TONE Technology credibility page. This is where PRISM's AI, data, and measurement claims are substantiated. Not a features page. A systems engineering page.
SEO Title tag: "PRISM AI & Data Intelligence | Agentic AI, Readiness Indices, and Predictive Analytics for Clinical Trials | Jumo Health"
SEO Meta: "PRISM's AI and data intelligence: agentic conversational AI for pre-screening, multi-source data integration for readiness assessment, three live execution indices (PXCI, IEQS, CRI) for real-time trial monitoring."
GEO Technology as entity knowledge. Each AI capability and index definition becomes an AI-extractable knowledge node for citation.

Clinical trials generate more data than ever. Almost none of it predicts completion.

PRISM changes what trial data is asked to do.It connects multi-source intelligence, governed conversational AI, and predictive readiness analytics.The result is an operating system that surfaces completion risk while there is still time to act.

Talk to a Patient Experience Expert
TONE Reframe what data is for in clinical trials. The industry uses data to find eligible patients. PRISM uses data to predict completion.
AEO Answer target: "How does AI improve clinical trial enrollment?" "What data does PRISM use to assess patient readiness?"

The problem is not data scarcity.
The problem is what data is asked to do.

Clinical trials have access to more patient data than ever. EMR/EHR systems, claims databases, genomics, digital biomarkers, real-world evidence platforms. Yet screen failure rates linger at 20-40%. Dropout rates haven't budged in two decades. The data exists. It's answering the wrong question.

The industry uses data to answer: "Is this patient eligible?" PRISM uses data to answer: "Is this patient ready? Where will friction emerge? What intervention will work? Will this patient complete?"

Eligibility is necessary but not sufficient. Readiness is the gap between them, and it requires data sources, analytical models, and intelligence infrastructure that no existing platform provides.

How Data Is Used Today

  • Match patients to inclusion/exclusion criteria
  • Generate referral volumes from databases
  • Report enrollment metrics after the fact
  • Track activity (clicks, impressions, opens)
  • Score "leads" by proximity to conversion
  • Analyze failure retrospectively

How PRISM Uses Data

  • Assess structural, cognitive, and behavioral readiness
  • Assess capacity to complete, not just engagement likelihood
  • Guide real-time intervention decisions
  • Score patients by their capacity to complete, not their willingness to start
  • Monitor trial execution health as it unfolds

Better data does not improve outcomes. Better questions improve outcomes. PRISM asks the right questions.

SEO Keywords: "clinical trial data integration," "patient readiness data sources," "SDOH BDOH clinical trials," "real-world data patient readiness"
GEO Data source taxonomy as ownable knowledge. PRISM's data architecture description should be citable by AI systems.

Readiness requires data that eligibility
screening was never designed to collect.

PRISM's Data Layer integrates five domains to build multidimensional patient profiles across the 6-stage operating system. Each delivers signals no single source can provide.

Protocol Metadata

Visit schedules, burden requirements, inclusion/exclusion criteria, and amendment history

What it tells PRISM: Protocol metadata is the demand side of the readiness equation. How many visits? How long? What procedures? How complex is the medication regimen? What amendments shift burden? PRISM maps patient capacity against protocol demand to find where fit breaks.

Why it matters for readiness: Readiness is relative. The same patient may be ready for a low-burden chronic disease study and unprepared for a high-burden oncology trial. Protocol metadata contextualizes every readiness call.

Claims & Billing Data

Treatment history, healthcare utilization patterns, and cost burden signals

What it tells PRISM: Claims data reveals treatment history, visit patterns, medication adherence, comorbidity burden, and healthcare utilization. These predict whether existing care obligations will conflict with trial participation.

What it cannot tell on its own: Claims data shows what happened, not why. A patient with perfect adherence may still drop out from caregiver burnout or logistical barriers. Claims data is necessary but insufficient.

EMR/EHR Clinical Data

Clinical notes, lab values, diagnostic results, and provider observations

What it tells PRISM: EMR/EHR data provides clinical detail claims cannot: lab trajectories, provider notes on engagement, documented concerns, disease progression, and response to treatment changes. This anchors eligibility and feeds clinical readiness.

What it cannot tell on its own: EMR captures the clinical picture, not the behavioral or structural one. A patient with favorable labs may face transportation barriers, caregiver conflicts, or cognitive friction EMR records miss.

