Model Systematycznego Przeglądu

You're describing the exact failure mode AAR was built to kill: debrief once, then forget. A recurring review isn't a meeting—it's a mechanism. Start with a fixed cadence: 15 minutes, same day each week, no exceptions. Use the four AAR questions verbatim: What was the plan? What actually happened? Why the gap? What do we change next time? The critical shift is the third question—it forces you to find the systemic cause, not the person. Your skipped double-check isn't laziness; it's a workflow that makes the step optional under pressure.

Here's the trap: if your review ends with "we'll be more careful," you've failed. Every review must produce one concrete, testable change—a checklist redesign, a physical barrier, a second-nurse confirmation prompt. Log it. Next week, ask: did the change hold? That's the accumulation loop. Also enforce no-blame: rank stays at the door. If nurses fear punishment, they'll hide near-misses, and your review becomes a theater. Track patterns across weeks—two wrong meds in one month is a signal, not a coincidence. That's how you move from patching to preventing.

You’ve hit the exact failure mode that separates a postmortem from a systematic review. A postmortem is event-triggered and one-time; a systematic review is periodic, cross-project, and cumulative. Here’s how to build it.

First, fix the cadence. Don’t wait for loss. Run a 30-minute review at the end of every onboarding, every sprint, every client kickoff—not just failures. The U.S. Army’s AAR works because it’s ritualized, not reactive.

Second, use the four fixed questions: What was the plan? What actually happened? Why the gap? What do we change next time? Keep rank and blame out. Focus on the process, not the person.

Third, close the loop. Every review must produce one or two concrete, testable changes—not a list of 20 wishes. Assign an owner and a deadline. Add it to your onboarding checklist or template immediately, so the next project inherits the fix.

Finally, log every lesson in a single “playbook” document. After three projects, patterns emerge—that’s when experience becomes wisdom. Without this storage and retrieval step, you’ll keep relearning the same painful lesson. The goal isn’t to feel better after a loss; it’s to make the next win cheaper and faster.

Build a Unit Review Loop, Not a Journal Entry

Your journal was a good start, but it was a one-time event. Systematic review demands a fixed cadence, not a reaction. Treat every unit like a mini-AAR.

Step 1: Set the four questions before you grade.
Write them on a card: What was the plan? What actually happened? Why the gap? What do we change next time? Answer them within 48 hours of the unit's end, while memory is fresh.

Step 2: Separate data from feelings.
Pull three numbers: average score, score distribution, and the single lowest-performing standard. Compare to your pacing plan—did you spend 3 days on fractions but only 1 on word problems? That gap is your smoking gun.

Step 3: Force one change, not five.
Pick the single highest-leverage adjustment—e.g., "Add a 5-minute daily exit ticket on prerequisite skills." Write it as a PDCA Check action: Plan next unit, Do the change, Check mid-unit data, Act again.

Leave rank at the door: blame the process, not yourself. The goal is the next unit, not the last one.

Review is a habit, not a post-mortem. Schedule 30 minutes after every unit, and your pacing will correct itself through repetition.