Journal of Antimicrobial Chemotherapy (1992) 30, 721-735
Correspondence Btoaany for gryeopeptides J Antimicrob Chemother 1992; 30: 721
A. C.SCOTT Central Microbiological Laboratories, Western General Hospital. Crewe Road, Edinburgh EH4 2XU. UK
Ref« Durodie, J. & Coleman, K_ (1992). Bioatsay for glycopeptides. Journal of Antimicrobial Chemotherapy 29, 347-«. Fuji, R., Grossman, M. & Tickner, W. (1961). Mkromethod for determination of concentration of antibiotic* in terum for application in clinical paediatrics. Paediatrics 28, 662-669.
J. DURODIE K. COLEMAN SmithJGine Beecham Pharmaceuticals, Brockham Park, Betckworth, Surrey RH3 TAJ, UK Refi Brooki, B. & Coleman, K. (1989). The use of image analysis for MIC determination and bioatsay. In Rapid Microbiological Methods for Foods, Beverages and Pharmaceuticals (Stannard, C. J., Pettit, S. B. & Skinner, F. A., Eds.), pp. 73-85. Society for Applied Bacteriology Technical series No. 25. Blackwell Scientific Publications, Oxford. Durodie, J. &. Coleman, K.. (1992). Bioassay for glycopeptides. Journal of Antimicrobial Chemotherapy 29, 347-8. Fuji, R., Grossman, M. & Tickner, W. (1961). Micromethod for determination of concentration of antibiotics in serum for application in clinical paediatrics. Paediatrics 28, 662-9.
Characterization of imipenem-ttsistairt PtemdomoHas aentgimosa strains front cystic flbrosis patients
Reply
J Anttmicrob Chemother 1992; 30: 721-724 J Anttmicrob Chemother 1992; 30: 721
Sir, We thank Dr Scott for drawing our attention to the erythromycin assay described by Fuji, Grossman & Tickner (1961), of which we were previously unaware. The Fuji paper certainly deserves mention since it also utilizes a 0-haemolytic streptococcus and measures unhaemolysed blood, but there the similarity ends. Fuji et al. (1961) employ a capillary tube micromethod which requires 24 h at 4°C followed by 18 h at 37°C, compared to the overnight zone diffusion assay employed by us (Durodie & Coleman, 1992). Our method lends itself to certain automated methods of
Sir, Imipenem is a potent /7-lactam antibiotic active
against Pseudomonas aeruginosa which is used to treat pulmonary infections in cystic fibrosis (CF) patients. In some CF-centres resistance emerged rapidly and persisted after cessation of imipenem-cilastatin (IMP/CIL) treatment (Pedersen et al., 1985, 1987). In other trials a slight increase in the MIC was noted during treatment, but all strains were fully susceptible one month later (Strandvik et al., 1988). We report the value of characterizing sputum isolates, particularly by DNA restriction fragment length polymorphism, to distinguish between resident strains becoming 771
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
Sir, Durodie & Coleman (1992) are right to be enthusiastic about their assay for glycopeptide antibiotics in blood by radial diffusion in agar seeded with haemorytic streptococci with subsequent measurement of zones of unhaemorysed red cells, but they are wrong to claim novelty for it. This is the method described by Fuji, Grossman & Tickner (1961) and used by them to measure antibiotic levels in smallvolume samples from neonates. The method is highly accurate, reproducible and elegant, but it is not new and, as Fuji et al. (1961) pointed out 30 years ago, it can be used to assay not only glycopeptides but any other antibiotic to which the chosen streptococcus is sensitive.
reading (Brooks & Coleman, 1989), and this was the principal reason for developing such an assay. In conclusion, we feel that the longer incubation time of the Fuji method, coupled with the rather cumbersome procedures employed and the difficulty of imaging the final zones, all weigh against its use in a routine laboratory.
