principal investigators and senior site personnel found that 55% of sites reported time from site selection to full activation now exceeds five months, with 39% reporting that startup timelines are longer than they were two years prior. Contract and budget delays were cited by 66% as a frequent source of that burden ^[3]. Those are not scientific problems. They are operational ones, and the technology layer is where most of them begin.

The architecture of the eClinical technology stack is not an IT question. It is a financial and regulatory one.

The Core Systems: What Each One Does

The four foundational systems in a modern clinical trial are the Clinical Trial Management System (CTMS), the Electronic Data Capture platform (EDC), the Electronic Trial Master File (eTMF), and the Randomization and Trial Supply Management system (RTSM). Each addresses a distinct category of trial data and regulatory requirement.

CTMS

The CTMS is the operational hub. It handles site and investigator management, subject enrollment tracking, study timelines, milestones, budgets, monitoring visit coordination, and site payments ^[4]. Unlike systems that capture clinical data, the CTMS captures operational data: which sites are activated, how enrollment is pacing against plan, which monitoring visits are overdue, and where budget variances are accumulating.

A well-implemented CTMS gives sponsors and CROs real-time oversight across a study portfolio. Without one, or with one that is not properly integrated with downstream systems, operational intelligence defaults to spreadsheets, email threads, and individual project managers' recollections. During a regulatory inspection, that gap is visible. Whether the CTMS serves as the formal system of record for specific regulated documents, such as monitoring visit reports, depends on sponsor SOPs and implementation: some organizations route those documents through the eTMF only, while others generate them in the CTMS and auto-file to the eTMF.

EDC

The EDC is where clinical data is entered, validated, and stored. Site staff complete electronic case report forms (eCRFs) directly in the EDC, which performs real-time data validation, generates and tracks queries, and maintains a full audit trail of every entry, correction, and signature ^[5].

FDA 21 CFR Part 11 governs electronic records and signatures used in FDA-regulated clinical investigations. Under the regulation, systems must enforce role-based access controls, maintain computer-generated, time-stamped audit trails, and confirm that records cannot be altered without leaving a traceable record ^[6]. The FDA's 2003 guidance on Part 11 scope and application clarified a risk-based enforcement discretion posture: the agency focuses inspections on systems generating records that directly support regulatory submissions, not on every electronic tool an organization uses ^[7]. The underlying predicate rules, including GCP requirements for data integrity, remain enforceable regardless.

In the EU, GDPR creates additional obligations for handling personal data in clinical trials, including data minimization and pseudonymization requirements ^[8]. That said, while GDPR provides certain data subject rights, including the right to erasure in some contexts, those rights are subject to exemptions for scientific research under Article 89 and to legal obligations arising from national and EU pharmaceutical regulations. Trial data supporting a regulatory submission cannot simply be deleted on a participant's request where record retention obligations apply.

eTMF

ICH E6(R3), adopted by the International Council for Harmonisation at Step 4 on January 6, 2025, defines essential records as those that, individually and collectively, permit the evaluation of a trial's conduct and the quality of the data produced ^[9]. These documents serve to demonstrate compliance of the investigator, sponsor, and monitor with GCP and applicable regulatory requirements. Together they form the Trial Master File.

The scope of an eTMF includes the protocol and all amendments, investigator brochures, informed consent forms, IRB and ethics committee approvals, safety reports, monitoring visit reports, financial agreements, and regulatory correspondence. Maintaining these documents in an electronic system, under validated access controls and with complete audit trails, can meet FDA 21 CFR Part 11 requirements when the system is properly validated, configured, and governed by appropriate SOPs, and aligns with EMA GCP expectations under ICH E6(R3), which the EMA adopted as effective July 23, 2025 ^[6],[9].

The Drug Information Association (DIA) TMF Reference Model provides a standardized taxonomy for organizing documents within an eTMF. It is an industry reference standard, not a regulatory requirement; no regulation mandates its use. In practice, it has become the dominant framework for eTMF structure, and organizations that adopt it can build automated completeness tracking, document routing, and quality control workflows more reliably than those working from custom-built structures ^[11].

A critical point that the Reference Model does not resolve: eTMF completeness is not a closeout task. ICH E6(R3) expects essential records to be available to the sponsor and investigative site throughout the trial, and that expectation implies ongoing maintenance rather than batch-filing before an audit ^[9].

