Clinical patterns

Sign in / sign out

Identity on a workstation twelve people share, in a corridor, mid-shift, with gloves on. Every attributed action in the system — every acknowledgement, every override — depends on this being both correct and fast.

Stable · v1.0 Pattern Attribution WCAG 2.2 AA AcuteLine

Overview

Clinical authentication has an unusual property: the security control and the safety control point in opposite directions. Security wants short sessions and frequent re-authentication. Safety wants a clinician never to be locked out of a patient's data during a resuscitation.

This pattern resolves that by separating reading from attributing. Access to view is protected but generous; the ability to write something into a patient's record with your name on it is protected strictly. Nothing here ever stands between a clinician and information they need to act.

Fast

Sign-in on a shared workstation is measured in seconds and works with gloves. Anything slower gets defeated by shared logins, which destroys attribution entirely.

Unambiguous

Who is signed in is visible at all times, not buried in a menu. A clinician acting under a colleague's session is an attribution failure with clinical consequences.

Never obstructive

No authentication state ever hides clinical information already on screen or blocks an emergency pathway. Locking degrades to read-only, never to blank.

Sign in

Shared workstation sign-in
Sign in to AcuteLine
ED workstation 4 · tap your badge, or sign in below
Caps Lock is on

Emergency access

Break-glass

Every clinical system needs a path for the case where authentication is broken, the directory is unreachable, or the clinician in front of the patient is not the one with an account. The design question is never whether to have it, but how to make it fully available and fully visible.

Emergency access · granted and declared

Locking, not logging out

Idle timeout is where most clinical systems get this wrong. Blanking the screen during a resuscitation is a safety failure; leaving a writable session open in a corridor is a security one. The system does neither.

Locked — readable, not writable
HARLAND, Arthur J.
MRN 44 812 907
Date of birth
1964-03-11 · 61 y
Location
ED · Bay 4
Session locked — viewing only
S. Whitfield's session locked after 10 minutes idle. Clinical data stays visible. Tap a badge to resume, or sign in as someone else.
While lockedBehaviour
Clinical data on screenStays visible. Never blanked, never blurred.
AlarmsFully visible and audible. Locking never suppresses a signal.
Navigation and viewingPermitted — reading is not an attributed action.
Any attributed writeBlocked until identity is re-established.
ResumeBadge tap or credentials. Same user returns to the same context.
Switch userNew identity, and the patient context is re-confirmed.
Why blanking is the wrong default

A blanked screen during an arrest costs seconds that matter, and reliably produces the workaround that defeats the whole control: a shared account nobody logs out of. Read-only locking protects attribution — which is the actual clinical risk — without ever putting a login prompt between a clinician and the information in front of them.

Switching user

The highest-risk transition in this pattern, because the person changes while the patient context does not.

Do
You have an unsaved assessment
S. Whitfield started an assessment on HARLAND, A. It cannot be recorded under your name.

The draft cannot cross an identity boundary without a decision.

Don't
Signed in as A. Bergström
Your previous work has been restored.

Someone else's draft, silently re-attributed. The audit trail now says Bergström wrote something Whitfield wrote.

Showing who is signed in

Do's and don'ts

Do
Session locked — viewing only
Clinical data stays visible. Tap a badge to resume.

Locking degrades to read-only. Nothing is hidden; only attributed writes are suspended.

Don't
Screen locked — sign in to continue

A blanked screen during an arrest. This is what produces the shared account nobody logs out of.

Accessibility

Outcomes of use

What this contributes to, in the terms of Usability & context of use. These are attributes believed to contribute to an outcome; the outcome itself is settled by observing real use in a specified context, not by this page.

Clinical safety notes

Risk controls carried by this pattern

Trace these in your risk file (ISO 14971) and usability engineering file (IEC 62366-1).

NotJustAnyMed.Tech Design System · Sign in / sign out · v1.0 · draft for review
Reference applications named in this system are fictional; all patient data shown is fabricated.