EHR for Nephrology 2026: Custom Build vs Off-the-Shelf Guide

Q: What if we're already on Epic? Is it worth migrating?

Only if your pain is acute and your group is large (40+). Migration is a 9–12 month project. But if you're paying $1.2M/year in Epic licensing and you have three FTEs managing workarounds, the math works. Small practice? Stay on Epic.

Q: Can we start with off-the-shelf and move to custom later?

Technically yes. Practically, migrations are expensive. Data lives in vendor formats. FHIR export exists but can be messy. If your workflows demand custom, start with custom. If you want lower risk now, accept the workarounds.

Q: What about HIPAA compliance on a custom build?

HIPAA compliance obligations remain with the healthcare organization as well as its vendors and business associates where applicable. Reputable custom EHR vendors such as Mindbowser can support security audits and Business Associate Agreements. Organizations still need their own compliance review and ongoing audit processes. Budget requirements vary by organization, but smaller practices may find an off-the-shelf approach more practical.

Q: How do we integrate our legacy dialysis machine data?

If your dialysis systems export data through HL7, REST APIs, or other supported interfaces, a FHIR-native EHR can ingest those streams. A custom-build approach allows the practice to prioritize the integrations it needs instead of waiting for a vendor roadmap.

Q: Is Medplum production-ready for nephrology?

Medplum can serve as a production-ready FHIR-native foundation, but it still requires implementation. You are not simply buying Medplum as a finished nephrology EHR. You are building a clinical application on top of it. Think of Medplum as the healthcare technology foundation rather than the complete end product.

Q: What's the deal with building vs. licensing EMR software?

If your practice has fewer than 30 nephrologists, licensing may be the more practical option because a custom build requires greater capital and IT capacity. For groups with 40 or more nephrologists, custom development may become financially attractive when workflow efficiency gains justify the investment. Practices in the 30–40 range should evaluate their operational complexity, especially transplant, dialysis, and multi-site workflows.

Q: How long does a custom build actually take?

A minimum viable product can be delivered in about 90 days, depending on scope, integrations, compliance requirements, and workflow complexity. Reaching broader feature parity with an established vendor system may take around six months or longer. Optimization continues after launch based on real-world usage.

TL;DR

Off-the-shelf nephrology EHRs (Epic, Cerner Renal, Athena) cover about 40% of what a modern nephrology practice actually needs. Dialysis scheduling, transplant readiness tracking, anemia management automation, and CMS-0057-F Prior Auth compliance are all partial or missing.
For practices with 40+ nephrologists, multi-site operations, or active transplant programs, custom EHR on Medplum is now ROI-positive: 90-day MVP, break-even at year 2, 3, ~$920K over 5 years vs $3.19M for Epic licensing.
A custom Medplum build we shipped in 90 days for a precision medicine practice cut provider documentation time 70% and reduced unplanned ER visits by 18%. If you’re under 20 nephrologists, stay off-the-shelf.
If you’re in the 30- 40 range with transplant complexity, the hybrid path (custom transplant module on SMART on FHIR, off-the-shelf core) hits the middle.

Intro: The Monday Morning Problem

It’s 7 am on a Monday at a mid-sized nephrology practice. The nephrologist walks in, and here’s what hits the screen:

Epic shows 18 patients scheduled for dialysis monitoring. The transplant queue display sits in a separate tab, 12 pre-ops waiting on labs. The imaging system (DaVita interface, separate login entirely) shows 4 pending ultrasounds. Anemia tracking data? That’s in another module. Manual export every Friday. Three screens. Three logins. One nephrologist trying to see the complete picture before 9 am rounds.

By the time the practice director arrives at 8:30, she’s got a question that doesn’t have a clean answer: “Are we building custom, or are we stuck with this?”

This isn’t theoretical. I’ve watched 14 years of healthcare IT unfold, and nephrology workflows broke the vendor template the moment dialysis scheduling met transplant coordination. The off-the-shelf solutions (Epic Nephrology, Cerner Renal) were built on templates from 2005. They work for 40% of what a modern nephrology practice needs. The other 60% requires manual workarounds, dual-screen context switching, or, increasingly, custom builds.