RTSM

RTSM systems, also referred to as Interactive Response Technology (IRT), serve two connected functions: assigning participants to treatment arms using validated randomization algorithms, and tracking investigational product supply and logistics across sites ^[12].

In blinded or double-blind trials, the RTSM is the only system that knows who received which treatment. It must confirm that treatment allocations remain confidential, that drug supplies at sites do not fall to zero or accumulate excess waste, and that every dispensing event is attributed to an authorized user in a timestamped, auditable record compliant with FDA 21 CFR Part 11 ^[6]. Where RTSM functionality supports GMP-regulated investigational product supply, manufacturing, release, or distribution processes, EMA Annex 11 expectations for computerized systems may also be relevant ^[10].

Under 21 CFR 312.57 and 312.59, sponsors are required to track the disposition of investigational drugs: how much was shipped, administered, returned, or destroyed ^[13]. RTSM systems automate this accountability. In decentralized or hybrid trials, where direct-to-patient drug shipment has become more common, RTSM supply forecasting and inventory management functions become operationally essential.

How These Systems Connect, and Where They Break Down

In an integrated eClinical environment, these four systems exchange data automatically. An enrollment event in the CTMS triggers a randomization call to the RTSM. EDC data flows into CTMS dashboards. Documents filed in the eTMF are linked to monitoring visit reports generated through the CTMS. Each system draws from a shared subject identifier so that patient records are consistent across platforms.

Practical reality at many organizations falls short of that design. A few integration failure patterns appear repeatedly:

Subject ID mismatch. When the CTMS and EDC are from different vendors without validated interface configuration, subject identifiers can diverge. A patient enrolled in the CTMS under one identifier may appear under a different format in the EDC, making cross-system reconciliation manual rather than automatic.

eTMF metadata lag. Monitoring visit reports generated in the CTMS are often auto-filed to the eTMF. When that integration is not properly configured, documents accumulate in a staging queue and are filed in batches, creating the appearance of contemporaneous filing that the audit trail does not support.

RTSM and EDC reconciliation gaps. The RTSM records when a patient received an investigational product. The EDC records when the study visit occurred. When these systems do not share a validated data feed, study teams must manually reconcile dispensing records against visit records, a process that introduces error risk at every step.

These failures are not exotic. They reflect what happens when integration is treated as an IT configuration task rather than a validated, tested interface with documented specifications and ongoing quality oversight.

ICH E6(R3) formalizes the expectation that sponsors use fit-for-purpose computerized systems with validated performance, access controls, audit trails, and ongoing monitoring throughout the trial ^[9]. The guideline explicitly requires that sponsors' quality management systems address computerized system governance, including oversight of systems operated by CROs and vendors. That shared accountability is a significant shift from previous GCP language that focused primarily on the sponsor's own systems.

System Validation: The Obligation Under the Infrastructure

ICH E6(R3) explicitly requires sponsors to establish procedures for validating computerized systems, including access controls, audit trails, and ongoing performance monitoring ^[9]. FDA's 21 CFR Part 11 and the 2007 FDA guidance on computerized systems used in clinical investigations set out complementary expectations for electronic records and electronic source data, including validation, user authentication, and audit trail requirements for systems supporting regulated clinical records ^[5],[6]. While FDA's 2003 Part 11 scope guidance exercised enforcement discretion on certain specific Part 11 provisions, it explicitly confirmed that the underlying predicate rule data integrity obligations remain in full force. For drug accountability and record retention, 21 CFR 312.57 and 312.59 establish what sponsors must track; how those records are maintained electronically is governed by Part 11 and E6(R3) ^[13]. The practical consequence: core trial systems must be validated, documented, and maintained throughout the study lifecycle. Under ICH E6(R3), that obligation extends to systems operated by CROs and technology vendors on the sponsor's behalf ^[9].

Validation requirements span the full system lifecycle: user requirements specifications, installation and operational qualification, performance qualification, change control procedures, periodic review, and retirement documentation. For a sponsor running multiple integrated systems, that means maintaining validation documentation for each platform and for each validated interface between platforms.

Responsibility for validation does not transfer entirely to a vendor when a sponsor licenses a software-as-a-service platform. Sponsors retain accountability for confirming that vendor-managed systems meet their documented user requirements and GCP expectations. This is an area where sponsors frequently underestimate the documentation burden, and where inspectors regularly identify gaps.

