HeMYTHology: 10 myths of hematology

Nov. 1, 2003

Are you a believer?

Here are 10 generally accepted concepts from the hematology lab, and the reality checks for each. What is your own myth factor?

MYTH #1: Band counts are important.

REALITY: It is hard to make a case for the importance of a parameter that has no clear-cut definition. Various attempts at defining bands exist in literature and textbooks but, in actual practice, there is a very wide range of opinions. Like the definition of art, it is one of those I-know-it-when-I-see-it type of things. So problematic are bands that the College of American Pathologists (CAP) could not reach a consensus with its participants using Kodachrome examples. Thats why your master list says grans/bands. And some institutions have done the same thing, combining neutrophils and bands into one category. Do not be surprised if your emergency room doctors and pediatricians are reluctant to abandon bands.

MYTH #2: Moving averages are averages. 

REALITY: This is a partly true statement, which in classic True/False rules makes it false. Moving averages, Xb or Bulls algorithm is actually a complex equation that shaves off outliers, uses information from the last batch to figure the current one and cannot be done without a calculator. Since it uses data from the previous batch to figure the current batch, it may take a while for the graph to rebound back to normal if it has been out. 

MYTH #3: Linearity verification must be performed periodically on hematology analyzers.

REALITY: The current CAP checklist does not address the issue of linearity at all, just calibration so, as far as CAP is concerned, the linearity performed when the instrument is installed should suffice. The Clinical Laboratory Improvement Act (CLIA) refers to reportable range, but does not specify frequency. Other than state and federal regs, should you be concerned? If an instrument is kept in good repair and calibration according to manufacturer specs, virtually nothing will significantly affect the linearity. Linearity is the personality, while calibration is the attitude, which we know can change. 

MYTH #3: On a stained blood smear, one platelet equals 15,000×109/L. 

REALITY: This is another partly true statement. On a well-made wedge smear from a normal specimen where the red blood cells (RBCs) are just touching, the rule holds true: In a 100x field, one platelet equals about 15,000×109/L in the specimen. But on a smear from a very anemic blood, going far back into the smear where the RBCs touch, will give an inflated estimate of platelets. In other words, your view must reflect the RBC count to get a valid 15,000 rule. The opposite holds true for a polycythemic blood: The observed field will have to appear more crowded to get an accurate estimate. 

MYTH #5: Hematology analyzers must be recalibrated at regular intervals.

REALITY: This is really a trick statement, and the key word is recalibrated. Analyzers do not have to be recalibrated per se, but the calibration does have to be verified at regular intervals. In other words, if your calibration procedure shows your Acme analyzer within tolerances when you run the calibrators, you do not have to hit the recal button. Most of us hit the button anyway, even if the adjustment is insignificant. It makes us feel as if we are getting our moneys worth. 

MYTH #6: If your instrument flags the diff, you must check the slide. 

REALITY: This is probably the most widely practiced misuse of lab resources, and my personal least-favorite myth. Many man-hours are wasted in labs by checking slides that do not necessarily have to be checked. The solution? Do your own study of your instrument, and see if the flags work for your needs and your own definition of abnormal. Quick example: If your lowest level diff flag (left shift +, Imm grans/bands +, etc.) almost always flags normal patients, and does not catch any abnormals by itself, it is a useless flag for your lab and your definition of abnormal. There are too many laboratories that are not using the sophisticated technology of modern analyzers intelligently or efficiently. My experience is that technologists and analyzers have an error rate (both may miss a small percentage of abnormal cells) and they are about even, if you fine-tune things. 

MYTH #7: Hematology commercial controls must be parallel tested before being put in use.

REALITY: I cannot think of a real-life scenario where parallel testing of assayed controls would be of much value. The main reason? Heme controls come with insert values, a list of expected ranges for your instrument, which make parallel testing moot. Let us say you receive a new lot of controls, and parallel test them that is, test them on the analyzer after running the current lot, so you know the instrument is performing properly, and thus know the new lot is also all right. Then you put the new lot away for two weeks. When you later put the controls into use, you find that the RBC values are all out of the 2sd range on the low side, as the insert sheet tells you. What do you do now? Whether you have parallel tested or not, you still have to decide whether you have a control material problem or an instrument problem. Parallel testing makes sense and is necessary when you receive unassayed control, as in coagulation controls. Many labs still do some fine-tuning on their commercial controls anyway, such as readjusting the mean and SD, but for the most part, it is just that: fine-tuning. 

MYTH #8: If you never have outliers, the system is working. 

REALITY: Actually, the opposite is true. If you never have outliers for a particular system, chances are one of two things is happening. First, your quality control (QC) limits are set so broadly that they would virtually never be out of range. Secondly, the QC just is not measuring what you thought it was or measuring in the way you think it was. Either one means the QC is not detecting errors or problems. If a QC tool is never out, it is not doing you any good. 

MYTH #9: Very low white blood cells (WBCs) must have manual

REALITY: When the WBCs are in the ultra-low range below say, 0.6×109/L this is the time when automated instruments really can be useful. The number of WBCs on a wedge smear with an 0.6 WBC is statistically too small, and it takes too long to count; the results usually are not worth the effort. Since a good automated instrument (with good low-end sensitivity) will still be counting hundreds of cells, even on a low count, the automated diff will give you better information than a manual diff. Of course, you can make buffy coat smears, but again, the effort involved vs. useful info must be considered. Many labs make use of TFTC too few to count and reserve buffy coats or manual diffs for the insistent physician. 

MYTH #10: Microhematocrits must be QCd. 

REALITY: We are now firmly in the area of controversy. My position on this issue is this: The microhematocrit is a primary method that is, a method against which other methods, controls, reagents, instruments and calibrators are standardized and, as such, cannot be QCd in the traditional sense. In the case of microhematocrit, if the RPM is checked and the packing time is standardized, that is it. If you are comparing instrument HCTs to the microhematocrit every day, then you are not QCing the manual crits; you are doing the opposite. 

Depending on your own myth factor, you may decide to re-examine your procedures and policies. You might even discover some myths of your own.

Roy Midyett, MT(ASCP) is hematology supervisor at Presbyterian Hospital, Whittier, CA.

Additional reading:
1. CAP Hematology Glossary. 2003.
2. Brown B. Hematology: Principles and Procedures. 1993.
3. Coulter STKS Reference Manual. 1991.
4. CAP TODAY. Summer 1994.
5. CAP Hematology and Coagulation Checklist. 2001.
6. Code of Federal Regulations [CLIA 88] 67.

November 2003: Vol. 35, No. 11

© 2003 Nelson Publishing, Inc. All rights reserved.