Most patient education searches start with the wrong question.

Teams ask which vendor is best before they have decided what they are actually buying, and "patient education vendor" covers at least four different kinds of product: clinically reviewed content libraries, software that organizes and assigns that content, video and multimedia platforms, and broader patient-engagement systems where education is one module among many.

That ambiguity has cost consequences.

AVIA Marketplace's patient education category describes a market shift away from solutions focused only on content creation and toward solutions that function as delivery mechanisms, and advises health systems to decide whether they need better content, a delivery mechanism, or both.

AVIA's consumerism lead also cautions that purchasing decisions should sit inside the organization's digital strategy so a digital education solution does not become one more siloed thing patients receive. The first decision is category and implementation model. Vendor shortlisting comes second.

This guide is written for that first decision, and for the procurement work that follows.

What is a patient education vendor?

A patient education vendor supplies healthcare organizations with some combination of clinically reviewed education content, software to organize and deliver it, multimedia and video, personalization, workflow or EHR integration, analytics, and patient-facing access. Very few vendors do all of it well, and some do only one part.

Four rough categories are useful for building a shortlist.

Content-first vendors. The core product is the clinical education content itself: condition, procedure, and medication material, licensed for use inside systems the organization already runs.

Education platforms. Content plus the software layer that assigns, personalizes, tracks, and reports on it.

Patient-engagement and care-journey platforms. Education is one component of a wider workflow that may include navigation, reminders, messaging, monitoring, or intake.

Specialty education platforms. Depth in one area, such as therapy and rehabilitation, medication education, or a specific service line.

The boundaries are not clean. KLAS notes that some organizations produce their own educational content and house it in a vendor's solution, such as a portal or interactive patient system, rather than licensing the vendor's library at all. Decide which of these patterns describes your intended end state before you read a single vendor deck.

Why vendor selection is harder than comparing library size

Library counts are the easiest number to publish and the least informative one to evaluate. The harder variables are clinical governance, comprehension, accessibility, language access, workflow fit, analytics, privacy, and implementation, and they differ sharply across the four categories above.

Independent market research supports treating this as a portfolio decision rather than a single purchase.

In its market report focused on patient education, KLAS found that while many organizations use a single patient education provider, several run multiple solutions to meet specialty needs such as orthopedics, pediatrics, and rehabilitation, and that customers across measured vendors want more specialty content, more data analytics, and more language options without added expense.

Note that this report predates Elsevier's acquisition of Mytonomy and WebMD Ignite's acquisition of Healthwise, so its vendor groupings should be read with those changes in mind.

KLAS also reported a finding that should temper any library-size conversation: organizations saw positive results such as increased engagement and satisfaction, while many patients found the content lengthy and not user-friendly.

Volume is not comprehension. The criteria below are ordered roughly by how much damage a bad answer causes.

1. Evaluate clinical content quality and governance

Start here, because everything downstream inherits whatever the content gets wrong.

Ask who writes the material, who reviews it clinically, and what credentials the reviewers hold. Ask what evidence hierarchy the editorial process uses, how often each topic is reviewed, and how quickly a major guideline change or safety communication reaches published patient-facing content.

Ask whether revision history is available and auditable, whether your own clinical teams can approve, modify, or suppress specific topics, and whether medication information goes through a separate review track.

If the vendor uses AI to summarize, simplify, or translate clinical content, ask exactly where human clinical review sits in that pipeline and who signs off before patients see the output.

Vendors do publish governance detail, and it is worth reading closely rather than accepting the summary.

Wolters Kluwer states that UpToDate patient education topics are developed in-house and edited by the same physicians who edit its clinical topics, so the patient material is consistent with the evidence-based content clinicians use for decisions.

WebMD Ignite states that all of its 25,000-plus health education pieces meet standards for accuracy and accessibility. Both are vendor statements. Treat them as claims to verify against a documented editorial policy, named reviewer credentials, and a review calendar you can inspect.

"Evidence based" and "clinically validated" on a slide are not governance. A documented editorial process, with dates and names, is.

