How to Build a Clinical Trial Site Profile That Sponsors Trust

Introduction

When 83 decision-makers at Nordic biopharma companies and CROs were asked what they most wanted to know about a site they had never worked with, 71% put its recruitment and retention track record first ^[1]. The survey was small and regional, but it describes a pattern most feasibility teams will recognize: when a sponsor cannot vouch for a site from experience, it looks for evidence it can check.

That evidence has to be findable first. In interviews with 21 staff at pharmaceutical companies and CROs in four Nordic countries, sponsors described identifying sites mainly through databases of previous trial performance, and, according to the interviewees, saw visibility in those databases as important to global-level identification ^[2]. From the sponsor's chair, a capable site with no structured record of its capabilities may look much like a site with none.

A clinical trial site profile is the structured record that closes that gap. In this guide the term covers the sponsor-facing record of a site's expertise, patient access, qualifications, infrastructure and past performance. Protocol-specific feasibility answers are a separate step. Regulatory points are labeled by jurisdiction, and advice that is not a requirement is marked as a recommendation.

How Sponsors Select Sites

Selection is a sponsor duty. ICH E6(R3), the international good clinical practice guideline (Annex 1 adopted in January 2025, consolidated with Annex 2 in June 2026), makes the sponsor responsible for selecting investigators and institutions and expects each investigator to be qualified and to have adequate resources and facilities ^[3]. In the United States, 21 CFR 312.53 requires sponsors of investigational new drug (IND) studies to select investigators qualified by training and experience, and to obtain a signed Form FDA 1572 and a CV or other statement of qualifications before the investigator begins ^[4]. Other regions apply their own national rules.

In practice, the sponsor or CRO works through a sequence like this:

1. The study team turns the protocol into site requirements: target population, enrollment goal, therapeutic area, investigator expertise and special equipment.
2. Candidate sites come from past-performance databases, investigator networks, CRO relationships and referrals. In the Nordic interviews, performance databases were the main source and electronic health record (EHR) data was rarely used ^[2].
3. Shortlisted sites receive a feasibility questionnaire covering patient availability, staffing, infrastructure and experience with similar studies.
4. The strongest responses may lead to a pre-study site visit.
5. Selected sites supply regulatory documents and move to contracting.

A 2024 task force of sponsor, CRO and site representatives divides site feasibility into four assessments: site profile, site capability, site performance and protocol-specific questions, and recommends gathering the first three before the protocol is final ^[5]. Three of its authors work for a clinical trial technology vendor, which the paper discloses. For practical site-facing use, this guide groups the first three into one sponsor-facing record and calls it the site profile.

What sponsors weigh is more specific than most sites assume. In the 2018 Nordic survey, 42% of respondents ranked "first patients ready for inclusion" as the most preferred site strength, ahead of good data entry and reporting (25%) and reachable personnel with backup (23%) ^[1]. Equipment and facilities came first for 1%, which suggests facilities work as a threshold: sponsors assume them, then decide on patients and people.

What the Evidence Shows

Sponsors are cautious for a measurable reason. A Tufts Center for the Study of Drug Development analysis of about 16,000 sites in 151 Phase II and III trials run from 2008 to 2010 found that 11% of activated sites, meaning sites ready to begin recruitment, enrolled no patients, with regional rates from 7% in Western Europe to 20% in Latin America ^[6]. With all sites in the study as the denominator, 48% under-enrolled ^[6]. These figures are more than a decade old, and no newer public benchmark of comparable scale turned up while preparing this article, so they indicate scale rather than a current rate.

Sponsors also have reason to doubt self-reported patient counts. Overestimated study populations were a leading cause of recruitment delay in the Nordic survey ^[1]. In the interview study, sponsors and CROs said most sites based their counts on past experience alone, and that sites able to support estimates with EHR data seemed more reliable and more attractive ^[2]. The authors note that whether those estimates were actually more accurate was not established, and legal limits on EHR queries constrain the practice ^[2].

Reputation is a weak substitute for evidence. In the survey, 61% of respondents had selected a site that under-recruited because an opinion leader was associated with it ^[1]. A 2016 Tufts CSDD interview study of 26 executives reported sites averaging 70% familiar and 30% new across the 15 companies that supplied the figure ^[7].

