Answering your questions on masks for AFB, blood culture contamination rate, preventing cross--contamination, CBC dilution, and pregnancy tests.

Daniel M. Baer

Masks for AFB 

Q I work in a small (125-bed) hospital laboratory that offers routine bacteriology and fungal cultures. AFB cultures are sent to a reference laboratory. Bacterial contamination of the fungal cultures is assessed microscopically. All fungi are referred to another laboratory for identification. My concern is that all specimen processing occurs on an open bench. What is the risk for a laboratory-acquired infection? Should we wear a TB mask during specimen processing?

A The magnitude of risk for a laboratory-acquired infection varies with the prevalence of disease in the patient population and the effectiveness of the infection control or safety measures employed by the laboratory. Exposure of laboratory employees to infectious agents generally occurs one of three ways: 1) from activities that generate aerosols (inhalation of agents); 2) from percutaneous (sharps) and mucous membrane (splashes) exposures; or 3) from ingestion. Most routine bacteriology work involves agents of moderate risk and can, therefore, be performed on the open bench. The primary risk is associated with a percutaneous or mucous membrane exposure. However, work with infectious agents that are transmitted by aerosols (e.g., Mycobacterium tuberculosis, systemic mycoses, Brucella species, Francisella species) requires the use of a biological safety cabinet (BSC) and other biosafety level 3 practices. [1] The risk of a laboratory-acquired infection from these agents is reduced in your laboratory because a ll specimens for AFB culture and identification of filamentous fungi are referred to another laboratory. However, the initial processing of specimens that may contain these agents for bacteriology or fungal cultures does pose some risk and should be performed in a BSC. [1]

The use of a TB mask (N95 particulate respirator) by the individual processing specimens is not recommended because potential aerosols are not contained and, therefore, pose a risk to other laboratory employees. To minimize the potential risk of infection to employees, the laboratory should define and implement the engineering and work practice controls appropriate to the biosafety level of the laboratory. [2] These safety precautions are discussed in the CDC-NIH guideline. [1]

David L. Sewell, PhD, ABMM 

Director of Microbiology

Veterans Affairs Medical Center

Portland, OR

References

(1.) CDC-NIH. Biosafety in microbiological and biomedical laboratories. 4th ed. US Government Printing Office, 1999.

(2.) College of American Pathologists. Standards for Laboratory Accreditation. 1999.

Blood culture contamination rate

Q I read with great interest the article in the May 2000 issue about blood culture contamination rates. We have been struggling with a contamination rate [greater than]3.0%.

My facility is a 42-bed hospital with a 120-bed attached nursing home. We collect from about 60 to 190 blood cultures each month, and our contamination rate varies from 3.2% to 5.1%.

The average age of the patients with contaminated blood cultures is about 75 years. I have found no trends associated with who collected the culture (e.g., nursing vs. lab staff), or with the department in which it was collected (e.g., ER, outpatient, inpatient).

Our medical director feels that the small total number of blood cultures collected here and the elderly population that we serve is skewing our results. While I agree, administration is always looking for performance improvement measures and continues to insist that we improve our contamination rate.

Are there any guidelines for small hospitals serving primarily elderly patients? Are there published thresholds for facilities such as mine? Do you have any suggestions regarding this issue?

A The article in the May 2000 issue of MLO is an up-to-date review of the blood culture contamination problem. [1] In addition, CAP Today recently published a long article about preventing blood culture contamination. [2]

Your administrator is correct to be concerned about the blood culture contamination rate. Several studies have shown that a false-positive blood culture increases hospitalization by an average of five days, and increases pharmacy and laboratory costs by more than $1,000. Together, this can cost a hospital $4,000 to $5,000 in additional, unnecessary costs.

The CAP Q-Tracks program recently published data from 155 institutions showing an average overall blood contamination rate of 2.94%. Analysis of these data shows that the use of tincture of iodine as a disinfectant and collection of higher blood culture volumes correlated with a lower contamination rate.

Skin flora cause contaminated cultures. Tincture of iodine is a more effective skin disinfectant then povidone-iodine (Betadyne). Disinfection technique is important prior to venipuncture. First, the skin should be cleaned for one minute using 70% isopropyl alcohol. Next, a 2% tincture of iodine should be applied and allowed to dry for two minutes before phlebotomy. Complete drying of the disinfected area is important.

Contamination is reduced if blood is collected directly into the culture bottle rather than using an intermediary syringe and needle. Filling of the bottle to its recommended volume is also important.

The training and experience of individuals who collect blood culture samples has a direct effect on the contamination rate. Those institutions with the best contamination rates have dedicated blood culture collection teams. The important factor is that the team members are well-trained and experienced.

