College of American Pathologists
Printable Version

  Feature Story


cap today

Sketching out changes in the microbiology checklist

November 2002
Vida Foubister

Many CAP checklist questions for lab inspectors can be traced to requirements set by the Clinical Laboratory Improvement Amendments of 1988.

But when Thomas A. Merrick, MD, education commissioner for the CAP Commission on Laboratory Accreditation, instructed laboratorians earlier this year on how to inspect the microbiology laboratory, he chose to focus on questions that might not be covered by the CLIA ’88 regulations.

"The emphasis of late has been on clinically relevant questions that address the very important interface between the microbiology lab and the medical staff, pharmacy, and infection control committee," says Dr. Merrick, director of the microbiology department at Presbyterian/St. Luke’s Medical Center, Denver.

Working with CAP staff, Dr. Merrick crafted 13 vignettes that highlight recent changes to the microbiology checklist. This includes questions that he and the CAP’s technical specialists are often asked to clarify during inspections and others that have been raised in discussion groups on the American Society for Microbiology’s Web site.

Now for the first vignette:

1.  Review of the laboratory’s quality control records show that the gram-stain procedure is checked and recorded for each new batch of stains and at least weekly against known gram-positive and gram-negative control organisms. The inspector notes that the laboratory is also performing other bacteriology stains and is checking these stains weekly. Is this acceptable?

The laboratory inspector can choose to:

  • Not cite a deficiency because the laboratory is in compliance.
  • Not cite a deficiency and informally make an oral recommendation.
  • Not cite a deficiency but make a written recommendation.
  • Cite a deficiency.
Two microbiology checklist questions apply. MIC.21540, a Phase I question, asks, "Is the gram-staining procedure checked and recorded for each new batch of stains and at least weekly against known gram-positive and gram-negative control organisms?" The lab here is checking the gram-stain procedure at least weekly, so the answer to this checklist question is yes.

The other, MIC.21560, is a Phase II question. It asks, "Are all bacteriology stains checked with a positive control and negative control (when appropriate) for intended reactivity each day of use or weekly, whichever is less frequent?"

Centers for Medicare and Medicaid Services staff reviewed the checklist and found this question to be "less stringent than the CLIA requirements," Dr. Merrick says. Future versions of the checklist will specify that gram stains can be checked weekly but that all other bacteriological stains, such as for Legionella, Nocardia, and Bordetella, must be checked daily or on each day of use.

Of the lab in the vignette, Dr. Merrick says he would not cite it as deficient because it is compliant with the question as written. "But obviously I would make a strong recommendation to them that the next version of the checklist will require them to show that they check these daily," he says.

This change notwithstanding, Dr. Merrick expects there will be relief from excessive QC testing in other areas within the next year.

An American Society for Microbiology membership survey found the rates of QC failures for common laboratory reagents, such as catalase, oxidase, indole, and coagulase, were less than one percent, and it was presented with a subsequent survey to the Clinical Laboratory Improvement Advisory Committee and the Centers for Disease Control and Prevention. CLIAC and CDC agreed with ASM’s recommendation that new lots of commercial reagents with a 99 percent or greater success rate need to be tested only once. "However, we must wait until the regulations are published in the Federal Register before we can make a change in the checklist," Dr. Merrick adds. Surveys are being done now on QC failure rates for miniaturized or automated bacterial identification systems and on laboratory protocols for media QC that might result in less frequent quality testing of these products within the next year.

Two changes to the laboratory-general checklist about the review of QC data have already taken effect. Instead of weekly review of QC data by a supervisor, the lab must prove oversight review at least monthly. In addition, there’s no longer a requirement to label reagents with date received or date opened-only the date prepared or reconstituted, expiration date, and storage requirements are needed. "Those are changes that will help us cut back on some of our unnecessary paperwork," Dr. Merrick says.

2.  While reviewing the laboratory’s performance on a recent CAP Survey, the inspector notices that the laboratory submitted a result of "gram-positive cocci present" from a sterile body site and the result was graded as acceptable. As the inspector questions the supervisor and reviews the procedure manual, she discovers that the gram-positive cocci isolated from a body site are routinely identified to the genus and species level. Should the laboratory be confronted with this discrepancy?

