The CY 2027 OPPS Proposed Rule: Surviving the IPO List Exodus with Outpatient CDI

CDI Services

The CY 2027 OPPS Proposed Rule: Surviving the IPO List Exodus with Outpatient CDI

Jul 13, 2026 by Liyakath Ali 0 Comments , , , ,

The Centers for Medicare & Medicaid Services (CMS) recently released the Calendar Year (CY) 2027 Hospital Outpatient Prospective Payment System (OPPS) proposed rule on July 2, 2026. While the proposal includes a modest 2.4% increase in OPPS payment rates, the underlying regulatory shifts represent a massive disruption to hospital margins.

Between aggressive site-neutral payment expansions and sweeping changes to the Inpatient-Only (IPO) list, hospitals must fundamentally rethink their Clinical Documentation Improvement (CDI) strategies. In 2027, protecting your revenue will rely entirely on the precision of your outpatient data management.

The IPO List Exodus: 637 Procedures Moving to Outpatient The most alarming takeaway from the CY 2027 proposed rule is the continuation of CMS’s three-year phase-out of the IPO list. For CY 2027, CMS intends to remove 637 procedures from the IPO list, assigning them instead to clinical ambulatory payment classifications (APCs).

When highly complex procedures—especially those in orthopedics and cardiology—shift from inpatient to outpatient settings, the documentation requirements change drastically. Clinicians must meticulously capture comorbidities, severity of illness (SOI), and risk of mortality (ROM) to justify inpatient admissions when clinically necessary or to secure maximum outpatient reimbursement. Legacy, inpatient-focused CDI programs are simply not equipped to handle this volume of ambulatory scrutiny.

Margin Squeezes: 340B Cuts and Site-Neutral Imaging The margin pressure doesn’t stop at the IPO list. The CY 2027 OPPS proposed rule targets two other massive revenue drivers:

  • 340B Drug Payments: CMS proposes to slash Medicare payments for 340B-acquired drugs to average sales price (ASP) minus 33.4%. Furthermore, CMS plans to accelerate the 340B remedy budget neutrality adjustment, increasing the annual reduction from 0.5% to 3% effective January 1, 2027.
  • Site-Neutral Imaging: CMS is proposing to expand site-neutral payment policies to include imaging services without contrast when performed in excepted off-campus provider-based outpatient departments.

With imaging and pharmacy revenues facing such steep proposed cuts, maximizing reimbursement through perfectly coded, highly specific clinical notes is an operational imperative.

Defending Your Revenue with Doc-U-Aide Ai2 Fighting algorithmic payer denials with manual, retrospective audits is no longer a viable strategy. As procedures migrate to the outpatient space, health systems need a robust Clinical Data Foundation that natively supports concurrent, point-of-care documentation integrity.

This is the core architecture behind Doc-U-Aide Ai2. By unifying documentation workflows, hospitals can ensure that no clinical detail is lost between the inpatient and ambulatory domains. Furthermore, integrating our ambient clinical intelligence tool, Srava, ensures that complex medical necessity criteria are captured accurately by the physician in real time, long before the claim drops.

With Doc-U-Aide Ai2, your organization moves from a reactive audit posture to a proactive, defensible data strategy.

The Bottom Line The CY 2027 OPPS proposed rule makes one thing abundantly clear: outpatient CDI is no longer the future; it is the immediate present. Hospitals that fail to modernize their data management and documentation technologies will face severe financial penalties as the IPO list vanishes.

Prepare your health system for 2027. Contact us today to learn how Saince is transforming CDI and clinical data management for the modern regulatory landscape.

AI in CDI Needs Governance, Not Blind Automation
CDI Services

AI in CDI Needs Governance, Not Blind Automation

Jul 6, 2026 by Liyakath Ali 0 Comments , , , ,

AI Is Changing CDI, But Speed Alone Is Not the Goal

Artificial intelligence is rapidly entering clinical documentation integrity, medical coding, revenue cycle management, physician documentation, risk adjustment, and health information management workflows.

For CDI leaders, the opportunity is significant.

AI can help review charts faster, identify documentation gaps earlier, prioritize high-risk cases, surface clinical evidence, and reduce the manual burden on CDI specialists, coders, physicians, and revenue-cycle teams.

But there is also a serious risk.

AI can scale documentation ambiguity just as fast as it scales productivity.

That is the central challenge for healthcare organizations adopting AI-enabled CDI tools. The goal should not be to generate more queries, more coding suggestions, more diagnosis prompts, or more automated alerts. The goal should be to improve the accuracy, completeness, compliance, and defensibility of the clinical record.

In other words, the future of AI in CDI is not blind automation.

The future is governed intelligence.

What Is AI-Enabled CDI?

AI-enabled CDI refers to the use of artificial intelligence, machine learning, natural language processing, large language models, and advanced analytics to support clinical documentation integrity workflows.

These tools may help CDI teams:

  • Identify missing or unclear diagnoses.
  • Detect documentation gaps.
  • Suggest potential HCCs, CCs, or MCCs.
  • Prioritize charts for review.
  • Support outpatient CDI and risk adjustment workflows.
  • Analyze provider documentation patterns.
  • Flag potential clinical validation concerns.
  • Support coding accuracy.
  • Improve revenue-cycle efficiency.
  • Reduce administrative workload.

Used appropriately, AI can be a powerful CDI productivity and quality tool.

Used poorly, it can become a compliance problem.

That is because CDI is not simply a data extraction exercise. CDI requires clinical reasoning, coding knowledge, regulatory awareness, compliant communication, physician engagement, and an understanding of the patient’s complete clinical story.

AI can assist that process.

It should not replace it.

The Problem With Blind CDI Automation

Blind automation occurs when healthcare organizations allow AI tools to generate documentation suggestions, diagnosis prompts, or physician queries without sufficient governance, evidence review, or human accountability.

In CDI, that is dangerous.

A diagnosis should not be suggested simply because an algorithm finds a keyword.

A physician query should not be generated simply because a model predicts a coding opportunity.

An HCC should not be captured simply because it appears in a prior note or problem list.

A CC or MCC should not be treated as valid unless the clinical evidence supports the condition and the provider documentation accurately reflects the patient’s status.

The danger is not that AI makes CDI faster.

The danger is that AI can make weak CDI faster.

  • That can lead to:
  • Unsupported diagnoses.
  • Over-querying providers.
  • Leading or non-compliant queries.
  • Inaccurate final-coded claims.
  • HCC capture without clinical validation.
  • Increased audit exposure.
  • Higher denial risk.
  • Provider distrust.
  • Compliance concerns.
  • Polluted clinical data.
  • Inaccurate quality and risk reporting.

If AI accelerates these problems, the organization may not discover the damage until months or years later through payer audits, risk adjustment reviews, clinical validation denials, quality reporting discrepancies, or compliance investigations.

CDI Is About Documentation Integrity, Not Query Volume

One of the biggest mistakes health systems can make is measuring AI-enabled CDI success by query volume alone.

More queries do not necessarily mean better documentation.

More alerts do not necessarily mean better CDI performance.

More suspected diagnoses do not necessarily mean better risk adjustment.

A strong CDI program should focus on the quality, relevance, compliance, and clinical defensibility of documentation interventions.

Better metrics include:

  • Percentage of queries supported by clear clinical evidence.
  • Provider response quality.
  • Reduction in ambiguous documentation.
  • Improvement in coding accuracy.
  • Reduction in clinical validation denials.
  • Improvement in HCC documentation specificity.
  • Improvement in CC/MCC documentation accuracy.
  • Reduction in retrospective coding conflicts.
  • Audit outcomes.
  • Provider trust and adoption.
  • Evidence traceability from chart review to query to final documentation.

AI-enabled CDI should help teams ask better questions, not simply ask more questions.

Why Provider Accountability Still Matters

The physician or qualified provider owns the clinical story.

AI can summarize evidence, detect patterns, and recommend review opportunities. CDI specialists and coders can interpret documentation gaps and coding implications. But the provider is ultimately responsible for documenting the patient’s condition accurately.