Practitioner commentary points the same way. Two Global Life Sciences Alliance partners wrote that feasibility has become a high-volume questionnaire exercise that captures site capacity poorly ^[8], and that capable sites are overlooked because of visibility and communication problems ^[9]. The pieces share authors and are opinion without quantitative data.

Treat these findings as directional: the Nordic data are small regional samples, and sponsors elsewhere may weigh the factors differently.

What a Sponsor-Ready Site Profile Contains

Each row answers a question a feasibility team would otherwise ask by email. The Basis column shows where the expectation comes from.

A checkable profile is easiest to see in an example. The weak entry reads: "Extensive diabetes experience. Excellent enrollment. Large patient database." Nothing in it can be verified. The entry below is illustrative, with invented figures:

Every figure carries a denominator, a date and a source, so a feasibility reader can test it without a phone call.

Building and Maintaining the Profile

This section is editorial advice, not a set of regulatory requirements.

1. Name an owner and a reviewer. One person answers for accuracy, and a second reads every update before it goes live.
2. Assemble the evidence file before writing: a study list with targets, actuals and dates; current CVs and training records; applicable delegation records; and facility documentation.
3. Derive patient counts with a written method. Where local law and institutional policy allow EHR queries, record the criteria and date. Where they do not, label the figure an estimate and state its basis.
4. Write the narrative last. Describe therapeutic expertise with a controlled therapeutic-area hierarchy, mapped where a platform supports it to standard vocabularies such as ICD-10, SNOMED CT, MeSH or MedDRA, and make each strength a summary of completed studies.
5. Cross-check names, addresses, laboratories, subinvestigators and the applicable IRB/IEC or ethics committee against the regulatory file. For US IND studies, that means the current Form FDA 1572 ^[4]. Check the institution's own web pages as well.
6. Publish one master record to a small number of platforms, and set review triggers: a study closeout, a change of principal investigator or coordinator, new equipment, an inspection, and at least once a year for CVs and training dates. No regulation or guideline reviewed for this article sets an update interval for a discovery profile, so the cadence is a judgment call.

Answering feasibility questionnaires

Feasibility work is usually unpaid, according to the task force ^[5], and commentators say higher-performing sites assess their own workload and staffing before they respond ^[8]. A standing feasibility pack built from the profile makes that practical: verified study history, dated patient counts, current staffing and concurrent studies, and standard answers on equipment. Each questionnaire then needs three fresh inputs: the protocol's eligibility criteria applied to local data, the coordinator hours the study would take, and a decision to proceed or decline.

For what follows selection, see Kitsa's Clinical Trial Startup Workflow Checklist: Site Activation Guide.

Preparing for the pre-study site visit

A sponsor may send a monitor to test whether the profile describes the site that exists. Be prepared to provide or make available the documents behind each claim: investigator CVs and training records, applicable delegation records ^[11], dated patient counts with their method ^[3], laboratory, pharmacy and equipment documentation, and IRB details. Which documents a sponsor requests varies. If every figure in the profile traces to a document, the visit confirms the profile. If one does not, the visit questions it.

For how essential records are organized under the revised guideline, see Kitsa's How ICH E6(R3) Changes Clinical Trial Documentation.

Showing continuity when staff change

In the 2022 survey by the Society for Clinical Research Sites (SCRS), which had about 500 respondents overall and mostly executive-level leaders, 58% reported higher turnover at their company over the previous two years, 29% the same and 13% lower ^[12]. Read this as direction rather than an industry rate. Reachable personnel with backup was a top-ranked strength for 23% of Nordic respondents ^[1], so name the backup for each major role and keep delegation records current ^[11].

If your site is new to sponsors

The practical route is later-phase work, where patient access counts for more: a large patient population was first-ranked by 54% of respondents for Phase III against 33% for early-phase trials ^[1]. A newer site can substitute documentation for history, through dated EHR-backed counts ^[2], prior experience recorded on investigators' CVs, and a transparent account of capacity.

Regulatory and Documentation Considerations

A profile is a discovery document, not a regulatory record, but each claim in it should reconcile with one.

United States, IND drug studies (binding regulation). 21 CFR 312.53 requires the sponsor to select qualified investigators and to obtain the signed Form FDA 1572 and a CV or other statement of qualifications for the investigator ^[4]. The form lists the investigator's name and address, the research sites, clinical laboratories, the reviewing IRB and subinvestigators ^[4]. Both the form and the rule are specific to US drug studies under an IND. Device studies fall under a separate regulation, 21 CFR 812.43 ^[13], and other regions use their own documents.