Study Startup: The Phase Where Systems Are Tested First

Before a single patient is enrolled, a trial must activate its sites. This requires IRB or ethics committee approvals, executed clinical trial agreements, budget negotiations, staff training, system access provisioning, and regulatory submissions where applicable. The median activation time at cancer centers, reported in a 2018 Association of American Cancer Institutes benchmarking survey, was 167 days. The National Cancer Institute's guidance sets a target of 90 days ^[14]. Actual activation times at NCI-designated cancer centers in 2024 still ranged from 78 days to 313 days ^[14], a gap that reflects site-specific infrastructure differences and administrative process quality.

The ICON 2025 industry survey found that site pre-selection decline rates, where sites decline participation before formal selection, rose from 35% in 2021 to 47% in 2023, driven primarily by complex protocols, narrow eligibility criteria, and competition for qualified investigators ^[3]. Separately, a 55% reduction in active US investigators between 2018 and 2023 has reduced the available site base across therapeutic areas ^[14].

These pressures translate directly to technology demands. Contract lifecycle management, budget negotiation workflows, IRB submission tracking, regulatory document exchange, and training documentation often live in disconnected tools. The accumulated administrative latency is not primarily a scientific problem. It is a system design problem, and the sponsors best positioned to address it are those who have integrated their startup workflows with their CTMS rather than managing them in parallel.

Protocol Design as a Downstream System Problem

Protocol quality determines how well every downstream system performs. An ambiguous eligibility criterion generates screening failures that the EDC records but that no RTSM can prevent. An overly complex visit schedule creates eCRF completion burdens that degrade data quality. A poorly defined endpoint forces a protocol amendment that cascades simultaneously through the CTMS, eTMF, IRB re-submissions, and site training workflows.

A peer-reviewed 2024 study by Getz, Smith, Botto, and colleagues at Tufts CSDD, drawing on data from 950 protocols and 2,188 amendments collected from 16 pharmaceutical companies and CROs, found that the proportion of Phase I-IV trials requiring at least one substantial amendment has risen from 57% in 2015 to 76% ^[15]. The mean number of amendments per protocol increased 60%, from 2.1 to 3.3 ^[15]. A follow-up Tufts CSDD analysis reported that the average total time to implement a substantial amendment, from identifying the need to receiving final ethics board approval, now averages 260 days, and that investigative sites operate under differing protocol versions for an average of 215 days during that period ^[16].

The direct cost of a substantial protocol amendment ranges from $141,000 for a Phase II protocol to $535,000 for a Phase III protocol, and protocols with at least one amendment take an average of three additional unplanned months to complete compared to unamended protocols ^[17]. These figures come from Tufts CSDD analysis published in Therapeutic Innovation and Regulatory Science (Getz et al., 2016), and they do not include the downstream burden on eTMF teams reconciling amended documents across sites, EDC teams updating validation rules, or site staff undertaking re-consent and re-training.

Protocol design decisions made earlier, with better access to site feasibility data, historical operational benchmarks, and cross-document consistency checks, reduce amendment frequency and its downstream costs. Technology cannot fix a flawed protocol after the fact, but it can surface design risks before they reach the site network.

Regulatory and Documentation Considerations

ICH E6(R3)

ICH E6(R3) was endorsed by the ICH Assembly at Step 4 on January 6, 2025, representing the first full overhaul of the Good Clinical Practice guideline since the R2 addendum in 2016 ^[9]. The EMA adopted it as effective on July 23, 2025. The FDA published its final guidance reflecting E6(R3) on September 9, 2025. FDA guidance documents are non-binding, meaning they represent the agency's current expectations and recommendations but do not themselves establish legally enforceable obligations ^[18].

Several E6(R3) provisions have direct implications for clinical research systems:

Risk-proportionate quality management. The guideline formally moves away from one-size-fits-all monitoring, requiring sponsors to identify critical-to-quality factors for each trial and concentrate oversight where the risks to participant safety and data reliability are greatest ^[9]. For CTMS and EDC platforms, this means real-time data availability is more important than periodic batch reporting.

Data governance. A substantially expanded Section 4 on data governance sets expectations for data integrity, traceability, security, and the documentation of computerized system fitness for purpose across sponsor and investigator functions ^[9]. These expectations apply to sponsor-operated systems, CRO-operated systems, and vendor-provided platforms alike.