"If you report isolates to the species level, you must do so for the Survey specimens as well," says Dr. Merrick. "But if the organism is one that you normally identify to genus or presumptive genus level on a patient sample—for example, grampositive bacilli resembling Corynebacteria—with referral to a reference lab for further identification if clinically indicated, it is acceptable to report likewise for a Survey specimen."

The relevant microbiology checklist question is MIC.00350, a Phase II question that asks, "Are organisms in proficiency testing specimens identified to the same level as those from patient samples?"

For the lab in vignette No. 2, Dr. Merrick would look closely at the Survey results and cite a deficiency only if this level of reporting was found to be a trend. "If this is a one-time event and the supervisor has a reasonable explanation for it, then I would make a recommendation only. But it’s certainly something to be taken very seriously."

A related microbiology checklist question, MIC.01100, asks, "For areas where graded proficiency testing is not available, are other procedures used to validate performance at least semiannually?" Examples include fastidious, difficult-to-grow microorganisms, such as Helicobacter pylori, or hazardous organisms, such as Coccidioides sp.

In those situations a laboratory must perform an alternative form of proficiency testing, Dr. Merrick says. That can include ungraded PT programs, split sample analysis with other laboratories, split sample with established in-house materials, or clinical validation by chart review. For more ideas on how to meet this requirement, he recommends a new NCCLS document, GP29-A, titled "Validation of laboratory tests when proficiency testing is not available."

3.  The laboratory has purchased a new bacterial antigen kit that contains positive and negative onboard controls. Staff members are uncertain about the level of complexity of the test system, and this information is not in the package insert. They are running positive and negative liquid QC with each new lot of kits.

There are two updated questions in the microbiology checklist about rapid bacterial antigen tests. One is MIC.21760, a Phase II question, which asks, "For direct bacterial antigen tests on patient specimens that do include internal positive and negative controls, and are not classified as high complexity under CLIA ’88, is a positive and negative external control tested with each new kit lot number or separate shipments of a given lot number?" High-complexity tests that include an extraction phase must be checked each day of use with a positive organism, Dr. Merrick says.

The second is MIC.21770, also a Phase II question, which asks, "For direct bacterial antigen tests on patient specimens that do not include internal positive and negative controls, is a positive and negative control tested with each run of patient specimen?"

The laboratory must first determine the test’s complexity level. If the manufacturer hasn’t provided this information, it can be found on the Food and Drug Administration’s Web site. "In this case I would cite a deficiency until they have done that," Dr. Merrick says. If the test complexity is not defined, it becomes high complexity by default.

Another question that’s fairly new to the microbiology checklist and poses difficulties for labs, MIC.21620, a Phase II question, asks, "Are new reagent lots checked against old reagent lots or with suitable reference material before or concurrent with being placed in service?" Suitable reference materials include previously tested patient samples with known readings or control materials that have been tested on the old reagent lot. "The purpose of this parallel testing is to determine if a new lot has the same specificity and sensitivity as the last lot," explains Dr. Merrick. "This way if there’s a change in patient results over time, you know it is due to a change in the patient and not the sensitivity level of the new lot."

4.  The inspector reviews quality control records for the antimicrobial susceptibility tests and notes that a few weekly control tests are circled. The supervisor says these results were out of control, but when asked to provide evidence of corrective action, he says the control organism was faulty and a new one had been ordered. In the meantime, results from the susceptibility tests were being reported.

The relevant microbiology checklist question, MIC.21910, is a Phase II question that asks, "For antimicrobial susceptibility testing of either disk or dilution type, are control organisms tested with each new lot or batch of antimicrobials or media and each day the test is performed thereafter?"

