Indian Journal of Medical Microbiology Home 

[Download PDF]
Year : 2011  |  Volume : 29  |  Issue : 3  |  Page : 209--212

Towards a rational antimicrobial testing policy in the laboratory

N Banaji1, S Oommen2,  
1 Department of Microbiology, IGMC and RI, Pondicherry - 605 009, India
2 Department of Microbiology, PSG Institute, Coimbatore, Tamil Nadu - 641 004, India

Correspondence Address:
N Banaji
Department of Microbiology, IGMC and RI, Pondicherry - 605 009


Antimicrobial policy for prophylactic and therapeutic use of antimicrobials in a tertiary care setting has gained importance. A hospital«SQ»s antimicrobial policy as laid down by its hospital infection control team needs to include inputs from the microbiology laboratory, besides the pharmacy and therapeutic committee. Therefore, it is of utmost importance that clinical microbiologists across India follow international guidelines and also take into account local settings, especially detection and presence of resistance enzymes. This article draws a framework for rational antimicrobial testing in our laboratories in tertiary care centers, from the Clinical and Laboratory Standards Institute guidelines. It does not address testing methodologies but suggests ways and means by which antimicrobial susceptibility reporting can be rendered meaningful not only to the treating physician but also to the resistance monitoring epidemiologist. It hopes to initiate some standardization in rational choice of antimicrobial testing in laboratories in the country pertaining to nonfastidious bacteria.

How to cite this article:
Banaji N, Oommen S. Towards a rational antimicrobial testing policy in the laboratory.Indian J Med Microbiol 2011;29:209-212

How to cite this URL:
Banaji N, Oommen S. Towards a rational antimicrobial testing policy in the laboratory. Indian J Med Microbiol [serial online] 2011 [cited 2020 Oct 22 ];29:209-212
Available from:

Full Text


The concept of antimicrobial policy for a hospital has gained recognition over the years. However, it brings to our minds the policy of prophylactic and therapeutic 'use' of antimicrobials. In fact, it is defined as 'a set of rules under which an antimicrobial may or may not be prescribed in a setting'. It logically precedes that there is a need to formulate a set of rules under which an antimicrobial may or may not be tested in a setting, that is, a rational antimicrobial testing policy.

With the advent of newer antimicrobials and bacterial enzymes inactivating numerous antimicrobials, it is imperative that susceptibility testing accounts for both. It also needs to include input from the infection control committee, the pharmacy and the therapeutic committee, besides the microbiologist. Many laboratories and hospitals still make do without the current Clinical and Laboratory Standards Institute (CLSI) guidelines mainly due to financial constraints or non-availability of the document, adding to the confusion of choosing the appropriate antimicrobial. It is the microbiologist who needs to bear in mind the guidelines for the choice of antimicrobials to be tested. Rational testing choices, when combined with strict adherence to recommended methodology and quality control can generate meaningful epidemiologic data for rational therapeutic policies. Broadly speaking, a rational susceptibility testing protocol needs to take into account and address the following:

The class and number (one or more) of antimicrobialsThe identity of the bacteriumThe site of infectionDefine routine and selective reportingDetermine special requests by physiciansDetect resistance enzymesInclude therapy-related suggestions

Antimicrobial classification

Pharmacologic considerations of antimicrobials classifies antimicrobial agents as β-lactams, glycopeptides, tetracyclines, quinolones, macrolides, aminoglycosides, folate pathway inhibitors, lipopeptides and single-drug classes.[1]

Each class may include sub-classifications and a variety of agents. It is appropriate that our test balances the various classes and subclasses and does not exhibit preferential bias towards any class or choose multiple agents from a single class. However, knowledge of a specific agent within a class being available in the hospital formulary will help.

For example, while choosing different generation cephalosporins, a representative antimicrobial may be chosen keeping in mind the current recommendations. One may test only cefuroxime amongst the 2nd generation cephalosporins, while the option to test between ceftriaxone and cefotaxime may be skewed towards testing cefotaxime so as to simultaneously detect the CTX-M production, especially if combined with cefotaxime and clavulanic acid. Similarly, while choosing from quinolones and macrolides, a representative agent may be chosen, in accordance with the current recommendations.

The authors would further recommend testing and reporting of antimicrobials within a specific drug class as per the established hierarchy of activity, for example, reporting of ampicillin should precede over ampicillin-sulbactam, a 1st and 2nd generation cephalosporin should be followed by 3rd and 4th generation cephalosporins, aztreonam and carbapenems, in that order. This ensures that misreporting of sensitivity patterns does not occur, for example, reporting a lower antimicrobial as sensitive while the higher is resistant. One may also be able to pick up emerging and newer mechanisms of resistance.

