Choose and Test a Countermeasure

The team has one clear finding: accessioning cannot keep pace during the 0600-0800 and 1400-1600 arrival peaks because staffing is flat. A charge technologist proposes a first test, sized the way the Institute for Healthcare Improvement's (IHI) Model for Improvement recommends: as small as one change, in one place, over a short time, before wider rollout. For two weeks, one technologist's break and administrative block moves from 0700 to 0830, so both peaks have two technologists actively accessioning specimens instead of one.

Before this case pilot starts, its case-local balancing limits are explicit: overtime may not increase by more than 0.5 hour per technologist per pay period, missed breaks may not exceed one, and stat-potassium recollection may not rise above the 2.3% baseline. Crossing any limit stops or adjusts the pilot, triggers investigation, and requires rollback under the plan if the risk cannot be controlled.

The two-week result: 90th-percentile peak-window TAT fell to 54 minutes, inside the predicted range, and the share of stat BMPs over 45 minutes dropped from 17.1 percent to 9.8 percent. Overtime and the recollection rate held steady. One technologist reported that the shifted break felt rushed at shift handoff, an unexpected finding the Do stage of PDSA is built to capture even when it is not part of the original prediction.

That mixed result, a real gain plus one real complaint, is what a PDSA Act decision has to weigh: adopt as-is, adapt, or abandon. A quality indicator exceeding threshold, or one moving in the right direction, still requires documented review, not a simple close-out; a favorable trend on the target metric alone is not evidence that nothing else changed for the worse.

A countermeasure earns a wider rollout by holding its target measure and its balancing measures over a real test window, not by looking good on the one number someone was already tracking.