9 EHR Implementation Challenges US Teams Face, and How to Solve Each One
EHR/EMR

9 EHR Implementation Challenges US Teams Face, and How to Solve Each One

Pravin Uttarwar
CTO & Founder, Mindbowser
TL;DR
  • Most US organizations starting an EHR project are replacing a system, not leaving paper: by 2021, 96% of non-federal acute care hospitals and 78% of office-based physicians had a certified EHR (ONC). So the hard parts are data, interfaces, people and rules.
  • The nine challenges below are the conditions every project runs into. Each one gets a fix you can plan before go-live and a dated US source.
  • The decisions teams get wrong (skipped planning, thin training plans, over-customization) have their own page: EHR implementation mistakes.
  • Cost ranges live in our EHR software cost guide. This page names the cost lines; that one prices them.
  • When a challenge is structural, such as an interface your vendor won’t support, the fix is to build around your EHR.

Almost every US practice I talk to about an EHR project is replacing a system, not leaving paper behind, and that changes which problems hurt. The federal money that paid practices to adopt EHRs is gone too: under 42 CFR 495.310, “no payments may be made after CY 2021” in the Medicaid incentive program.

Today’s Medicare Promoting Interoperability Program is, in CMS’s own words, “a quality program”: hospitals submit data and must earn a minimum score, and a hospital that isn’t a meaningful EHR user gets a smaller annual Medicare payment update (42 CFR 412.64(d)(3)). So the federal money now shows up only as a reduction when a hospital falls short.

So you fund the project yourself, you inherit years of data from the old system, and you go live inside a web of interfaces and federal rules that didn’t exist the last time your organization chose an EHR. KLAS’s Arch Collaborative found that just 38% of organizations said their post-pandemic implementations hit the mark (KLAS, February 2025).

If you only read one number: 142 primary care physicians in one Wisconsin health system spent 5.9 hours of an 11.4-hour workday inside the EHR, and 44.2% of that time went on clerical and administrative tasks (Arndt et al., Annals of Family Medicine, 2017). That’s the time your new system has to give back, or at least not add to.

High Costs: Budget for the Whole Lifecycle

The license is rarely what surprises a budget. The lines that do are the ones nobody priced at contract time.

Plan for each of these before you sign:

  • the subscription or license itself
  • implementation services from the vendor or a partner
  • interfaces to labs, pharmacies, payers, your HIE (health information exchange) and devices
  • data migration and validation
  • training time, plus the clinical hours you backfill while staff train
  • the productivity dip after go-live (challenge 3 has the evidence)
  • post-go-live support and the upgrades that follow

Current ranges by system type, practice size and build route are in our EHR software cost guide. If you’re a small practice comparing products, start with EHR options for private practice.

The fix is a lifecycle budget with the dip in it. I’d give the dip its own line, with a number of weeks next to it, and carry the budget past go-live.

Staff Resistance: Bring Clinicians In Before the Contract

When a clinician pushes back on a new EHR, they’re usually doing the math on their own day: more clicks, slower notes, a template that doesn’t match how they examine a patient.

The evidence points the same way. In KLAS’s 2025 implementation research, change management that starts too late or stays too narrow is one of the recurring reasons projects miss the mark, along with training that gets cut or deferred (KLAS, February 2025). The fix is to give frontline staff a real say in the decision:

  1. Put practicing clinicians on the selection team, with a vote, before the contract is signed.
  2. Name a clinical champion per department and give them protected time.
  3. Build a super-user group (nurses, front desk and billing too, not just physicians) who test workflows before anyone else sees them.
  4. Show each group the one workflow that gets better for them. People move when they see their own Tuesday clinic get easier.

If you’re worried resistance means your team is the problem, it usually means the plan is.

Training and the Go-Live Productivity Dip

Plan for the dip; don’t hope it away.

Training is where most budgets get trimmed, and the data says that’s a false saving. In a KLAS Arch Collaborative study of more than 72,000 clinicians across 156 organizations, 38.4% of clinicians who said their initial training didn’t prepare them also said the EHR doesn’t enable high-quality care, compared with 10.3% of those who felt well trained (Longhurst et al., Applied Clinical Informatics, 2019).

It’s an association, so it can’t show cause, but the gap is wide. KLAS’s 2025 research adds that just 43% of physicians and 33% of nurses feel their initial EHR training was sufficient (KLAS, 2025).