That principle becomes even more important as AI becomes more sophisticated.

A polished AI-generated suggestion can look authoritative even when the underlying clinical evidence is incomplete or ambiguous. A documentation prompt may appear reasonable even when the diagnosis is not clinically supported. An automated query may seem efficient even when it risks leading the provider.

That is why provider accountability must remain central to AI-enabled CDI.

The best AI CDI workflows should preserve the provider’s role by:

  • Presenting evidence clearly.
  • Avoiding leading language.
  • Showing why a documentation clarification may be needed.
  • Allowing the provider to confirm, reject, or clarify.
  • Making the final clinical judgment explicit.
  • Maintaining an audit trail.
  • Supporting education rather than pressure.

The goal is not to make physicians rubber-stamp AI suggestions.

The goal is to help physicians document the clinical truth more efficiently and accurately.

Outpatient CDI and HCC Risk Adjustment Need Special Governance

AI governance is especially important in outpatient CDI and HCC risk adjustment.

Risk adjustment programs depend on accurate documentation of chronic conditions, disease burden, complications, comorbidities, and clinical status. But diagnosis capture without clinical support can create serious compliance exposure.

For outpatient CDI, the right question is not merely:

“Can this condition be captured?”

The better question is:

“Is this condition current, clinically supported, documented by the provider, and defensible under audit?”

This is where AI can be helpful if designed correctly.

A governed AI-enabled outpatient CDI workflow can help identify:

  • Chronic conditions that may need annual documentation.
  • Conditions appearing in prior history but not assessed in the current encounter.
  • Medication evidence that may suggest a condition requiring provider clarification.
  • Lab or imaging results that may support further documentation review.
  • Documentation inconsistencies between the problem list, assessment, plan, and orders.
  • Potential HCC opportunities requiring clinical validation.
  • Missing specificity in diagnosis documentation.

But AI should not convert these signals into final diagnosis capture without provider confirmation and evidence review.

In risk adjustment, defensibility matters as much as discovery.

What Good AI Governance in CDI Looks Like

A strong AI governance framework for CDI should include clinical, operational, compliance, technical, and revenue-cycle controls.

At minimum, healthcare organizations should define:

1. Clear Use Cases

AI should be deployed for specific CDI use cases, such as outpatient HCC review, inpatient CC/MCC documentation gaps, clinical validation support, query prioritization, or provider education.

Vague AI deployment creates vague accountability.

2. Evidence-Linked Recommendations

Every AI suggestion should be traceable to specific evidence in the medical record. CDI professionals should be able to see what triggered the recommendation and whether the evidence is clinically relevant.

No black-box diagnosis prompts.

3. Compliant Query Safeguards

AI-supported queries must follow compliant query practice. They should be non-leading, clinically supported, clearly written, and designed to clarify documentation rather than influence the provider toward a preferred financial outcome.

4. Human Review

AI should assist CDI specialists, coders, physicians, and HIM professionals. It should not bypass them.

Human review is essential for clinical interpretation, compliance judgment, and provider communication.

5. Provider Accountability

The provider must retain ownership of the final clinical documentation. AI should make the evidence easier to understand, not replace clinical judgment.

6. Audit Trails

AI-enabled CDI workflows should capture the reasoning path from source documentation to AI suggestion to CDI review to provider query to final documentation outcome.

This is essential for audit readiness.

7. Feedback Loops

Organizations should continuously review AI recommendations, false positives, provider responses, denial outcomes, coding results, and audit findings.

AI governance is not a one-time implementation task. It is an operating discipline.

8. Cross-Functional Oversight

AI in CDI should not be owned by one department alone.

Effective governance should include CDI, HIM, coding, compliance, medical staff leadership, quality, revenue cycle, IT, privacy, security, and clinical operations.

The Role of Clinical Data Management

AI-enabled CDI also depends on the quality of the underlying clinical data.

Many healthcare organizations still struggle with fragmented documentation across EHR notes, scanned documents, faxes, dictated reports, lab results, imaging reports, external records, prior authorizations, referral documents, discharge summaries, and patient-submitted files.

When clinical data is scattered, AI has an incomplete view of the patient.

That can create documentation gaps, missed evidence, duplicate work, and unreliable recommendations.

This is why clinical data management matters.

Before AI can support CDI effectively, organizations need a reliable way to capture, organize, normalize, retrieve, and govern clinical documentation across sources.

This is where a clinical data fabric becomes valuable.

A clinical data fabric helps connect documentation sources so that CDI, coding, quality, risk adjustment, and revenue-cycle teams can work from a more complete and trusted clinical record.

For Doc-U-Scribe, this creates a broader strategic role: not just transcription, but clinical data management.

For Doc-U-Aide, it creates a stronger CDI foundation: AI-assisted documentation review built on accessible, organized, and evidence-rich clinical data.

The Saince Perspective: Governed AI for Documentation Integrity

At Saince, our view is simple:

AI should make CDI more intelligent, not merely more automated. Doc-U-Aide is designed around the idea that outpatient CDI and documentation integrity require evidence, clinical validation, compliant workflows, and audit readiness.

The purpose of AI in CDI should be to help teams identify meaningful documentation opportunities, support accurate HCC and risk adjustment workflows, improve coding integrity, and strengthen the defensibility of the clinical record.

But AI should not become a shortcut around clinical judgment. AI should help CDI professionals and providers work better together. It should help organizations create documentation that is accurate, complete, timely, compliant, and clinically meaningful.

That is the difference between automation and integrity.

Why This Matters for Healthcare Leaders

For CFOs, AI-enabled CDI can improve revenue-cycle performance, reduce leakage, and support audit readiness.

For CMOs, it can improve documentation quality, provider alignment, and clinical accountability.

For compliance leaders, it can reduce the risk of unsupported diagnosis capture and non-compliant query practices.

For HIM leaders, it can strengthen coding accuracy, documentation governance, and clinical data quality.

For CDI leaders, it can elevate the CDI function from retrospective chart review to proactive documentation intelligence.

For CIOs, it creates a need for governed AI systems that are transparent, integrated, secure, and auditable.

For quality leaders, it supports more accurate data for measurement, reporting, benchmarking, and patient-risk analysis.

AI in CDI is not just a technology decision. It is a governance decision.

Conclusion: The Best CDI AI Will Be Evidence-Based, Governed, and Auditable

AI will change clinical documentation integrity.

But healthcare organizations should be careful about what they are optimizing for.

  • If they optimize only for speed, they may create faster documentation noise.
  • If they optimize only for query volume, they may create provider fatigue.
  • If they optimize only for HCC capture, they may create audit exposure.
  • If they optimize only for automation, they may weaken trust.

The better path is governed intelligence.

AI-enabled CDI should be evidence-linked, compliant, clinically reviewed, provider-accountable, auditable, and aligned with the organization’s documentation integrity strategy.

That is how healthcare organizations can use AI to strengthen clinical documentation rather than simply accelerate it. Because the ultimate goal of CDI is not more codes.

It is a more accurate clinical record. And in the AI era, that distinction matters more than ever.

Medical Transcription

Why ‘AI Accuracy’ Is a Marketing Claim — And What Actually Guarantees Clinical Document Quality

Jun 15, 2026 by Liyakath Ali 0 Comments ,

Every AI vendor in healthcare claims high accuracy. The numbers sound impressive: 97%, 98%, 99%. What those numbers rarely tell you is what they were measured against, under what conditions, and what happens when the conditions in the demo differ from the conditions in your department.

In clinical document processing, accuracy is not a marketing claim. It is a patient safety specification. A misread medication dosage, a wrongly transcribed diagnosis, a missed allergy — these are not acceptable error rates to optimize around. This piece is about what accuracy actually means in clinical AI, why the standard claims are misleading, and what a responsible accuracy guarantee looks like.

The Benchmark Problem

When an AI vendor tells you their system achieves 97% accuracy, the most important follow-up question is: 97% accuracy on what?