Correspondence
Figure. Pulsed-field gel electrophoresis with • contour clamped homogeneousfield(CHEF) of DNAs from P. aentginoM isolates from CF patients. DNAs were digested with Dral and electrophoresis conditions consisted of pulse times of 20 sec during 12 h and 5-15 sec during 17 h at 150 V. Lanes 1 and 9: high molecular-weight lambda concatemer ladder. Patient 2 (lanes 2-6): lane t pulsotype A susceptible to imipenem (IMP); lane 3: pulsotype A resistant to IMP; lane 4: pulsotype B susceptible to IMP; lane 5: pulsotype B resistant to IMP; lane 6: pulsotype C resisUnt to IMP. Patient 1 (land 7-8): lane 7: pulsotype D susceptible to IMP; lane 8: pulsotype D resistant to IMP.
resistant after IMP/CIL treatment, and exogenous resistant strains newly acquired by the patient. Sputum samples from CF patients in our hospital generally were taken and analysed once every three months, or in some cases once every month. All P. aeruginosa colonies isolated from each sample were tested for their antibiotic susceptibility, serogroup (16 antisera from Diagnostics Pasteur, Marnes-laCoquette, F) pyocin-type (Fyfc, Harris & Govan, 1984), plasmid DNA profile (Plesiat, Alkhalaf & Michel-Briand, 1988) and total DNA profile by restriction fragment length polymorphism ("pulsotype") (Dral DNA hydrolysis followed by pulsed-field gel electrophoresis, CHEF, Smith, Klco & Cantor, 1988). Two children each carrying for eight to eleven months a single type of P. aeruginosa which was susceptible to IMP/CIL (MIC = 1 mg/L) were treated with IMP/CIL (50 mg per kg/day as a 30 min intravenous infusion four times a day) and another antibiotic for two weeks.
Patient one (19 years old) was treated with IMP/CIL-tobramycin therapy. Two strains of P. aeruginosa were isolated after treatment. One was identical to the pretreatment isolate, and the second differed only in that it was resistant to imipenem (MIC = 16 mg/L) (Table). The two strains persisted for eleven months. In the second patient (23 years old), the strain (pulsotype A, pyocinotype 13/h) became IMP-resistant after IMP/CIL-riprofloxacin therapy and a second strain (pulsotype B, presence of two plasmids, pyocinotype 5/j), susceptible to IMP, was identified. Seven months later the patient was given two weeks of IMP/CIL-amikacin therapy. The second strain then developed resistance to imipenem and a third IMP-resistant strain (pulsotype C, one plasmid, pyocinotype 5/0 was detected: during the following four months the patient harboured three different IMP-resistant strains (Table). Three months later the patient received a lung transplant but did not survive. These observations confirmed that IMP-resistance may appear rapidly after
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
40-S
Correspondence
723
Table. Evolution of P. aeruginosa strains in the sputum of two patients given imipcnem/cLlasUtin therapy
Patient 1
2
Total no. of isolates tested
Genotype
Pyodnotype
MlC(ing/L) nnipcodn
23 4 7
D D D
V V V
> 16
10
21
A
6
4 10
A
10/h 13/h 5/j
6 7 6
A
specimens
none after treatment
11 11
11 10
none after first treatment course
8 7
after
4
No. of
6
2nd treatment course
B
13/h 5/j
B C
5/f
i 1 16
16
> 16 > 16
V, variable
IMP/CIL treatment, in spite of combination therapy with another antibiotic. However, the origin of the resistant strain may vary. Precise characterization of the P. aeruginosa strains was required to determine the origin of the strains. All the isolates from both patients were polyagglutinable, in agreement with Fomsgaard et al. (1988) for strains isolated from long-term colonized CF patients. Pyocin typing was unreliable in patient 1 and only 20% of the strains contained plasmids (Piesiat et al., 1988). In contrast, all strains may be characterized by restriction fragment length polymorphism. Pulsed-field dectrophoresis of Dral fragments of related strains may differ by only a few bands (Grothues et al., 1988), although our DNA profiles were very stable. We therefore used several techniques to establish relatedness among isolates. This accurate identification can be used to determine if coexistence of susceptible and resistant strains results from a mutation within a population derived from one strain (patient 1) or from the implantation of a new strain (patient 2 after first course of therapy). Thus cross-infection of CF patients from the environment should be prevented even when a strain is predominant in the bronchopulmonary tract (Hjelm, Branstrom & Warren, 1990; Ogle & Vasfl, 1991). Furthermore, the coexistence of three different resistant strains (patient 2 after second course of therapy) would be impossible to detect without identification by pulsotype. Acknowledgement. The "Association Francaise de Lutte contre la Mucoviscidose" is