The dip itself can last longer than people expect. At one US academic ophthalmology department, visits per provider were 16.9% lower on average across the 4 years after go-live, and no physician had returned to the pre-EHR volume by the final year (Lam et al., BMC Health Services Research, 2016). That’s 8 physicians in one specialty, so read it as a warning sign worth planning for. Support staffing in that department stayed the same throughout, so plan your staffing for the dip rather than assuming it fixes itself.

If something has to give in the budget, I’d cut a feature before I’d cut training. What works:

  • role-based training on your own workflows, using your own templates and patients in a test environment
  • reduced schedules for the first weeks after go-live, agreed with scheduling before the date is set
  • at-the-elbow support on the floor during go-live, then refresher sessions once people know what they don’t know
  • a named owner for training after go-live, because new hires arrive every month

Data Migration: Decide What Moves, What’s Archived, What Stays Read-Only

Most of the risk in migration sits in one decision: what not to move. My advice is to argue about that list early, with clinicians in the room.

Old problem lists, duplicate patients and free-text fields that the new system can’t map turn into errors the day clinicians start trusting the new chart. Split your data three ways:

  • Discrete data that moves: active problems, medications, allergies, recent results, immunizations, demographics. Map field by field, run trial loads, then validate and reconcile counts against the source.
  • History that’s archived: older encounters and scanned documents, kept searchable but not converted.
  • The legacy system, read-only: keep it available read-only for 30-90 days, set by your retention and audit needs. Records-retention law can require you to keep the data itself far longer, in an archive rather than the live system (see the state-law point below).

Retention is a separate question from migration. HHS says the HIPAA Privacy Rule doesn’t set a retention period for medical records; state law generally does (HHS HIPAA FAQ). So the archive has to outlive the project.

We do this work. Mindbowser moved 180 production databases; the cloud migration case study has the details. Our EHR data migration guide walks through the full process.

Interoperability: Plan Every Interface, and the Information-Blocking Rules

Your new EHR has to talk to systems you don’t control: the reference lab, the pharmacy network, the hospital down the road, your payers.

Hospitals show how far the US still has to go. In 2023, 70% of non-federal acute care hospitals engaged in all four domains of interoperable exchange (send, receive, find and integrate), but only 43% did so routinely. Seventy-one percent had routine access to the outside information they needed, yet only 42% said clinicians routinely used it when treating patients. Small (38%), rural (36%) and critical access hospitals (37%) were routinely interoperable less often (ONC Data Brief No. 71, May 2024).

The 21st Century Cures Act prohibits information blocking (45 CFR Part 171). HHS’s Office of Inspector General can penalise developers of certified health IT, HIEs and health information networks under its 2023 penalty rule, and disincentives for providers took effect on July 31, 2024 (89 FR 54662). On September 3, 2025, HHS announced it would increase enforcement resources against information blocking (HHS press release).

The fix:

  1. Inventory every interface before you pick a go-live date: labs, e-prescribing, HIE, payers, imaging, devices, patient apps.
  2. Name the standard each one uses: HL7 v2 messages, FHIR R4 APIs (a newer web-API standard) with US Core profiles, or C-CDA documents.
  3. Test every interface with real messages in a test environment before cutover, not after.

This is the part of an implementation we build. We build lab integrations, including Quest and LabCorp, over HL7 and FHIR. We integrate e-prescribing with Surescripts and DoseSpot, eligibility and claims with Stedi, and we’re a Health Gorilla partner for lab-network access. ConnectHealth connects your EHR to other EHRs, labs and payers over FHIR, HL7 and X12. For the wider picture, see our healthcare interoperability solutions.

Privacy and Security: Meet Today’s Rule, Prepare for the Proposed One

Start with the HIPAA Security Rule in force today. It requires a risk analysis, access controls, audit controls and business associate agreements (45 CFR Part 164, Subpart C).

HHS proposed a major update on January 6, 2025 (90 FR 898). It would make safeguards such as encryption and multi-factor authentication explicit requirements. As of October 2026, no final rule has been published in the Federal Register. My view: build to the proposed rule now. Encryption and MFA are cheaper to design in than to retrofit, and if the final rule changes, you’ve lost nothing.

An implementation raises your exposure in two places: during migration, when PHI (protected health information) moves between systems, and at every new interface. I’d run the risk analysis for the new setup before go-live, sign a BAA with every vendor and partner that touches PHI, and turn on audit logging from day one.

