9^th European Congress of Clinical Microbiology and Infectious Diseases

21-25 March, 1999, Berlin, Germany
Poster # P1196

Quality Assessment (QA) of Antimicrobial Susceptibility Testing in Russia

O.U. STETSIOUK, G.K. RESHEDKO, O.I. KRETCHIKOVA, E.L. RJABKOVA

Department of Clinical Pharmacology and Antimicrobial Chemotherapy, Smolensk State Medical Academy, Smolensk, Russia;
Interregional Association for Clinical Microbiology and Antimicrobial Chemotherapy (IACMAC), Russia

The PDF format poster (261 kb)

• Abstract
• Introduction and purpose
• Materials and methods
• Results and discussion
• Conclusion
• Acknowledgments

ABSTRACT

Objective: To assess quality and to determine the overall accuracy of antimicrobial susceptibility testing in Russian laboratories.

Methods: Eighteen laboratories performed the following QA tests:

1. test control strains of E.coli ATCC 25922 and S.aureus ATCC 25923 once each day for 10 working days;
2. test strains granted by WHO/CDC: K.pneumoniae WHO-1, S.aureus WHO-2, E.faecium WHO-3 and Enteric WHO-5 only once to defined sets of antimicrobials;
3. identification of Enteric WHO-5.

Results: Participating laboratories used different media (Russian AGV-media and Mueller-Hinton) and antimicrobial disks from different producers both domestic and imported. Interpretation of the results they performed according to "Methodical Rules" by Ministry of Health of the USSR dated 1983 and NCCLS 1995-96 recommendations.

Results of testing of E.coli ATCC 25922 and S.aureus ATCC 25923 to different antimicrobials were within the NCCLS range in 30-94% and 41-91% respectively. Smaller zone diameters were obtained in 6-64% and 4-53% and larger - in 0-7% and 1-44% respectively.

Testing of WHO strains to different antimicrobials showed that agreement in the interpretation of results was attained in 33-100% for K.pneumoniae WHO-1, 67-100% for S.aureus WHO-2, 20-92% for E.faecium WHO-3 and 0-100% for Enteric WHO-5. Correct identification of Enteric WHO-5 (Enterobacter cloacae) was performed in 10/17 centres.

Conclusions: The above results elucidate the problems with the internal quality control in most Russian laboratories. Personnel education and procedures standardisation are needed for the improvement of the present situation.

INTRODUCTION AND PURPOSE

The impact of antimicrobial susceptibility testing (AST) on effective antibacterial therapy is of increasing importance. The results of AST in vitro serve as a basis for administration of optimal antibiotic regimen for current patient, for empirical choice of the drugs for the treatment of definite type of infection and for creating of hospital formulary.

Strict standartisation of the testing procedure and regular performing of quality control tests are essential for achieving of repeatable clinically relevant results. The purpose of the current study was to assess the quality and to determine the overall accuracy of AST in Russian microbiology laboratories.

MATERIALS AND METHODS

Eighteen Russian microbiology laboratories were asked to perform the following QA tests:

1. Test control strains of E.coli ATCC 25922 and S.aureus ATCC 25923.to defined sets of antimicrobials once each day for 10 working days.
2. Test strains granted by WHO/CDC: K.pneumoniae WHO-1, S.aureus WHO-2, E.faecium WHO-3 and Enteric WHO-5 to defined sets of antimicrobials once only.
3. To perform identification of Enteric WHO-5.

RESULTS AND DISCUSSION

The participating laboratories used different media (Russian AGV-media and Mueller-Hinton) and antimicrobial disks from different producers both domestic and imported. Interpretation of the results was performed according to "Methodical Rules" by Ministry of Health of the USSR 1983 and NCCLS 1995-96 recommendations.

The results of testing of E.coli ATCC 25922 and S.aureus ATCC 25923 to different antimicrobials were within the NCCLS range in 30-94% and 41-91% respectively. Smaller zone diameters were obtained in 6-64% and 4-53% and larger - in 0-7% and 1-44% respectively. The results are presented in the tables 1-2.