greatly acknowledged for support. C. Bailly and M. Thouverez are thanked for their skilful assistance. C. GODARD J. LEROY Y. M1CHEL-BRIAND* Laboratoirt de Baclkriologie, Faadli ae Medicine. Centre Hospitaller Jem Mtnjoi. 25030 Besancon. France Reft Fomsgaard, A., Conrad, R. S., Galanos, C , Shand, G.H. &. Heiby, N. (1988). Comparative immunochemistry of Hpo-polysaccharides from typaMe and polyigglutinable Pseudomonas aeruginosa strains isolated from patients with cystic fibrosis. Journal of Clinical Microbiology 26, 821-6. Fyfe, J. A. M., Harris, G. & Govan, J. R. W. (1984). Revised pyocin typing method for Pseudomonas aeruginosa. Journal of Clinical Microbiology 20, 47-50. Grothues, D., Koopman, U., von der Hardt, H. & Tflmmler, B. (1988). Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. Journal of Clinical Microbiology 26, 1973-7. Hjelm, L.N., Branstrom, A. A. A Warren, R. L. (1990). Detection of restriction fragment length polymorphisms in clinical isolates and serially P""fE1'^ Pseudomonas aeruginosa strains. Journal of Clinical Microbiology 28, 2178-82. Ogle, J.W. & Vasil, M. L. (1991). Genetic heterogeneity in strains of Pseuaomonas aeruginosa from patients with cystic fibrosis. Journal of Clinical Microbiology 29, 663-4. 'Corresponding author.
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
Time of observation (months)
Intipenem treatment
724
Correspondence
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
et at., 1988). Prescriptions for prophylaxis are labelled ADP. Prescriptions for empirical therapy (ADE) treat a presumed infection before culture results become available. Prescriptions for documented therapy (ADT) are directed to a known (cultured) pathogen, primary or after ADE. Continuing AD therapy beyond 72 h in the presence of a negative culture result or in the absence of cultures is defined as continued empirical therapy (ADET). To allow separate evaluation of empirical therapy and subsequent continuation of empirical therapy or documented therapy with the same drug, we split up prescriptions, after the culture results became known or after a maximum of 72 h. For each prescription, the flow chart (Figure) is read down from top to bottom, except if the records are insufficient for categorization (stop at category VI) or if the criteria for infection (Gamer et al., 1988) are not met (stop at category V, unjustified). AD presciptions can be inappropriate for several reasons at the same time and therefore can be placed in more than one 471-80. (sub)category (categories IV down to II). Prescriptions for therapy are considered inappropriate if errors are made in dose (category Ila), interval (category lib) or route Optimizing antimicrobial therapy. A method (category He). As correct length of therapy is for antimicrobial drug use evaluation frequently arbitrarily established, a duration of / Antimicrob Chemother 1992; 30: 724-727 treatment which differs markedly from that proposed in a leading infectious diseases textSir, Reasons for monitoring antimicrobial drug book is considered either too long (AD) usage are to improve medical care, to limit the development and spread of resistance Table L Antimicrobial Drag (AD) Evaluation Categories, present study and to contain costs. In the most authoritative classification of good antimicrobial drug use, usage is classified as appropriate, probably I. Agree with the use of antimicrobial therapy/ prophylaxis, the prescription is definitely appropriate, inappropriate—less expensive or appropriate. toxic drug recommended, inappropriate— II. The AD therapy/prophylaxis prescription is modified dose recommended and unjustified inappropriate due to: (Kunin, 1973). We modified the original a. improper dosage criteria in order to be able to evaluate each b. improper dosage interval parameter of importance associated with AD c. improper route use (Table I)- To facilitate the selection into III. The AD therapy/prophylaxis prescription is inappropriate due to: (sub)categories, we arranged them in a flow a. excessive length chart (Figure). We used the evaluation in oneb. duration too short month target programmes in our hospital. A AD therapy/prophylaxis prescription is limited number of AD courses were summar- IV. The inappropriate due to: ized by one researcher internist to allow suba. more effective alternative agent (Aa): sequent evaluation by two independent experts specify in infectious diseases, as described by others b. less toxic Aa: specify (VolgeT et aJ., 1988). The results of the evaluac. less expensive Aa: specify d. less broad spectrum Aa: specify tion were presented in an educational report to V. The AD therapy/prophylaxis prescription is prescribers. unjustified: use of any antimicrobial is not The terms prescription, course, prophylaxis, indicated. infections were strictly defined, based on authoritative literature (Moss et at., 1981; Gamer VI. Records insufficient for categorization.