Most AI accuracy benchmarks are measured on clean, printed content in controlled conditions — clear PDF documents, high-quality scans, professionally recorded audio in quiet environments. Under those conditions, modern AI achieves genuinely impressive accuracy. But those conditions don’t describe the documents that create the most work for your HIM team.

What the Benchmarks Don’t Show

  • Handwritten physician notes from an ED where the physician wrote the note at 3am after a 14-hour shift
  • Third-generation fax copies where the original was itself a fax of a fax
  • Telehealth recordings from a rural clinic where the patient’s connection was unstable and the audio quality fluctuated
  • Patient intake forms where the patient used a pencil on a pre-printed form that was then photocopied
  • Dictation from a non-native English speaking physician with a distinctive accent

These are not edge cases. They are the daily reality of HIM processing in most health systems. Ask any vendor for their accuracy numbers specifically on these document types, under real operating conditions. The honest answers will be significantly different from the headline numbers.

75–84% AI-only accuracy on real handwritten clinical notes92–97% AI-only accuracy on clean printed faxes and forms99%+ HITL-validated accuracy across all content types

The Human-in-the-Loop Model: What It Is and Why It Matters

Human-in-the-Loop (HITL) processing is not an admission that AI fails. It is a principled architecture for achieving clinical-grade accuracy across all content types, including the ones where AI alone is insufficient.

The workflow works like this:

  1. The AI processes the document first — fast and inexpensive. It extracts what it can identify with high confidence and flags sections where confidence is below a defined threshold.
  2. Flagged sections route to a trained Data Quality Specialist — someone with medical vocabulary training who reviews and corrects the AI’s uncertain outputs.
  3. The validated output — AI-processed content plus human corrections — routes to the EHR staging queue.
  4. The provider or HIM specialist reviews the staged data and accepts it into the permanent record with a single click.

The result: the speed and scale advantages of AI, with human expert validation where it matters most. On handwritten notes — the hardest problem — the HITL model achieves 99%+ validated accuracy because the specialist only reviews the uncertain sections (typically 20–30% of the text), not the entire document.

Why Your Workforce Is the Competitive Moat

Pure-tech AI vendors often position their fully automated approach as a feature: ‘No humans needed. Straight-through processing.’ In most clinical workflows, this should be a warning sign, not a selling point.

Consider what ‘no humans needed’ actually means for a document like a physician’s handwritten ICU progress note, or a 30-minute telehealth recording with a complex patient presenting with multiple comorbidities. An AI system that routes those directly to the EHR without human review is making a bet that any errors it introduces are acceptable. In clinical documentation, that bet is not acceptable.

The Workforce Advantage Organizations that have medical transcription staff have a profound advantage in the HITL model: these individuals are trained clinical documentation specialists. They already understand anatomy, pharmacology, disease process, and clinical context. Retraining them as Data Quality Editors for AI output validation is a matter of weeks, not months — and produces validators who understand the clinical significance of what they are correcting, not just the text.

The Right Questions to Ask Any Vendor

About Accuracy

  • What content types is your accuracy benchmark measured on? Show me the methodology.
  • What is your accuracy specifically on handwritten physician notes from an ED and an ICU?
  • What happens when accuracy falls below your threshold — who is responsible for errors?
  • Can I see a pilot with my actual document types, not the vendor’s demo documents?

About the HITL Model

  • Do you offer Human-in-the-Loop validation? If so, who are the humans and what are their clinical qualifications?
  • Is HITL mandatory for certain content types, or optional? Who decides?
  • What is your validated accuracy with HITL versus AI-only, broken down by content type?

About Liability

  • When a processing error reaches the EHR, who is responsible? What is the escalation path?
  • What audit trail do you maintain — what did the AI output, what did the human change, and when?
  • Does your integration model require human approval before data enters the permanent record?

The Staging Model: The Right Integration Architecture

One more accuracy-related topic that is often glossed over in vendor conversations: how does the processed data actually get into the EHR?

Some vendors position real-time, fully automated write-back to the EHR as a feature. Epic and Cerner specifically restrict this for third-party applications for good reason: if an AI system writes a wrong dosage directly into the active medication list, that is a patient safety incident with no human having had the opportunity to catch it.

The responsible integration model is Data Staging: processed data is proposed to the EHR in a review queue, and a provider or HIM specialist explicitly accepts it before it becomes part of the permanent record. This adds approximately 30–90 seconds of staff time per document — and provides a liability shield, an audit trail, and the ability for a human to catch any AI error before it reaches the patient record.

When evaluating any clinical document processing platform, ask specifically how the data gets into your EHR and whether a human must approve it before it is written to the permanent record. If the answer is ‘it writes automatically with no human approval,’ that should be a significant concern.

The Honest Accuracy Commitment

At Doc-U-Scribe, we publish our accuracy benchmarks by content type — measured on real clinical documents from production environments, not controlled demo conditions. We require HITL validation for handwritten notes and other content types where pure AI accuracy is insufficient for clinical standards. And we integrate via data staging: no data enters the permanent record without a human approval click.

This approach is not the cheapest or the fastest for low-complexity, high-volume clean document workflows. But it is the approach that clinical documentation actually requires when the content is complex, the stakes are high, and the errors are not acceptable.

About Doc-U-Scribe Doc-U-Scribe processes all eight clinical content types with Human-in-the-Loop validation and data staging integration. We offer accuracy benchmarking on your own documents — send us 50 handwritten notes for a free accuracy comparison between AI-only and HITL-validated processing. Contact us at docuscribe.com.

Medical Scribe

Your Telehealth Recordings Are a Compliance Time Bomb. Here’s How to Defuse It.

Jun 5, 2026 by Liyakath Ali 0 Comments , ,

If your health system has an active telehealth program, someone in your organization is sitting on a compliance problem they probably don’t know about yet. It will surface — through a HIPAA audit, a malpractice discovery request, or a state medical board inquiry. The question is whether you find it first or a regulator does.

The problem is not the telehealth itself. The problem is what happens after the session ends.

The Anatomy of the Telehealth Documentation Gap

A typical telehealth workflow at most health systems looks like this: the clinician opens Zoom Health, Teams, or Doximity and conducts the encounter. The platform records the session. The session ends. The recording is saved to the platform’s cloud storage.

And then — nothing. The recording sits there. No connection to the EHR. No clinical note generated from the session. No retention schedule applied. No HIM department visibility.

Meanwhile, the clinician documents the encounter — sometimes. Studies of telehealth encounter documentation rates show that physicians create clinical notes for telehealth visits at a rate 15–25% lower than for equivalent in-person visits. The encounter happened. The clinical decision was made. The prescription was issued. But the documentation may not reflect it.

150+ avg telehealth sessions per week at a mid-size health system30 days Zoom default recording retention (some tiers)7–10 yrs HIPAA retention requirement for adult medical records

The Three Specific Risks

Risk 1: HIPAA Retention Violation

Telehealth session recordings that contain PHI — which is essentially all clinical telehealth sessions — are subject to HIPAA’s medical records retention requirements. For adult patients, most states require retention of 7–10 years. For pediatric patients, records may need to be retained until the patient reaches the age of majority plus an additional period.

Zoom’s default retention policy for recorded meetings can be as short as 30 days in some account configurations. If your clinical team is using Zoom Health without explicit retention configuration and your IT team is not managing the recordings, you may have hundreds of telehealth encounters that are past their 30-day expiration — gone, with no way to recover them.

The Discovery Scenario

A patient files a malpractice claim relating to a telehealth encounter from eight months ago. The plaintiff’s attorney issues a discovery request for all records relating to that encounter, including any recordings. If that recording no longer exists because Zoom deleted it at 30 days, and if no other documentation accurately captures the encounter, the organization faces significant liability exposure.

Risk 2: Undocumented Clinical Encounters

A telehealth visit where the physician issued a medication change, ordered a diagnostic test, or made a treatment recommendation without creating a corresponding clinical note is a documentation gap that creates both compliance and clinical risk. For the billing team, it is lost revenue — an encounter with no note cannot be billed. For the CDI team, it is a gap in the severity of illness profile. For the medicolegal team, it is an undocumented clinical decision.