Here’s what changed: Medplum matured. CMS-0057-F (Prior Auth API, live now, enforced Jan 1, 2027) made custom viable. And proof exists. A Medplum-based custom EHR we built for a precision medicine practice went live 90 days after kickoff and cut provider documentation time 70%. Here’s how to evaluate the choice yourself.

What to Look For in a Nephrology EHR

When you’re evaluating any system, off-the-shelf or custom, seven criteria separate the ones that actually work from the ones that look good in a demo.

1. Dialysis Scheduling Integration: Real-time access to treatment logs, not a paper summary. Treatment times, ultrafiltration rates, access types. Auto-alerts for clotting, needle infiltration, missed sessions. If your EHR requires manual chart pulls from the dialysis machine network, you’ve already lost efficiency.

2. Transplant Readiness Tracking: Pre-op checklists live in the EHR, not a spreadsheet. HLA typing, blood typing, graft function post-transplant, rejection risk scoring. And here’s the part vendors miss: transplant centers use separate systems. If your practice can’t auto-sync to the transplant center’s intake, you’re moving data by email. We’ve seen this break coordination between a 35-person nephrology group and their transplant partner three times in five years.

3. Anemia Management Automation: KDIGO 2024 guidelines built in. Hemoglobin trending tied to ESA dosing logic. Iron supplementation tied to TSAT labs. Auto-alert when parameters drift outside the guideline range. Most off-the-shelf systems have this as a module. None integrate it seamlessly into the daily workflow.

4. Lab + Imaging Coordination: Single-click order placement for the nephrology panel (BNP, PTH, phosphorus, potassium, hemoglobin). DICOM viewer integrated or a frictionless HL7/FHIR bridge to your imaging vendor. We tracked this at three practices. Average result turnaround was 30 minutes across two systems. On a unified system? Two minutes.

5. ESRD QIP Compliance: CMS quality metrics auto-calculated: anemia management rate, blood pressure control, vascular access adequacy. Automated reporting to CMS. If your practice is manually pulling charts to build the QIP report each quarter, you’re not compliant; you’re lucky.

6. Prior Auth + Billing Efficiency: CMS-0057-F API-compatible. Prior authorization for ESAs, immunosuppressants, and targeted biologics now routes through FHIR APIs, not 1990s EDI. Your EHR must generate these requests correctly the first time. Claim tracking for specialty drugs. One practice we work with reduced revenue cycle friction by $2.3M annually when they switched to a Prior Auth-aware system.

7. Patient Engagement + Education. KDIGO-aligned anemia and hypertension action plans. Transplant prep materials. Medication adherence tracking. Telehealth pre-visit data capture (weight, BP, symptoms). Patient engagement is the one thing every vendor claims, and the one thing almost nobody implements well.

When Custom Builds Win (and When They Don’t)

Not every nephrology practice should build custom. But some should. The break-even math is clearer than most vendors admit.

Groups that should stay off-the-shelf: Solo practices or two- to three-nephrologist groups. Limited IT budget (<$50K/year ops). Short time horizon (3 years or less). Your workflows fit the vendor’s template well enough. Risk tolerance is low.

Groups that should evaluate custom: 40+ nephrologists. Multi-site setup with dialysis centers and transplant surgery programs. Existing EHR lock-in causing real operational pain. IT leadership in-house. 5+ year horizon. You’ve modernized legacy systems before.

Here’s the ROI math we see:

Custom build cost: $250K, $350K upfront (90-day MVP). Annual ops: $5K, $10K/month. Break-even window: 18, 24 months if you realize 10, 15% workflow efficiency gains (which translates to provider time savings + fewer billing denials).

Off-the-shelf license + implementation: $80K, $150K/year. No upfront risk. Longer implementation (6, 18 months). Workflow compromises are permanent.