Pedersen, S. S., Prcssler, T., Heiby, N., Bentzon, M.W. & Koch, C. (1985). Imipenem-cilastatin treatment of multi-resistant Pseudomonas aeruginosa lung infection in cystic fibrosis. Journal of Antimicrobial Chemotherapy 16, 629-35. Pedensen, S. S., Presster, T., Jensen, T., Rosdahl, V.T., Bentzon, M.W. Heiby, N. et aJ. (1987). Combined imjpenem/cilastatin and tobramycin therapy of multi-resistant Pseudomonas aeruginosa in cystic fibrosis. Journal of Antimicrobial Chemotherapy 19, 101-7. Plenat, P., Alkhalaf, B. & Micael-Briand, Y. (1988). Prevalence and profiles of plasmidi in Pseudomonas aeruginosa. European Journal of Clinical Microbiology and Infectious Diseases 7, 261-4. Smith, C. L., Kfco, S. R. & Cantor, C R. (1988). Pulsed-fieM gel electrophoresis and the technology of large DNA molecules. In Genome Analysis. A Practical Approach (Davies, K. E., Ed.), pp. 4 1 72. IRL Press, Oxford. Strandvik, B., Malmborg, A. S., Bergan, T., Michalsen, H., Storrosten, O. T. &. WreUind, B. (1988). Imipenetn/cilastatin, an alternative treatment of Pseudomonas infection in cystic fibrosis. Journal of antimicrobial Chemotherapy 21,
Correspondence Btoaany for gryeopeptides J Antimicrob Chemother 1992; 30: 721
A. C.SCOTT Central Microbiological Laboratories, Western General Hospital. Crewe Road, Edinburgh EH4 2XU. UK
Ref« Durodie, J. & Coleman, K_ (1992). Bioatsay for glycopeptides. Journal of Antimicrobial Chemotherapy 29, 347-«. Fuji, R., Grossman, M. & Tickner, W. (1961). Mkromethod for determination of concentration of antibiotic* in terum for application in clinical paediatrics. Paediatrics 28, 662-669.
J. DURODIE K. COLEMAN SmithJGine Beecham Pharmaceuticals, Brockham Park, Betckworth, Surrey RH3 TAJ, UK Refi Brooki, B. & Coleman, K. (1989). The use of image analysis for MIC determination and bioatsay. In Rapid Microbiological Methods for Foods, Beverages and Pharmaceuticals (Stannard, C. J., Pettit, S. B. & Skinner, F. A., Eds.), pp. 73-85. Society for Applied Bacteriology Technical series No. 25. Blackwell Scientific Publications, Oxford. Durodie, J. &. Coleman, K.. (1992). Bioassay for glycopeptides. Journal of Antimicrobial Chemotherapy 29, 347-8. Fuji, R., Grossman, M. & Tickner, W. (1961). Micromethod for determination of concentration of antibiotics in serum for application in clinical paediatrics. Paediatrics 28, 662-9.