The number of antimicrobial agents being tested is restricted by the size of the surface of the testing medium (not more than 6 on a 90 mm diameter petridish). Often laboratories cannot conform to this standard on a daily basis, leading to misinterpretation or over-representation of a particular sensitivity. With the advent of newer antimicrobial automated systems which generate an extensive list of antimicrobials tested, often from a single class of agents, release of some antimicrobial results may be restricted.

The concentration of the antibiotic disc also plays an important role in interpretation of results. Some antibiotics are available at different concentrations, using the correct one is of utmost importance. Inadvertent placement of 10 μg of gentamicin instead of 120 μg can lead to mis-reporting of resistance in case of Enterococci spp.

Identity of the bacterium

The gram reaction of the bacterium combined with the antibacterial spectrum of the agent is singularly the most important consideration.

For example, it is irrational to test β-lactams and β-lactamase inhibitors, carbapenems for methicillin-resistant Staphylococcus aureus (MRSA), cephalosporins, cotrimoxazole and clindamycin for enterococci, aztreonam for gram-positive bacteria, linezolid and vancomycin for gram-negative bacteria (exception being vancomycin for Chrysobacterium spp.) 1st and 2nd generation cephalosporins for Pseudomonas spp., Salmonella spp. and Shigella spp., aminoglycosides for Salmonella spp. and Shigella spp., tigecycline for Pseudomonas and Proteus spp. [2] It may be preferable to restrict reporting of piperacillin-tazobactam for ESBL producers in serious patients despite no recommendations from the CLSI. [3] This selective non-reporting should be extended to Enterobacter, Citrobacter and Serratia spp., which become rapidly resistant to in vitro susceptible 3rd generation cephalosporins when on therapy.[2] Many a times an antimicrobial may be selectively reported, for example, rifampicin, in a patient with an MRSA infection, of prostheses with a comment that the agent cannot be used singly but in conjunction with another effective antimicrobial.[2]

On isolation of Salmonella spp., besides nalidixic acid as a marker for increased ciprofloxacin MICs, it is important to test for chloramphenicol and azithromycin, with the emergence of strains sensitive to chloramphenicol and increasing therapeutic use of azithromycin.

It is often impractical to include an antimicrobial from the CLSI list as it is more important to take into account the current antimicrobial resistance trends and prevalence in a given community or hospital prior to setting up the antimicrobial panel. For example, CLSI recommends testing for ampicillin in case of Enterobacteriaeceae but it is often found that resistance to ampicillin is over 73% in E. coli in low income country settings. [4] Moreover, commonly isolated species of Klebsiella, Citrobacter, Enterobacter, Serratia, Morganella, Providencia and P vulgaris are inherently resistant to ampicillin. Decision on incorporating testing of this antimicrobial for all Enterobacteriaceae isolates as a routine basis or to choose it on a case-to-case or site-to-site basis has to be made by the reporting microbiologist as per the prevalent susceptibility patterns of the bacteria.

Site of the infection

For an antimicrobial to be effective it has to penetrate to the site of infection, in adequate concentration. An antimicrobial that does not reach the bacterium cannot act on it unlike those that achieve high concentrations at the site, for example, in urine. Moreover, the pH of the environment also determines the efficacy of the antibiotic at the particular site, for example, Proteus spp. might be sensitive to nitrofurantoin in vitro but is rendered ineffective in the increased alkaline environment due to urinary tract infection with these bacteria. [5]

An antimicrobial chosen for the same bacterium, for example, Haemophilus influenzae when isolated from two distinct sites, namely, the lung (sputum) and cerebrospinal fluid (CSF), varies in action depending on its ability to cross the blood-brain barrier. Thus antimicrobials, such as 1st or 2nd generation cephalosporins (except cefuroxime), clindamycin, tetracycline and macrolides, although effective in respiratory infections, should not be reported for meningitis isolates, while daptomycin should not be reported for lower respiratory tract isolates. Hence, it is of extreme significance to have different panel of antibiotics for specific body sites, such as CSF, urine and stool.

Define routine and selective testing

It is desirable to lay down a panel of testing as the first line of therapy and only on determination of in vitro resistance to the first line is the second line tested. This obviates the use of broad spectrum antimicrobials when not warranted.

For example, reporting of carbapenems may be restricted to ESBL and AmpC producers only. Also, linezolid and glycopeptides may not be reported in a methicillin-sensitive strain of S aureus.

