Feature Story

Do’s and don’ts of managing specimen workups
Time to talk turkey in microbiology labs July 2004
William Check, PhD

Mary York, PhD, D(ABMM), president of MKY Consulting, Walnut Creek, Calif., presented her top-10 list of do’s and don’ts for managing microbiology specimen workups. She spoke at the second symposium on cost, quality, and clinical relevance.

1. Report what you know even if you are not 100 percent sure. Say "probable staphylococci," not "Gram-positive cocci in clusters"; say "probable streptococci/entero cocci," not "Gram-positive cocci in chains"; say "pseudomonad" if it is oxidase-positive and indole-negative.

2. Don’t perform irrelevant susceptibilities. You are telling the doctor you think he should take action. One instance: Don’t do susceptibilities on isolates from catheter tips. It is the blood culture that should govern therapy.

3. Do rapid identifications. She quotes NCCLS document M35 A: "If identification fits, there is no need to do more."

4. Do direct tests from blood culture bottles. The goal is to identify genus in four hours and species in 18 hours. Gram-positive cocci in clusters, for example, can be identified with two drops of culture fluid in rabbit plasma and cefoxitin and colistin disks.

5. Don’t work up coagulase-negative staph in one of two blood culture sets.

6. Look only for pathogens significant for that body site. For example, S. pneumoniae and MRSA are irrelevant in throats and reporting them will provoke unnecessary therapy.

7. Don’t work up yeasts in lower respiratory aspirate specimens (bronchoalveolar lavage, sputum, transtracheal aspirates). Yeasts are normal flora of the upper tract.

8. Don’t treat all fungal requests the same; consider the site. Pathogens in CSF are Cryptococcus and Coccidioides. Vaginal and urine specimens require yeast-only cultures.

9. Don’t work it up just because it is hemolytic. For example, don’t hunt for and report S. aureus in sputum; low numbers don’t correlate with pneumonia. Exceptions include beta-strep, Pseudomonas, or S. aureus in a wound; Listeria; Clostridium perfringens; and Vibrio.

10. Don’t work up mixed normal body site microbiota. Unless enteric Gram-negative rods are the single predominant organism, report "mixed enteric rods present." For example, for peritoneal fluid containing E. coli, enterococci, and Bacteroides, report "mixed aerobic and anaerobic GI microbiota."

William Check is a medical writer in Wilmette, Ill.