Characterization of imipenem-ttsistairt PtemdomoHas aentgimosa strains front cystic flbrosis patients
Reply
J Anttmicrob Chemother 1992; 30: 721-724 J Anttmicrob Chemother 1992; 30: 721
Sir, We thank Dr Scott for drawing our attention to the erythromycin assay described by Fuji, Grossman & Tickner (1961), of which we were previously unaware. The Fuji paper certainly deserves mention since it also utilizes a 0-haemolytic streptococcus and measures unhaemolysed blood, but there the similarity ends. Fuji et al. (1961) employ a capillary tube micromethod which requires 24 h at 4°C followed by 18 h at 37°C, compared to the overnight zone diffusion assay employed by us (Durodie & Coleman, 1992). Our method lends itself to certain automated methods of
Sir, Imipenem is a potent /7-lactam antibiotic active
against Pseudomonas aeruginosa which is used to treat pulmonary infections in cystic fibrosis (CF) patients. In some CF-centres resistance emerged rapidly and persisted after cessation of imipenem-cilastatin (IMP/CIL) treatment (Pedersen et al., 1985, 1987). In other trials a slight increase in the MIC was noted during treatment, but all strains were fully susceptible one month later (Strandvik et al., 1988). We report the value of characterizing sputum isolates, particularly by DNA restriction fragment length polymorphism, to distinguish between resident strains becoming 771
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
Sir, Durodie & Coleman (1992) are right to be enthusiastic about their assay for glycopeptide antibiotics in blood by radial diffusion in agar seeded with haemorytic streptococci with subsequent measurement of zones of unhaemorysed red cells, but they are wrong to claim novelty for it. This is the method described by Fuji, Grossman & Tickner (1961) and used by them to measure antibiotic levels in smallvolume samples from neonates. The method is highly accurate, reproducible and elegant, but it is not new and, as Fuji et al. (1961) pointed out 30 years ago, it can be used to assay not only glycopeptides but any other antibiotic to which the chosen streptococcus is sensitive.
reading (Brooks & Coleman, 1989), and this was the principal reason for developing such an assay. In conclusion, we feel that the longer incubation time of the Fuji method, coupled with the rather cumbersome procedures employed and the difficulty of imaging the final zones, all weigh against its use in a routine laboratory.
Correspondence
Figure. Pulsed-field gel electrophoresis with • contour clamped homogeneousfield(CHEF) of DNAs from P. aentginoM isolates from CF patients. DNAs were digested with Dral and electrophoresis conditions consisted of pulse times of 20 sec during 12 h and 5-15 sec during 17 h at 150 V. Lanes 1 and 9: high molecular-weight lambda concatemer ladder. Patient 2 (lanes 2-6): lane t pulsotype A susceptible to imipenem (IMP); lane 3: pulsotype A resistant to IMP; lane 4: pulsotype B susceptible to IMP; lane 5: pulsotype B resistant to IMP; lane 6: pulsotype C resisUnt to IMP. Patient 1 (land 7-8): lane 7: pulsotype D susceptible to IMP; lane 8: pulsotype D resistant to IMP.
resistant after IMP/CIL treatment, and exogenous resistant strains newly acquired by the patient. Sputum samples from CF patients in our hospital generally were taken and analysed once every three months, or in some cases once every month. All P. aeruginosa colonies isolated from each sample were tested for their antibiotic susceptibility, serogroup (16 antisera from Diagnostics Pasteur, Marnes-laCoquette, F) pyocin-type (Fyfc, Harris & Govan, 1984), plasmid DNA profile (Plesiat, Alkhalaf & Michel-Briand, 1988) and total DNA profile by restriction fragment length polymorphism ("pulsotype") (Dral DNA hydrolysis followed by pulsed-field gel electrophoresis, CHEF, Smith, Klco & Cantor, 1988). Two children each carrying for eight to eleven months a single type of P. aeruginosa which was susceptible to IMP/CIL (MIC = 1 mg/L) were treated with IMP/CIL (50 mg per kg/day as a 30 min intravenous infusion four times a day) and another antibiotic for two weeks.