Computerized system validation and access controls. E6(R3) explicitly requires validated systems with access controls, audit trails, backup procedures, and ongoing operational monitoring ^[9]. These requirements apply to EDC, eTMF, RTSM, and CTMS platforms, and to the interfaces between them, each of which must also be validated.

Essential records. Appendix C of E6(R3) updates the essential records framework and adds a documentation requirement for the assessment of fitness for purpose of non-trial-specific computerized systems, a category that covers many commercial SaaS platforms used in trial operations ^[9].

21 CFR Part 11 and the Predicate Rule Framework

FDA 21 CFR Part 11 governs electronic records and electronic signatures in FDA-regulated settings. The regulation requires validated systems, role-based access controls, timestamped audit trails, and the ability to produce accurate and complete copies of records for inspection ^[6]. The 2003 FDA guidance on Part 11 scope and application clarified that the agency takes a risk-based approach to enforcement: inspectors focus on systems that generate records used to support regulatory submissions or that directly affect patient safety and product quality, not on every electronic tool in an organization's portfolio ^[7]. The underlying predicate rules, including GCP data integrity requirements under 21 CFR Part 312, remain enforced in full regardless of whether specific Part 11 provisions are exercised with discretion.

The practical implication for sponsors: validating the core systems (EDC, eTMF, RTSM, CTMS interfaces) to Part 11 requirements is not optional, but organizations need not treat every peripheral system as requiring the same level of formal validation documentation as their EDC.

The DIA TMF Reference Model

The DIA TMF Reference Model, first published in June 2010 and maintained through subsequent versions, defines a standardized taxonomy for organizing trial documents ^[11]. It is an industry standard and is widely used by eTMF vendors as the default structure for document classification. It is not a regulatory requirement. No ICH guideline, FDA regulation, or EMA guidance mandates its use by name. Sponsors who deviate from it are not automatically non-compliant; sponsors who use it inconsistently, applying it to some studies and not others, may create reconciliation challenges during portfolio audits or CRO transitions.

Decentralized Clinical Trials and What They Change Technically

The FDA's final guidance on conducting clinical trials with decentralized elements, published September 18, 2024, does not create new regulations ^[19]. It clarifies that existing regulatory requirements, including Part 11, GCP, and investigational product accountability under 21 CFR 312.57 and 312.59, apply equally to trials with or without decentralized elements ^[13],[19]. Sponsors can incorporate telehealth visits, local healthcare providers near participants' homes, direct-to-patient drug shipment, and digital health technology-based data collection, but they must do so within the same compliance framework that applies to conventional trials.

For the technology stack, decentralization adds meaningful operational complexity. RTSM systems must support direct-to-patient drug shipment and track inventory at locations other than traditional investigative sites, including patient homes, local pharmacies, and community healthcare providers. EDC systems must receive and validate data from wearable sensors, home visit records, and remote source data inputs. The eTMF must capture documentation from local healthcare providers and decentralized service vendors who may not operate within the trial's standard site infrastructure. The CTMS must maintain enrollment and visit completion visibility across a more geographically distributed network than a traditional multi-site trial.

When decentralized elements are added without corresponding system configuration, the result is reduced operational visibility rather than the increased access those elements were intended to provide. Sponsors piloting decentralized designs should map their technology requirements to each decentralized element before study activation, not during.

AI and Automation in the Clinical Research System Stack

AI is entering clinical research operations through several practical routes, none of which replace the regulatory framework that governs data integrity and trial conduct.

Risk-based monitoring benefits from AI-driven anomaly detection applied to EDC data: surfacing implausible values, identifying sites with unusual query resolution patterns, and flagging enrollment trends likely to miss milestones before they do. These capabilities depend on clean, structured data in the underlying systems. AI does not resolve data quality problems; it surfaces them more quickly when the data infrastructure supports systematic analysis.

Protocol design is another active area. Tools trained on historical trial data can identify eligibility criteria likely to generate excessive screen failures, flag visit schedule complexity that predicts high dropout, and check proposed protocol language for internal inconsistencies before submission. A protocol that does not require a substantial amendment saves up to $535,000 per amendment avoided for a Phase III study, plus the 260-day implementation timeline and three unplanned months of study duration ^{[15],[16],[17]}.

