Where Does the Time Go?

A hospital core laboratory pulls a baseline month of stat basic metabolic panel (BMP) turnaround time (TAT), defined as specimen receipt in the laboratory to result release. Volume for the month is 850 stat BMPs. Mean TAT is 42 minutes against a 45-minute target set by the laboratory director, so the monthly summary looks fine on its face. The same data set has a 90th-percentile TAT of 71 minutes, 26 minutes past target, and 145 of the 850 results, 17.1 percent, run past 45 minutes.

One average and one tail number describe the same month and point in different directions. The mean says the process is basically working. The 90th percentile says roughly one in ten stat results runs badly late. Both numbers are computed correctly from the same data; the question is which one a laboratory should act on, and why they can disagree at all.

Seeing that disagreement clearly, before choosing a fix, means asking where a specimen waits, loops back for recollection, or sits in a queue without becoming a released result. Lean and Six Sigma are two structured ways to answer that question and to test a change safely, without guessing and without quietly dropping a required control in the name of speed.

A mean inside target does not mean the process is in control; check the tail before deciding nothing is wrong.