Tracking the contamination rate and feedback to phlebotomists also reduces contamination rates. Some hospitals maintain data on contamination rates for each person who collects blood culture specimens. This information is reported back to the individual. In one hospital, such feedback correlated with a drop in contamination from 2.65% to 1.4%.

We are not aware of data correlating hospital size or patient age with contamination rates. A CAP Probe exercise involving 640 hospitals came to the same conclusions as the Q-Tracks study, [3] but did not study the variables of age or hospital size. One might think that large academic medical centers would have higher rates because of lack of control and training of the many caregivers who collect blood cultures.

Daniel M. Baer, MD

Professor Emeritus of Laboratory Medicine

Oregon Health Sciences University

Portland, OR

References

(1.) Ernst D. Controlling blood culture contamination rates. MLO. May 2000;32:36-47.

(2.) Paxton A. Nipping contamination in the blood. CAP Today. May 2000;14:1, 56-70.

(3.) Schifman RB, Strand CL, Meier FA, Howanitz PJ. Blood culture contamination: A college of American Pathologists Q-Probe study involving 640 institutions and 497, 134 specimens from adult patients. Arch Pathol Lab Med. 1998;122:216-221.

Preventing cross-contamination

Q CLIA '88 requires that "effective measures be taken to prevent cross-contamination between gynecologic and nongynecologic cytology specimens during the staining process" and that "nongynecologic specimens that have a high potential for cross-contamination be stained separately from other nongynecologic specimens." What technique can be used to determine if a fluid sample for cytologic evaluation has a "high potential" for cross-contamination?

A Toluidine Blue is a nuclear stain commonly used in cytopathology to determine the cellularity of body cavity fluids and to search for abnormal cells. It can be used as a rapid method to separate highly cellular (or potentially malignant) specimens from fluids of low cellularity, thereby preventing cellular "floaters" in the staining dishes. Prior to the staining process, a wet prep consisting of one drop of sediment admixed with one drop of 0.5% Toluidine Blue stain can be quickly scanned by the technologist to determine if it is safe to stain the specimen with other cases.

The 0.5% Toluidine Blue stain is prepared by combining 0.5 gm Toluidine Blue, 20 mL 95% ethanol, and 80 mL distilled water. The stain should be filtered, stored in a dark glass container, and kept in the refrigerator to prevent fungal contamination.

The Toluidine Blue staining technique is relatively simple. After centrifuging the specimen, pour off the supernatant and place one drop of the sediment on a glass slide. Add one drop of 0.5% Toluidine Blue stain and mix the two drops with a pipet tip, applicator stick, or the edge of a coverslip. Cover with a small coverslip so that the sample remains in the central portion of the slide. Let the prep sit for a few seconds before examining with a microscope.

The Toluidine Blue wet prep should be read within 1/2 hour unless the coverslip is sealed with clear fingernail polish or petroleum jelly (or kept in a covered Petri dish). In addition, as a wet prep (and not an alcohol fixed/stained smear), the microscope slide should be handled using universal precautions.

Jacalyn L. Papillo CT(ASCP)

Chief Technologist

Cytolopathology

Fletcher Allen Health Care 

Burlington, Vermont

Suggested reading

(1.) Clinical Laboratory Improvement Act of 1988, Subpart K, Sec. 493.1257 (a)(2). Federal Register. 1993; 58:(141)

(2.) Keebler CM, Somrak TM. A Manual of Cytotechnology. 7th ed. Chicago, IL: American Society of Clinical Pathologists Press; p.426, 433, 438. 1993

(3.) Bibbo M. Comprehensive Cytopathology, 2nd edition. Philadelphia, PA: WB Saunders; p. 908. 1996

CBC dilutions

Q When making dilutions to verify CBC values which fall outside of the reportable range for our analyzer, how close should the original value be to the diluted in order to report the original number?

A In February 2000's CAP Today, Dr. Albert Rabinovitch succinctly answered a question on linearity requirements for hematology analyzers. He stated, "Hematology analyzers function primarily as particle counters. The concept of linearity applies only to chemistry in relation to spectrophotometric absorbance/transmission and analyte concentrations. Technically, only the hemoglobin measurement on a CBC instrument is such a test." [1] Nevertheless, most of us faithfully perform linearity verification on our hematology analyzers, using a commercial product or patient samples, and incorporate the findings into our operating procedures.

Instrument manufacturers state an operating range and a reportable range. The commercial linearity material can be used to verify this range, if it adequately covers the stated linear interval. We use a commercial product but have also validated the manufacturer's range by diluting patients with WBCs greater than the reportable range but within the operating range. Unless one has established and validated a reportable range, beyond what the manufacturer states, it's risky to expect that the original and diluted value will consistently match.

