TEXT

Top Abstract Text References

Microscopic examination remains the initial diagnostic test in the processing of specimens in the clinical microbiology laboratory. The timely report of a Gram stain result gives the physician important information about the presence and cause of infection. The Gram stain has a broad staining spectrum (2, 3) and classifies bacteria as either gram positive or gram negative. The preparation and analysis of a Gram stain is a highly complex procedure that requires extensive experience in order for a correct result to be reported (11).

(Part of this research was presented at the 99th General Meeting of the American Society for Microbiology, 30 May to 3 June 1999 [D. L. Church, Abstr. 99th Gen. Meet. Am. Soc. Microbiol., abstr. C-177, 1999].)

Despite the clinical importance of the Gram stain, there are few available standards for the reading and interpretation of this test. The assignment of semiquantitative and quantitative values to the number of cells and bacteria seen is clearly arbitrary, since published criteria for general use vary dramatically (1, 9, 10, 14, 16, 17). Although all aspects of Gram stain interpretive criteria need to be standardized for all specimen types, the lack of standard semiquantitative and quantitative criteria for the analysis of cells and bacteria is particularly problematic. Although laboratories may report only semiquantitative Gram stain criteria, it is clinically important to have a standardized schema for assigning quantities of cells and bacteria to individual semiquantitative scores. Since several clinical studies have shown that the presence of moderate to heavy amounts of pus and/or the presence of bacteria on Gram stains correlates with the presence of infection (4, 8, 12), information reported from specimen Gram stains needs to be accurate.

Prior to the introduction of standardized Gram stain interpretation criteria by our external proficiency-testing program (Clinical Microbiology Proficiency-Testing [CMPT] program, University of British Columbia, Vancouver, British Columbia, Canada) 4 years ago, there was a lack of interlaboratory consistency in this regard among microbiology laboratories participating in the Gram stain surveys. It was therefore of interest to evaluate the degree of standardization of semiquantitative and quantitative Gram stain reporting criteria that exists amongst microbiology laboratories in Alberta, Canada, and to also determine if they were routinely utilizing the CMPT quantitative Gram stain criteria as opposed to other previously described schemata for reporting smears from clinical specimens.

Alberta has participated in the CMPT program since 1992. The CMPT program developed a simulated Gram stain program and began to routinely send smear challenges on alternate surveys as educational exercises in 1995. All simulated Gram stain slides are manufactured by the CMPT program according to specifications previously outlined for each challenge by the program's expert committee (5). Although it was recognized that these slides were an invaluable test of proficiency, there was a lack of consistency in the semiquantitative and quantitative interpretation of cells and bacteria on Gram stain challenges when laboratories were asked to use the criteria they utilized in-house. The committee agreed to not grade these new challenges until standardized criteria for reporting the numbers of cells and bacteria on Gram stain surveys were implemented. The CMPT program provided all laboratories with their recommended Gram stain-interpretive criteria (Table 1) (7), which were previously developed from published schemata that varied widely (1, 9, 10, 13, 14, 16, 17). For example, some published criteria grade the numbers of cells according to observations using a low-power field (×100) (1, 9), in contrast to the oil immersion field (×1,000) recommended by others (13), and there is no consensus as to how many fields need to be examined, although clearly more than 10 and perhaps as many as 20 to 40 fields must be read for each slide.

The CMPT program encourages laboratories to report both semiquantitative and quantitative criteria on Gram stain surveys for cells and bacteria, although laboratories are not graded on these scores.

The Laboratory Proficiency Testing Program, College of Physicians & Surgeons of Alberta, surveyed all microbiology laboratories in the province in November 1997 regarding the interpretative criteria being routinely used to read Gram-stained smears, including the CMPT slides. Laboratories were requested to provide a copy of their laboratory's criteria for Gram stain interpretation of various elements which included definitions for both semiquantitative (i.e., grades 1+, 2+, 3+, and 4+) and quantitative reporting of cells (i.e., numbers of polymorphonuclear and epithelial cells per oil immersion field) and bacteria when 10 or more oil immersion fields are read (oil immersion field of ×1,000). The results of the survey were analyzed against the CMPT program's interpretive criteria to determine how many laboratories were compliant with the recommendations of our external proficiency program. Data were entered into a spreadsheet program, Microsoft Excel version 6.0, and analyzed using standard descriptive methods.