2. Test for health literacy and actual comprehension

AHRQ's Health Literacy Universal Precautions Toolkit takes the position that organizations should structure health information and services so that everyone can understand and use them, simplifying communication and confirming understanding with every patient rather than trying to predict which patients will struggle. That reframes the vendor question from "what reading level do you target?" to "how do you know patients understood?"

Evaluate plain language, sentence length, familiar terminology, action-focused instructions, information hierarchy, and whether visuals carry meaning or decorate the page. Then ask the harder questions: has this content been tested with patients, in which populations, and what changed as a result?

Does the material support teach-back, where a clinician asks the patient to explain the plan in their own words? AHRQ is explicit that "Do you understand?" and "Does that make sense?" are not teach-back questions, because patients tend to answer yes either way.

Reading-level targets are a useful communication benchmark, and many vendors publish one.

Wolters Kluwer offers two tiers of patient material, with The Basics written in simple language to answer the four or five most important questions and Beyond the Basics running longer and more detailed, and states that patient information from Lexidrug is written for a fifth-to-seventh-grade reading level and delivered in English, Spanish, and 18 additional languages.

Useful comparators, but no single grade-level target is a universal federal requirement for patient education. Treat readability formulas as one input and patient testing as the real evidence.

3. Treat accessibility as a core procurement requirement

Accessibility belongs in the requirements document, not in a feature-comparison footnote, and for many US healthcare organizations it now carries a dated compliance obligation.

HHS's Section 504 rule sets standards for the accessibility of web content and mobile applications of organizations that receive HHS federal financial assistance.

On May 7, 2026, HHS's Office for Civil Rights issued an interim final rule extending the compliance dates by one year, moving the deadline for recipients with 15 or more employees from May 11, 2026 to May 11, 2027, and for recipients with fewer than 15 employees from May 10, 2027 to May 10, 2028.

HHS framed the extension as a response to reports that a significant number of recipients, including community health centers and hospitals of varying sizes, would not be able to meet the original deadlines for conforming to WCAG 2.1 AA. The extension changed the dates, not the requirement.

Analysis of the rule also notes that it permits alternative methods of achieving accessibility where those methods provide an equal or greater level of access than WCAG 2.1 Level AA.

Which rules apply to your organization is a legal determination, not a vendor claim. Recipients of HHS funding, public entities covered by ADA Title II, and private organizations with no federal funding do not all carry identical obligations, and some carry more than one. Work that out with counsel, then use WCAG as the technical reference for testing.

For vendor evaluation, request an accessibility conformance report and then verify it against the artifacts patients actually receive: screen-reader compatibility across the real assignment and viewing flow, keyboard navigation and visible focus, contrast, zoom and text resizing without loss of content, properly structured and tagged PDFs, captions and transcripts on video, audio description where visual content carries instruction, meaningful alt text, large-print output, and audio alternatives.

Ask what was tested, by whom, with which assistive technology, and when.

One qualitative point from community discussion is worth keeping in the room during these conversations. A healthcare social worker described the difficulty of finding usable education materials for a patient with low vision who had no smartphone, no home internet, and limited digital literacy.

That is firsthand practitioner experience rather than prevalence data, and it is exactly the case a demo never covers. If your population includes patients like that, print, large print, audio, and caregiver-mediated formats are requirements, not fallbacks.

4. Evaluate multilingual content and language access

Section 1557 of the Affordable Care Act sets the frame.

HHS states that the rule requires covered entities to take reasonable steps to provide meaningful access to each individual with limited English proficiency who is eligible to be served or likely to be encountered in their health programs and activities, and applies to any health program or activity receiving HHS funding, including hospitals that accept Medicare and providers receiving Medicaid payments.

A multilingual content library contributes to that obligation. It does not discharge it. Meaningful language access is an operational program covering interpretation, qualified translation, notices, and workflow, evaluated against the population you actually serve.

For the vendor specifically, ask which languages are covered and at what depth, because coverage often differs between articles, videos, subtitles, and printed handouts. Ask who performs translation, whether a qualified human translator reviews clinical content, and how translated material stays in parity with the English source when the source is updated.