Patient one (19 years old) was treated with IMP/CIL-tobramycin therapy. Two strains of P. aeruginosa were isolated after treatment. One was identical to the pretreatment isolate, and the second differed only in that it was resistant to imipenem (MIC = 16 mg/L) (Table). The two strains persisted for eleven months. In the second patient (23 years old), the strain (pulsotype A, pyocinotype 13/h) became IMP-resistant after IMP/CIL-riprofloxacin therapy and a second strain (pulsotype B, presence of two plasmids, pyocinotype 5/j), susceptible to IMP, was identified. Seven months later the patient was given two weeks of IMP/CIL-amikacin therapy. The second strain then developed resistance to imipenem and a third IMP-resistant strain (pulsotype C, one plasmid, pyocinotype 5/0 was detected: during the following four months the patient harboured three different IMP-resistant strains (Table). Three months later the patient received a lung transplant but did not survive. These observations confirmed that IMP-resistance may appear rapidly after
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
40-S
Correspondence
723
Table. Evolution of P. aeruginosa strains in the sputum of two patients given imipcnem/cLlasUtin therapy
Patient 1
2
Total no. of isolates tested
Genotype
Pyodnotype
MlC(ing/L) nnipcodn
23 4 7
D D D
V V V
> 16
10
21
A
6
4 10
A
10/h 13/h 5/j
6 7 6
A
specimens
none after treatment
11 11
11 10
none after first treatment course
8 7
after
4
No. of
6
2nd treatment course
B
13/h 5/j
B C
5/f
i 1 16
16
> 16 > 16
V, variable
IMP/CIL treatment, in spite of combination therapy with another antibiotic. However, the origin of the resistant strain may vary. Precise characterization of the P. aeruginosa strains was required to determine the origin of the strains. All the isolates from both patients were polyagglutinable, in agreement with Fomsgaard et al. (1988) for strains isolated from long-term colonized CF patients. Pyocin typing was unreliable in patient 1 and only 20% of the strains contained plasmids (Piesiat et al., 1988). In contrast, all strains may be characterized by restriction fragment length polymorphism. Pulsed-field dectrophoresis of Dral fragments of related strains may differ by only a few bands (Grothues et al., 1988), although our DNA profiles were very stable. We therefore used several techniques to establish relatedness among isolates. This accurate identification can be used to determine if coexistence of susceptible and resistant strains results from a mutation within a population derived from one strain (patient 1) or from the implantation of a new strain (patient 2 after first course of therapy). Thus cross-infection of CF patients from the environment should be prevented even when a strain is predominant in the bronchopulmonary tract (Hjelm, Branstrom & Warren, 1990; Ogle & Vasfl, 1991). Furthermore, the coexistence of three different resistant strains (patient 2 after second course of therapy) would be impossible to detect without identification by pulsotype. Acknowledgement. The "Association Francaise de Lutte contre la Mucoviscidose" is
greatly acknowledged for support. C. Bailly and M. Thouverez are thanked for their skilful assistance. C. GODARD J. LEROY Y. M1CHEL-BRIAND* Laboratoirt de Baclkriologie, Faadli ae Medicine. Centre Hospitaller Jem Mtnjoi. 25030 Besancon. France Reft Fomsgaard, A., Conrad, R. S., Galanos, C , Shand, G.H. &. Heiby, N. (1988). Comparative immunochemistry of Hpo-polysaccharides from typaMe and polyigglutinable Pseudomonas aeruginosa strains isolated from patients with cystic fibrosis. Journal of Clinical Microbiology 26, 821-6. Fyfe, J. A. M., Harris, G. & Govan, J. R. W. (1984). Revised pyocin typing method for Pseudomonas aeruginosa. Journal of Clinical Microbiology 20, 47-50. Grothues, D., Koopman, U., von der Hardt, H. & Tflmmler, B. (1988). Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. Journal of Clinical Microbiology 26, 1973-7. Hjelm, L.N., Branstrom, A. A. A Warren, R. L. (1990). Detection of restriction fragment length polymorphisms in clinical isolates and serially P""fE1'^ Pseudomonas aeruginosa strains. Journal of Clinical Microbiology 28, 2178-82. Ogle, J.W. & Vasil, M. L. (1991). Genetic heterogeneity in strains of Pseuaomonas aeruginosa from patients with cystic fibrosis. Journal of Clinical Microbiology 29, 663-4. 'Corresponding author.
