Key Highlights
- A new electronic health record is bought to improve care. For a stretch after go-live it often does the opposite, and the cost is not shared evenly. It lands on the hospital’s finances, on the staff who have to work in the system, and on the patients they care for.
- The hit is measurable. Productivity commonly falls on the order of 20 to 40 per cent in the first weeks as clinicians see fewer patients while they adjust (Medical Economics), and for many organisations the disruption is not a passing phase but a new, lower baseline.
- Poor adoption is a workforce problem too. The EHR is a named contributor to burnout for 43 per cent of nurses and 62 per cent of physicians who report it, and burnout drives the turnover that costs a hospital most (KLAS, 2025).
- The good news is that the biggest lever sits with the hospital. Training and support are the part of a rollout it fully controls, and the evidence says they move the outcome more than anything else.
A system meant to help that first makes things worse
A hospital replaces its EHR to standardise care, join up information, and work more efficiently. Then go-live arrives, and for a run of weeks the system seems to do the reverse. Clinics fall behind, order entry takes longer than it did on the old platform, and the help desk is swamped. This is not about which vendor was chosen. Whether a hospital runs Epic, Oracle Health, or MEDITECH, the pattern repeats, because it is an adoption problem, not a software one.
The gap sits between two things that are easily confused: a system being live and a system being used well. Clinicians are asked to relearn how they work while still caring for patients, and the training that would carry them across is usually the first line cut when timelines tighten. What follows is predictable, and its cost falls in three places.
The cost to the hospital
The first bill is operational and financial. Productivity commonly falls sharply in the first weeks after an EHR go-live, on the order of 20 to 40 per cent, as clinicians see fewer patients while they adjust, and the help desk is swamped at the same time (Medical Economics). That means fewer patients seen at precisely the moment the organisation can least afford it.
How long the dip lasts is the part leaders underestimate. With good preparation and support, most organisations climb back within a few months. Without it, recovery drags on far longer, and some never fully return to where they were. A rocky go-live is not a bad month. It can become the permanent baseline the hospital operates from.
The cost to staff
The second bill is paid by the people using the system, and it is heavy. Physicians already spend close to two hours on the EHR and desk work for every hour of direct patient care (Sinsky et al., Annals of Internal Medicine, 2016), and the after-hours “pyjama time” that follows them home runs to around 86 minutes a night for family physicians (Arndt et al., 2017). A poorly adopted system adds to that load rather than easing it.
The consequence is burnout, and the EHR is squarely in the frame. KLAS attributes some level of burnout to around a third of nurses and physicians, and among those affected, the EHR is named as a contributor by 43 per cent of nurses and 62 per cent of physicians; only 18 per cent of physicians report a strong or elite EHR experience at all (KLAS, 2025). This is not only a wellbeing issue. Burnout drives people out of the door, and every clinician who leaves is expensive to replace, in recruitment, onboarding, and the institutional knowledge that walks out with them. Poor adoption quietly becomes a retention and recruitment problem.
The cost to patients
The third bill is the one that matters most, because it is paid in safety. When a system fights the workflow, clinicians do what people always do under pressure: they invent their own way round it. Shadow spreadsheets, paper notes, and unofficial workarounds appear, and every one of them is a patient-safety and data-integrity risk, because the record no longer holds the full picture.
The danger concentrates in the first weeks live, when unfamiliar tasks surface fastest and fluency is lowest. The high-risk workflows are exactly the ones people fumble when they are still learning: medication administration and reconciliation, patient identification, allergy documentation, order entry. A hesitation or a mis-click in any of those is not a productivity statistic. It is a potential harm to a patient, and it is why the quality of adoption is a clinical concern, not an administrative one.
Planning a move to a new clinical system, or trying to recover from a rocky one? Our healthcare playbook covers how hospitals scale training and protect care through a go-live.
The one cost you can avoid
Of the forces behind a hard go-live, most are set during the technical build and can only be shaped, not fixed, once clinicians are on the floor. Training and support are different. They are the part of the transition a hospital owns outright, right up to go-live and beyond, and the evidence that they matter is unusually clear. The KLAS Arch Collaborative, drawing on more than a million clinician surveys, found a 90-point difference in Net EHR Experience Score between clinicians who felt their training prepared them well and those who did not, on a scale from -100 to 100.
Two things make that preparation stick: hands-on practice before go-live, and content the hospital builds itself. Practice has to be safe, because no one should learn on real patients. Assima’s object-based cloning produces an anonymised, editable copy of the actual EHR, so clinicians rehearse complete tasks, exceptions included, with no patient data exposed. When the hospital’s own people build those lessons, the content reflects how care runs on the ward rather than in a generic reference build. At CHU Montpellier, the hospital’s team built more than 150 lessons across every staff profile and cut classroom time per course from three hours to 40 minutes. At Zuyderland, a Dutch group of four hospitals, a team of its own nurses built around 300 exercises and prepared 2,500 staff for a new clinical system while, their manager reported, care to patients continued uninterrupted.
Once live, support inside the system carries the rest. Assima Assist walks a clinician through a step at the moment they are stuck, and Assima In-App Search brings the hospital’s own instruction onto the screen they are already on, so an unfamiliar task does not become a phone call or a paused clinic.
Conclusion
A rocky EHR go-live is often treated as an unavoidable rite of passage. It is not, at least not in how deep the dip runs or how long it lasts. Those are set largely by how well people were prepared, and preparation is the part of the whole programme the hospital most fully controls. Get it right and the disruption is brief and expected, and the cost to finances, to staff, and to patients stays small. Get it wrong and the system inherits the blame for months, sometimes permanently, for a gap that was never about the software. The technology was only ever going to be as good as the workforce’s ability to use it.
See it on your own systems. Book a demo and watch Assima shrink the EHR go-live dip.
Frequently Asked Questions
Let’s Answer Some of Your Questions.
Because clinicians are learning an unfamiliar system while still caring for patients, so tasks that were automatic on the old platform now take longer. Productivity typically falls 20 to 40 per cent in the first month. With structured support it usually recovers within 60 to 90 days; without it, recovery can take 6 to 12 months or longer.
It adds to an already heavy load and feeds burnout. Physicians spend close to two hours on the EHR for every hour of patient care, and the EHR is named as a burnout contributor by 43 per cent of nurses and 62 per cent of physicians who are affected. Because burnout drives turnover, poor adoption becomes a retention problem, and replacing an experienced clinician is costly in recruitment, onboarding, and lost knowledge.
When staff cannot use the system smoothly, they create workarounds like shadow spreadsheets and paper notes, each of which is a safety and data-integrity risk. Errors concentrate in the first weeks after go-live, particularly in high-risk tasks such as medication reconciliation, patient identification, and allergy documentation.
Its depth and length can be reduced sharply. The dip is driven largely by how well clinicians were prepared, which is the part a hospital controls. Realistic practice before go-live and support inside the system afterwards keep the dip brief and expected rather than deep and prolonged.
The hospital itself, with the right tooling. Vendors teach the system well in general, but only the hospital knows how its own workflows run. At CHU Montpellier and Zuyderland, the hospitals’ own teams built the simulations while Assima supplied the platform, so the content stayed accurate and owned by the people who use it.