Mindbowser holds a SOC 2 Type II report, and Mindbowser signs a BAA with every client that handles PHI. More on HIPAA-compliant software development and on business associate agreements.

Need Custom EHR Workflows? Let's Build Around Your Existing Clinical Systems.

Usability and Documentation Burden: Test With Real Clinicians

Primary care physicians in the Arndt study spent 5.9 hours a day in the EHR; 44.2% of that was clerical and administrative work, and inbox management took another 23.7% (Arndt et al., 2017). Poor usability also carries safety risk. Of 1.735 million patient safety reports from 2013 to 2016 in and near Pennsylvania, 557 used language linking an EHR usability issue to possible patient harm. Data entry (27%), alerting (22%) and interoperability (18%) were the most common categories (Howe et al., JAMA, 2018).

Certified products are tested for this. ONC’s safety-enhanced design criterion requires user-centered design processes, and the regulation itself sets a minimum of 10 test participants for each covered capability (45 CFR 170.315(g)(3)(ii)). That testing uses the vendor’s test workflows. Your workflows need their own round:

  • run your own usability sessions on your top 10 workflows, with the clinicians who’ll use them
  • fix order sets, templates and default settings before go-live, while changes are cheap
  • after go-live, use EHR event logs to find where time goes, the way the Arndt team did

Ambient AI documentation is one answer to the burden; our guide to AI in EHR covers what it needs to work.

Regulatory Compliance: Know Which Rule Binds Whom

Half the regulatory anxiety I hear is about rules that bind your vendor, not you. Sorting that out first saves weeks.

RuleWho it bindsWhat it means for your implementationKey date
Standardized API criterion, 45 CFR 170.315(g)(10) (2015 Edition Cures Update)Developers of certified health ITCertified products must answer patient and population data requests through FHIR APIs, using the standards in § 170.215 and US CoreIn force
ONC HTI-1 (89 FR 1192)Developers of certified health ITUSCDI v3 became the certification baseline on January 1, 2026 (ONC). Ask your vendor for its updated certificationPublished January 9, 2024
EHI export, 45 CFR 170.315(b)(10)Developers of certified health ITCertified products must export all of a patient population’s electronic health information in a computable format. This is your exit route (challenge 9)In force
Information blocking, 45 CFR Part 171Providers, developers, HIEs and networksYour own policies on releasing data are covered, not just your vendor’sProvider disincentives from July 31, 2024
HIPAA Privacy and Security RulesCovered entities and business associatesRisk analysis, access and audit controls, BAAsSecurity Rule update proposed January 6, 2025
CMS Interoperability and Prior Authorization rule, CMS-0057-F (CMS)Medicare Advantage, Medicaid, CHIP and Exchange plan payersPayer APIs your EHR can call for patient data and prior authorizationAPI requirements generally from January 1, 2027

If you’re building your own EHR, certification becomes your question. If your build sits on a certified platform, the platform is certified; certifying your EHR built on it is a separate step: scope it from day one. Check any product’s status on ONC’s Certified Health IT Product List. Our USCDI v3 compliance guide covers the data requirements, and our explainer on ONC’s HTI-1 final rule covers the rest of that rule.

Vendor Dependency: Plan Your Exit Before You Sign

The best time to plan leaving a vendor is before you choose one. Once your data, interfaces and workflows live inside a product, every renewal is negotiated from a weaker position.

Ask these before you sign:

  • What do you commit to on uptime and support response, in writing?
  • Which APIs can we use, at what volume, and what do they cost to use?
  • How do we run the full patient-population EHI export, in what format, and how long does it take?
  • What help do you provide at termination, and for how long?
  • Is the product on the Certified Health IT Product List today, and for which criteria?

The EHI export criterion gives you a way out: certified products must be able to export all the electronic health information they store, in a computable format, with a public link to the format (45 CFR 170.315(b)(10)). Ask to see it run on test data before you sign.

For the shortlist itself, see our guide to comparing EHR vendors. If a partner is building part of your system, our guide to evaluating an EHR development partner covers the contract questions.

Which Challenges Hit Hardest by Organization Type?

The same nine challenges weigh differently depending on who you are, and only one of those differences is well measured. ONC’s 2023 data shows small (38%), rural (36%) and critical access hospitals (37%) routinely interoperable less often than hospitals overall (43%). If you run one, put interfaces (challenge 5) at the top of your list.