Risk 3: No HIM Governance

Most HIM departments have no visibility into the volume, content, or retention status of their organization’s telehealth recordings. If your medical records team cannot produce a telehealth recording in response to a legal request, they need to know that the recording either exists (and where) or does not exist (and why). Right now, most cannot answer that question.

The Solution Architecture

Solving the telehealth documentation problem requires three capabilities working together:

1. Automated Ingestion from Telehealth Platforms

The platform needs to pull recordings automatically from your telehealth provider — not rely on clinicians or staff to manually upload them. Native API integrations with Zoom Health, Microsoft Teams Health, Doximity, and Teladoc are essential. The recording goes from the telehealth platform to your clinical document processing platform without any human action required.

2. Transcript and Clinical Note Generation

Every recording gets processed into a speaker-diarized transcript — identifying which speaker is the clinician and which is the patient. Natural language processing extracts diagnosis mentions, medication changes, follow-up instructions, and referrals. A draft SOAP note is generated from this content.

This draft note is then sent to the provider for review through their existing EHR workflow. The review and approval takes approximately 90 seconds to 2 minutes — far less than writing the note from scratch.

3. Retention-Compliant Filing

The recording, the transcript, and the approved note are all filed to the patient’s chart with a retention schedule that matches your organization’s medical records retention policy. HIM now has visibility into the recording, the documentation, and the retention timeline.

The Operational ROI

Beyond compliance, the telehealth documentation workflow generates measurable revenue recovery. Consider a health system with 150 telehealth sessions per week where 20% currently go undocumented:

  • 30 undocumented encounters per week × average professional fee of $120 = $3,600 in unbilled revenue per week
  • $3,600 × 50 working weeks = $180,000 in annual revenue leakage from incomplete documentation alone
  • Most organizations recover this revenue within 60 days of implementing automated telehealth documentation

The provider time savings add to this: if a provider spends 8 minutes documenting a telehealth encounter manually, and the AI-generated draft reduces that to 2 minutes, 6 minutes of provider time is recovered per encounter. At 150 sessions per week, that is 15 hours of provider time per week returned to clinical care.

Implementation Considerations

What to Look For in a Telehealth Documentation Platform

  • Direct API integrations with your specific telehealth platform(s) — manual upload workflows will not be adopted consistently
  • Speaker diarization that correctly identifies clinician vs. patient speech — critical for note accuracy
  • HITL validation for low-confidence transcription segments — ambient noise, accents, and medical terminology all affect accuracy
  • Provider review workflow that surfaces within the existing EHR — not a separate application requiring another login
  • Configurable retention policies that match your state and organizational requirements
  • Audit trail showing processing date, AI confidence scores, and human review actions

The Phased Approach

Most organizations benefit from a phased implementation: start with new recordings going forward (prospective), demonstrate the workflow, then assess the retrospective backlog of existing recordings that may need processing for compliance.

The prospective workflow goes live in days. The retrospective review of existing recordings is a project that can be sized and scheduled once the prospective workflow is running smoothly.

Free Pilot Offer

Doc-U-Scribe offers a free Telehealth Documentation Pilot: send us 10 existing recordings from any platform, and we will return speaker-diarized transcripts and draft SOAP notes within 24 hours. No setup required. BAA signed upfront. See the workflow before you commit to anything.

Uncategorised

Is your hospital budgeting on a number that doesn’t exist?

Jun 3, 2026 by saince_admin 0 Comments , , , , , , , , , , ,

The FY2027 IPPS Proposed Rule: What the Payment Numbers Don’t Tell You — and Why CDI Has Never Mattered More

On April 14, 2026, CMS published the FY2027 Inpatient Prospective Payment System (IPPS) and Long-Term Care Hospital Prospective Payment System (LTCH PPS) Proposed Rule in the Federal Register (Vol. 91, No. 71, CMS-1849-P). At 576 pages, this is one of the most consequential annual rulemakings in healthcare finance — and like every year, the headline number risks obscuring what hospitals will actually experience.

Here is what CDI Managers, RCM Directors, and hospital finance leadership need to understand before October 1, 2026.

The Payment Rate: One Number, Four Different Realities

CMS is proposing a market basket rate-of-increase of 3.2 percent for FY 2027, based on IGI’s fourth quarter 2025 forecast of the 2023-based IPPS market basket. Before that number generates any budget optimism, it needs to be reduced by the 0.8 percentage point productivity adjustment mandated under the Affordable Care Act.

That leaves a 2.4 percent net applicable percentage increase — but only for hospitals that meet both conditions: submitting quality data under Section 1886(b)(3)(B)(viii) of the Act, and demonstrating meaningful use of certified EHR technology under Section 1886(b)(3)(B)(ix).

Hospitals that do not meet both conditions face materially different outcomes, as set out directly in the proposed rule:

  • Hospital that submits quality data but is not a meaningful EHR user: 0.0% net update
  • Hospital that does not submit quality data but is a meaningful EHR user: 1.6% net update
  • Hospital that submits neither: −0.8% net update

This is not a footnote. It is the operative payment reality for a significant number of hospitals. CMS’s own impact analysis estimates the aggregate effect of all proposed changes would increase payments to acute care hospitals by approximately $1.9 billion in FY 2027 — but that aggregate masks wide variation by hospital type, geography, and program participation.

For RCM Managers building FY2027 revenue projections: the applicable rate for your hospital depends entirely on your quality reporting and EHR attestation status. Model accordingly.

The Programs That Move Money: What’s Actually Changing

Hospital Readmissions Reduction Program (HRRP)

CMS is proposing to add a new measure to the HRRP: the Hospital 30-Day, All-Cause, Risk-Standardized Readmission Rate Following Sepsis Hospitalization. This measure would have an early look beginning with the FY 2028 program year, with full use beginning in the FY 2029 program year.

The HRRP’s existing financial reach is significant. The proposed rule’s impact analysis shows that for FY 2027, 2,358 out of 2,832 eligible hospitals — more than 83 percent — are projected to receive a readmission penalty. The average penalty as a share of payments across all hospitals is estimated at 0.48 percent. For the hospitals penalized most heavily (those with Medicare utilization between 50 and 65 percent), the average change in payments is projected at −6.8 percent, and for those with utilization above 65 percent, −7.8 percent.

For CDI and quality leaders: sepsis documentation is now a readmission risk variable. The specificity of the sepsis diagnosis, the documentation of the clinical course, and the accuracy of discharge disposition coding all feed directly into whether a readmission is attributed to your hospital under this measure. CDI programs that have not yet built a sepsis documentation workflow should treat this proposal as the starting gun.

Hospital Value-Based Purchasing (VBP) Program

The proposed rule includes no net financial impact to the VBP program for the FY 2027 payment year — the program is budget neutral by statute. However, CMS is proposing to modify five condition-specific and procedure-specific mortality measures beginning with the FY 2032 program year, including:

  1. Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate Following Acute Myocardial Infarction (AMI)
  2. Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate Following Heart Failure
  3. Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate Following Pneumonia
  4. Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate Following COPD
  5. Hospital 30-Day, All-Cause, Risk-Standardized Mortality Rate Following CABG Surgery

Additionally, CMS is proposing to modify three claims-based complication measures beginning with the FY 2028 payment determination: Excess Days in Acute Care after Hospitalization for AMI, for Heart Failure, and for Pneumonia.

These are not distant concerns. The FY 2028 measures use performance data from periods that begin well before October 2026. Documentation of AMI, heart failure, pneumonia, and COPD encounters today — the specificity of the principal diagnosis, the capture of relevant complications and comorbidities, the clinical accuracy of the discharge summary — determines where your hospital lands on these measures in the coming payment years.