To clarify this question and further define out-of-control results, the checklist says: "The overall performance of the susceptibility test system should be monitored by testing appropriate reference strains each day the test is performed. However, the frequency of test monitoring may be reduced to weekly, including the testing of new lots or batches of antimicrobials or media, if the laboratory can document satisfactory performance with daily control tests as suggested by NCCLS guidelines." Performance is considered satisfactory if all reference strains were tested for 30 consecutive test days and this is documented, and, for each drug/microorganism combination, no more than three of the 30 values (zone diameter or minimum inhibitory concentrations) are outside the accuracy ranges. These limits can be established by the laboratory or by NCCLS guidelines M2-A7 or M7-A5.

The laboratory in vignette No. 4 first must be able to give the inspector documentation showing that it has completed the 30-day, daily QC tests. And, if new antimicrobials have been added since these tests were done, it needs to document that these new panels have been tested.

When a weekly test is found to be out of control because of an obvious error such as the wrong drug or control stain having been used, positive contamination, or an incorrect atmosphere of incubation, the QC test should be repeated. "If it’s OK, then the lab can continue with weekly quality control testing," Dr. Merrick says.

However, if there’s not an obvious explanation for out-of-control results, the lab must return to daily QC testing long enough to define the source of the adverse result and to document its resolution. This documentation should include testing of appropriate reference organisms for five consecutive test days and proof that for each drug/microorganism combination all five of the disk diameters and minimum inhibitory concentration values are within the accuracy ranges.

"The important point is that during the retesting period, the out-of-control antibiotic should not be reported on patient samples," says Dr. Merrick. Another microbiology checklist question, MIC.21480, a Phase II question, directly addresses that. It asks, "Are the results of controls reviewed for acceptability before reporting patient results?"

Because the laboratory in the vignette has not done this, Dr. Merrick would cite two deficiencies: one for not having evidence of corrective action and another for reporting results on out-of-control tests.

5.  Review of routine antimicrobial susceptibility test results shows that the laboratory is reporting second- and third-generation cephalosporins on all gram-negative isolates, even when the organism is susceptible to a first-generation cephalosporin. Should the inspector cite a deficiency?

Two Phase I microbiology checklist questions apply. MIC.21900 asks, "Are guidelines established for the number and type of antibiotics reported for organisms isolated from different sites of infection?"

To address this question, the laboratory should refer to the latest supplement of NCCLS M100-S11 titled "Performance standards for antimicrobial susceptibility testing," which contains current information on antibiotic selection, interpretation, and QC testing. But each laboratory should set its own guidelines.

"Selection of the most appropriate antimicrobial agents to test and report is a decision best made by each local laboratory in consultation with infectious disease practitioners, the pharmacy, and infection control committee of the medical staff," Dr. Merrick says. Cascade reporting, or reporting no more than four agents, improves the clinical relevance of test reporting and minimizes the selection of multiresistant nosocomial strains by overuse of broad-spectrum antibiotics.

Also implicit in microbiology checklist question MIC.21900 is the requirement that laboratories only report antimicrobial agents that are effective at the infection site where the organism was isolated. This means reporting drugs such as penicillin, cefotaxime, or ceftriaxone and meropenem for a Streptococcus pneumoniae isolate found in a cerebrospinal fluid specimen from a patient with meningitis and not antimicrobial agents that have poor penetration of the meninges and CSF.

The second checklist question, MIC.21950, asks, "Does the procedure manual address unusual or inconsistent antimicrobial testing results?" A new NCCLS document, M39-P, includes a section on data verification that addresses the intent of this question.

"It’s important for a laboratory to have a procedure for cumulative antimicrobial susceptibility test data, and it should address in its procedure manual how it’s going to handle unusual or inconsistent test results," Dr. Merrick says. These supplementary protocols can include checking for transcription errors, satisfactory growth, or the absences of subtle contamination on plates or trays and reproducibility of results by repeating the test or using an alternative test method.

For the laboratory reporting second- and third-generation cephalosporins on all gram-negative isolates, Dr. Merrick says he wouldn’t cite a deficiency if the medical staff has insisted on this information. However, in circumstances where this is not the case, he would strongly recommend that the laboratory decide which group A agents to report routinely and which group B agents to report only selectively in consultation with the medical staff. "They’re not doing anyone a favor by reporting antibiotics that aren’t necessary," he adds.