The machine-translation question deserves a direct answer: the Section 1557 implementing rule addresses machine translation and, as summarized by the American Translators Association, requires review by a qualified human translator where the underlying text is critical to an individual's rights, benefits, or meaningful access, or where accuracy is essential. Then ask how patient language preference is captured, stored, and honored automatically at the point of assignment.

Cost is a live issue here. KLAS reported that customers want more language options without added expenses, which suggests language coverage is frequently a paid add-on. Get the pricing in writing during evaluation.

5. Test EHR and clinical workflow integration

Integration is where demos are most persuasive and most misleading. An available integration is not the same thing as an efficient workflow.

Vendors do publish concrete integration detail.

WebMD Ignite states that its education can be distributed through Epic, Oracle/Cerner, and MEDITECH, and that Ignite on FHIR uses SMART on FHIR technology to integrate into Epic and into clinician workflow, serving content based on the patient's profile and clinical encounter.

In the therapy and rehabilitation segment, Medbridge states that its programs support EMR integrations, with additional integrations supported via Redox. AVIA similarly lists flexible integrations with EHR systems and other care platforms among the capabilities of leading solutions.

None of that tells you whether a nurse can find and assign the right material in the middle of a discharge. Test that yourself, with your own clinicians, in your own build:

• How many clicks and how many seconds from the encounter to an assigned, patient-appropriate resource?
• Can assignment be triggered from structured clinical data such as diagnosis, procedure, or order, rather than free-text search?
• Does completion or acknowledgment return to the clinical record, and in what form?
• Can locally authored content live alongside the licensed library in the same search and assignment flow?
• What breaks when you change your own EHR configuration, upgrade a version, or restructure a discharge workflow?
• Is single sign-on in place for staff, and are clinician favorites or templates supported?

Ask for named reference sites running your exact EHR, your version, and your workflow. "We integrate with Epic" is a starting point for a conversation, not an answer.

6. Compare the patient delivery experience

Education that reaches no one produces no benefit, so map delivery routes against your actual population rather than an idealized one. Portal, mobile web, native app, SMS, email, bedside screen or in-room television, QR code, print, video, interactive module, and caregiver sharing are all legitimate channels, and most organizations need several.

Vendors position differently here. WebMD Ignite describes reaching patients with education in the Epic portal, putting education into clinical workflow, and automating educational outreach by SMS. Medbridge describes giving patients access by text, email, or QR code. Wolters Kluwer describes content that integrates into workflows and is accessible through patient portals, email, and print.

Do not assume smartphone ownership, reliable internet, or comfortable digital literacy. Ask what the experience is for a patient with none of the three, and whether authorized caregivers can receive the same education on the patient's behalf.

7. Evaluate video and multimedia carefully

Video is a strong format and an oversold one. The evidence supports the first half of that sentence more confidently than vendors usually acknowledge.

A systematic review of video-based educational interventions for patients with chronic illnesses, published in the Journal of Medical Internet Research, found that video-based tools were most effective for patient knowledge, improving it in 30 of 40 reported outcomes, while roughly half of health behavior outcomes and patient self-efficacy outcomes improved, and a minority of health care use outcomes and disease severity outcomes improved.

The authors also reported that 48% of the superiority and noninferiority randomized trials and 54% of the pre-post studies carried moderate or high risk of bias, and concluded that there is robust evidence video-based tools improve patient knowledge across several chronic illnesses, with less consistent improvement in disease severity and health care use.

The defensible claim is that video can improve understanding and knowledge. Claims that a video platform reduces readmissions require a specific intervention studied in a specific population, and you should ask for that study by name.

Format-wise, assess clinician-presented video, animation, demonstrations, procedure preparation, medication education, rehabilitation exercise video, and interactive modules against your service lines. Also ask how video is captioned, transcribed, translated, and updated, because multimedia is the most expensive content type to keep current.

