The Healthcare IT Stack: From EMR to Revenue Cycle — Who Owns What and Why It's Sticky

Executive Summary

Healthcare IT is a $200B+ global market defined by regulatory complexity, life-or-death data sensitivity, and switching costs that make enterprise software look transient. The dominant EHR (Electronic Health Record) vendors — Epic, Oracle Health (formerly Cerner), and a handful of specialty players — have entrenched positions that took decades to build and will take decades to dislodge. Revenue cycle management (RCM), clinical decision support, data interoperability, and AI-powered diagnostics layer on top of the core EMR. For investors, the space offers durable incumbents with near-zero churn, emerging disruptors with AI tailwinds, and one of the largest addressable markets in software. Understanding the stack — who owns which layer, why each is sticky, and where value is migrating — is the starting point.

Industry Overview

The 30-Second Version

Hospitals and health systems run on two core software systems: the EHR (which captures all clinical data — diagnoses, medications, lab results, imaging orders) and the RCM system (which converts clinical encounters into billable claims and collects payment from insurers and patients). Epic dominates EHR at large academic medical centers; Oracle Health (Cerner) serves government and community hospitals; Meditech and CPSI serve smaller facilities. RCM is a $25B+ software and services market with fragmentation across dozens of vendors. The entire stack is held together by HL7 FHIR standards that are still incompletely implemented across the industry.

The Depth

U.S. healthcare IT spending is approximately $150B annually in 2026, growing at 8–10% CAGR. Global healthcare IT is $380B+. The HITECH Act of 2009 mandated EHR adoption across U.S. hospitals, effectively funding a 10-year installation wave that Epic and Cerner captured. By 2026, EHR penetration among U.S. hospitals is 96%+. The growth frontier is no longer installation — it's optimization, interoperability, data monetization, and AI-augmented clinical and administrative workflows.

The stack has five layers:

1. Core EHR/EMR: Clinical documentation, CPOE (computerized physician order entry), medication management
2. Revenue Cycle Management: Claims coding, billing, collections, prior authorization, denial management
3. Ancillary Clinical Systems: Radiology (PACS/RIS), laboratory (LIS), pharmacy (pharmacy IS), imaging AI
4. Data & Analytics: Population health management, clinical data warehouses, quality reporting
5. Patient Engagement: Patient portals, scheduling, telehealth, remote monitoring

Each layer has distinct vendor landscapes, pricing models, and switching dynamics. An average health system the size of a 500-bed hospital runs 100–300 software applications across these layers, many of which the EHR vendor (Epic) is progressively trying to consolidate.

How Money Flows Through the System

The Revenue Cycle Explained

Healthcare revenue cycle is the end-to-end process of converting a clinical service into a collected payment. A single patient visit generates the following workflow:

1. Scheduling & registration: Patient demographics, insurance verification (eligibility check via payer API)
2. Clinical documentation: Provider documents the encounter in EHR; diagnosis codes (ICD-10) assigned
3. Charge capture: Clinical services mapped to procedure codes (CPT/DRG)
4. Coding: Medical coders review documentation and assign accurate codes; AI-assisted coding tools (3M, Nuance/Microsoft, Optum) increasingly handle routine encounters
5. Claims submission: Electronic claim (837 EDI format) submitted to payer (insurance company)
6. Adjudication: Payer processes claim; approves, denies, or requests additional information
7. Remittance posting: Explanation of benefits (835 EDI) posted back to RCM system
8. AR management: Denied claims appealed, patient balances billed, collections managed
9. Payment posting: Final cash collected; revenue recognized

Denial rates at U.S. hospitals average 5–10% of claims submitted. Each denial that is not overturned is revenue leakage. The industry loses $100B+ annually to claims denials and write-offs. This is where RCM software and services vendors compete to add the most recoverable value.

EHR Pricing Model

Epic's pricing is opaque but follows a pattern:

• Implementation fee: $30–200M for large health system implementations (one-time)
• Annual license/subscription: $5–20M+ per year for large systems, scaled to bed count or patient volume
• Interface fees: Additional charges for each third-party system connected via Epic's API
• Module fees: Each additional Epic module (MyChart patient portal, Healthy Planet population health, Cheers CRM) carries additional licensing

Oracle Health (Cerner) has transitioned toward cloud-based subscription pricing. Its base offering for a community hospital starts at $1–5M annually. The U.S. Department of Defense and Veterans Affairs contracts — worth $10B+ over their lifetimes — are Oracle Health's largest single relationships.

Key Business Models and Their Economics

The highest quality business model in healthcare IT is EHR software at scale: Epic is privately held, but based on disclosed financial data and industry analysis, it generates $4–5B+ in annual revenue with estimated EBITDA margins of 25–35%. It has never lost a major health system account to a competitor once fully implemented. That is an extraordinary statement about switching costs.

RCM services (outsourced) carry structurally lower margins because human labor is core to the model. R1 RCM (acquired by Accenture in 2024) operated at 25–35% gross margins because clinical coders, billers, and denial management specialists are significant cost centers. The AI displacement opportunity — using large language models to automate coding and denial management — is the primary investment thesis for clinical AI vendors in 2026.