6.  The laboratory’s quality improvement plan includes a monitor of blood culture contamination rates, and the results from last quarter showed that six percent of blood cultures were contaminated. The supervisor said she was not involved with the study but felt that nursing staff drew most of the contaminated cultures. Is this problem a laboratory concern?

Yes, says Dr. Merrick, and he would cite the laboratory for a deficiency. "The laboratory needs to work with nursing staff and phlebotomists on ways to reduce contamination," he says. Each laboratory should also maintain blood culture statistics that are reviewed regularly by the laboratory director.

The relevant checklist question, MIC.22630, is a Phase II question: "Are sterile techniques for drawing and handling of blood cultures defined, made available to individuals responsible for specimen collection, and practiced?"

This question’s importance is highlighted by studies that show a false-positive blood culture increases hospitalization by an average of five days and can increase pharmacy and laboratory costs by $1,000 if it’s a coagulase-negative staphylococci that requires vancomycin therapy.

A good quality improvement study for the lab in No. 7 would be to find ways to lower its blood culture contamination rate. Dr. Merrick suggests setting a threshold of 2.9 percent, the overall contamination rate found in a CAP Q-Tracks study of 158 institutions (2001 Annual Summary Report, Blood Culture Contamination Q-Tracks, page7).

7.  The laboratory recently received nasal swabs and a few pulmonary specimens designated as "rule out anthrax." You ask to see the laboratory procedures for handling organisms that may require special engineering or work practice controls, and the supervisor says nothing is in writing but they always use universal precautions.

"There are two issues here, one dealing with safety to laboratory workers and the second dealing with security procedures that a laboratory and hospital should have in place to deal with a suspected or real bioterrorist attack," says Dr. Merrick.

Two checklist questions address this. MIC.19060, a Phase II question, asks, "Have policies and procedures been developed to minimize the occupational risk of exposure to infectious agents handled in the microbiology laboratory, in accordance with current recommendations regarding the biosafety levels for working with different organisms?" MIC.19160, a Phase I question, asks, "Are engineering and work practice controls appropriate to the biosafety level of the laboratory defined and implemented?"

These two checklist questions relate to the safety of laboratory workers. Though routine handling of Bacillus anthracis specimens carries a low risk, "all of us need to be aware of laboratory-acquired infections," Dr. Merrick says. The CDC recently provided a list of steps laboratories should follow if B. anthracis is suspected. And an article from the CDC, titled "Laboratory-acquired meningococcal disease—United States, 2000," published in the March 13 issue of the Journal of the American Medical Association, discusses laboratory-acquired meningococcal disease and the precautions each laboratory should take with spinal fluid and blood culture isolates of Neisseria meningitidis.

No checklist question addresses security issues, but most hospitals have developed bioterrorist plans that include security throughout the hospital, including the laboratory. Dr. Merrick refers laboratories for guidance to the joint CDC and National Institutes of Health document, "Biosafety in microbiological and biomedical laboratories." The most recent edition can be found on the CDC Web site at

Dr. Merrick says he would cite the laboratory in No. 8 for two deficiencies and advise it to develop safety and security policies.

8.  The laboratory’s procedure manuals are well written and include all of the pertinent information about the laboratory’s scope of testing. A face sheet at the front of the manual is signed by all of the people working in the department. It says all persons are knowledgeable about the manual’s contents. Is this adequate?

Checklist question MIC.12140, a Phase II question, asks, "Does the laboratory have a system documenting that all personnel are knowledgeable about the contents of procedure manuals relevant to the scope of their testing activities?"

In this case, Dr. Merrick would not cite a deficiency, "but I’d make a strong recommendation to the laboratory to do more than just sign the front of the manual." A good comprehensive competency assessment program includes directly observing patient test performance and performance of instrument maintenance and function checks; monitoring the recording and reporting of test results; reviewing worksheets, quality control, proficiency testing, and preventive maintenance records; evaluating test performance by testing previously analyzed specimens, internal blind unknowns, or external proficiency test samples; and evaluating problem-solving skills.