8. Determine how personalization actually works

"Personalized" spans a wide range. It helps to grade it:

1. A clinician selects a generic resource manually.
2. Content is mapped to a diagnosis or procedure.
3. Content adapts to care stage, such as pre-op, discharge, or follow-up.
4. Delivery adapts to language, channel, and format preference.
5. Assignment is triggered automatically from structured clinical data.
6. Content adapts to patient behavior or progress over time.

Then ask what drives it. Which patient data feeds the logic, and does any of it constitute PHI? Can a clinician override an automated assignment? Can the patient tell why they received a given resource? Where generated text is involved, who reviews it before it reaches a patient, and is that review logged?

KLAS has noted interest in AI tools that can translate medical information into a patient's preferred language, tone, or comprehension level. That capability is genuinely useful and genuinely risky. Require a described governance path for any generated clinical explanation rather than accepting "AI-powered" as a feature.

9. Understand analytics before accepting engagement claims

Build a measurement hierarchy and make the vendor state which tier their metrics occupy.

• Delivery: assigned, sent, delivered.
• Engagement: opened, viewed, watch percentage, completed.
• Understanding: questions asked, teach-back results, knowledge checks.
• Behavior and process: appointment preparation, follow-up completion, adherence proxies.
• Clinical and utilization: complications, readmissions, emergency visits, condition-specific outcomes.

A 90% video completion rate is a delivery-and-engagement metric. It does not establish clinical improvement.

Where education does move clinical outcomes, effect sizes vary widely by intervention and population: a systematic review and meta-analysis of therapeutic patient education in obesity and diabetes found significant improvement in HbA1c and body weight alongside substantial heterogeneity in reported outcomes.

Expect that variability, ask which outcomes the vendor's platform has been studied against, and instrument your own measurement rather than inheriting a dashboard.

Practically, ask whether you can export raw event data, join it to your own clinical data, and define your own cohorts. Recall that KLAS customers across vendors were asking for more data analytics, which suggests out-of-the-box reporting frequently falls short of what health systems need.

10. Map privacy and HIPAA responsibilities to the actual data flow

This section needs precision, because the shorthand version is wrong.

HIPAA obligations follow the data. A vendor is not automatically a business associate because it sells software to a covered entity; what matters is whether the vendor creates, receives, maintains, or transmits PHI on the covered entity's behalf.

Where it does, HHS states that covered entity and business associate customers must obtain satisfactory assurances in the form of a business associate agreement that the vendor will appropriately safeguard the PHI it creates, receives, maintains, or transmits, and that subcontractors handling PHI on a business associate's behalf are themselves business associates.

There is also no such thing as a HIPAA-certified product.

HHS is direct about this: HHS and its Office for Civil Rights do not endorse private consultants' or education providers' seminars, materials, or systems, and do not certify any persons or products as Privacy Rule compliant, and OCR does not endorse, certify, or recommend specific technology or products.

HHS also notes that the HIPAA Rules do not endorse or require specific technologies, but set standards for how covered entities and business associates may use or disclose ePHI, requiring risk analysis and reasonable safeguards.

Work the data flow instead:

• Does the system receive PHI, and which elements?
• Does it create or maintain PHI, including assignment and viewing records tied to identifiable patients?
• Does it transmit PHI, and to whom?
• Is the vendor therefore a business associate, and is a BAA in place covering the actual services?
• Which subcontractors touch the data, and are they under BAAs?
• Where is data stored and processed?
• What retention and deletion controls exist, and who can invoke them?
• How is access controlled and audited?
• What is the incident response and breach notification process, with what timelines?
• Can you export your data, and the configuration around it, at termination?

11. Compare implementation and ongoing support

The contract is the easy part. Name the implementation owner on both sides, then get specifics on integration timeline, content migration, local branding, administrator configuration, local clinical review, accessibility testing, staff training, support SLAs and escalation, change management for clinical teams, release management and update cadence, and analytics configuration.

Clinician-facing community discussion is consistent on the two practical failure modes: time and discoverability. Nursing and clinical threads repeatedly describe limited visit and shift time and difficulty locating the right education quickly.

