If Your EHR Goes Down: What Physicians Should Do During Downtime

If your EHR is down right now, do this first: stop what you're doing and confirm it's actually down.

Not your computer. Not just you. The system. Try logging in from another workstation, check with the person next to you, and look at your secure chat or overhead announcement. Then activate your downtime protocol immediately. Don't wait 10 minutes to see if it comes back on its own. That's how orders get lost.

When the EHR Goes Down: Stay Calm, Confirm the Scope, and Start the Downtime Plan

If you're in clinic and your templates won't load, or you're on rounds and orders won't enter, your first job is triage the tech, not the patient. Is this local to your pod, your whole clinic, or the entire health system? A frozen computer is an IT ticket. A system-wide outage is a patient safety event.

Activate the downtime plan now. Every hospital and large practice has one. Most people have never opened it.

If you're in this situation, here's who you notify, in this order:

1. Your immediate team. Charge nurse, clinic manager, office manager. Say out loud: "I'm on downtime procedures."
2. IT / Help Desk. Get a ticket number and an ETA, even if the ETA is "we don't know."
3. Operational leadership. Nursing supervisor, medical director on call, administrator on call. They need to trigger the formal downtime communication.
4. Your patients. A quick, "Our computer system is temporarily down, so we're switching to our backup process to keep you safe. It might take a little longer today," goes a long way.

Before you touch paper, do a 60-second safety sweep. Who on your census cannot wait?

• Who has time-sensitive meds due in the next 2 hours? Insulin, heparin drips, antibiotics, anti-epileptics.
• Who has pending critical labs or imaging you were waiting to act on?
• Who is mid-discharge or mid-admission?
• What high-acuity situations are active right now? ED boarders, laboring patients, post-op checks.

The mindset shift: this is not the time to get creative. Do not start documenting full notes on random printer paper, in your personal notebook, on sticky notes that will fall off. Use only your site's approved downtime tools. If you don't know where they are, ask your charge nurse right now. That binder you ignored in the supply room? That's gold today.

What to Do While the System Is Down: Document, Prescribe, Order, and Communicate Safely

Once downtime is declared, everyone should be using the same workaround. If you improvise your own, reconciliation will be a nightmare and something will get missed. I promise.

Here's the core workflow I tell residents and new attendings to follow:

Get the paper packet. Downtime forms are not all the same. You want: downtime progress note / encounter form, manual order sheets, medication administration record (MAR), downtime prescription pads, patient ID labels or wristband printers if they still work, and the critical lab/result callback log.

Document like someone else has to read it at 2 a.m. Because they will. For every patient you touch, write: full name + DOB + MRN, date/time (and write "DOWN TIME" at the top), what you saw, what you decided, what orders you gave verbally or on paper, what meds were given, who you called, and what still needs to happen.

Orders and prescribing on paper:

• Use the approved manual order sheets. No verbal orders unless you absolutely have to, and then use closed-loop read-back. "Give 4mg Zofran IV now" -> nurse repeats "4mg Zofran IV now" -> you confirm.
• Allergy and med rec? You don't have alerts now. That banner that saves you? Gone. Ask the patient again. Check the last printed MAR if you have it. Check the paper face sheet. Double-check.
• High-risk meds - anticoagulants, insulin, opioids, chemo, pediatric dosing - get a second set of eyes. This is policy at most places for a reason.
• Prescriptions going outside? Handwritten or printed downtime scripts need full info. No shortcuts.

Communication is where downtime breaks down. I've seen more harm from bad handoffs during downtime than from the outage itself.

Use closed-loop for everything. If the lab calls a critical potassium on the phone, you write it down, read it back, document time, who called, and who you notified. Keep a single running log for the unit so critical results aren't on five different scraps of paper.

And handoff. If you're leaving while still on downtime, your sign-out needs to be painfully explicit: "Patient in Room 4 had Rocephin ordered on paper at 10:15, not given yet because pharmacy is on manual verify. Needs dose before 12:00. Allergy to penicillin confirmed verbally."

