One recruitment standard across every site in your network

Why is one recruitment standard hard to hold across a network?

Tufts CSDD reported in its March/April 2026 Impact Report that a growing share of clinical trials run at academic medical center- and network-affiliated sites.^⁠[3] For-profit organizations, including sites and site networks, took about half of the $7.6 billion in reported industry payments to US research sites in 2022, and from 2019 to 2022 their payments grew 11.8% a year, against 4.9% for nonprofit institutions.^⁠[4]

Investor-backed networks grow by adding sites, and the deals continued in 2025. BayPine agreed in April to acquire CenExel, an 18-location US network, and in August THL agreed to buy Headlands Research from KKR, which founded it in 2018.^⁠[6],[7] An acquired site can bring its own EHR, coordinators and pre-screening habits. Without a network standard, recruitment quality depends on which site a patient reaches.

Site results already vary widely. In a 2013 Tufts CSDD analysis of more than 150 studies and nearly 16,000 sites, 11% of sites in a typical trial enrolled no patients and 37% enrolled fewer than planned.^⁠[8] Capacity is tight too: in WCG's 2024 survey of 852 sites worldwide, 46% said their top challenges restricted their capacity to take on new studies.^⁠[9]

What changes day to day at each site?

Bond runs the same three stages at every site: identify, engage and consent. The network agrees once on how each protocol's criteria are read. Coordinators at each site then check that reading against a sample of their own records before go-live, as described in how we validate eligibility logic.

1. 1

   Identify

   Identify reads structured and unstructured EHR records against the protocol's criteria, ranks candidates, and shows the chart evidence behind each criterion decision. Every site reviews the same kind of list.^⁠[1]

2. 2

   Engage

   Engage voice and text agents contact likely matches, ask the pre-screening questions the chart cannot answer, and book screening visits into the site's calendar. Patients are told AI assistance is used and can reach a person at any time: the agent transfers the call live to a coordinator or books a human callback, whichever the site prefers.^⁠[1],[2]

3. 3

   Consent

   Consent explains the IRB-approved consent form in plain language and answers patient questions, with escalation to staff. The PI and delegated staff still obtain consent.^⁠[1]

Central, local or mixed outreach

• Central. One approved script per study across all sites, with escalations going to one network team.
• Local. Each site's script carries its own name and details, with escalations to its coordinators.
• Mixed. Central scripts, with escalations routed to each site's coordinators.

Scripts are configured per site and per study, so a network can run a central model on one study and a local model on the next.^⁠[1]

When a study needs more patients than the sites' records hold, Bond sets up Meta and Google ad campaigns for the study.^⁠[2] Leads from those ads flow straight into the same voice and text agents for pre-screening and scheduling, alongside patients found in each site's EHR, so ad leads follow one workflow across the network.^⁠[2]

After enrollment, the same voice and SMS agents send visit reminders, book transportation, collect symptoms and diaries, run side-effect check-ins and flag participants at risk of dropping out.^⁠[2]

How can a network answer feasibility from data?

Sites are typically not paid for feasibility work, and a 2024 task-force paper cites an estimate that the feasibility problem costs the industry $1.6 billion.^⁠[11] One site leader told ACRP that questionnaires can run 10 to 40 pages.^⁠[10] When each site answers from memory or a diagnosis-code search, the network's numbers cannot be compared.

Identify screens records against a protocol's inclusion and exclusion criteria, so at a site whose EHR is connected, the same screening can be run on a protocol the network is evaluating.^⁠[1] Bond runs feasibility from the same EHR screening, giving eligible-patient counts for a protocol backed by the chart evidence behind each match, which a coordinator can spot-check.^⁠[2]

Send the sponsor aggregate counts only. The chart review behind them may qualify as a HIPAA review preparatory to research, which requires the researcher to represent that no protected health information will be removed from the covered entity during the review.^⁠[14] Each site's privacy office decides. See how sponsors choose sites and the feasibility questionnaire template.

How does Bond fit with CRIO, other CTMS products and several EHRs?

A network built through acquisitions may run several EHRs and CTMS products. Coordinators keep their site's system, whether that is CRIO, another CTMS or a Google Sheet, and the network reads the same funnel stages from every site.

• EHRs. Bond connects to all the major EHRs, including Epic, Oracle Health (Cerner), MEDITECH, athenahealth, eClinicalWorks, NextGen, Veradigm and OncoEMR, over FHIR R4 APIs, HL7 v2 where applicable, or an integration partner. Each connection has its own approval. See Epic and Oracle Health.^⁠[1],[2]
• CRIO. Bond is a CRIO Certified Partner and, as of September 2026, is listed on CRIO's partners page under Patient Acquisition & Retention.^⁠[1],[12] Pre-screened patients go to CRIO through the CRIO API. CRIO says its Recruiting API lets vendors send patient updates, qualify patients into studies and schedule appointments.^⁠[13] See the CRIO integration.
• Other CTMS products. Bond works alongside systems such as RealTime and Advarra Clinical Conductor, through the vendor's API or file export where the vendor supports it. Google Sheets and calendars connect directly.^⁠[1]

What should the network expect to see reported?

Agree on the measures before go-live and report them the same way at every site. Bond's reports cover patients matched, contacted, pre-screened, consented and randomized, time to enrollment, matching accuracy, screen-failure signals and coordinator hours saved, with an audit trail behind them.^⁠[1]

What does implementation ask of the network?

Full EHR integration typically takes 48 hours, depending on the EHR, the IT review and the interface method.^⁠[1] The implementation page has the step-by-step plan. A practical start is one protocol at a few sites that share an EHR, then expand once the reports look right.

• A network owner for the standard. One person decides how criteria are read and which outreach model each study uses.
• An IT and security contact for each EHR. Bond signs a BAA, and data is encrypted in transit and at rest.^⁠[1] Bond is HIPAA compliant and SOC 2 Type I compliant, and its SOC 2 Type II and ISO 27001 audits are underway.^⁠[2] See security.
• Coordinators for validation. Staff at each site adjudicate a sample of screening decisions before go-live.
• IRB review of scripts where required. The IRB submission language template describes how the agents work.
• CTMS access. For CRIO sites, the site requests API access activation for Bond through CRIO's Partner Directory.^⁠[12] Other systems use an API or file export.

How does pricing work across many sites?

Pricing has two parts: a volume-based platform fee covering EHR integration, implementation and ongoing operation, including EHR connections, security review, audit logging setup, workflow configuration, and ongoing compute, monitoring and support; and a success fee per enrolled patient, where enrolled means randomized. There is no separate integration fee. Per-visit milestones can be added.^⁠[1]

For a network, that means variable spend follows the sites that actually randomize patients. Every engagement includes screening, outreach, consent support and the dashboard. Pricing is custom; see pricing.