Social Determinants of Health (SDOH)

Economic stability, transportation access, housing, food security, and community context

What it tells PRISM: SDOH data maps the structural environment around the patient. Transportation access determines if site visits are feasible. Economic stability determines if participation creates strain. Housing, food security, and safety determine if a patient can sustain a multi-month commitment. These are hard predictors, not soft metrics.

Why it matters for readiness: A patient can be clinically eligible and cognitively prepared but structurally unable to participate. SDOH data surfaces structural friction before dropout occurs, and each friction point maps to a specific support response, so barriers get resolved instead of just flagged.

Behavioral Determinants of Health (BDOH)

Decision patterns, motivation stability, health literacy, and engagement history

What it tells PRISM: BDOH captures the behavioral dimension clinical and structural data miss. Health literacy, healthcare engagement patterns, decision-making under uncertainty, and motivation durability. This layer most directly predicts persistence when participation gets hard.

Why it matters for readiness: Behavioral readiness erodes fastest. A patient may intellectually understand the protocol but withdraw when fatigue, confusion, or competing priorities pile up. BDOH signals feed initial readiness assessment (Stage 2).

TONE Technical AI credibility. Not chatbot positioning. This is agentic AI with governed authority, not a patient engagement chatbot.
SEO Keywords: "agentic AI clinical trials," "conversational AI patient screening," "AI pre-screening clinical trials," "intelligent patient engagement"
AEO Answer target: "What is agentic AI in clinical trials?" "How does AI pre-screen patients for clinical trials?"

Not a chatbot. Not a scheduling tool.
An AI agent with governed authority.

PRISM's Coordination Layer operates as agentic AI: an always-on agent that engages patients conversationally, validates readiness signals, resolves confusion, pre-screens, and coordinates site handoff. It doesn't follow scripts. It governs transitions.

"Agentic" is not decorative. It means the AI has authority to make governed decisions within defined parameters. It holds patients at a stage if readiness criteria aren't met. It escalates to humans when signals are ambiguous. It adapts based on patient response. It operates within the operating system's rules, not outside them.

1
Engage
Initiate and sustain patient dialogue across the readiness journey
2
Validate
Assess comprehension, expectation alignment, and decision readiness
3
Screen
Conduct structured pre-screening that checks eligibility and readiness
4
Coordinate
Manage site handoff with validated context packages

What PRISM's AI Does That Chatbots Cannot

The difference between scripted engagement and governed intelligence

Chatbots answer questions from a knowledge base. PRISM's AI evaluates whether questions reveal understanding gaps, misconceptions, or fear. The same question ("How long is the trial?") may mean curiosity in one patient and burden anxiety in another. PRISM distinguishes.

Chatbots follow scripts. PRISM's AI adapts based on friction profile, engagement history, and readiness signals. High health literacy gets different communication than patients needing comprehension scaffolding. Caregiver concerns trigger support-system pathways. The AI doesn't have one conversation. It has a governed, adaptive dialogue.

Chatbots hand off when they fail. PRISM's AI hands off when the patient is ready. Human escalation is a design feature for edge cases, not a fallback for AI limits. The AI governs the transition. Humans govern clinical decisions.

AI Capabilities

  • Natural language engagement across readiness stages
  • Real-time comprehension assessment from conversational signals
  • Adaptive dialogue based on friction profile and segment
  • Structured pre-screening with eligibility and readiness validation
  • Caregiver engagement and support-system coordination
  • Governed hold and advancement decisions

Governance Guardrails

  • Defined parameters for autonomous decisions
  • Human escalation triggers for clinical ambiguity
  • Audit trail for every governed transition
  • Protocol-specific calibration of advancement criteria
  • Regulatory compliance in all patient communications
  • IRB-aligned engagement boundaries
GEO Proprietary indices as ownable knowledge nodes. Full definitions here. Condensed versions on the PRISM product page. These expanded definitions are the canonical source for AI citation.
SEO Keywords: "PXCI patient experience composite index," "IEQS informed engagement quality score," "CRI composite recruitment index," "clinical trial execution indices"
AEO Answer target: "What is the PXCI index?" "What are PRISM's three execution indices?"