Hospital-Acquired Condition (HAC) Reduction Program

CMS is proposing to adopt five modified risk-standardized mortality measures into the HAC Reduction Program beginning with the FY 2028 payment determination, replacing three current measures. Also proposed: removal of three eCQMs beginning with the FY 2030 payment determination — the Venous Thromboembolism Prophylaxis (VTE-1), the Intensive Care Unit VTE Prophylaxis (VTE-2), and the Discharged on Antithrombotic Therapy (STK-02) measures.

Hospital Inpatient Quality Reporting (IQR) Program

Three new quality measures are proposed for the IQR:

  1. Excess Days in Acute Care After Hospitalization for Diabetes — beginning with the FY 2029 payment determination
  2. Advance Care Planning eCQM — beginning with the FY 2030 payment determination
  3. Hospital Harm–Postoperative Venous Thromboembolism eCQM — beginning with the FY 2030 payment determination

The Advance Care Planning eCQM, in particular, represents a documentation domain that many CDI programs have not yet operationalized. Advance care planning conversations are clinical interactions that must be documented in the medical record with specificity — who participated, when, what was discussed — to qualify as a reportable measure. This is a CDI function, even if it has not traditionally been framed that way.

The MDH Expiration: A $258 Million Hit to Rural Hospitals

One of the most immediately impactful provisions in this proposed rule affects a specific and financially vulnerable hospital category. The Medicare-Dependent Small Rural Hospital (MDH) program is set to expire for discharges occurring on or after January 1, 2027, under current law.

Under the MDH program, qualifying hospitals received the higher of the IPPS Federal rate or a hospital-specific rate. Beginning January 1, 2027, absent a change in law, these hospitals will be paid entirely at the standard IPPS Federal rate.

CMS estimates that the expiration of temporary changes to the low-volume hospital payment policy — which is related but distinct — would decrease aggregate low-volume hospital payments by approximately $258 million in FY 2027, affecting approximately 589 providers that are expected to no longer qualify under the post-expiration criteria.

For hospitals in this category, revenue cycle leadership needs to model this change immediately. The MDH differential has been a meaningful component of net revenue for qualifying hospitals, and the loss of it beginning in January — not October — creates a mid-year budget exposure that requires proactive planning.

MS-DRG Recalibration and Budget Neutrality

The proposed rule includes annual MS-DRG reclassification and recalibration for FY 2027, consistent with CMS’s ongoing methodology. The proposed MS-DRG Reclassification and Recalibration Budget Neutrality Factor is 0.998687 — slightly less than 1.0, meaning the recalibration itself has a small aggregate negative effect on standardized amounts, offset by the budget neutrality requirement under statute.

For CDI programs, the practical significance of annual MS-DRG recalibration is this: the relative weight of any given DRG — which directly determines payment — changes every year. A diagnosis your team has been reliably capturing at a certain weight may carry a different weight in FY 2027. CDI programs that do not systematically review annual weight changes against their most frequently coded diagnoses are operating with a lag that costs real revenue.

New Technology Add-On Payments: $836 Million in FY 2027

CMS is proposing to continue New Technology Add-On Payment (NTAP) status for 41 technologies in FY 2027, with an aggregate estimated total impact of approximately $836 million. Technologies range from CAR T-cell therapies to cardiac monitoring devices to novel antimicrobials.

For CDI programs at hospitals using any of these technologies, the NTAP linkage is direct: the add-on payment is only triggered when the appropriate ICD-10-PCS procedure code — or in some cases, the ICD-10-CM diagnosis code — is present on the claim. Documentation must be specific enough to support the code. Vague or non-specific operative or procedure notes do not get the payment. This is precisely where concurrent CDI review adds financial value that is entirely separate from DRG optimization.

The TEAM Model and CJR-X: New Mandatory Episodes Beginning October 1, 2027

CMS is proposing to test and expand two episode-based payment models with direct implications for hospital documentation and care coordination:

TEAM (Transforming Episode Accountability Model): A mandatory 5-year model that would begin October 1, 2027 for acute care hospitals in mandatory CBSAs. TEAM holds hospitals financially accountable for the cost and quality of care across episodes beginning with one of five surgical procedures: coronary artery bypass graft, lower extremity joint replacement, major bowel procedure, surgical hip/femur fracture treatment, and spinal fusion. The episode extends 90 days post-discharge.

CJR-X (Comprehensive Care for Joint Replacement Expanded): CMS is proposing to expand the CJR model nationally, beginning October 1, 2027, covering lower extremity joint replacement episodes across both inpatient and outpatient settings.

For CDI and RCM leaders at hospitals in mandatory TEAM markets, this is not a FY2027 issue — it is a FY2027 preparation issue. Risk stratification in episode-based models depends entirely on how accurately a patient’s clinical complexity is documented and coded at the index admission. Comorbidity capture, principal diagnosis accuracy, and procedure specificity all influence the benchmark against which your hospital’s episode performance will be measured. Incomplete documentation at the index admission means your risk-adjusted target price will be set too low — and your hospital will bear the financial consequence of that inaccuracy across the entire 90-day episode.

Wage Index: Transition and Discontinuation of the Low Wage Index Policy

The proposed rule continues the FY 2027 wage index update using wage data from cost reporting periods beginning in FY 2023, with the proposed 2022 Occupational Mix Survey adjustment applied. CMS is also proposing a transition policy for the discontinuation of the low wage index hospital policy — a policy that had provided wage index increases to hospitals in areas with below-average wage indexes.

Hospitals in affected wage index areas need to model the impact of this transition on their labor-related payment components. The wage index directly affects the labor-related share of the standardized amount (proposed at 66.0 percent for FY 2027 for hospitals with wage indexes above 1.0, and 62 percent for those at or below 1.0), which means wage index changes translate dollar-for-dollar into changes in operating payments.

What This Means for CDI Programs: Five Priorities Before October 1

The FY2027 IPPS Proposed Rule, read in full, points to five clear operational priorities for CDI programs between now and the October 1 effective date:

1. Audit your quality data submission status and EHR attestation. The difference between a 2.4% update and a 0.0% or negative update is not a clinical issue — it is an administrative and compliance issue. Confirm your hospital’s status for both requirements and escalate any risk immediately.

2. Build a sepsis documentation workflow now. The proposed addition of the Sepsis Readmission measure to the HRRP means that sepsis documentation quality will carry readmission risk weighting beginning in the FY 2028 program year. Concurrent CDI review of sepsis encounters — ensuring that the clinical criteria for sepsis are documented, the source of infection is specified, and the clinical response is captured — needs to begin now, not when the measure is finalized.

3. Review your FY 2027 MS-DRG weight changes against your top diagnoses. Every year, MS-DRG weights shift. Every year, some hospitals discover post-implementation that their most frequently coded diagnoses have changed weight. Schedule a systematic pre-implementation review before October 1.

4. Map your NTAP-eligible technologies to your CDI query and coding workflows. For each of the 41 proposed NTAP technologies used at your hospital, confirm that your CDI and coding processes reliably capture the required codes. The $836 million aggregate NTAP pool is only accessible to hospitals whose documentation supports the required specificity.

5. Prepare for TEAM if you are in a mandatory CBSA. Begin now with comorbidity capture audits for CABG, joint replacement, bowel resection, hip/femur fracture, and spinal fusion cases. Your risk-adjusted target prices under TEAM will be set based on how accurately these patients’ complexity is documented at the index admission.

What This Means for RCM Managers and Directors

Model net revenue at your actual applicable percentage increase, not the headline market basket figure. The 3.2% market basket is reduced to 2.4% for qualifying hospitals — and potentially to 0.0% or below for those with quality reporting or EHR attestation gaps.

Account for MDH expiration in January if applicable. The MDH payment differential ends January 1, 2027, not October 1. Budget models that treat FY2027 as a uniform full-year rate will be wrong.

Quantify your HRRP penalty exposure. With 83 percent of eligible hospitals projected to receive a readmission penalty averaging 0.48 percent of payments, HRRP is a material revenue risk for most hospitals. CDI investment that reduces preventable readmission attribution is a net revenue strategy, not just a quality initiative.

