Tips on technology.

Ektachem linearity

Q Although the Ektachem instruments have a unique concept with dry-slide technology, isn't there a linearity problem?

A One reason the Ektachem's chemistry tests may appear to be nonlinear is that lyophilized specimens may not produce linear results. This has been shown on the CAP Linearity Proficiency Survey. For example, in the 1993 LN2B survey, nonlinear results were produced for the following analytes: total bilirubin, glucose, triglyceride, cholesterol, HDL cholesterol, albumin, amylase, and LDH. Sodium, potassium, chloride, creatinine, urea, uric acid, total protein, calcium, magnesium, phosphorus, AST, alkaline phosphatase, and hemoglobin were linear by this method.

The topic of this nonlinearity was the subject of a recent paper by Kroll and Chesler.[1] They showed that for HDL cholesterol, 54% of test methods run on various instruments in the CAP survey were nonlinear. The Ektachem was one of these instruments. Lyophilized specimens were used in the surveys. This problem is the result of matrix effect.

Another reason for believing that the Ektachem is nonlinear in its response is that reflectance spectrometry does not follow Beer's law. The response of the photometer is nonlinear, following the Kubelka-Munk formula, 2R[(1-R).sup.2]=ac/s, where R=reflectance, a=molar extinction coefficient, c=concentration, and s=light scattering coefficient. This relationship between concentration and reflectance is complex, requiring computer curve fitting. For this reason, the Ektachem and other reflectance photometers are calibrated using liquid serum samples that have been assayed at multiple levels by a reference method.

Using liquid human serum samples, the situation is quite different. The VA National Center for Laboratory Accuracy and Standardization (VANCLAS) periodically sends fresh liquid serum samples to all of the VA laboratories in North America for analysis.[2] Multiple samples covering a wide range of values are analyzed over several days for various analytes. Analysis of the participants' test results at the VANCLAS central laboratory provides each participant with information about within-day precision, total overall precision, average percent bias (accuracy of calibration), and linearity of its analyses. Based on this information, the Ektachem's analyses are, in fact, linear.

References

1. Kroll MH, Chesler R. Nonlinearity of high density lipoprotein cholesterol determinations is matrix dependent. Clin Chem. 1994; 40: 389-394.

2. Naito HK, Kwak Y-S, Hartfiel JL, et al. Matrix effects on proficiency testing materials: Impact on accuracy of cholesterol measurement in laboratories in the nation's largest hospital system. Arch Pathol Lab Med. 1993; 117: 345-351.

RF latex test

Q Can you provide references for performing rheumatoid factor (RF) latex tests in body fluids - synovial fluid in particular? We have been doing this test for a long time, and a rough estimate of positive results, in my experience, is about 95%. The reaction is from weak positive to 3+ agglutination (unfortunately, we are not doing titers automatically). The kit's package insert specifies only serum, and nothing is said about other specimens.

Recently we had a patient with a 2+ agglutination in synovial fluid but negative in serum. Since the manufacturer recommended only serum, is it valid to do RF latex testing on body fluids? In the case above, is the positive RF in synovial fluid a false positive? Or did the patient start to show RF in the fluid before it leaked out into the plasma? Do you think RF latex testing procedures should be modified for body fluid specimens? Our procedure requires 1:20 dilution with glycine buffer before mixing a drop of specimen and a drop of latex for 2 minutes. Quality control is performed accordingly.

A There are a number of old references regarding latex tests for rheumatoid factor in synovial fluids. A good one that discusses the technical problems and clinical usefulness is by Sibley et al.[1] In the Sibley procedure, the synovial fluid (SF) for rheumatoid factor was centrifuged at 3,000 rpm for 10 minutes and stored at -70 [degrees] C until the test was performed. Serum was drawn within 7 days of joint aspiration. RF in both serum and SF was determined by the latex test method of Singer and Plotz[2] after heat inactivation of 56 [degrees] C for 30 minutes. (This was done to inactivate the Clq component of complement, which can cause false positive agglutination.) Visible agglutination at a dilution of 1:80 or greater was considered a positive result by the Sibley procedure.

The latex test for RF was developed by the manufacturers for assay of RF in serum specimens. I am not aware of any latex test kit for RF that has been approved by the Food and Drug Administration for assay of RF in synovial fluids.

The RF latex test is affected by the protein concentration and pH of buffer. Usually a glycine buffer at pH 8.2 is used.[3] A buffer with a pH of almost 7.4 would be closer to the isoelectric point of gamma globulin, which is pH 6.6. Latex particles coated with gamma globulin may spontaneously aggregate in dilute solutions without proteins.