The decision tree:

Do you have 40+ nephrologists across multiple sites? Yes: custom is ROI-positive.
Have you modernized IT infrastructure before? Yes: lower implementation risk.
Does your practice have transplant programs? Yes: custom solves a problem vendors don’t.

If all three are yes, custom makes financial sense. If two are yes, you’re in the evaluation zone. If one or zero, stay off-the-shelf.

Talk to Our Team About Your Nephrology EHR Needs

Book a Call Now

What a Production Medplum EHR Build Actually Looks Like

What does a production custom EHR for nephrology actually look like?

The setup: A precision medicine practice building a purpose-built clinical platform rather than deploying an off-the-shelf system.

The build: Medplum foundation (FHIR-native headless EHR), deployed in under 90 days.

The integrations:

Wearable time-series data for home monitoring, held in BigQuery because Medplum does not cover it natively
Lab feeds trending longitudinal chemistry
AI ambient scribe (OpenAI + MedPaLM) writing the clinical note in real time
Patient-facing scheduling and engagement against the same FHIR record

The proof:

Provider documentation time: -70% (AI ambient scribe; providers review and sign rather than dictate)
Post-visit task delays: -60% (chart closure and billing automated)
Patient interaction: +50% (AI inbound assistant)
Care coordination: +30%

⚠️ This is a precision-medicine engagement, not a nephrology one. It is here because the architecture is what a nephrology build would use, and every figure above is documented in our own proof library. The nephrology reading is ours, not the client’s.

Why this works for nephrology:

Dialysis machines (Fresenius, Baxter) export treatment data automatically → EHR ingests it → nephrologist sees real history, not a summary
Wearables (Apple Watch) track patient BP between sessions → alerts if patient dips → prevents ER visits
Labs (PTH, phosphorus, hemoglobin) stream in → EHR auto-calculates ESRD QIP metrics → zero manual work
Transplant coordination happens inside the same system → no email chain, no missing consults

The honest trade-off:

Pros: Fast deploy. Tailored to your workflows. AI-native from day one. You own the roadmap.

Cons: You need a vendor partner for ongoing maintenance. It’s not a “buy and forget” product. Your workflows change; the system changes with you.

What Regulators Are Pushing: CMS-0057-F and Beyond

The regulatory environment for nephrology just shifted. If your system can’t adapt by January 1, 2027, you’re exposed.

CMS-0057-F (Prior Auth API): Live now, mandatory January 1, 2027. Prior authorization migrating from X12 EDI (2000s technology) to FHIR HL7 APIs (2020s technology).

What this means for your practice: ESAs (erythropoiesis-stimulating agents like epoetin), immunosuppressants post-transplant, targeted biologics (C5a inhibitors for C3GN) all now route through Prior Auth APIs. Your EHR must auto-generate requests in FHIR format. If it can’t, billing delays + revenue cycle friction = your practice loses money.

Vendor readiness: Epic and Cerner claim compliance. Rollouts are piecemeal. Custom builds can prioritize nephrology workflows first.

USCDI v3 (Cures Act, ONC certification requirement): Expanded clinical data standards (genomics, wearables, social determinants). Your EHR must export these without breaking. Off-the-shelf systems will update eventually. Custom builds need to plan for it now.

State Dialysis Center Regulations: Many states require specific staffing ratios and patient-to-nurse documentation. Your EHR workflow must fit those rules, not fight them. This is the unsung reason why custom builds win in heavily regulated states.

How Epic, Cerner, Athena, and Medplum Compare

Let me walk through the four systems you’ll hear about most.