Factor in low-volume hospital payment policy changes. If your hospital has been receiving the low-volume hospital payment adjustment, confirm whether you qualify under the post-January 2027 criteria — which revert to the original statutory methodology from the FY 2005-2010 period.

The Bottom Line

The FY2027 IPPS Proposed Rule will be finalized in August 2026 and takes effect October 1 — with some provisions, including the MDH expiration, taking effect January 1, 2027. The preparation window is open now.

The organizations that will navigate this rule successfully are those that read past the headline number and address the specific operational levers the rule actually contains: quality program reporting compliance, documentation-driven sepsis readmission risk, NTAP code capture, MS-DRG weight changes, and — for hospitals in mandatory markets — the episode-based accountability that begins with TEAM in FY2028.

Clinical documentation is the foundation of every one of these levers. It determines your applicable payment rate category, your readmission measure performance, your quality measure reporting results, your NTAP eligibility, and your risk-adjusted episode benchmarks. In the FY2027 environment, CDI is not a billing function. It is the clinical data infrastructure that determines whether your hospital captures the revenue it earns and avoids the penalties it can prevent.

Saince provides AI-powered CDI software designed to help hospitals achieve documentation accuracy at scale — from concurrent CDI and physician query management to MS-DRG impact analysis and quality measure documentation workflows. To learn how Saince can help your team prepare for FY2027, contact us at sales@saince.com.

Source: Federal Register, Vol. 91, No. 71, April 14, 2026. Medicare Program; Hospital Inpatient Prospective Payment Systems for Acute Care Hospitals (IPPS) and the Long-Term Care Hospital Prospective Payment System and Policy Changes and Fiscal Year (FY) 2027 Rates; Requirements for Quality Programs; and Other Policy Changes. CMS-1849-P.

The 8 Clinical Content Types Your EHR Cannot Handle — And What to Do About Each One
Medical Scribe

The 8 Clinical Content Types Your EHR Cannot Handle — And What to Do About Each One

May 26, 2026 by Liyakath Ali 0 Comments , , ,

Every HIM Director knows the feeling. You open the queue on Monday morning, and before you can touch the structured work — the coding queries, the CDI reviews, the compliance reports — you have to wade through the pile. The faxes that arrived over the weekend. The telehealth recordings sitting in a Zoom folder someone emailed you about. The handwritten notes from the ICU that were scanned and sent over as image files. The patient intake forms that front desk couldn’t get to on Friday.

This is the pile that gets no respect in healthcare IT conversations. Vendors talk about EHR optimization, clinical decision support, population health analytics. Nobody talks about the pile. But the pile is where your team’s time goes, where burnout starts, and where patient safety risks hide.

The reason the pile exists is structural: EHRs were designed to manage structured, discrete data — lab values, vital signs, medication orders, coded diagnoses. They were not designed to ingest, classify, and extract meaning from the unstructured content that represents 80% of all clinical information a health system generates. That gap is the pile.

This article breaks down each of the eight content types that HIM departments commonly face, the specific processing challenges each one creates, and the approaches that are actually working in production environments today.

Why This Matters More Than Ever

Three trends are converging to make the unstructured data challenge more acute than ever for HIM:

  • Telehealth expansion has created a new category of unmanaged clinical content: video recordings, audio logs, and session transcripts that exist outside any EHR workflow
  • Regulatory scrutiny is increasing — HIPAA auditors are specifically asking about telehealth recording retention, and organizations that cannot demonstrate compliant workflows are at risk
  • Staffing shortages are making manual document processing unsustainable — HIM teams are smaller and facing higher volumes simultaneously
80% of clinical data is unstructured$300B annual cost of dirty data in US healthcare70% of medical communications arrive by fax

Content Type 1: Inbound Faxes

The Challenge

Despite everything the healthcare industry has done to modernize clinical communication, approximately 70% of medical information exchange still occurs via fax. A fax arrives as a PDF or TIFF image — a photograph of a document, to be precise — and requires a trained human to read, classify, identify the patient, extract the relevant clinical data, and manually enter that data into the appropriate EHR fields.

For a mid-sized health system processing 200–500 inbound faxes per day, this manual workflow consumes thousands of labor hours per year and is a primary driver of HIM burnout. It also creates clinical risk: a fax misclassified as routine when it contained urgent lab results, or a referral routed to the wrong department because the patient name was ambiguous.

What Actually Works

Intelligent Document Processing (IDP) platforms now achieve 94–97% auto-filing accuracy on clean, printed fax content. The workflow: the fax arrives, AI classifies the document type (referral, lab result, prior auth, prescription refill), extracts the patient demographics and key clinical data, matches to the correct MPI record, and stages the structured data for EHR routing — all in seconds.

The important caveat: AI accuracy degrades on faxed-of-faxes (third-generation copies), handwritten content within faxes, and unusual document formats. A Human-in-the-Loop (HITL) validation step — where a trained specialist reviews low-confidence extractions — is essential for maintaining the accuracy levels that clinical documentation requires.

Key Metric

Teams implementing automated fax processing reduce manual fax handling time by 60–70% on average, with the remaining staff time redirected to higher-value CDI and coding work.

Content Type 2: Scanned Documents

The Challenge

Scanned documents are the legacy problem that never went away. Decades of paper records, converted to PDF or TIFF through departmental scanners, live in document management systems as what HIM professionals call ‘dumb images’ — files that an EHR can store but cannot search, cannot index by clinical concept, and cannot use to trigger decision support.

A scanned operative report, for example, contains the surgeon’s technique, the implant specifications, the post-operative instructions, and the anesthesia record. All of that clinical information is invisible to any analytics tool unless a human re-keys it into structured fields.

What Actually Works

Modern OCR (Optical Character Recognition) combined with Natural Language Understanding (NLU) can extract and structure the clinical content from most clean scanned documents with high accuracy. The resulting output — tagged clinical entities, ICD-10 and CPT code suggestions, extracted patient demographics — can be attached to the document and indexed in the EHR, making decades of scanned content searchable by concept for the first time.

The practical limitation remains handwritten content within scanned documents, which requires a different approach covered in Content Type 5 below.

Content Type 3: Telehealth Session Recordings

The Challenge

Telehealth exploded during the pandemic and has stabilized at a level that has fundamentally changed clinical documentation requirements. Most health systems now have hundreds of telehealth sessions per week — many of which are being recorded by the telehealth platform (Zoom Health, Microsoft Teams, Doximity, Teladoc) and stored in a cloud folder that HIM has no visibility into, no retention control over, and no connection to the EHR.

This creates three simultaneous problems. First, a HIPAA compliance risk: telehealth recordings containing PHI must be retained under the same medical records retention standards as any other clinical documentation. Second, a revenue cycle risk: physicians are creating clinical notes for telehealth visits at lower rates than in-person visits, leaving encounters undocumented and unbilled. Third, a medicolegal risk: if a patient’s telehealth session recording is subpoenaed and the organization cannot produce it because Zoom deleted it after 30 days, that is a significant liability.

What Actually Works

Platforms that can ingest recordings directly from telehealth providers (via API integration with Zoom, Teams, Doximity) and automatically produce structured clinical output are the only scalable solution. The processing pipeline: audio extraction from the video file, speaker-diarized transcription identifying which speaker is the clinician and which is the patient, natural language processing to extract diagnoses and medication mentions, and generation of a draft SOAP note for provider review.

The provider reviews the AI-generated note in under two minutes, corrects any errors, and signs it. The recording is then filed with the encounter, the note is filed in the EHR, and the billing record is complete. Total provider burden per telehealth encounter: approximately 2 minutes additional time for documentation review.

Compliance Note

HIPAA requires telehealth recordings containing PHI to be retained under the same standards as other medical records — typically 7–10 years for adult patients. Organizations should audit their current telehealth recording storage and retention practices before their next HIPAA review.

Content Type 4: Clinical Video Files

The Challenge

Beyond telehealth, health systems generate a significant volume of clinical video content that belongs in the medical record: surgical procedure recordings, endoscopy videos, wound documentation photographs and videos, radiology-adjacent imaging, and clinical training recordings that reference specific patient cases. These files typically live on surgical system hard drives, camera memory cards, or departmental shared drives — disconnected from the EHR and from any structured clinical workflow.