International (guideline). ICH E6(R3) is a harmonised guideline that regulators implement through their own frameworks. It states that the investigator should be qualified by education, training and experience, should be able to demonstrate the potential to recruit the agreed number of eligible participants within the agreed period (for example, using retrospective or current data), and should have sufficient time, qualified staff and adequate facilities for the duration of the trial ^[3]. It also says computerised systems should be fit for purpose, with risk-based validation where appropriate ^[3]. The recruitment clause is the closest GCP provision to a profile's enrollment and patient-count fields.

United States (non-binding guidance). FDA's 2009 investigator guidance says investigators remain responsible for supervising delegated tasks, should keep a list of the qualified people who hold significant delegated duties, and should have sufficient time to conduct and supervise the trial ^[11].

Consensus recommendation. CTTI recommended in 2018 that investigator qualification move beyond one-size-fits-all training and credit prior experience and protocol-specific preparation, because there is little evidence that GCP training alone sufficiently qualifies investigators and their delegates ^[10]. Pair GCP dates with studies and roles.

What Inaccurate Profile Data Costs

FDA's fiscal year 2024 data record 609 clinical investigator inspections: 484 with no action indicated, 110 with voluntary action indicated and 15 with official action indicated ^[14]. The most common observations included failure to comply with Form FDA 1572 requirements, inadequate or inaccurate case history and study records, and protocol deviations ^[14]. Those findings concern conduct at sites. The data do not show that inaccurate profiles cause them, and no study located for this article tests that link. That a site placed in studies it cannot staff may be more exposed is a hypothesis.

The sponsor-side cost is better documented. Overestimated study populations were a leading cause of recruitment delay in the Nordic survey ^[1], and ICH E6(R3) and FDA's investigator guidance both treat adequate staffing and sufficient investigator time as conditions of conducting a trial properly ^[3][11].

AI and Automation Perspective

Automation depends on structured data, which is a reason to build the profile in structured fields. Sponsors in the Nordic interviews found sites through performance databases ^[2], and structured profiles are what such databases and matching software can read. Automation can plausibly help by drafting a site's study history from registry records for staff to confirm, flagging inconsistencies with the regulatory file, and producing EHR-based patient counts that a person reviews.

The limits are practical. Legal obstacles to EHR queries were a recurring theme in the Nordic interviews ^[2], and an automatically produced count may not reflect a specific protocol's eligibility criteria. ICH E6(R3) expects computerised systems to be fit for purpose, with risk-based validation where appropriate ^[3]. Models that rank sites on historical performance may also inherit the visibility gap, since a site with no recorded history gives them nothing to rank; that is an inference, not a measured finding. The 2024 task force reported that, in its experience, the Shared Investigator Platform had not delivered the real-time site data it was meant to provide ^[5], so a site's own verified profile is one controllable source of current, site-supplied information, and one named person should own it.

How Kitsa Fits In

Kitsa Site Network lets research sites claim a profile, keep their information and capabilities current, link affiliations and credentials, and become visible to sponsors searching for sites. On the sponsor side, KScout supports site selection, and KScreener handles EHR-connected patient pre-screening at the site.

Key Takeaways

• Sponsors verify before they trust. For unfamiliar sites, recruitment and retention history was the most valued information in the Nordic survey ^[1].
• Visibility comes before evaluation. Sponsors identify sites through performance databases, and interviewees saw database visibility as important to global-level identification ^[2].
• Show the method behind every patient count. Sponsors perceived EHR-supported counts as more reliable, although whether they were more accurate was not established ^[2].
• Treat facilities as a threshold. Equipment and facilities were the first-ranked site strength for 1% of respondents, against 42% for patients ready for inclusion ^[1].
• Pair GCP dates with experience. CTTI found little evidence that GCP training alone qualifies investigators and their delegates ^[10].
• Reconcile the profile with applicable regulatory and study records: for US IND studies, Form FDA 1572 ^[4], and, where significant duties are delegated, the applicable delegation record or qualified-person list ^[11].