That is qualitative pain-point evidence rather than measured prevalence, and it points to the same conclusion as AVIA's warning about siloed tools: a feature demonstration is not a workflow test. Put a real nurse in a real build with a real patient scenario and time it before signing.

12. Understand the full cost, not only the license price

Model the total cost across base subscription, licensing model (per user, per site, or enterprise), implementation, integration work, premium content tiers, specialty modules, additional languages, video packages, analytics, customization, API access, content migration, and support. Ask directly which integrations, languages, modules, or media formats carry incremental fees, since that is where budgets break in year two.

Public pricing is rarely available for enterprise patient education agreements. Where a vendor does not publish it, get the full fee schedule in the RFP response rather than estimating from third-party sources.

The patient education vendor landscape

The categories below are descriptive, not ranked. There is no independent testing behind any ordering here, and none should be inferred.

Enterprise clinical content and patient information providers.

WebMD Ignite now includes Krames and Healthwise, following WebMD Health Corp.'s February 2024 acquisition of Healthwise's operating assets, including content, products, technology, and client relationships; its stated model centers on content delivered through EHRs, portals, CRM, and SMS.

Wolters Kluwer offers UpToDate patient education in two reading tiers edited by the same physician editors as its clinical content, plus UpToDate Educate for patient engagement with delivery through workflows, portals, email, and print.

Elsevier offers patient education within its Clinical Solutions portfolio, including PatientPass.

KLAS's market report grouped these established vendors by relative strengths, noting strong written content from Healthwise and Krames, smooth content-sharing workflows from Wolters Kluwer's integration, and digestible patient content from Elsevier.

Digital education and multimedia-focused platforms. Mytonomy is the clearest example, and its ownership has changed.

Elsevier completed its acquisition of Mytonomy on May 5, 2026 and is integrating it into its Clinical Solutions portfolio alongside PatientPass; Mytonomy's platform was described at the time as used by more than 70 hospitals and 200 clinics across the United States, delivering video-based content across oncology, cardiology, orthopedics, maternity, behavioral health, pediatrics, and chronic disease management.

KLAS had previously characterized Mytonomy as a newer vendor known for engaging short-form video content. Elsevier has said both Mytonomy Cloud for Healthcare and PatientPass will continue to be supported, with deeper integration into its Clinical Solutions portfolio over time.

So the product persists, but the roadmap is now Elsevier's: evaluate it on that basis rather than as an independent vendor, and ask what the planned integration means for your renewal terms and support commitments.

Specialty platforms.

Medbridge is the common example in therapy and rehabilitation, combining a library of video-based exercises and patient education resources on an engagement platform, with programs delivered by text, email, or print and home exercise programs, remote therapeutic monitoring, and patient-reported outcomes collection.

Other providers specialize in medication education, specific procedures, or single service lines. Specialty depth is frequently the reason organizations end up running more than one vendor.

Engagement and care-journey platforms. In AVIA's framing, these are solutions where education is one component of broader engagement, navigation, monitoring, or communication. The buying question shifts accordingly: you are evaluating a platform strategy, not a content license.

For every vendor on your list, record only verifiable facts: product focus, content formats, stated integrations, stated languages, and documented functionality. Awards, testimonials, and marketplace profiles are marketing and market research respectively. Neither is a controlled study.

Where HeyGen fits, and where it does not

HeyGen is not a patient education content vendor, and it should not be shortlisted as one. There is no clinical content library, no editorial board, and no clinical review process. What it provides is a video production and localization layer for education your own clinicians have already written and approved, which is a different line item in a different part of the budget.

That distinction matters most in three places in the framework above.

Language coverage is the clearest one. Translating an existing library of clinician-recorded video is normally a reshoot-and-revoice project, which is why vendor language counts are often thin on the video side while article coverage is broad. HeyGen's video translator takes an existing video or a YouTube link and produces versions in 177+ languages with lip sync, voice preservation, and subtitles, and videos in a collection can use a multilingual embedded player that drops into a page or an LMS.

