Decentralized clinical trials have been described as the future of drug development for long enough that calling them the future feels slightly ironic. The technology for remote monitoring, direct-to-patient drug delivery, and telehealth-enabled visits has been available for years. The regulatory guidance from both FDA and EMA has been progressively more permissive. And yet the vast majority of Phase 2 and 3 trials still run through traditional investigator site networks.
This is not a technology adoption lag story. It is a structural tension story. The site-based model exists because sites provide something that pure decentralized enrollment cannot easily replicate: a local clinical relationship between a trained investigator and a patient, with the institutional infrastructure that backs it. Understanding why that infrastructure persists, and where it genuinely impedes enrollment versus where it enables it, is important before drawing conclusions about what hybrid trial design should look like.
What Decentralized Enrollment Actually Promises
The core argument for DCT enrollment approaches is geographic reach. Traditional site networks concentrate patients in academic medical centers and urban research practices. A patient with a qualifying condition who lives two hours from the nearest activated site faces a participation burden that has nothing to do with their medical eligibility or their interest in contributing to research. Decentralized enrollment, whether through direct patient recruitment, telemedicine screening visits, or mobile research units, can reach those patients.
For conditions where the eligible population is geographically dispersed, this is a real and meaningful advantage. Rare diseases where prevalence in any single metro area might not support a traditional site-only approach, or chronic conditions where the trial burden is modest enough that frequent site visits are not required, are the clearest use cases for fully or predominantly decentralized designs.
The secondary argument is speed. Recruiting patients directly, without the activation timeline required for traditional sites, can theoretically compress the front end of enrollment. Sponsors have run trials where direct-to-patient digital outreach generated screened candidates faster than any site network would have.
What the Site Model Actually Provides
The case for site-based enrollment is less frequently articulated, partly because the industry has spent the last decade focused on DCT innovation and partly because site-based enrollment's strengths are harder to quantify.
The most important is data quality and procedural compliance. Phase 3 trials submitted to regulatory agencies require extremely high confidence in the procedures used to collect primary endpoints. A nurse practitioner who conducts study procedures dozens of times per year, in a clinical environment with consistent documentation and oversight, produces data with a character that remote procedures have not yet demonstrated equivalently across all endpoint types. For trials where the primary endpoint is a clinical assessment, a procedure outcome, or a lab result from a specific assay, site-based collection is often not just preferable but required.
The second is the investigator relationship. A PI who has a clinical relationship with the patient brings context that a direct-to-patient digital approach cannot. That context matters for eligibility determination, for adverse event management, and for patient retention. Sites report that patient dropout rates are meaningfully lower when participants are enrolled through a treating physician who continues to be involved in their care during the trial. That relationship is hard to replicate at scale through telemedicine.
Third is the infrastructure for complex protocols. Trials that require imaging, biopsies, specialized laboratory processing, or frequent clinical assessments cannot be run in a fully decentralized model without mobile clinical units or participant home visits that add their own costs and operational complexity. For these trials, the site's physical infrastructure is not a relic but a functional requirement.
Where Hybrid Design Makes Sense
Hybrid trial designs try to capture reach benefits from decentralized approaches while preserving the data quality and investigator oversight of the site model. In practice, hybrid means different things in different contexts.
The most common hybrid pattern is decentralized screening with site-based procedures. Patients are identified and initially screened remotely, potentially through direct digital outreach, telehealth screening visits, or patient-reported pre-qualification tools. Patients who pass pre-screening are then referred to a nearby activated site for formal eligibility confirmation and enrollment procedures. This approach can improve candidate identification without requiring every study procedure to happen remotely.
A second pattern is the satellite site model, where a hub site with full investigator and procedural infrastructure activates affiliated community practices or clinics as satellite collection points. Patients receive study visits closer to home, but their investigator relationship and procedural oversight remain connected to the hub. This extends geographic reach without fully removing the site-based accountability structure.
We work primarily in the site-based and hybrid space, so our perspective here is not neutral. But our experience is that sponsors who design hybrid models with a clear view of which trial elements require site-based infrastructure and which can genuinely be decentralized tend to get the benefits of both models. Sponsors who apply DCT components as an overlay on a fundamentally site-based design, without redesigning the protocol to support it, often find that the hybrid adds coordination complexity without meaningfully improving enrollment.
The Enrollment Implication
For enrollment specifically, the most consequential question about DCT versus site models is: where does the eligible population actually live, and how does that map to your site network?
If the eligible population is concentrated in the catchment areas of your activated sites, and your sites have good pre-screening capability, site-based enrollment will generally outperform direct patient recruitment because the patients are already in a clinical relationship. The issue is not finding them; it is working up their eligibility systematically once you have found them.
If the eligible population is geographically dispersed or concentrated in communities that are underrepresented in traditional site networks, decentralized recruitment channels can genuinely expand the available patient pool. But expanding the candidate pool only helps if there is a qualified site to receive those candidates for formal enrollment. Direct patient recruitment that generates interest but has no downstream enrollment infrastructure to convert that interest is a funnel with a missing bottom step.
The practical takeaway is that enrollment planning should start with the geography of eligibility, not with a preference for either model. For the trials where we work on pre-screening, the question we start with is whether the site network, as currently constituted, physically overlaps with the population that carries the qualifying condition profile. That geographic question shapes everything downstream, including whether hybrid approaches are worth the additional coordination overhead they introduce.
What Is Not Changing
The tension between decentralized reach and site-based rigor is not going away, and neither side of it is wrong. DCTs improve access for underserved populations and can be genuinely faster for appropriate indications. Site-based trials provide oversight, data quality, and investigator relationships that remain the standard for most pivotal programs.
What is changing, gradually, is the protocol design sophistication around identifying which elements of a given trial genuinely require site-based execution and which do not. As that design thinking improves, hybrid models will likely become more precise and less of a hedge. For now, most Phase 2 and 3 trials will continue to run primarily through site networks, and improving how those networks identify and engage eligible patients remains the highest-leverage place to work on enrollment.