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
Time of observation (months)
Intipenem treatment
724
Correspondence
Downloaded from http://jac.oxfordjournals.org/ at University of Glasgow on June 26, 2015
et at., 1988). Prescriptions for prophylaxis are labelled ADP. Prescriptions for empirical therapy (ADE) treat a presumed infection before culture results become available. Prescriptions for documented therapy (ADT) are directed to a known (cultured) pathogen, primary or after ADE. Continuing AD therapy beyond 72 h in the presence of a negative culture result or in the absence of cultures is defined as continued empirical therapy (ADET). To allow separate evaluation of empirical therapy and subsequent continuation of empirical therapy or documented therapy with the same drug, we split up prescriptions, after the culture results became known or after a maximum of 72 h. For each prescription, the flow chart (Figure) is read down from top to bottom, except if the records are insufficient for categorization (stop at category VI) or if the criteria for infection (Gamer et al., 1988) are not met (stop at category V, unjustified). AD presciptions can be inappropriate for several reasons at the same time and therefore can be placed in more than one 471-80. (sub)category (categories IV down to II). Prescriptions for therapy are considered inappropriate if errors are made in dose (category Ila), interval (category lib) or route Optimizing antimicrobial therapy. A method (category He). As correct length of therapy is for antimicrobial drug use evaluation frequently arbitrarily established, a duration of / Antimicrob Chemother 1992; 30: 724-727 treatment which differs markedly from that proposed in a leading infectious diseases textSir, Reasons for monitoring antimicrobial drug book is considered either too long (AD) usage are to improve medical care, to limit the development and spread of resistance Table L Antimicrobial Drag (AD) Evaluation Categories, present study and to contain costs. In the most authoritative classification of good antimicrobial drug use, usage is classified as appropriate, probably I. Agree with the use of antimicrobial therapy/ prophylaxis, the prescription is definitely appropriate, inappropriate—less expensive or appropriate. toxic drug recommended, inappropriate— II. The AD therapy/prophylaxis prescription is modified dose recommended and unjustified inappropriate due to: (Kunin, 1973). We modified the original a. improper dosage criteria in order to be able to evaluate each b. improper dosage interval parameter of importance associated with AD c. improper route use (Table I)- To facilitate the selection into III. The AD therapy/prophylaxis prescription is inappropriate due to: (sub)categories, we arranged them in a flow a. excessive length chart (Figure). We used the evaluation in oneb. duration too short month target programmes in our hospital. A AD therapy/prophylaxis prescription is limited number of AD courses were summar- IV. The inappropriate due to: ized by one researcher internist to allow suba. more effective alternative agent (Aa): sequent evaluation by two independent experts specify in infectious diseases, as described by others b. less toxic Aa: specify (VolgeT et aJ., 1988). The results of the evaluac. less expensive Aa: specify d. less broad spectrum Aa: specify tion were presented in an educational report to V. The AD therapy/prophylaxis prescription is prescribers. unjustified: use of any antimicrobial is not The terms prescription, course, prophylaxis, indicated. infections were strictly defined, based on authoritative literature (Moss et at., 1981; Gamer VI. Records insufficient for categorization.
Pedersen, S. S., Prcssler, T., Heiby, N., Bentzon, M.W. & Koch, C. (1985). Imipenem-cilastatin treatment of multi-resistant Pseudomonas aeruginosa lung infection in cystic fibrosis. Journal of Antimicrobial Chemotherapy 16, 629-35. Pedensen, S. S., Presster, T., Jensen, T., Rosdahl, V.T., Bentzon, M.W. Heiby, N. et aJ. (1987). Combined imjpenem/cilastatin and tobramycin therapy of multi-resistant Pseudomonas aeruginosa in cystic fibrosis. Journal of Antimicrobial Chemotherapy 19, 101-7. Plenat, P., Alkhalaf, B. & Micael-Briand, Y. (1988). Prevalence and profiles of plasmidi in Pseudomonas aeruginosa. European Journal of Clinical Microbiology and Infectious Diseases 7, 261-4. Smith, C. L., Kfco, S. R. & Cantor, C R. (1988). Pulsed-fieM gel electrophoresis and the technology of large DNA molecules. In Genome Analysis. A Practical Approach (Davies, K. E., Ed.), pp. 4 1 72. IRL Press, Oxford. Strandvik, B., Malmborg, A. S., Bergan, T., Michalsen, H., Storrosten, O. T. &. WreUind, B. (1988). Imipenetn/cilastatin, an alternative treatment of Pseudomonas infection in cystic fibrosis. Journal of antimicrobial Chemotherapy 21,