This criterion is often debatable vis-a-vis the turnaround time. In laboratories not using automated systems there is a delay of an extra day to identify the susceptibility patterns of these selective agents, which may be a case of too little too late. Hence an in-house policy, wherein the testing of the first, second or third line antimicrobial for select cases (e.g. patients on ventilators or those with prolonged hospital stay) could be done simultaneously, but their results may be withheld in case the first line appears to be susceptible.

Determine special requests by physicians

As we define selective testing, in the event of in vitro resistance to second line agents or failure of therapy, it is permissible to test for an agent not included in the laboratory's panel but requested by the physician. At this point one should keep in mind the cluster of agents for which cross-resistance or cross-susceptibility are nearly complete, for example, test for ceftriaxone on request need not be done in case test for cefotaxime has been carried out in the first place.

A scenario could also arise due to co-morbidities in the patient restricting the use of certain antimicrobials, therefore necessitating the use of those not routinely tested, for example, a pan-resistant Acinetobacter spp. sensitive only to colistin, by itself can contribute to increased morbidity. Relatively safer alternatives, such as tigecycline, may be offered to the clinician by the microbiologist. Therefore, it is of utmost importance that, microbiologists leave the bench for the bedside and have an ongoing dialogue with the clinicians, especially while treating resistant infections.

 Detect resistance enzymes and resistance genes

Most of the inactivating enzymes produced by bacteria can be detected phenotypically in vitro by positioning the antimicrobial discs in a standardized manner. This makes the reporting of resistance enzymes easier and time saving. The phenotypically resistant bacterium may be preserved for confirmation of genotypic resistance for epidemiologic purposes. [Table 1] summarizes the standardized tests that must be included in susceptibility testing. [2]{Table 1}

Therapy-related comments

The treating physician may find it challenging to choose an appropriate antimicrobial in the presence of resistance enzymes and the discovery of newer agents. Including a therapy-related comment in the susceptibility report would draw their attention to rational choices.

Apart from routine comments, such as 'enterococci are inherently resistant to cephalosporins and cotrimoxazole' or 'carbapenem is the drug of choice for extended spectrum lactamase infections', specific therapy-related comments may be added. For example, when presented with a nalidixic acid-resistant but ciprofloxacin-susceptible isolate, a comment may be inserted, namely 'in view of nalidixic acid resistance, ciprofloxacin if used is recommended at a higher dose than normal'.


The CLSI recommends a wide array of antimicrobials that may be tested against nonfastidious bacteria. The choice of antimicrobials to be tested being the microbiologist's, one needs to adopt rational practices at the laboratory bench as well as while reporting. This rational testing policy can be achieved only if consider the class of antimicrobial, identity of the bacterium and resistance enzyme produced if any, site of infection, and report selectively alongwith recommendations for therapy.[9]


1Chamber HF, Deck DH. In: Basic and Clinical Pharmacology. 11 th ed. Katzung BG, Masters SB, Trevor AJ, editors. New Delhi: Tata-McGraw Hill; 2009.
2National Committee for Clinical Laboratory Standards (NCCLS): 2011, Performance standards for antimicrobial susceptibility testing; twenty-first informational supplement. NCCLS document M100-S21, vol. 31, no. 1. NCCLS, Wayne, PA, USA, Jan 2011.
3Paterson DL, Bonomo RA. Extended Spectrum â-Lactamases: A Clinical Update. Clin Microbiol Rev 2005;18:657-86.
4Jordi V, Tibor P. Update on Antibacterial Resistance in Low-Income Countries: Factors Favoring the Emergence of Resistance. Open Infect Dis J 2010;4:38-54.
5Mackie and Mc Cartney's Practical Medical Microbiology: 15 th ed. Edinburgh, UK: Churchill Livingston; 1996.
6Thomson KS. Controversies about Extended-Spectrum and AmpC -Lactamases. Emerg Infect Dis 2001;7:333-6.
7Singhal S, Mathur T, Khan S, Upadhyay DJ, Chugh S, Gaind R, et al. Evaluation of Methods for AmpC â-Lactamases in Gram Negative Clinical Isolates from Tertiary Care Hospitals. Indian J Med Microbiol 2005;23:120-4.
8Picão RC, Andrade SS, Nicoletti AG, et al. Metallo-beta-lactamase detection: Comparative evaluation of double-disk synergy versus combined disk tests for IMP-, GIM-, SIM-, SPM-, or VIM-producing isolates. J Clin Microbiol. 2008 Jun;46(6):2028-37. Epub 2008 Mar 5.
9González-López JJ, Coelho A, Larrosa MN, Lavilla S, Bartolomé R, Prats G. First detection of plasmid-encoded blaOXY beta-lactamase. Antimicrob Agents Chemother 2009;53:3143-6.