What Actually Works

For procedural video, the primary value of AI processing is in the audio track: surgeon narration of technique, anesthesia record verbalized during the procedure, nursing documentation spoken aloud. Speaker-diarized transcription of this audio, combined with procedure code extraction, provides a structured clinical record that can be attached to the surgical encounter.

The video file itself — after audio processing — can be stored in a HIPAA-compliant clinical media repository with EHR linking, making it retrievable for quality review, surgical outcome tracking, and medicolegal purposes.

Content Type 5: Handwritten Physician Notes

The Challenge

Handwritten notes are the hardest problem in clinical document processing, and any vendor who tells you otherwise is not being honest with you. The variability of individual physician handwriting, combined with the speed at which clinical notes are typically written, produces documents that push the limits of even the most advanced AI recognition systems.

The practical accuracy range for pure AI-only handwriting recognition on real clinical notes from emergency departments and intensive care units is 75–85%, depending on the legibility of the specific physician’s handwriting. At 80% accuracy, one in five words is wrong. In a clinical context, a misread medication dosage or a wrongly transcribed diagnosis code is not an acceptable error.

What Actually Works

The only approach that achieves clinically acceptable accuracy on handwritten notes is a combination of AI and human validation — what is called Human-in-the-Loop (HITL) processing. The AI processes the note first (fast, inexpensive), identifies high-confidence extractions, and flags ambiguous sections. A trained clinical documentation specialist — someone with medical vocabulary training, not a general transcriptionist — reviews and corrects the flagged sections before the output routes to the EHR.

This hybrid approach achieves 99%+ validated accuracy because the human expert only reviews the sections where the AI is uncertain — typically 20–30% of the text — rather than transcribing the entire note from scratch. It is faster than pure manual transcription and more accurate than pure AI.

Industry Honesty Always ask AI vendors for their accuracy benchmarks specifically on handwritten clinical notes — not on printed documents, not on clean dictation. Benchmark tests on handwritten ED and ICU notes from actual clinical environments consistently show accuracy 10–20 percentage points lower than vendors advertise for clean content.

Content Type 6: Patient Paper Forms

The Challenge

Despite the proliferation of patient portal self-service tools, a significant percentage of patient-facing documentation still arrives on paper: intake questionnaires, health history forms, consent documents, release of information requests, and HIPAA acknowledgments. Each of these forms contains structured data fields — patient demographics, chief complaints, medication lists, insurance information — that must be manually re-entered into the EHR.

For a practice seeing 50 patients per day, manual form processing can consume 2–3 hours of front desk time — time that could be spent on patient interaction, scheduling, and care coordination.

What Actually Works

Template-aware extraction — where the processing system knows the structure of your specific forms — achieves 92–97% accuracy on printed patient forms. The system recognizes each form type, maps the handwritten or printed entries to the corresponding EHR fields, matches the patient to the Master Patient Index (MPI), and stages the structured data for one-click acceptance by a staff member.

The key differentiation from generic OCR is the template-awareness: the system needs to be configured with your specific form designs to achieve high accuracy. This configuration typically takes days, not months, and can accommodate hundreds of different form templates.

Content Type 7: Voice and Audio Files

The Challenge

Physician dictation has historically been the core use case for medical transcription — and remains a significant volume workflow for many health systems. Beyond structured dictation, audio files in clinical settings include bedside recording devices, voicemail messages with clinical instructions, audio from patient home monitoring devices, and podcast-format provider communications.

Modern ambient AI (Dragon Medical, Nuance DAX) has significantly automated the structured dictation workflow. However, these tools are optimized for in-EHR, real-time use by the dictating physician. They do not process audio files that arrive after the clinical encounter, audio from devices outside the EHR environment, or audio from non-physician clinical staff.

What Actually Works

AI transcription of audio files using models trained on medical vocabulary achieves 88–96% accuracy on clearly-recorded physician dictation. Combined with ICD-10 and CPT code suggestions from the transcript, this produces a structured clinical note that requires only provider review and signature.

For audio with background noise, multiple overlapping speakers, or non-standard clinical vocabulary, the HITL layer is again essential for achieving acceptable accuracy.

Content Type 8: PDF and Word Files

The Challenge

External clinical documents — referral packets, specialist consult letters, hospital discharge summaries, external lab results — frequently arrive as PDF or Word files. Unlike faxed documents, these files contain selectable text that can be extracted without OCR. However, that text is typically unstructured narrative that requires NLP to extract the discrete clinical data elements of interest.

What Actually Works

Full-text extraction combined with clinical NLP entity recognition can classify these documents, identify the key clinical concepts (diagnoses, medications, procedures, follow-up instructions), and tag the document with structured metadata that makes it searchable within the EHR. The document itself is filed to the patient chart; the structured entities are available to CDI and analytics tools.

The Integration Reality

All eight of these content types ultimately need to connect to your EHR. The integration landscape has two primary standards:

  • HL7 v2 ORU messages for high-volume, reliable document routing — the standard that labs and radiology have used for decades and that every major EHR supports
  • FHIR DocumentReference for modern EHR connectivity, allowing the source document (the original fax, the original recording) to be linked to the patient chart alongside the structured extracted data

The practical reality: do not let any vendor promise ‘seamless auto-writing’ of structured data directly into the active medical record. Epic and Cerner specifically restrict direct third-party writes to the legal medical record for liability reasons. The correct integration model is Data Staging — structured data is proposed to the EHR, and a clinician or HIM specialist reviews and accepts it. This creates a liability shield (the human remains responsible for the data) while eliminating the tedious manual entry work.

Where to Start

The practical recommendation for any HIM department beginning this journey: don’t try to solve all eight content types at once. Identify the one or two that are causing the most operational pain and the most burnout risk, run a structured pilot on those, demonstrate ROI, and expand.

For most departments, the answer is fax automation — the volume is highest, the ROI is most visible, and the setup is typically fastest (48–72 hours to connect to an existing fax number). Telehealth documentation is the second most common urgent need, driven by compliance concern.

The goal is not to replace your HIM team. The goal is to redirect their expertise — from manual data entry to data quality validation, from document indexing to CDI querying, from printing faxes to clinical content governance. That transition, done well, improves both staff retention and departmental value.

About Doc-U-Scribe

Doc-U-Scribe is the Intelligent Clinical Data Foundation — a single platform that handles all eight clinical content types with Human-in-the-Loop validation built into every workflow. We offer free pilots for each content type. Contact us at docuscribe.com to schedule a demonstration with your actual document types.

Medical Transcription

Beyond Words: Why Generic AI Scribes Fail the Psychiatric Litmus Test

May 15, 2026 by Liyakath Ali 0 Comments , , ,

The Crisis of “Generic” Intelligence

The market is currently flooded with “ambient AI” tools promising to solve physician burnout. For a standard primary care visit (sore throat, flu shot), these generic tools perform adequately. But Behavioral Health is not a generic specialty. In psychiatry, the “data” isn’t just the words spoken; it’s the affect, the pauses, the emotional nuance, and the complex multi-speaker dynamics of a therapy session.

When behavioral health providers use a generic AI scribe, the “AI-driven efficiency” often turns into a new form of administrative labor.

The Cost of “Bad” AI in Behavioral Health

Using the wrong tool for psychiatry creates significant downstream problems:

  • The Edit-Athon (People Costs): If an AI scribe fails to understand the specific vocabulary of a Mental Status Exam (MSE) or misses the context of a History of Present Illness (HPI), the psychiatrist must spend an hour “fixing” the note. This defeats the purpose of the technology and actually increases digital fatigue.
  • Clinical Hallucinations (Patient Care): Generic AI models are prone to “hallucinating”—filling in gaps with logical but incorrect information. In mental health, an incorrectly transcribed mood descriptor or a missed nuance regarding medication adherence can have serious clinical consequences.
  • The Disconnect (Patient Experience): If a provider is constantly checking a screen to see if the AI is “getting it right,” the therapeutic alliance—the very heart of psychiatric care—is broken.