One could approach this from a more analytical method, looking at tolerance limits at specific points outside the reportable range, establishing confidence intervals and creating a protocol where if the dilution matches the original value within two SDs, the original value is reported. [2,3] However, this will require more rigorous analysis and review.

A variety of interferences can impact a result outside the reportable range. One must understand the counting methodology and limitations of the analyzer. After the sample is diluted, the interference may be significant, resulting in a final value that cannot match the original value. A few things to consider: RBC results and MCV may be falsely elevated due to the WBCs being counted simultaneously with RBCs and platelets; a spun hematocrit is approximately 2% to 5% higher than a calculated hematocrit; Hemoglobin may be falsely elevated due to turbidity; MCH and MCHC values may be inaccurate due to the affected parameters from which they're calculated; falsely elevated WBC values may be seen in samples with cryofibrinogen, cryoglobulins, nucleated red blood cells, micromegakaryocytes, or any particle sized in the range that is counting WBCs. [4]

My recommendation is to report a diluted WBC, diluted platelet, a spun hematocrit, and MPV (depending on which parameter is outside reportable range). RBCs can be reported if they are corrected for out of range WBCs. A manual hemoglobin can be performed. The reportable range identifies the values where the instrument is accurate. Diluting a sample to achieve a value in the reportable range is the best route to turning out a quality, clinically significant, analytically accurate result, a result that has instrument and methodology limitations already factored into the equation.

--Juanita Petersen, MT(ASCP), MBA

Manager, Core Laboratory

Oregon Health Sciences University Hosptial

Portland, OR

References

(1.) CAP Today. Queries and comments. February 2000; 14(2):71-72.

(2.) NCCLS. Performance Goals for the Internal Quality Control of Multichannel Hematology Analyzers. Approved Guideline H26-A, Wayne PA, NCCLS 1996

(3.) NCCLS. Evaluation of the Linearity of Quantitative Analytical Methods. Proposed Guideline EPG-P. Wayne PA, NCCLS 1986

(4.) Coulter Hematology Analyzers, Hematology Procedures for Abnormal Bloods. Coulter International Corporation. April 1999.

Pregnancy test

Q The doctors in our emergency department (ED) want to perform the pregnancy tests for their department. How are these tests used to diagnose ectopic pregnancy? Also, the doctors claim that the pregnancy test is exempt from CLIA '88. Is this true?

A Ectopic pregnancy is a critical diagnosis in the ED. It is the leading cause of fetal death in the first trimester and accounts for 15% of maternal deaths, usually because of massive hemorrhage. [1] It must be considered for any woman of childbearing age who presents with abdominal pain. Early diagnosis reduces the chance of serious complications and subsequent infertility. One skipped menstrual period is not a reliable means of diagnosing pregnancy, particularly in an ED situation. Tests for beta human chorionic gonadotropin (HCG) and ultrasound are the preferred methods for diagnosing pregnancy, normal or ectopic. The combination of a sensitive beta HCG test and ultrasound has a positive predictive value of 88% to 99% for ectopic pregnancy. [2]

Almost all ectopic pregnancies produce HCG levels [greater than] 40 mIU/mL. Therefore, use of a test that detects 25 mIU/mL is necessary to rule out pregnancy. When ectopic pregnancy is suspected but HCG is negative, a second follow-up test is recommended in 48 hours. The HCG test cannot differentiate between an ectopic and normal, viable pregnancy; however, serial quantitative testing may do so if the level of HCG climbs rapidly to very high concentrations. Concentrations of HCG usually double every 48 hours in early normal pregnancies, but HCG levels rise at similar rates with some ectopic pregnancies, too.

Most of the single-use immunologic pregnancy tests are categorized as waived tests under CLIA. You can check the CLIA complexity category of any test kit by going to www.cdc.gov/phppo/dls/testcat.

--Daniel M. Baer, MD

Professor Emeritus of Laboratory Medicine

Oregon Health Sciences University

Portland, OR

References

(1.) Atrash HK, Friede A, Houge CJR. Ectopic pregnancy mortality in the United States, 1970-1983. Obstet Gynecol. 1987;70:817-822.

(2.) Olshaker JS. Emergency department pregnancy testing. J Emerg Med. 1996;14:59-65.

Editor's note: Material for this column appeared in previous editions of MLO.

Daniel M. Baer is professor emeritus of laboratory medicine at Oregon Health Sciences University in Portland, OR, and a member of MLO's editorial advisory board.