9.  While observing the specimen set-up bench, the inspector notes that a urine specimen container is unlabeled but is accompanied by a request form. The form includes patient information but nothing about type of infection or organisms expected. Is this considered a deficiency?

Yes, says Dr. Merrick. "The urine container was not labeled and I don’t think it should be accepted."

There are three relevant microbiology checklist questions. The one Phase II question, MIC.13050, asks, "Are there documented procedures for the proper collection, transporting and handling of culture specimens?" The two Phase I questions, MIC.13100 and MIC.13200, ask, "Are there criteria for establishing specimen acceptability?" and "Do requests for isolation include source of specimen and, when appropriate, type of infection and/or organism expected?"

The first two questions are straightforward: Every laboratory should have a specimen collection manual that spells out the criteria for specimen acceptability. The final question, which relates to getting information about the type of infection and/or organism expected, is trickier. "If your laboratory is like my laboratory and most of the ones I’ve inspected, we’re lucky if we get any of that information," Dr. Merrick admits. "But the point of this is to develop a request form that includes space for that information and to make an effort to get more information. We all do a better job of working up the specimen if we have that available to us."

10.  A positive blood culture result was called to a nursing floor on an evening shift, but the attending physician did not hear about it until the next day and appropriate treatment was not started until then. Is this a deficiency?

"This may be one of the more important questions in all the scenarios I’ve presented," says Dr. Merrick. "It addresses the postanalytical portion of our testing."

Microbiology checklist question MIC.22620, a Phase II question, asks, "Are initial positive cultures reported in accordance with laboratory policy for alert values?"

The CAP does not set standards for which results should be called. Each lab, in consultation with its medical staff, needs to develop a list of analytes that require immediate notification of a physician or other clinical personnel responsible for patient care when test results fall within established alert or critical ranges. "These values represent those results that prompt rapid clinical attention to avert significant patient morbidity or mortality," Dr. Merrick says.

Other issues that need to be considered: how to contact appropriate health care providers and document these calls, when to release sensitive laboratory results over the phone, and what results should be called to infection control personnel.

Dr. Merrick would cite a deficiency if the laboratory did not have a policy in place to handle results within alert or critical ranges.

11.  Stools submitted for ova and parasitic exam are examined by a direct wet mount and concentration procedure, but permanent stains are not performed because, in the supervisor’s opinion, they never yield positives that are not seen by the other two techniques. How should the inspector respond?

This is "definitely a deficiency," Dr. Merrick says. "Permanent stains are the most sensitive method for identifying intestinal protozoa."

Checklist question MIC.52100, previously a Phase I and now a Phase II question, asks, "Does the microscopic examination of all stools submitted for an ova and parasitic examination include a concentration procedure and a permanent stain?"

If the specimen is received fresh, the laboratory should perform a direct wet mount to look for motile forms of intestinal protozoa. If there’s going to be delay, then it’s appropriate to use a stool preservative—but both concentration and permanent stains must be performed.

12.  The hematology laboratory processes all malaria smears. The procedure describes the preparation of both thick and thin smears. The supervisor says they prepare both but never look at the thick smear because nobody knows how to read them. Is this a deficiency?

Two microbiology checklist questions, MIC.52200 and MIC.52260, both Phase I, apply: "Are both thick and thin films made to provide thorough examination for blood parasites?" and "Are an adequate number of fields examined under oil immersion (e.g. 300 fields)?"

"Our parasitology consultants feel that it’s important to screen both thick and thin blood smears, and they should be examined immediately to detect parasites," Dr. Merrick says. "One can detect smaller numbers of parasites on the thick smear." Therefore, he would cite a deficiency because laboratories that don’t feel comfortable reading a thick smear must send it out to others who can analyze it and report back the results.

Vida Foubister is a writer in Mamaroneck, NY. Address questions about the CAP Laboratory Accreditation Program to