Narrative-first, evidence-driven

Back to playbooks

This is a small, repeatable playbook for helping clinicians who are drowning in admin drag.

Goal: give minutes back without creating new harms.

Core loop:

Map the drag → pick 1–2 targets → apply guardrails → ship a small fix → measure → repeat

What’s inside

• Admin Drag Map (MD / Clinic)
  Where time goes (documentation, prior auth, inbox, etc.) and what’s driving it.

• AI for Care (with Guardrails)
  What AI can do safely, what it must never do, and the always-required rules.

• Triage and Prioritization
  How to pick the first target (frequency × minutes × judgment × risk).

• Standard Plays (Minimum Viable Interventions)
  A small menu of safe interventions (delegation + standardization first, automation second).

• Scoreboard (Prove It Helped)
  3–5 metrics to prove this is working.

• Field Notes (MD / Clinic)
  A 5-minute weekly capture that turns lived pain into the next fix.

• One Ask (Clinic Ops / IT / Leadership)
  A short request you can send to leadership/ops/IT to remove a bottleneck.

How to use (the 30-minute version)

1. Do a quick admin drag map (10 min)
2. Use triage to pick one target (5 min)
3. Apply guardrails (2 min)
4. Choose one standard play and ship a minimum viable fix (10–60 min)
5. Update the scoreboard weekly (2 min)
6. Capture a field note weekly (5 min)

Success criteria

This is working if:

• clinician after-hours (“pajama time”) is down
• inbox time is down
• note closure time is down
• prior auth/referral cycles get shorter
• nothing gets less safe, less humane, or less accountable

Back to playbooks