Table 1. Distribution (%) of the results (zone diameters) of ATCC strains with respect to NCCLS (1998) acceptable ranges (Normal - % of results within NCCLS range, Low - % of zone diameter results smaller than NCCLS range, High - % of zone diameter results larger than NCCLS range).

Table 2. Distribution (%) of the results (zone diameters) of ATCC strains with respect to NCCLS (1998) acceptable ranges (Normal - % of results within NCCLS range, Low - % of zone diameter results smaller than NCCLS range, High - % of zone diameter results larger than NCCLS range).

From our point of veiw there are three main causes explaining the prevalence of smaller zone diameters of quality control strains:

1. Low content of b-lactam antibiotics in sensitivity disks due to the poor stability of these drugs especially in case of improper storage;
2. Unacceptable high Ca²⁺, Mg²⁺ and thimidine contents in Russian AGV-medium that leads to false resistance to aminoglycosides, quinolones and anti-folates;
3. Non-standardised (heavy) inoculum used.

The high frequency of larger S.aureus ATCC 25923 zones around oxacillin disk could be due to oxacillin 10 mcg disks used in some of the participating laboratories.

Testing of WHO/CDC strains to different antimicrobials showed that agreement in the interpretation of results with those of CDC was attained in 33-100% for K.pneumoniae WHO-1, 67-100% for S.aureus WHO-2, 20-92% for E.faecium WHO-3 and 0-100% for Enteric WHO-5. The results are presented in the tables 3, 4, 5, 6.

Table 3. Agreement (%) and the rates of errors (%) of the results of WHO/CDC QA strains interpretation (by sensitivity category) with respect to CDC results interpretation.

Table 4. Agreement (%) and the rates of errors (%) of the results of WHO/CDC QA strains interpretation (by sensitivity category) with respect to CDC results interpretation.

Table 5. Agreement (%) and the rates of errors (%) of the results of WHO/CDC QA strains interpretation (by sensitivity category) with respect to CDC results interpretation.

Table 6. Agreement (%) and the rates of errors (%) of the results of WHO/CDC QA strains interpretation (by sensitivity category) with respect to CDC results interpretation.

The obtained results of testing of WHO/CDC quality control strains with unknown susceptibility patterns confirm that testing of clinical strains in Russian laboratories in routine practice produce a lot of errors. There is a trend to over-detection of the antimicrobial resistance, whereas false-susceptible results are relatively rare.

The main reason of the majority of problems with AST is that the only official document regulating the AST procedure in Russia is the "Methodical Rules" by Ministry of Health of the USSR revised last time in 1983.

This document advocates use of Russian AGV-medium for testing, composition of which differs considerably from Mueller-Hinton agar and does not meet WHO requirements for cations and thimidine contents, etc.

These guidelines include only disk-diffusion procedure of 23 "old" antimicrobials for non-fastidious bacteria only. No recommendations have been developed for dilution tests and screening of specific types of resistance (MRSA, penicillin-resistance in pneumococci, ESBL-production in gram-negative bacteria).

CONCLUSION

• The above results elucidate the problems with the internal quality control in most Russian laboratories.
• Personnel education and procedures standardisation are needed for the improvement of the present situation.
• There is an urgent demand for National guidelines for AST in Russia, harmonised with the international practices (WHO, NCCLS, SFM, DIN, BSAC, etc.).

ACKNOWLEDGMENTS

We are indebted to Dr. Rosamund Williams and Dr. John Stelling from Laboratory Training and Support, Division of Emerging and Other Communicable Diseases Surveillance and Control, World Health Organization and Dr. Fred C. Tenover from Nosocomial Laboratory Branch, CDC, for their kind permission to use the schedule and the WHO/CDC control strains for QA in Russia.

We are grateful for our colleagues from microbiology laboratories of Russia who participated in this QA project.

The skilful technical assistance of Boris Makushkin, data processing co-ordinator, is gratefully acknowledged.