For everyone else, use Figure 1: mark the challenges that bind you, then plan the ones with the most money or the most patients behind them first. A small practice will usually feel cost and the productivity dip most. A digital-health company building its own EHR will feel regulation and interfaces most.

When Do These Challenges Point to Building Around Your EHR?

Some of these challenges are structural. No amount of training fixes an interface your vendor won’t support, a specialty workflow its templates can’t hold, or a patient app it doesn’t offer.

That’s where building comes in, and it usually means building around the EHR you have. We build custom EHRs and the software around existing ones: integrations, data migration, SMART on FHIR apps and custom clinical layers. Mindbowser is an Epic Vendor Services member, and we build integrations with Epic, Oracle Health and athenahealth. We also offer EHR-to-EHR migration as a service. You own the code and the data.

When the EHR itself is the product (a new care model, a specialty nobody serves well), a custom build is the third route. A precision medicine practice we built for had its initial platform live in under ninety days. Starting with a short discovery and a prototype gets you better results. Our guide to ready-made or custom EHR helps you choose the route, and our custom EHR development page shows how we build. If Epic is your EHR, see our Epic integration work.

What’s Changing Next

Three items to put on your implementation calendar, each checked at the source on October 25, 2026:

  • HIPAA Security Rule update: proposed January 6, 2025; no final rule published yet. Design to the proposal.
  • CMS-0057-F payer APIs: requirements generally begin January 1, 2027, so payer data and prior-authorization APIs become something your EHR can call.
  • ONC’s HTI-5 proposal: published December 29, 2025 as a deregulatory proposal to the certification program (Federal Register); still a proposed rule as of today. Watch which certification criteria it changes before you lock your vendor’s roadmap into your plan.

Information-blocking enforcement is the fourth thing to watch. HHS said in September 2025 it would step it up; the first public actions will show what it means in practice.

Where to Start With Your EHR Implementation

Go back to Figure 1 and mark the three challenges that bind your organization hardest. Give each one an owner and a line in the budget before you sign anything.

Then read the pages that cover the rest of the work: the EHR implementation guide for phases, team and timeline, the mistakes to avoid for the planning traps, and the EHR data migration guide if you’re moving off a legacy system.

What I still can’t fully answer is why the productivity dip lasts years in some studies and weeks in others. If your organization measured its own, I’d like to compare notes.

Frequently Asked Questions

Data migration is one of the most common. Moving records from an old system without duplicates, missing results or mismapped fields takes careful mapping, trial loads and validation, and the old system usually stays available read-only for a period after go-live.

Four of the most common are cost (including the productivity dip after go-live), staff resistance, interoperability with labs, pharmacies and other providers, and privacy and security requirements under HIPAA. Each needs an owner and a budget line before the contract is signed.

No single study ranks one issue first for every provider, but documentation burden is one of the best measured. Physicians already spend a large share of their day in the EHR, much of it on clerical work, so a new system that adds clicks or slows notes meets resistance. Usability testing with your own clinicians on your own workflows is the fix.

For a custom EHR build, an initial platform can be live in under 90 days. Replacing an existing EHR takes longer as the number of sites, interfaces and records to migrate grows. Our EHR implementation guide breaks timelines down by system type and phase.

The biggest drivers are the system type, the number of providers and interfaces, how much data you migrate, and the training time you backfill, including the productivity dip after go-live. Our EHR software cost guide has current ranges by route, from subscription products to custom builds.

Pravin Uttarwar

Pravin Uttarwar

CTO & Founder, Mindbowser

Connect Now

Pravin Uttarwar is CTO & Founder at Mindbowser. He has 16+ years of experience as a developer and technology leader, with deep expertise in healthcare platform architecture, AI/ML strategy, and build-vs-buy decision frameworks.

His career spans founding and growing Mindbowser from a startup to a 150+ person healthcare technology company, while maintaining hands-on technical depth across system architecture, remote team operations, and developer experience.

Share This Blog

Read More Similar Blogs

Let’s #Transform Healthcare,# Together.

Partner with us to design, build, and scale digital solutions that drive better outcomes.

Location

Global Tech Teams LLC, 525 Washington Blvd, Industrious at Newport Tower, Jersey City, NJ 07310, United States.

Contact

+1 408 786 5974
contact@mindbowser.com
BOOK A QUICK CONSULTATION

Have a Healthcare Project in Mind?

Let’s discuss your goals, workflows, and next steps in a focused consultation call.

Calendar icon Schedule a Call

Contact form