College of American Pathologists
Printable Version

  President’s Desk





cap today



March 2007
Limping cows in the laboratory

Thomas M. Sodeman, MD

I don’t care about floors. I care about cows. Are they falling down? That’s all I need to know.

—Temple Grandin, PhD, Animals in Translation

In February of last year, R. Bruce Williams, MD, who chairs the CAP Commission on Laboratory Accreditation, brought a book to the Council on Laboratory Accreditation meeting in San Diego. Animals in Translation is written by Temple Grandin, PhD, who works with the government and the fast food industry. An animal scientist, she creates systems that ensure animals in livestock-handling facilities are treated humanely. Her concepts are radically simple; she is relentlessly outcomes-oriented. Dr. Grandin, who is autistic, has insights into perception and motivation that make for a remarkable read.

When Temple Grandin audits a farm or a plant, she works with a very short list of critical control points, single measurable elements that cover a multitude of sins. Rather than creating a checklist to examine all the things that might compromise a cow’s ability to walk (bad flooring, poor diet, inadequate hoof care), she looks to see how many are limping. If too many cows are limping, the facility fails the audit. It’s up to the farmer to figure out why, and fix it.

In the quote that opens this column, Dr. Grandin is talking about process-oriented inspection systems that evaluate the cleanliness and integrity of flooring instead of whether the cows can walk on it safely. Auditors, she says, should be looking at outputs—how the animals are doing—rather than inputs—structures, schedules, training records, and equipment. “A good animal welfare audit has to measure the animals,” she writes, “not the plant.”

Dr. Grandin’s paperback struck a nerve that day, and we’ve been talking about limping cows ever since.

There’s a lot going on in the Laboratory Accreditation Program. The College is investing $9 million in new information systems that will integrate quality measures and continuously monitor performance throughout the accreditation cycle. We will soon be able to combine proficiency testing results, trend analysis, and on-site inspection findings. Those data will enable laboratory staff to monitor day-to-day quality and address underlying deficiencies long before they become significant concerns.

The leaders of our accreditation program are thinking about conceptual changes to our system, about becoming more outcomes-oriented. They will keep us posted as they move forward; anything they do will be data-driven and carefully framed. In the meantime, however, each of us can identify the critical control points—the cows that must not limp in our laboratories. The idea is to keep the list short and concrete, and make the response autonomic.

Critical control points are outcomes that measure the whole process. Physician/nurse satisfaction with laboratory performance is a big one; survey them regularly. Are the values you produce consistent with the clinical picture? Are your statistics sound? Does your turnaround time meet their patient care needs? Is your off-site laboratory work timely and accurate?

Unhappy physicians and nurses are limping cows and investigation is required. Can they not believe the values? Are the results not timely? Are they not posted to the chart correctly? There are dozens of possibilities. The pathologist’s task is to deconstruct that unhappiness, find out why the colleague/nurse satisfaction cow is limping, and fix it.

Laboratory errors that affect patient care and complaints of any kind are limping cows that call for an immediate investigation. Disappointing proficiency test results are also markers for trouble. Label and order accuracy are indicators to track, as is the number of samples misplaced en route to the laboratory.

If all of your instruments are meeting their quality control numbers, a lot of things are happening correctly. Inconsistent QC results could mean a number of things. Maybe the equipment was not functioning properly. Perhaps a reagent was inappropriate, out of date, or improperly stored. Maybe the technologist wasn’t properly trained in use of the equipment. It could be that the refrigerator wasn’t cold enough and the material went bad. Rather than set up elaborate checklists for each of these, monitor the instrument performance. If disappointed, work backward.

Laboratory tests that trigger treatment decisions are the ultimate critical control points, but until now we have had few opportunities to link our results to patient benefit. That will change as molecular testing and pharmacogenomics become increasingly available. For example, last year, the College partnered with the American Society of Clinical Oncology to write guidelines for HER2 testing. New research had shown a response rate of 50 percent when HER2-positive breast cancer was treated with a powerful agent, trastuzumab (Herceptin). However, laboratory tests to detect HER2 were running 25 percent false-positive and the drug was highly cardiotoxic. The College collaborated with the ASCO to develop clinical practice guidelines for improved HER2 testing accuracy. HER2 results will now enable clinicians to provide this treatment only to those who will benefit.

An outcomes-oriented approach to laboratory quality will enable pathologists to examine their quality assurance concerns, both anatomic and clinical, and ask hard questions. Are we measuring what is necessary to ensure a good outcome for our patients? Sound processes will always be important, but they are a means to an end. Quality laboratory medicine fosters optimal patient outcomes. The rest, we all know, is detail.

Dr. Sodeman welcomes communication from CAP members. Send your letters to him at