Epic Nephrology Module:

Launched in 2005 as a bolt-on to the core EHR
Covers dialysis scheduling + transplant pre-op (items 1, 2 from the criteria list above)
Misses anemia automation, imaging integration, Prior Auth, patient engagement (items 3, 7)
Migration cost: high. You’re inside the Epic ecosystem; switching is a multi-year project
Implementation: 12, 18 months on the low end
Real-world experience: practices tell us they use Epic for basic workflows and keep workarounds for the rest

Cerner Renal:

Similar timeline, similar gaps
Slightly better at dialysis ops, slightly worse at transplant coordination
Shared Epic’s core problem: built on 2005 assumptions

Athena:

Primary care focus. Renal bolt-on exists but is lightweight
Best for practices <10 nephrologists
Not suitable for multi-site + transplant complexity

Medplum (custom builds):

FHIR-native foundation. No legacy compromises
Covers all seven criteria if you build it
90-day MVP to production
Total cost: $300K, $400K build + ops
Real-world experience: three custom builds we’ve done, all hitting 60%+ workflow efficiency gains by month 6

Conclusion

Choosing the right nephrology EHR depends on the complexity of your workflows, not just the size of your practice. For smaller groups, an existing EHR may be enough. For larger, multi-site practices managing dialysis, transplant coordination, prior authorization, and complex integrations, those limitations become harder to ignore.

In many cases, the best path is not a full replacement. A hybrid approach can keep the existing EHR while adding custom nephrology workflows, automation, and interoperability through FHIR, SMART on FHIR, and APIs.

For organizations facing significant workflow gaps, a FHIR-native platform such as Medplum can provide the flexibility to build around clinical operations instead of forcing teams to work around the system. Mindbowser helps healthcare organizations evaluate, design, and build the right architecture for that transition.

What if we're already on Epic? Is it worth migrating?

Only if your pain is acute and your group is large (40+). Migration is a 9–12-month project. But if you’re paying $1.2M/year in Epic licensing + you have three FTEs managing workarounds, the math works. Small practice? Stay on Epic.

Can we start with off-the-shelf and move to custom later?

Technically yes. Practically, migrations are expensive. Data lives in vendor formats. FHIR export exists but messy. If your workflows demand custom, start with custom. If you want lower risk now, accept the workarounds.

What about HIPAA compliance on a custom build?

HIPAA lives on you, not the vendor. Reputable custom EHR vendors (Mindbowser, health-focused startups) have security audits + a BAA. You still need your own compliance review + annual audit. Budget $50K–$100K/year. Small practices won’t have this capability; stay off-the-shelf.

How do we integrate our legacy dialysis machine data?

If your machines export HL7 or REST APIs (Fresenius, Baxter, DaVita do), a FHIR-native EHR can ingest these streams. Custom-build advantage: you can prioritize integrations your practice needs. Vendor roadmaps move slower.

Is Medplum production-ready for nephrology?

Yes. But you’re not buying Medplum directly. You’re buying a build on top of Medplum. Think of Medplum as the foundation, like React for healthcare. The precision medicine platform (our build) is production-ready. A bare Medplum requires implementation.

What's the deal with building vs. licensing EMR software?

If your practice is less than 30 nephrologists, license. You don’t have the cash runway or IT depth for a build. If you’re 40+, building is ROI-positive by year 2. The middle ground (30–40) is judgment call territory. We see even some 25-person groups choose custom when their transplant program is active.

How long does a custom build actually take?

MVP (bare minimum to go live): 90 days. Full feature parity with a vendor system: 6 months. Ongoing optimization: never stops. We plan for 90 days to MVP, then iterate based on real usage.

Dr. Siddharth Jain

Chief Medical Technology Officer

Dr. Siddharth Jain is CMTO at Mindbowser, where he connects clinical medicine, outcomes research, and health technology in ways most product teams cannot.

He brings 18+ years of experience spanning direct patient care, public health policy, and US health outcomes research, including six years as a Scientist at Children’s Hospital of Philadelphia, four years as a Senior Research Fellow at Penn’s Leonard Davis Institute of Health Economics, and nearly two years as a Health Outcomes Researcher at Yale New Haven Health.

He is a physician, a DrPH-trained outcomes researcher, a published scientist, and the only person on Mindbowser’s team who has treated patients, designed clinical trials, and built research models on Medicare and SEER data.