FAQ

Glossary

• MedDRA: standardized terminology used to code and exchange regulatory medical information, including adverse events, indications and medical history.
• ICD-10: WHO's classification for coding diseases and related health conditions; coded diagnosis data may support site and patient feasibility where lawful and appropriate.
• SNOMED CT: clinical terminology used in electronic health records to code diagnoses, findings and procedures.
• MeSH: the US National Library of Medicine's controlled vocabulary for indexing biomedical literature.
• eCOA: electronic clinical outcome assessment, meaning a clinical outcome assessment implemented on an electronic data-collection platform. It may be a patient-reported, clinician-reported, observer-reported or performance outcome.
• eISF: electronic investigator site file used to organize and retain site-level essential records; external access depends on the system, permissions, and study arrangements.
• Form FDA 1572: the statement of investigator that sponsors obtain before an investigator begins an IND drug study in the United States.
• Pre-study site visit: a sponsor or CRO visit to assess whether a shortlisted site can run the study.

References

1. [1]Dombernowsky T, Haedersdal M, Lassen U, Thomsen SF. "Criteria for site selection in industry-sponsored clinical trials: a survey among decision-makers in biopharmaceutical companies and clinical research organizations." Trials 20:708, 2019. https://link.springer.com/doi/10.1186/s13063-019-3790-9
2. [2]Laaksonen N, Bengtström M, Axelin A, Blomster J, Scheinin M, Huupponen R. "Clinical trial site identification practices and the use of electronic health records in feasibility evaluations: An interview study in the Nordic countries." Clinical Trials, 2021. https://doi.org/10.1177/17407745211038512
3. [3]International Council for Harmonisation. "ICH E6(R3) Guideline for Good Clinical Practice, final consolidated guideline." ICH, 2026. https://database.ich.org/sites/default/files/ICH%20E6%28R3%29_Step4_FinalConsolidatedGuideline_2026_0616_.pdf
4. [4]U.S. Government. "21 CFR 312.53, Selecting investigators and monitors." Electronic Code of Federal Regulations. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-D/part-312/subpart-D/section-312.53
5. [5]Bruneau B, et al. "Redefining feasibility in clinical trials: Collaborative approaches for improved site selection." Contemporary Clinical Trials Communications 40:101291, 2024. https://doi.org/10.1016/j.conctc.2024.101291
6. [6]Getz K. "Enrollment Performance: Weighing the Facts." Applied Clinical Trials, 2012. https://www.appliedclinicaltrialsonline.com/view/enrollment-performance-weighing-facts
7. [7]Getz K, Lamberti MJ, Chakravarthy R. "New Benchmarks for Trial Initiation Activities." Applied Clinical Trials 25(12), December 2016. https://www.appliedclinicaltrialsonline.com/view/new-benchmarks-trial-initiation-activities
8. [8]McNerney D, Matheus C. "The Feasibility Gap: Why Strong Sites Miss Study Opportunities." Applied Clinical Trials, April 2026. https://www.appliedclinicaltrialsonline.com/view/feasibility-gap-strong-sites-study-opportunities
9. [9]McNerney D, Matheus C. "Why Capable Research Sites Are Still Being Overlooked in Clinical Trials." Clinical Researcher (ACRP), vol. 40, issue 4, August 2026. https://acrpnet.org/2026/08/11/why-capable-research-sites-are-still-being-overlooked-in-clinical-trials
10. [10]Clinical Trials Transformation Initiative. "Recommendations Now Available for Improving Investigator Qualification." CTTI, 2018 (page last modified 2025). https://ctti-clinicaltrials.org/topics/investigators/investigator-qualification/recommendations-now-available-for-improving-investigator-qualification/
11. [11]U.S. Food and Drug Administration. "Guidance for Industry: Investigator Responsibilities, Protecting the Rights, Safety, and Welfare of Study Subjects." FDA, October 2009. https://www.fda.gov/media/77765/download
12. [12]Society for Clinical Research Sites. "SCRS Site Landscape Survey 2022 White Paper." SCRS, 2023 (turnover question, Figure 4; also listed in the SCRS white paper library at myscrs.org/learning-campus/white-papers-guides). https://myscrs.org/wp-content/uploads/2023/03/SCRS-Site-Landscape-Survey-2022-White-Paper.pdf
13. [13]U.S. Government. "21 CFR 812.43, Selecting investigators and monitors." Electronic Code of Federal Regulations. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-H/part-812/subpart-C/section-812.43
14. [14]U.S. Food and Drug Administration. "Bioresearch Monitoring (BIMO) Fiscal Year 2024 Metrics." FDA. https://fda.gov/media/192870/download