A total of 28 of 32 (87.5%) laboratories completed and returned the Gram stain survey. Two laboratories that do not routinely perform Gram stains did not participate in the survey. There was a wide variability in the Gram stain reporting criteria used to grade both cells and bacteria. Many laboratories use semiquantitative criteria (grades 1+, 2+, 3+, and 4+) rather than quantitative criteria for reporting Gram stain results.

Table 2 outlines the interpretive criteria for reporting cells and bacteria used by all of the laboratories that completed the survey. There was better agreement among laboratories in interpreting the number of cells than the number of bacteria present. Overall there was good agreement among laboratories for the numbers of cells and bacteria per field used to report 1+ or 2+ grades. A total of 19 of 28 (68%) laboratories would have reported <1 cell and bacterium per field as 1+ (rare). A total of 24 of 28 (86%) laboratories would have reported the presence of between 1 and 5 cells as 2+ (few), and 27 of 28 (96%) would have called the presence of between 1 and 10 bacteria per field 2+ or few. The ranges of cells per field and bacteria per field were most variable amongst the 3+ and 4+ grades. Laboratories interpreted a moderate number of cells or bacteria per field to be anywhere between 2 and 50 cells for a 3+ (moderate) grade and 2 to >50 cells or bacteria per field for a 4+ (many or numerous or heavy) grade.

Only 11 of 32 (34.4%) Alberta laboratories routinely utilize the Gram stain reporting criteria specified by the CMPT program (Table 2). Many rural laboratories have adopted the CMPT criteria for routinely reporting Gram stains, which is an improvement, since many small sites did not previously use interpretive criteria for reading stains. Larger laboratories in urban centers routinely utilize other published reporting criteria for interpretation of clinical specimens (1, 9, 13, 14, 16) and may not convert to CMPT criteria for the purpose of reporting simulated Gram stain challenges on proficiency testing surveys. Several large laboratories indicated that they do not agree with the quantitative criteria specified by the CMPT program because it does not agree with reporting schemata that have been previously published (13).

Diagnostic microbiology laboratories should ensure that the Gram stain smears from clinical specimens are being consistently interpreted by using common interpretive criteria in-house for grading the presence of cells and bacteria. Physicians clinically rely in many cases on an initial Gram stain report to confirm the presence of infection by showing the presence of pus and bacteria in submitted clinical specimens. The results of the initial Gram stain should also be correlated to the culture results from individual clinical specimens to ensure that all of the bacteria of interest found in the smear are recovered. Accurate interpretation of Gram stains is therefore a critical performance parameter that should be monitored by the laboratory's internal and external quality assurance programs.

However, this study also highlights the need to develop standardized Gram stain-interpretive criteria that would be accepted and widely used by all microbiology laboratories within our province. Although individual laboratories may have a particular schema in place for reading Gram stains, the increasing transfer of patients in our province between regions has made it necessary to standardize many laboratory procedures across facilities (5, 6). The lack of standardization across the province in Gram stain reporting between laboratories located in different health care regions poses a potential problem for medical specialists in urban centers caring for patients transferred from rural regions. The results of this study will be used to educate laboratories in our province that are currently not using standard criteria for Gram stain interpretation. In addition, an educational program will be developed for technologists to enhance continuing education in the accurate interpretation of Gram stains.

Despite the clinical importance of the Gram stain, there are currently only a few published quantitative criteria for the numbers of cells and bacteria that have been clinically evaluated for interpretation of specific types of clinical specimens. Gram stain scoring criteria for the diagnosis of bacterial vaginosis have been extensively evaluated (10, 17), as have Q-scoring criteria for sputum specimens (1, 14, 16). Microscopic criteria for the diagnosis of urethritis, particularly Neisseria gonorrhoeae infection in men, have also been studied (9). Although these criteria are being used to interpret smears from other types of specimens, these schemata may not be generally applicable.

Several other aspects of the interpretation and reporting of Gram stains also need to be addressed. No standards exist for the descriptive naming of cellular and bacterial morphotypes from Gram stains. Clinical outcome trials also need to be performed to determine the most effective way to report Gram stain information to physicians. There have been few studies of the clinical outcomes of Gram stain reporting. Mizrachi et al. (15) showed that physicians were significantly more likely to follow up poor-quality sputum specimens if an interpretive comment was also provided on the report in addition to the numbers of epithelial cells and bacteria present.

International standards should be developed and implemented for Gram stain interpretation of all types of clinical specimens. The current lack of standardization of Gram stain reporting criteria not only poses a problem in grading the results of proficiency testing challenges but may also impact the quality of patient care.