Competitive Dynamics

Epic's Platform Strategy

Epic's competitive strategy mirrors ServiceNow in enterprise software: own the core system of record (EHR), then expand horizontally across every adjacent workflow. Epic's current module list includes:

• MyChart: Patient portal with 300M+ enrolled patients — the largest patient-facing health platform in the U.S.
• Cheers: CRM for patient acquisition and retention
• Healthy Planet: Population health management
• Cogito/SlicerDicer: Analytics and reporting
• Nebula: Genomics data management
• Haiku/Canto: Mobile EHR apps
• Cosmos: De-identified research database with 260M+ patient records

Cosmos deserves special attention. Epic's aggregated, de-identified patient database is a research and AI training asset with no comparable equivalent. Pharmaceutical companies, biotech firms, and health systems pay to access it. This data moat — built on Epic's dominant clinical install base — is increasingly monetizable as AI-driven drug discovery and population health analytics grow.

Oracle Health's Challenges

Oracle acquired Cerner in 2022 for $28.3B, betting that database and cloud infrastructure expertise could modernize an EHR platform that had fallen behind Epic. The integration has been turbulent. Multiple large health systems have defected to Epic post-acquisition, citing implementation concerns and product roadmap uncertainty. The VA and DoD contracts provide revenue stability but constrain Oracle Health's public narrative. Oracle is betting on a cloud-native EHR built on Oracle Cloud Infrastructure — execution risk remains high.

The Interoperability Push

The 21st Century Cures Act and CMS interoperability rules require EHR vendors and payers to support FHIR-based APIs by 2023–2024. This theoretically reduces Epic's moat by making data more portable. In practice, the data is more accessible but workflows remain deeply embedded. Changing an EHR is not a data migration problem; it is a workflow re-training problem for thousands of clinicians. Physicians who spend years learning an EHR's idiosyncrasies — order sets, SmartPhrases, documentation templates — face productivity losses of 20–40% during and after transitions.

AI Disruption Thesis

The most active investment area in healthcare IT in 2026 is ambient clinical documentation — AI that listens to a patient-physician conversation and generates the clinical note automatically. Nuance DAX (Microsoft), Suki AI, Abridge, and others are targeting physician burnout (documentation consumes 34–40% of physician time). Epic has partnered with both Nuance and has its own in-house AI documentation layer. The question is whether AI documentation displaces EHR vendors or gets absorbed by them. Evidence so far suggests Epic absorbs — it is the integration point, and AI documentation tools need EHR integration to be clinically useful.

The Public Market Landscape

Veeva deserves elaboration: it is the only pure-play public healthcare IT company with Epic-like switching costs. Its Vault platform (regulatory submissions, clinical trial management, commercial CRM for pharma reps) is deeply embedded in life sciences workflows. NRR runs 110–115%. It has consistently grown 15–20% annually with 25%+ operating margins — a rare combination.

What to Know Before You Walk Into Any Meeting on This Topic

1. KLAS Research is the industry bible: KLAS rates every healthcare IT vendor on customer satisfaction. A KLAS score is often a board-level decision factor when health systems choose software. Know the KLAS rankings for any category you're discussing.

2. Switching costs are qualitative, not just quantitative: The 3-year EHR migration timeline is only part of the cost. Physician productivity loss, staff retraining, and the risk of data loss during migration mean CFOs and CMIOs face career risk for failed EHR transitions. That career risk is a switching cost.

3. Payer-provider dynamics shape vendor success: Software that facilitates prior authorization, value-based care contracts, or quality reporting sits at the interface of payers and providers. Both are customers. Both have significant leverage. Vendors who can serve both sides (Cotiviti, Evolent) have more complex but higher-value business models.

4. The RCM services market is being disrupted by AI faster than EHR: Routine medical coding (E&M coding for office visits, DRG assignment for inpatient stays) is increasingly automated. AI-assisted coding accuracy matches or exceeds human coders for common procedure types. This compresses labor costs in RCM services and creates pricing pressure on outsourced RCM vendors.

5. Government payers are non-trivially different: Medicare and Medicaid reimbursement follows different rules than commercial insurance. Health systems with high Medicaid payer mix (safety-net hospitals, federally qualified health centers) have structurally lower revenue per visit and prioritize cost efficiency in IT purchasing.

6. Cybersecurity is now existential: The Change Healthcare cyberattack in February 2024 disrupted $100B+ in claims processing for weeks, causing hospital cash flow crises across the U.S. Healthcare IT vendors' security posture and business continuity plans are now top-3 procurement criteria. Vendors without enterprise-grade security are losing deals.

Glossary of Terms That Matter

• EHR / EMR (Electronic Health/Medical Record): Clinical software capturing patient health data; EHR implies interoperability, EMR is practice-specific
• CPOE (Computerized Physician Order Entry): Electronic ordering of labs, medications, imaging within EHR
• RCM (Revenue Cycle Management): End-to-end process from patient registration to payment collection
• ICD-10: International Classification of Diseases, 10th edition — standardized diagnosis codes
• CPT (Current Procedural Terminology): Procedure codes used for billing
• DRG (Diagnosis-Related Group): Medicare's inpatient payment classification system
• Prior Authorization: Payer requirement for pre-approval before certain procedures/medications
• Claims Denial: Payer refusal to reimburse a submitted claim; rates average 5–10%
• HL7 FHIR: Fast Healthcare Interoperability Resources — modern standard for health data exchange
• KLAS Research: Healthcare IT research firm tracking vendor performance and market share
• HITECH Act: 2009 legislation incentivizing EHR adoption; funded the mass EHR installation wave
• Meaningful Use: Program under HITECH defining EHR functionality standards for incentive eligibility
• Population Health Management: Analytics-driven approach to managing health outcomes for patient cohorts
• HIS (Hospital Information System): Broad term for the integrated software running a hospital's operations