Protect Patient Safety: Prioritize High-Risk Situations, Medication Reconciliation, and Test Tracking

Not all patients carry equal risk during downtime. You need to re-rank your list in your head.

If you're covering inpatient, ED, or L&D during an outage, watch these groups like a hawk: ICU patients on drips, ED patients with undifferentiated complaints, anyone on anticoagulation or insulin, post-op patients in the first 24 hours, patients awaiting dispo based on a pending test, and any patient with frequent PRN needs that normally auto-document.

Three safety traps I see over and over:

1. Medication mess. This is number one for a reason. Missed doses because no one knows when the last one was given. Duplicate doses because two people wrote it down. Active vs. discontinued confusion because the home med list is frozen in time. The EHR normally protects you from yourself. Now it doesn't.

2. Lost labs and imaging. You ordered a troponin an hour ago. Did it result? Who would know? Where would it go? During downtime, results often go back to a manual call system or a backup viewer. If your site has a lab hotline or radiology reading room number, use it. Start a paper list: patient, test, time ordered, time result expected, who is responsible for follow-up, callback number documented.

3. Disappearing orders. New orders are on paper. Old orders are invisible. Consults you called? Procedures? That discharge you entered 5 minutes before the crash that never transmitted?

If the outage goes past 60-90 minutes, call a safety huddle. Five minutes. Stand up. What has been done, what is pending, who is sickest, what are we worried about losing track of. This resets the team. Especially important on nights and weekends when you're short-staffed and tired.

After the EHR Comes Back: Reconcile, Backfill, Review, and Prevent the Next Breakdown

The system says it's back. Don't immediately start pounding in orders like normal. Verify it first. Can you log in? Can you see patients? Are orders actually transmitting to pharmacy and lab? Ask IT if it's a full restore or a read-only view first. I've seen teams pile in orders during a partial restore and lose half of them.

Then comes the part everyone hates: reconciliation. This is not optional busywork. This is patient safety work.

Do this per your policy, but in general:

• Scan or enter your downtime notes into the chart. Make sure they are labeled as downtime documentation with correct original date/times.
• Re-enter orders properly. Compare your paper order sheets to what's now in the EHR. Look for duplicates and misses. Pharmacy and nursing should be doing this too, but you need to check your own patients.
• Rebuild the MAR. Make sure every dose you gave on paper is now reflected electronically so the next shift doesn't re-dose.
• Close the loop on results. That critical lab you wrote on a paper towel? Make sure it's in the chart, acted on, and attested.
• Fix the discharge paperwork. Discharge summaries, med rec at discharge, prescriptions, patient instructions. If you discharged someone on paper, their after-visit summary and PCP follow-up letter never happened.

Run a final check: open each of your patients' charts and ask, "If someone reads this tomorrow, will it tell the full story of what happened today?" If there are gaps, fill them now while you still remember.

Then report. What slowed you down? Where did the downtime binder fail? Did you run out of printed face sheets? Was there no one at the pharmacy window? Speak up in the debrief. Downtime plans only get better when people who actually used them say what didn't work.

Build a Better Downtime Habit: Training, Templates, and Team Readiness

You cannot learn downtime during downtime. That's like learning CPR during a code.

If you're in a leadership role, or just someone who wants the next outage to suck less, do this next week, not next month:

Keep a small stash of your own templates ready. Don't rely on finding the binder at 2 a.m. Printed progress note template, order sheet, sign-out sheet, critical result log. Keep them where you actually know where they are. One folder in your workroom. One in your clinic manager's office.

Run a drill. 20 minutes. Once a year. "System is down for the next hour, let's practice." You'll find out immediately who knows where the downtime computer is and who doesn't.

And get role clarity. In a real outage, everyone should know: who activates the plan, who prints or distributes the census, who tracks new orders to pharmacy/lab, who logs critical results, who does reconciliation. If you don't assign that, everyone assumes someone else is doing it.

Downtime feels chaotic because it is. But it's manageable. The teams that do well aren't smarter - they just had a plan, used the same forms, communicated out loud, and cleaned up afterward methodically.