Document generation is a third entry point. Producing regulatory documents, including protocols, informed consent forms, investigator brochures, development safety update reports, and clinical study reports, from structured clinical data reduces authoring time and improves cross-document consistency. Any AI-generated regulatory document requires human expert review, validation, and sign-off before submission. The regulatory value is in consistency and efficiency, not in removing professional judgment from the process.

AI tools used in clinical research workflows must meet validation expectations consistent with ICH E6(R3). The FDA's January 2025 draft guidance, "Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products" (Docket No. FDA-2024-D-4689), proposes a risk-based credibility assessment framework for establishing the suitability of an AI model for a particular context of use, evaluated against the specific decision it is intended to support ^[20]. This guidance applies to AI tools producing data or analysis intended to inform regulatory submissions; it is currently in draft form and not yet final.

How Kitsa Fits Into This Infrastructure

Kitsa's platform is designed for teams that recognize the protocol and document layer as the upstream source of many downstream system problems.

KScribe, Kitsa's AI-powered regulatory document generation module, addresses one of the most persistent contributors to both protocol amendments and inspection findings: inconsistency across the clinical document set. When a protocol, informed consent form, investigator brochure, and development safety update report are generated from the same structured data source with cross-document consistency checks applied at authoring time, the gap between what the protocol says and what other documents reflect shrinks. Kitsa's platform is designed for secure regulatory document workflows, with SOC 2, HIPAA, and ISO 27001-aligned security controls and AWS Virtual Private Cloud deployment.

KScout supports site selection with data-driven feasibility intelligence, directly relevant to the startup-delay problem ICON's 2025 survey quantified: teams approaching sites at declining acceptance rates waste startup weeks, and the daily trial costs that accompany those weeks are real ^[2],[3].

Key Takeaways

• The eClinical technology stack, CTMS, EDC, eTMF, and RTSM, consists of purpose-built systems that must be individually validated and operationally integrated. Gaps in integration generate the manual reconciliation steps where data quality problems and audit trail deficiencies arise.
• ICH E6(R3), adopted January 6, 2025 (EMA effective July 23, 2025; FDA September 9, 2025), establishes explicit expectations for computerized system validation, data governance, access controls, and risk-proportionate quality management across the full trial technology stack.
• FDA 21 CFR Part 11 requires validated systems, audit trails, and access controls for electronic records used in regulated clinical investigations. The FDA's 2003 scope and application guidance confirmed a predicate-rule-first enforcement discretion approach, focusing inspection resources on systems whose records directly support submissions.
• Protocol amendment frequency has risen substantially: 76% of Phase I-IV protocols now require at least one substantial amendment, up from 57% in 2015, at direct costs of up to $535,000 per Phase III amendment and average implementation timelines approaching 260 days from identification to final ethics approval.
• FDA's September 2024 final DCT guidance establishes that all existing regulatory requirements apply equally to trials with decentralized elements, including RTSM drug accountability, EDC data integrity, and eTMF essential records, without exception for remote or hybrid designs.
• System validation is a shared responsibility. Sponsors retain accountability for confirming that vendor-managed and CRO-operated systems meet GCP expectations, with user requirement specifications, vendor qualification documentation, interface specifications, and validation protocols as the primary documentation artifacts.
• FDA's January 2025 draft AI guidance (FDA-2024-D-4689) proposes a risk-based credibility assessment framework for AI tools producing data to support regulatory submissions. It is currently in draft; organizations should monitor finalization and align their AI governance practices accordingly.