Sibley et al[1] studied a series of 140 patients and did not find it necessary to treat the synovial fluid with hyaluronidase. In some studies, however, the more viscous fluid from SF was treated with hyaluronidase to make the fluid clearer and the agglutination reaction easier to read.[3-5]

RF is present in serum in high titers in many rheumatoid arthritis (RA) patients. Seropositivity at the onset of rheumatoid arthritis is helpful in the clinical evaluation of the patient.[6]

In some instances, patients with a negative serum RF were observed to have RF in their synovial fluid, suggesting that RF is produced locally in the synovial fluid.[6] Thus, determination of synovial fluid RF may be of some value in clinically diagnosed cases of seronegative rheumatoid arthritis. In most seropositive patients, however, RF is also present in the synovial fluid, usually in titers similar to those found in serum.[4,5,7-10]

There have been reported seronegative RA patients who had high titer synovial fluid RF.[4,8-10] Sibley et al,[1] however, found only one RA patient who was seronegative but synovial fluid RF positive. MacSween et al[5] reported that determination of synovial fluid RF is of little value.

Christian[6] stated the synovial fluids from most seropositive subjects contain RF as determined by conventional serologic tests. In Christian's opinion, the practical importance of measurement of RF in synovial fluid is nil. Additionally, some patients with RF in their serum didn't demonstrate RF in their synovial fluid.[8,9]

An interesting study was done by Panush et al.[11] They measured synovial fluid RF of IgG, IgA, and IgM classes in RA and osteoarthritis (OA) patients. They found the IgG and IgM RF levels were increased in synovial fluids of patients with RA compared with those of patients with OA.

References

1. Sibley JT, Harth M, Burns D. The mucin clot test and the synovial fluid rheumatoid factor as diagnostic criteria in rheumatoid arthritis. J Rheumatol. 1983; 10: 889-893.

2. Singer JM, Plotz CM. The latex fixation test. Am J Med. 1956; 21: 888-892.

3. Krehl WA, Boisvert PL, de Forest GK, et al. The rheumatoid factor in serum and synovial fluid. Yale J Biol Med. 1957; 30: 30-37.

4. Huskisson EC, Hart FD, Lacey BW. Synovial fluid, Waaler-Rose and latex tests. Ann Rheum Dis. 1971; 30: 67-71.

5. MacSween RNM, Dalakes TK, Jasani MK, et al. Rheumatoid factor in serum and synovial fluid. Scand J Rheumatol. 1972; 1: 177-180.

6. Christian CL. Immunologic aspects of rheumatoid arthritis. In: Cohen AS, ed. Laboratory Diagnostic Procedures in the Rheumatic Diseases. Boston, Mass: Little Brown and Co; 1975: chap 4.

7. Schur PH, Sanderson J. Immunologic studies of the proteins of human synovial fluid. Arthritis Rheum. 1963; 6:115-129.

8. Rodnan GP, Eisenbeis CH, Crighton AS. The occurrence of rheumatoid factor in synovial fluid. Am J Med. 1963; 35: 182-188.

9. Bland JH, Clark L. Rheumatoid factor in serum and joint fluid. Ann Intern Med. 1963; 58: 829-836.

10. Cracchiolo A, Barnett EV. The role of immunological tests in routine synovial fluid analysis. J Bone Joint Surg. 1972; 54A: 828-840.

11. Panush RS, Bianco NE, Schur PH. Serum and synovial fluid IgG, IgA, and IgM anti-gamma globulins in rheumatoid arthritis. Arthritis Rheum. 1971; 14: 737.

Infant occult blood testing

Q Are the cards for occult blood testing (Hemoccult, Coloscreen, Seracul) adequate for infant testing? I understand the cards were meant for mass screening for colorectal cancer. I have little knowledge of the indications for occult blood testing in infants, so I am interested in knowing if the cards are sensitive enough for this purpose. Or should the old guaiac method be done?

A Test cards for stool occult blood are applicable for infant and adult testing, but they should not be used for testing gastric contents since gastric acid reduces the test's sensitivity. Clinical situations where stool occult blood testing might be performed in an infant include gastric ulcer and GI tract malformation such as Meckel's Diverticulum where bleeding might occur. The most likely use of the test, however, is for stool that appears to contain blood to see if the red color is due to blood or another material.

Urine cultures

Q We are looking at the appropriateness of urine culture orders, and we have just completed a 3-month QA study relating routine urinalysis findings to urine culture findings. Results of the study show there were no clinically significant data generated from the urine culture for patients with concurrent routine urinalyses that lacked evidence of pyuria. We believe that if the patient is asymptomatic and has a routine urinalysis with normal numbers of WBCs, it is unnecessary to perform a urine culture. Is this recommendation prudent? Are there any documented practice guidelines published to substantiate this practice?