For organizations that only need the audio track localized, AI dubbing covers that path, and the same workflow is available programmatically through the video translation API for teams localizing a library rather than a single asset. None of that satisfies Section 1557 on its own. Where translated material is critical to a patient's rights, benefits, or meaningful access, qualified human translator review still applies, and HeyGen output should route through the same clinical and linguistic sign-off as any other translated patient material.

Production capacity is the second. Producing new clinician-presented video is slow enough that most education programs ration it, and outdated video is worse than no video when a guideline changes. Generating a presenter-led explainer from an approved script through an AI video generator shortens the cycle from script to published asset, which is what makes quarterly content refreshes realistic rather than aspirational.

The clinical bottleneck stays exactly where it belongs: a clinician writes and approves the script, and nothing publishes without that sign-off.

The third is what HeyGen cannot do for you, which buyers should hear plainly.

It does not assign education inside Epic, Oracle Health, or MEDITECH, does not write completion back to the clinical record, does not provide clinically reviewed source content, and does not carry the analytics hierarchy described earlier.

Accessibility remains your responsibility too: captions and transcripts help, but audio description, alternative formats, and the accessibility of whatever player or portal hosts the video are yours to test.

And because HeyGen is a general-purpose video platform rather than a healthcare product, run the same data-flow analysis from section 10 before anything patient-identifiable goes near it. The safest default is to keep PHI out of scripts and source footage entirely and to confirm enterprise security, access control, and contracting terms directly with HeyGen rather than assuming a BAA is available.

Read that as a complement to a patient education vendor, not a replacement for one. A reasonable architecture is a governed content and delivery platform integrated with the EHR, plus a production and localization layer for the locally authored video that no library will ever cover. If you want to see the production side before deciding whether it belongs in your stack, HeyGen's free AI video tools are the low-commitment place to start.

RFP checklist

Related questions

What is a patient education vendor?

A patient education vendor supplies healthcare organizations with clinically reviewed patient education content, software to deliver it, or both. Offerings range from licensed content libraries to platforms that assign, personalize, translate, and track education, sometimes as part of a broader patient engagement system.

What should hospitals look for in patient education software?

Clinical content quality and documented editorial governance first, then health literacy and comprehension testing, accessibility conformance, language access depth, clinician workflow and EHR integration, patient delivery channels including non-digital options, analytics that distinguish delivery from outcomes, a clear PHI and business associate analysis, implementation and support commitments, and independent evidence behind any outcome claim.

How does patient education software integrate with an EHR?

Two broad patterns. In manual assignment, a clinician searches a library inside or beside the EHR and sends material to the patient or prints it.

In integrated assignment, content is triggered by structured data such as a diagnosis, procedure, order, or discharge event, often via SMART on FHIR or a vendor-specific interface, with delivery through the patient portal and an acknowledgment written back to the record. Ask which pattern the vendor runs in production on your EHR version, and test the time and clicks yourself.

Does patient education improve health outcomes?

Patient education reliably supports understanding and knowledge. Effects on behavior, adherence, utilization, and clinical outcomes vary considerably by intervention and population.

The JMIR systematic review of video-based education in chronic illness found knowledge improved in most reported outcomes while health care use and disease severity improved in a minority, with notable risk of bias across studies. Do not assume any given platform reproduces the results of a specific studied intervention.

Does a patient education vendor need to be HIPAA compliant?

The right question is whether the vendor will create, receive, maintain, or transmit PHI on your behalf. If it will, it is a business associate, a BAA is required, and its subcontractors handling PHI are business associates too. If it will not, the analysis and the contractual requirements differ.

No product is HIPAA certified: HHS and OCR state they do not certify persons or products as compliant and do not endorse specific technologies.

How should you evaluate patient education effectiveness?

Separate the tiers and measure each on its own terms. Delivery tells you the material was sent. Engagement tells you it was opened or viewed. Comprehension requires knowledge checks, patient questions, or teach-back.

Behavior requires process measures such as appointment preparation or follow-up completion. Clinical outcomes require condition-specific measures and a study design that can support the claim. Vendors most often report the first two tiers, so plan to instrument the rest yourself.