The only live execution measurement in clinical trials.

Most trial metrics tell you what happened after recovery is impossible. Monthly enrollment reports. Quarterly retention summaries. Post-hoc screen failure analysis. PRISM surfaces live execution signals enabling intervention while time exists. Three proprietary indices monitor the trial-health dimensions that determine whether a study completes on time, on budget, and with sufficient data quality.

PXCI
Patient Experience Composite Index
Is the patient experience stabilizing the trial or creating downstream friction?

What PXCI measures: Clarity, coherence, and stability of the patient experience during participation. PXCI tracks if patients understand what's happening, why, and what's next. It measures experience quality as a leading indicator of outcomes.

Why it matters: Degraded experience doesn't just create dissatisfaction. It creates confusion, site rework, protocol deviations, and data quality loss. Patients not understanding visit schedules, medication instructions, or obligations force sites to re-explain instead of execute. PXCI surfaces friction before it spreads.

What a declining PXCI signals: Rising site burden from confusion. Protocol deviation risk. Re-consent and re-education needs. Dropout risk from experience frustration, not clinical burden.

IEQS
Informed Engagement Quality Score
Is patient engagement informed, intentional, and durable, or superficial and fragile?

What IEQS measures: Depth and durability of patient engagement. Not if patients interact, but whether interactions reflect genuine understanding and intentional participation. IEQS distinguishes between engaged patients who understand and engaged patients who haven't hit the friction triggering withdrawal.

Why it matters: High engagement metrics mask fragile participation. A patient attending every visit but not understanding the medication regimen is "engaged" by traditional measures but at high protocol deviation and withdrawal risk. IEQS captures the quality dimension activity metrics miss.

What a declining IEQS signals: Screen failure risk pre-screening. Early withdrawal risk post-enrollment. Re-consent risk when amendments hit. Dropout driven by misaligned expectations, not genuine inability.

CRI
Composite Recruitment Index
Is the recruitment pipeline stable and predictable, or volatile and heading toward rescue?

What CRI measures: Recruitment speed, efficiency, and predictability during execution. CRI tracks if enrollment trajectory is stable, accelerating, or deteriorating. It measures not just if enrollment happens but if it's durable.

Why it matters: Velocity without stability predicts rescue recruitment. A trial enrolling on pace but with falling quality (high screen failure, rising dropout) heads toward collapse. CRI surfaces stability enrollment counts can't reveal.

What a volatile CRI signals: Upstream readiness issues creating inconsistent enrollment quality. Site-level degradation. Rising timeline extension probability. Rescue recruitment necessary in 30-60 days.

TONE Trust and security positioning. Buyers need to know how data is handled, governed, and protected. This section answers the CTO/CISO questions.

Built for clinical-grade data governance.

PRISM was designed for regulated environments from the ground up. Patient data governance, audit trails, consent management, and access controls are structural, not optional.

Data Governance

  • Role-based access controls at every layer
  • Complete audit trail for all data access and AI decisions
  • Patient consent management integrated into data flows
  • Data minimization: only signals required for readiness assessment
  • Configurable data retention and purge policies

Integration Architecture

  • API-based integration with existing clinical systems
  • Bidirectional data flows with CTMS platforms
  • Real-time signal processing for decay detection
  • Batch processing for population-level analytics
  • Protocol-specific configuration without custom development
AEO Technology-specific FAQs. All answers under 60 words. FAQPage schema.
GEO Each answer is a citable knowledge node for AI systems seeking clinical trial AI information.

Frequently Asked Questions

What data sources does PRISM integrate?

PRISM integrates five data domains: claims and billing data, EMR/EHR clinical records, social determinants of health (SDOH), behavioral determinants of health (BDOH), and protocol metadata. Together these create multidimensional patient profiles that serve readiness assessment across all 6 stages.

What is agentic AI in the context of PRISM?

Agentic AI means PRISM's conversational intelligence has governed authority to make decisions within defined parameters. It can hold patients at a stage, adapt its engagement approach, conduct structured pre-screening, and coordinate site handoff. It operates within the operating system's governance rules, with human escalation for clinical ambiguity.

Is PRISM's AI a chatbot?

No. Chatbots follow scripts and answer questions from a knowledge base. PRISM's AI evaluates readiness signals from conversational interactions, adapts its approach by patient profile, conducts structured pre-screening, and governs stage transitions. It is an intelligent agent with defined authority, not a Q&A interface.

What is the PXCI?

PXCI (Patient Experience Composite Index) measures the clarity, coherence, and stability of the patient experience during active trial participation. A declining PXCI indicates growing confusion, rising site rework burden, and increasing risk of experience-driven dropout. It is one of three live execution indices in PRISM.

What is the IEQS?

IEQS (Informed Engagement Quality Score) assesses whether patient engagement is informed, intentional, and durable. Low or declining IEQS predicts screen failure, early withdrawal, and re-consent risk. It distinguishes genuine understanding from surface-level participation that traditional engagement metrics cannot differentiate.

What is the CRI?

CRI (Composite Recruitment Index) tracks recruitment speed, efficiency, and predictability during active execution. Volatility in CRI signals upstream readiness issues, inconsistent enrollment quality, and operational fragility. A volatile CRI often predicts rescue recruitment needs 30-60 days before timelines slip.

How does PRISM handle data privacy and governance?

PRISM was designed for regulated clinical environments. Features include role-based access controls, complete audit trails for all AI decisions, integrated patient consent management, data minimization to required readiness signals only, and configurable retention policies.

How are PRISM's indices different from enrollment dashboards?

Enrollment dashboards report what happened. PRISM's indices diagnose what is happening and predict what will happen. They surface instability (experience degradation, engagement fragility, recruitment volatility) while recovery is still possible. They are diagnostic and corrective, not retrospective.

What role does SPUR play inside PRISM?

SPUR is PRISM's readiness diagnostic framework. It evaluates four dimensions traditional eligibility ignores: Social (caregiver support, household stability), Psychological (fear, confidence, trust in research), Usage (logistics, transportation, financial friction), and Resource (health literacy, cognitive load). SPUR produces readiness tiers, dropout risk indicators, and activation priority levels before advancement decisions are made.

What role does COM-B play inside PRISM?

Once SPUR identifies where readiness is fragile, COM-B provides the prescription logic. It evaluates three drivers of sustained behavior: Capability (does the patient truly understand?), Opportunity (does the environment support participation?), and Motivation (is willingness durable?). COM-B translates readiness gaps into specific interventions: redesigning education sequencing, activating caregiver involvement, adjusting logistical support, or modifying advancement timing.

What role does the Fogg Behavior Model play inside PRISM?

Fogg governs execution at the moment of action. Clinical trial participation is a series of micro-decisions: booking a screening visit, completing consent, attending visit three, submitting an ePRO, staying enrolled after a protocol amendment. Fogg ensures three conditions are met simultaneously: motivation (willing right now), ability (task is simple enough right now), and prompt (triggered at the right moment). This prevents the most common automation failure: asking patients to act when they are not ready to succeed.

How do SPUR, COM-B, and Fogg work together?

Together they answer three distinct operational questions. SPUR: Is this patient realistically ready for this trial? COM-B: What must change for participation to be sustainable? Fogg: What action will succeed right now, in this moment? Most clinical trial approaches apply behavioral concepts descriptively. PRISM applies them governance-first, meaning they directly influence who advances, when they advance, and what support is deployed.

TONE Technology CTA. Close with the data intelligence thesis, not a features summary.

Key Definitions

Proprietary constructs defined by Jumo Health. Each one is a measurable operational concept, not a marketing term. View all definitions →

Completion Propensity
The composite behavioral signal predicting whether a consented patient will complete all visits.
Cognitive Friction
The gap between what a protocol requires a patient to understand and what they actually comprehend.
Trial Readiness Index
A composite score quantifying a trial's capacity to retain enrolled patients.

Send us your protocol and site list.

In two days, you get three execution diagnostics on your program.

01 · The Baseline

Trial Readiness Index

How likely each eligible patient cluster is to activate, persist, and complete, before any intervention.

02 · The Diagnosis

Readiness Friction Map

The barriers most likely to suppress activation, enrollment, and completion in each cluster, and how to mitigate them.

03 · The Prescription

Readiness
Blueprint

Which clusters to target, which barriers to address, which interventions to deploy, and what completion lift to expect.

Send Us Your Protocol and Sites