Nineteen Days from BCMA Go-Live: A Conditional Go/No-Go
Northgate Hospital is 19 days from bar-code medication administration go-live on 14 units. During integrated testing, 12 of 400 medication orders could not be scanned because the wholesaler's labels do not meet the formulary symbology standard. Those two drugs account for roughly 9% of doses given in the ICU. The vendor describes the defect as a post-go-live issue and offers a manual override workflow. Pharmacy wants to delay go-live entirely. The CMIO argues that 97% scan success still beats the current paper process, which has no scanning at all. The steering committee meets in 48 hours and has asked you, the informatics nurse lead, for a recommendation with conditions.
1. What additional information do you need before recommending a go/no-go, and why does each item matter?
Read the worked answer
The 97% figure is not decision-grade on its own; it describes a rate, not a risk. Needed:
- Which two drugs. If they are high-alert (insulin, heparin, opioids, concentrated electrolytes) or look-alike/sound-alike, the override removes the last independent verification for exactly the doses where error is most harmful. If they are low-risk, the calculus changes entirely.
- Failure mode. Is the scanner rejecting a non-compliant label, or mis-mapping a readable one to the wrong product? A reject is an inconvenience; a mis-scan is a wrong-drug event waiting to happen.
- Can pharmacy apply compliant labels? Repackaging or over-labeling may close the gap in days. Volume, staffing, and expiry implications determine whether that is real or aspirational.
- What the override actually requires. A one-tap override with a drop-down reason is not equivalent to one requiring an independent second nurse, a witnessed check, and a hard-coded audit trail.
- Baseline harm rate on paper, so that 'better than paper' can be quantified rather than asserted.
- Policy and accreditation expectations for barcode verification at your organization, and how a documented, time-limited exception is treated.
- Readiness factors: super-user coverage, downtime procedures, training completion by unit.
Each item moves the question from a percentage to a residual-risk judgement.
2. Compare delaying go-live, proceeding with a documented manual override, and a limited or staged go-live. State your recommendation with rationale and the conditions that would reverse it.
Read the worked answer
Delay: safest in isolation, but a full-system delay also defers the benefit of scanning for the 97% of doses that work, and paper carries its own unmeasured error rate. Delay is proportionate only if the affected drugs are high-alert and cannot be relabeled.
Proceed with override: defensible only if the override is designed as a safety control rather than a bypass - independent double-check by a second nurse, structured reason capture, visible flag on the MAR, and daily reporting during the exception period.
Staged go-live: the strongest option if the affected volume is concentrated. Go live on the units where the two drugs are low-volume or absent; hold ICU until pharmacy-labeled stock is verified.
Recommendation: conditional proceed, structured as: (1) pharmacy applies compliant labels within 10 days with a named owner and daily count; (2) override requires witnessed double-check until labels are compliant; (3) daily override and near-miss report to the steering committee for 30 days; (4) a documented pause trigger - any wrong-drug or wrong-dose event involving an override, or override volume above a pre-set threshold, halts the rollout pending review.
Reversal conditions: a mis-scan rather than a reject; inability to relabel; or either drug being high-alert with high ICU volume. In those cases, delay ICU only and proceed elsewhere.
3. Describe how you would document and communicate this decision, including who owns residual risk.
Read the worked answer
Documentation: a go/no-go memo that records the options considered, the data reviewed, the option selected, the residual risk accepted in plain language, the named owner for each mitigation, and the date of automatic re-review. Enter the unlabeled-product defect on the risk register with a severity rating and an expiry date. Record any dissent from pharmacy, so that the record reflects the debate honestly.
Ownership must be split, not collective: pharmacy owns label compliance, the CMIO owns the override build and its audit trail, nursing owns adherence to the witnessed check, quality owns event surveillance. Diffuse ownership is how the exception becomes permanent.
Communication: a one-page job aid at every automated dispensing cabinet, a scripted huddle on each unit before go-live, an explicit statement that the double-check is mandatory rather than advisory, and a named contact for the first two weeks. Frontline staff need to hear the reason, not just the rule - an override perceived as paperwork gets bypassed.
Follow-through: a 14-day re-review already on the calendar at the time of approval, with the pause authority delegated to a named group rather than left to be rediscovered during a crisis.