FAQ

References

1. [1]National Library of Medicine, National Institutes of Health. "ClinicalTrials.gov: A 25-Year Journey to a Half-Million Registered Studies." NLM Director's Blog, April 2, 2025. https://nlmdirector.nlm.nih.gov/2025/04/02/clinicaltrials-gov-a-25-year-journey-to-a-half-million-registered-studies/
2. [2]Smith, Z.P., DiMasi, J.A., Getz, K.A. "New Estimates on the Cost of a Delay Day in Drug Development." Therapeutic Innovation and Regulatory Science. 2024 Sep;58(5):855-862. DOI: 10.1007/s43441-024-00667-w. PubMed PMID: 38773058.
3. [3]ICON plc. "ICON Survey Reveals Increasing Clinical Trial Startup Delays, Underscoring Need for Human-Centred Site Activation Solutions." ICON Press Release, December 2, 2025. https://www.iconplc.com/news-events/press-releases/icon-survey-reveals-increasing-clinical-trial-startup-delays
4. [4]Medidata Solutions. "CTMS 101: What is a Clinical Trial Management System?" Medidata Blog, 2025. https://www.medidata.com/en/life-science-resources/medidata-blog/clinical-trial-management-systems-ctms/
5. [5]U.S. Food and Drug Administration. "Computerized Systems Used in Clinical Investigations." FDA Guidance for Industry, May 2007. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/computerized-systems-used-clinical-investigations
6. [6]U.S. Food and Drug Administration. "21 CFR Part 11: Electronic Records; Electronic Signatures." Code of Federal Regulations Title 21, Part 11. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-A/part-11
7. [7]U.S. Food and Drug Administration. "Part 11, Electronic Records; Electronic Signatures: Scope and Application." FDA Guidance for Industry, August 2003. https://www.fda.gov/media/75414/download
8. [8]European Parliament and Council. "Regulation (EU) 2016/679 (General Data Protection Regulation)," Article 89 (safeguards for research purposes). Official Journal of the European Union, May 2016. https://eur-lex.europa.eu/eli/reg/2016/679/oj
9. [9]International Council for Harmonisation. "Integrated Addendum to ICH E6(R1): Good Clinical Practice E6(R3)." ICH Harmonised Guideline, Step 4 Final Adoption January 6, 2025. https://database.ich.org/sites/default/files/ICH_E6(R3)_Step4_FinalGuideline_2025_0106.pdf
10. [10]European Medicines Agency. "Annex 11: Computerised Systems." EudraLex Volume 4, EU Guidelines for Good Manufacturing Practice for Medicinal Products. https://health.ec.europa.eu/system/files/2016-11/annex11_01-2011_en_0.pdf
11. [11]Drug Information Association (DIA). "Trial Master File Reference Model." Published via CDISC. https://www.cdisc.org/standards/trial-master-file-reference-model
12. [12]Applied Clinical Trials. "IRT/RTSM: The Role of Randomization and Trial Supply Management in Clinical Trials." Applied Clinical Trials Online. https://www.appliedclinicaltrialsonline.com/view/irt-rtsm-role-randomization-trial-supply-management-clinical-trials
13. [13]U.S. Food and Drug Administration. "21 CFR Part 312: Investigational New Drug Application." §312.57 (Record retention) and §312.59 (Disposition of unused supply). https://www.ecfr.gov/current/title-21/chapter-I/subchapter-D/part-312
14. [14]Applied Clinical Trials. "Accelerating Clinical Trial Activation." Applied Clinical Trials Online, January 2026. https://www.appliedclinicaltrialsonline.com/view/accelerating-clinical-trial-activation
15. [15]Getz, K., Smith, Z., Botto, E., Murphy, E., Dauchy, A. "New Benchmarks on Protocol Amendment Practices, Trends and their Impact on Clinical Trial Performance." Therapeutic Innovation and Regulatory Science. 2024;58(3):539-548. DOI: 10.1007/s43441-024-00622-9
16. [16]Applied Clinical Trials. "Shining a Light on the Inefficiencies in Amendment Implementation." Applied Clinical Trials Online, 2026. https://www.appliedclinicaltrialsonline.com/view/shining-a-light-on-the-inefficiencies-in-amendment-implementation
17. [17]Getz, K.A., Stergiopoulos, S., Short, M., et al. "The Impact of Protocol Amendments on Clinical Trial Performance and Cost." Therapeutic Innovation and Regulatory Science. 2016;50(4):436-441. DOI: 10.1177/2168479016632271. PubMed PMID: 30227022.
18. [18]Association of Clinical Research Professionals (ACRP). "FDA Publishes ICH E6(R3): What It Means for U.S. Clinical Trials." ACRP Blog, September 2025. https://acrpnet.org/2025/09/16/fda-publishes-ich-e6r3-what-it-means-for-u-s-clinical-trials
19. [19]U.S. Food and Drug Administration. "Conducting Clinical Trials With Decentralized Elements: Guidance for Industry, Investigators, and Other Interested Parties." FDA Final Guidance, September 2024. https://www.fda.gov/media/167696/download
20. [20]U.S. Food and Drug Administration. "Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products." FDA Draft Guidance for Industry, Docket No. FDA-2024-D-4689, January 2025. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological