Q&A column, 7/15

Create PDF

• The height of the work surface plus the height of the centrifuge or printers should be 38 inches above the floor or slightly lower. An optimum solution would be an adjustable work surface so individual users can adjust the height accordingly. This will reduce the need to reach the hands above the shoulders when moving tubes in and out of the centrifuge. The keyboard, mouse, and monitor should be height adjustable to accommodate all users. If the horizontal surface is limited, consider placing the keyboard, mouse, and monitor on a monitor arm.
• Frequent changes in body position will prevent sustained or static postures. A brief stretching break every 40 to 60 minutes or more frequent mini breaks are recommended.
• Alternate between sitting and standing throughout the day.

Suggestions for problem No. 3

• If there is insufficient knee space for complete sitting, consider using a sit-stand chair. These chairs allow you to rest your buttocks on the seat and do not require the leg space beneath the counter that a typical chair or stool requires.
• Consider standing on an anti­fatigue mat to reduce the forces on the lower extremities and feet.
• Use a footrest while standing to reduce back pressure. Place one foot on the footrest to relieve pressure on that foot and alternate feet regularly.

The Eastman Kodak Company. Kodak’s Ergonomic Design for People at Work, 2nd ed. Hoboken, NJ: Wiley; 2004.

Sandra M. Woolley, PhD, CPE, Ergonomist, Occupational Safety, Mayo Clinic Systems Quality Office, Rochester, Minn.

[hr]

Q. What is ideal or acceptable month-to-month variation or lot-to-lot variation in mean values in various clinical chemistry tests such as lipid profiles and enzymes?

A. This is an interesting and important question. Unfortunately, it does not have a simple answer.

To start, we should define the terms more specifically. I am going to infer that the question refers to the mean of patient values (as opposed to quality control material), a parameter too few laboratories monitor on a regular basis. Also, I would propose that the mean value is not as good a marker as the median value; mean values will be affected by extreme values much more than median values. Finally, it may be best to restrict the analysis to outpatient values; again, hospitalized patients are likely to have many extreme values.

For some tests, like enzymes, electrolytes, and calcium, there should be little variation in the median values from outpatients over time. If the medians change, in all likelihood, the assays are not working properly, the reference intervals are no longer appropriate, and too many healthy patients will appear to be “abnormal.” It’s difficult to provide a single percentage of acceptable variation for all tests. It might be better to look at the percentage of patients who, with a given percentage change, will be called abnormal. For example, a five percent increase in alanine transaminase is unlikely to be terribly significant, whereas a five percent change in calcium would be significant.

It is important to point out that preferred laboratory practice is to check values on patient samples with each lot number change.1
As an example, in our laboratory, when we receive (and before we implement) new lots of reagents or calibrators, we typically check five to 10 patient samples, looking for significant (as defined in the previous paragraph), systematic differences. In some cases, even if these comparisons look acceptable, we follow the patient medians after implementation to ensure ongoing consistency.

For “lipid panels” (as well as some other measurands like hemoglobin A1c, creatinine, bilirubin), in addition to ensuring that there is no significant shift over time, one should also ensure accuracy. Physicians interpret values on these tests using national or international guidelines, so getting an accurate answer is as important as getting consistent answers over time. An excellent way to do this is to participate in one of the CAP’s Accuracy-Based Surveys, where commutable material is used and your laboratory’s results are compared to the reference method results. Benchmarks for agreement for each of the components are defined by the guidelines and by the individual Surveys themselves. For example, for total cholesterol, the two components of total error, bias and imprecision, should each be three percent or less, meaning that the maximum deviation from the true value should be less than 10 percent.2

1. College of American Pathologists. COM.30450 New reagent lot confirmation of acceptability. In: All Common Checklist. April 21, 2014.
2. College of American Pathologists. Accuracy-Based Lipid Survey Participant Summary (ABL-B); 2012.

Gary L. Horowitz, MD, Medical Director, Clinical Chemistry, Beth Israel Deaconess, Medical Center, Boston, Associate Professor, Pathology, Harvard Medical School, Chair, CAP Chemistry Resource Committee
[hr]

Dr. Kiechle is medical director of clini­cal pathology, Memorial Healthcare, Hollywood, Fla. Use the reader service card to submit your inquiries, or address them to Sherrie Rice, CAP TODAY, 325 Wau­ke­gan Road, Northfield, IL 60093; srice@cap.org. Those questions that are of general interest will be answered.