A Your QA study relating routine urinalysis findings to urine culture results is quite consistent and supports the findings of others who have performed similar studies.[1] As a result, some form of rapid screening, usually for pyuria, is increasingly practiced by more and more laboratories to determine the suitability of the specimen for culture.

I agree that if the patient is asymptomatic and has a normal routine urinalysis, then it is unnecessary to perform a urine culture. My question, however, is this: How often does the clinical laboratorian know whether or not the patient is symptomatic? If this is a critical part of the pathway, then it may, in some situations, be more costly to the laboratory to determine the symptoms of the patient than to set up the culture.

Your approach to this problem is commendable, and such an approach is appropriate for many clinical settings. To enhance the likelihood of success in your endeavors, I recommend the following: 1) Your clinical colleagues should be actively involved in the review of your QA data as well as in the development of the procedures and guidelines that flow from this. 2) Physicians should be given the right to override the routine procedures or guidelines for any patient or specimen, using an easy, straightforward mechanism.

Reference

1. Stamm WE. Measurement of pyuria and its relation to bacteriuria. Am J Med, 1983; 75 (July 28, 1983 suppl): 53-58.

Order of blood drawing

Q What is the proper venipuncture order of draw? Should I be including additional tubes, such as black? If so, what are some tests for which these tubes might be used? Is a royal blue top tube used for trace metals and drug testing? What is the best color tube to use for blood alcohol levels? When performing venipuncture on a pediatric patient, do some physicians want the EDTA tube drawn before any other color due to the amount of blood per the amount of anticoagulant (in cases of short draws)?

A The order in which vacuum tubes are filled during venipuncture is important so additives in one tube will not contaminate the specimen in a subsequent tube. NCCLS Standard H3-A3[1] recommends this order of draw: 1. blood culture tube; 2. nonadditive tube (red stopper); 3. coagulation tube (blue stopper); 4. additive tube (e.g., green, lavender, gray stopper, or other additive tube).

When several different additive tubes are filled, their order is also important:[2,3]

1. The blue stopper coagulation tube with sodium citrate is filled first. Black stopper buffered sodium citrate tubes for sedimentation rate tests can also be filled at this time.

2. Next the heparin-containing (green stopper) tube should be filled.

3. The EDTA-[K.sub.3] - containing (lavender stopper) tube follows.

4. The clot activator (gray and yellow stopper) tube (CAT) containing thrombin is next. This tube is for rapidly obtaining serum for Stat chemistry tests. The tube induces blood to clot in 3 to 5 minutes. Sherwood has a gel separator tube containing thrombin, the Stat-vac tube.

5. Gel separator (gold stopper) tubes (SST) follow. The Becton Dickinson SST, Sherwood Corvac, and Terumo AutoSep gel separator tubes have a clot activator - silica or ground glass particles - that promotes clotting.

6. The oxalate-fluoride - containing (gray stopper) tube is last.

Blood for alcohol tests can be drawn into any of the usual blood tubes including those with anticoagulants (heparin, EDTA, fluoride). Each lab has its own preferences.

If no plain tube is needed for the collection and a blue stopper coagulation tube is needed, a small red top tube should be used first and discarded so tissue thromboplastin does not contaminate the specimen during venipuncture.

The royal blue stopper tube, which is made from special lead-free glass, is for trace metal tests. It can be filled with red stopper non-additive tubes.

When filling tubes from a syringe, additive tubes should be filled before nonadditive tubes.[4]

References

1. National Committee for Clinical Laboratory Standards. Procedures for the Collection of Diagnostic Blood Specimens by Venipuncture. Approved standard H3-A3, Villanova, Pa: NCCLS; 1991.

2. Calam RR, Cooper MH. Recommended "order of draw" for collecting blood specimens into additive-containing tubes. Clin Chem. 1982; 28: 1399.

3. Becton Dickinson Vacutainer SST package insert. August 1993.

4. Becton Dickinson Lab Notes Newsletter. September 1990.

Edited by Bernard E. Statland, M.D., Ph.D., executive vice president, National Health Laboratories, and president and CEO, National Reference Laboratory, Nashville, Tenn.

Panelists (name following each answer indicates respondent panelist):

Daniel M. Baer, M.D., Veterans Affairs Medical Center, and professor of pathology, Oregon Health Sciences University, Portland, Ore.

Peter C. Fuchs, M.D., Ph.D., director of microbiology, St. Vincent Hospital and Medical Center, Portland, Ore. John A. Koepke, M.D., professor emeritus of pathology, Duke University Medical Center, Durham, N.C. Byron A. Myhre, M.D., Ph.D., professor of pathology and chief of clinical pathology, UCLA School of Medicine, Harbor General Hospital Campus, Torrance, Calif. Robert M. Nakamura, M.D., senior consultant and chairman emeritus, department of pathology, Scripps Clinic and Research Foundation, La Jolla, Calif.