The Srava Difference: Clinical Cognition, Not Just Transcription

At Saince, we believe psychiatry deserves better than a “one-size-fits-all” model. That’s why we developed Srava Clinical Cognition (integrated within our Saince Sara and Saince Prava ecosystems).

Srava is the first ambient AI specifically trained on psychiatric datasets. It doesn’t just “listen”—it understands. It recognizes the subtle behavioral descriptors required for an MSE and automatically structures the unstructured narrative of a 45-minute session into a clinically accurate draft.

By utilizing Srava, psychiatric providers can rediscover the “art of listening.” You focus 100% on the patient, while Srava works in the background to build the clinical data foundation of the encounter. The result? Higher revenue through more accurate documentation, significantly reduced “pajama time,” and a deeper connection with the patient.

Clinical Documentation

The EHR Blind Spot: Why “Dark Data” Extraction is the New Frontier of Revenue and Care Quality

May 11, 2026 by Liyakath Ali 0 Comments , , ,

The Digital Graveyard Problem

The industry has spent a decade moving paper to the EHR, but we have accidentally created a “Digital Graveyard.” Most EHRs are excellent at tracking structured data—vitals, lab results, and pharmacy orders. However, the most critical clinical insights—the nuance of a patient’s social history, the subtle progression of symptoms mentioned in a narrative note, or the specific care gaps identified in an external consult—are buried in unstructured text.

This is Dark Data. It represents roughly 80% of all clinical information. Because it isn’t “searchable” by standard EHR analytics, it effectively doesn’t exist for the purposes of quality reporting or risk adjustment.

The Financial and Clinical Impact of Data Blindness

Ignoring unstructured data isn’t just an IT oversight; it is a direct hit to the organization’s health:

  • Lost Revenue in Value-Based Care (VBC): In risk-adjustment models (like Medicare Advantage), your reimbursement is tied to the complexity of your patient population. If a physician mentions a chronic condition in a narrative note but doesn’t “check the box” in the EHR, that HCC (Hierarchical Condition Category) code is lost. That’s thousands of dollars in missing revenue for work your clinicians are already doing.
  • Compromised Patient Care: If a care gap (like a missed screening) is buried in a scanned PDF from an outside provider, your population health team won’t see it. This leads to missed opportunities for early intervention and poorer long-term outcomes.
  • Compliance & Audit Risk: Relying on manual review to find specific data points for a clinical audit is expensive and prone to error.

Turning Narrative into Intelligence with Saince Analyze

Saince Doc–U-Scribe transforms this Digital Graveyard into a Clinical Data Foundation. Using proprietary Natural Language Processing (NLP) through the Saince Analyze module, we “read” every dictation, consult, and scanned report.

The platform identifies clinical concepts, flags care gaps, and extracts HCC codes that would otherwise be missed. This isn’t just about storage; it’s about Data Activation. We push those extracted data points back into the EHR as structured fields, making them instantly visible for billing and clinical decision-making.

By building this foundation today, you aren’t just solving today’s revenue leak; you are creating the high-fidelity data asset required for the next generation of AI-driven medicine.

Building on our strategy, these two posts tackle the “Big Picture” infrastructure challenges and the “Specialty” clinical hurdles. They are designed to position Saince One as both a visionary enterprise architect and a deeply empathetic clinical partner.

The Legacy Debt Trap: Why Your 20th-Century Infrastructure is Sabotaging Your 2026 AI Ambitions
Medical Scribe

The Legacy Debt Trap: Why Your 20th-Century Infrastructure is Sabotaging Your 2026 AI Ambitions

May 5, 2026 by Liyakath Ali 0 Comments ,

The Hidden Weight of “Technical Debt”

For many health systems, the path to innovation is blocked by the ghosts of software past. As organizations grow through acquisitions or transition to modern EHRs like Epic or Cerner, they often leave behind a trail of “zombie” legacy systems. These are old databases and archives kept on “life support” simply because they contain historical patient records that might be needed for a legal request or a rare clinical look-back.

This isn’t just an IT nuisance; it is Legacy Debt, and the interest rates are staggering.

How Legacy Silos Hurt the Enterprise

Maintaining a fragmented landscape of old applications is a multi-front assault on your organization:

  • The Talent Drain (People Costs): Your high-value IT talent shouldn’t be spent maintaining servers for a 15-year-old software version that only three people know how to use. The labor cost of patching, securing, and supporting “zombie” systems is a massive, non-productive spend.
  • The “Data Scavenger Hunt” (Patient Care): When a clinician needs a patient’s historical oncology report or a specific lab trend from a previous provider, they shouldn’t have to log into three different portals. Delays in data retrieval lead to incomplete clinical context, redundant testing, and slower care delivery.
  • Security & Compliance Risk: Legacy systems are the “soft underbelly” of healthcare cybersecurity. They often lack modern encryption and are no longer patched by vendors, making them prime targets for ransomware that can paralyze an entire network.

The Saince Solution: Building the Unified Nexus

Saince One allows you to decommission the past to power the future. Through our Clinical Data Foundation, we provide a secure, Vendor Neutral Archive (VNA) that centralizes all historical, unstructured, and legacy data into a single, searchable repository.

Instead of paying multiple maintenance fees, you consolidate your data into the Saince Fabric Core. This doesn’t just save money; it creates a clean, high-fidelity data asset. By unifying your silos, you provide clinicians with a “single pane of glass” view of the patient’s entire history and ensure your organization is AI-ready. You cannot train a predictive model on data you can’t reach; Saince One makes that data accessible, actionable, and secure.

CDI Services

The Invisible Tax on Healthcare: Solving the Multi-Channel Crisis of Clinical Data Ingestion

Apr 30, 2026 by Liyakath Ali 0 Comments

The Hidden Cost of “Analog-Digital” Chaos

In most health systems today, there is a “hidden tax” levied on every patient encounter. It isn’t found on a balance sheet, but it is felt in every department. It’s the cost of manual intervention. Despite billions spent on Electronic Health Records (EHRs), a massive volume of clinical information still arrives at the “edge” of the organization through fragmented, multi-channel streams: high-resolution scans, digital and analog faxes, external clinical summaries, and unstructured narrative files.

When this data hits your network, the clock starts ticking. In a typical mid-sized hospital, an army of HIM professionals and administrative staff spends thousands of hours annually acting as human “routers”—manually opening files, identifying patients, classifying document types, and clicking through EHR screens to attach them.

How Manual Ingestion Hurts the Bottom Line

This manual approach doesn’t just feel slow; it is a financial and clinical liability:

  • Skyrocketing People Costs: As patient volumes grow, health systems are forced to hire more administrative staff just to keep up with the document backlog. This is a non-scalable model that eats into tightening margins.
  • The “Waiting Room” Delay: Imagine a specialist referral fax that arrives 15 minutes before the patient. If that document is sitting in a manual queue, the physician enters the exam room “blind,” without the context of previous labs or consults. This forces redundant questions, wastes expensive clinical time, and frustrates the patient.
  • Physician Burnout & “Pajama Time”: When ingestion is slow, documentation lags. Physicians are often left to complete “charting” long after their shift ends because the necessary external data wasn’t ready during the encounter.

The Saince Solution: Edge Ingestion & Orchestration

Saince Doc-U-Scribe solves this by applying the 80/20 Rule of Clinical Documentation. We don’t just “digitize” faxes; we orchestrate intelligence at the edge.

Through Saince Ingest and Saince Orbit, the platform utilizes AI-driven Optical Character Recognition (OCR) and automated classification to identify a document the moment it enters the ecosystem. It matches the patient, extracts the intent, and routes it to the correct EHR field or provider queue automatically. By automating the 80% of high-volume, low-complexity inputs, you reclaim staff hours, reduce turnaround time (TAT), and ensure the “Clinical Data Foundation” is ready before the patient even checks in.