Six-Week Infusion Pump Conversion With Three Educators

Priority is driven by risk of harm and by the size of the observed practice gap, not by the calendar or by who is easiest to schedule.

1. Sequence to the physical swap, unit by unit, so training and go-live are not separated by weeks of decay.
2. High-alert units first - ICU, ED, OB, pediatrics - because the pilot's 68% drug-library error rate is a harm signal, not a knowledge curiosity.
3. Place the four train-the-trainer seats deliberately: one night-shift charge nurse and one unit educator from the two highest-volume units, chosen for teaching skill rather than seniority.
4. Redesign the four-hour class into a 60-75 minute station rotation: drug-library selection, alarm limits, secondary programming, occlusion troubleshooting, each with return demonstration. Four hours of lecture will not produce programming accuracy.
5. Nights and weekends: 20-minute sessions at 0300 plus piggybacking on existing huddles. Never rely on "come in on your day off" for 190 night nurses.
6. Add a day 7-10 booster (spaced) because massed training alone decays before go-live.
7. Low-volume units get just-in-time training plus a validated super-user on every shift for the first two weeks.

Validation must answer one question: can this nurse program this pump with this drug library, under time pressure, without a critical error?

• Common standard, variable practice time. Everyone meets the same performance criterion; the amount of practice differs.
• Performance-based. Two consecutive correct attempts on a high-alert scenario (heparin infusion with a weight-based rate change) with no critical error and no prompting. Completing the e-module alone never counts.
• Build the checklist from failure modes seen in the pilot and in the near-miss reports - drug-library misselection, wrong channel, missed secondary - and have content experts review it.
• Include a "detects and corrects" item. Program an intentional library error and see whether the nurse catches it; the near-miss data suggest nurses did not recognize the error when it occurred.
• Check the raters. Dual-score 10% of validations to look for inter-rater disagreement.
• Document defensibly: scenario, date, rater, outcome, in the learning management system, so the record stands up at survey.
• Super-users complete the same validation plus a short session on the teach-back method; subject-matter expertise is not the same as teaching skill.

The instinct is right; the arithmetic is not. 640 nurses at roughly 60 minutes each is about 640 educator-hours. Three educators plus 40 super-user hours per unit cannot deliver that in six weeks without pulling nurses from patient care.

Proposal - redefine "prepared" as explicit guarantees:

• 100% of nurses in high-alert units validated by observed return demonstration before they touch the new pump.
• 100% of all nurses complete the e-module and a hands-on station before first independent use; anyone who has not is not assigned solo pump responsibility.
• 100% of shifts covered by a validated super-user for the first 14 days on every unit.
• A 72-hour rule: no high-alert drug programmed on the new pump without a second check until the unit's validation rate exceeds 90%.

Defense: present a one-page risk register - what is guaranteed, what residual risk remains, and what mitigates it - and ask the CNO to sign it. This is not a refusal; it is the version of the CNO's intent that can actually be executed. Forcing massed hands-on sessions in six weeks would consume the education capacity and still leave untrained night staff at go-live, which is the outcome the CNO is trying to prevent.