Antibiotic Allergy-clinical Relevance and Consequences Adrian Y. WU Department of Medicine, Queen Mary Hospital Antibiotic allergy is one of the most commonly encountered problems in clinical practice. The symptoms may range from mildly annoying to life threatening. Drug allergies can become a major problem for patients with the multiple drug allergy syndrome, and more so since the emergence of antibiotic resistant organisms often limits the choice of antibiotics that can be used. β-lactam antibiotics (penicillins and cephalosporins) are the most common causes of drug allergy, and also the best studied. This article reviews the management of patients with allergies to β-lactam antibiotics.
altering the antigenic determinants. It is thought that the genetic predisposition to producing active drug metabolites that react with serum proteins may affect the likelihood of becoming sensitised to drugs. Allergies to β-lactam antibiotics may involve many different immunological mechanisms, each producing a unique clinical presentation. These include anaphylactic reactions, Stevens-Johnson syndrome, erythema multiforme, morbilliform drug rashes, fixed drug eruptions, toxic epidermal necrolysis and many organ specific drug reactions.
Allergic reactions have certain characteristics that help distinguish them from other types of adverse drug reactions. These include:
Diagnosis of Drug Allergy A detailed history is of utmost importance when diagnosing drug allergies. The symptoms, drug history and the timing of reaction in relation to drug administration will help to pinpoint the offending drug and the type of reaction. A previous history of drug allergies or a family history of drug allergies is important since these factors greatly predispose the patient to developing allergies to other drugs. Sullivan suggested that in people allergic to penicillin, there was at least a tenfold increase in reaction rate to non-beta-lactam antibiotics. Other atopic diseases do not predispose the patient to drug allergies. A history of concomitant illness is also important. Quite frequently, viral rashes are mistaken as drug rashes caused by antibiotics given to treat the viral illness. Drugs given under some circumstances may also lead to a nonimmunologically mediated rash. For example, ampicillin/ amoxicillin given during infection caused by Epstein-Barr virus or cytomegalovirus, or when given to a patient with acute lymphocytic leukaemia, may lead to a drug rash. Similarly, penicillin given to a patient with syphilis may lead to the Jarisch-Herxheimer reaction. Under other circumstances, these patients will not react to these antibiotics. AIDS also predisposes patients to drug allergies. 50% of AIDS patients treated with trimethoprimsulfamethoxazole develop an allergic rash.
1. Prior sensitisation. Allergic reactions to drugs, like other immunologically-mediated reactions, are acquired. The likelihood of sensitisation depends on the genetic predisposition of the individual, the amount of drug given, and the length, frequency and route of exposure. In general, longer or more frequent exposure will more likely lead to sensitisation, and cutaneous is more likely to sensitise than intravenous, which in turn is more likely to sensitise than oral administration. Some cases of allergic reactions to drugs given apparently for the first time may be due to cross-sensitisation or previous occult exposure. 2. Consistency in symptoms. Allergic reactions lead to a limited and well-characterised set of symptoms. Different allergic mechanisms lead to different sets of symptoms that can often be distinguished clinically. Repeated exposure will invariably reproduce similar symptoms in any given individual. 3. The severity of symptoms is sometimes unrelated to the dose of drug given. Even minute amounts of a drug far below therapeutic dose can lead to a fatal reaction. This fact makes the use of a "test dose" a very dangerous proposition under some circumstances. 4. Temporal relationship. A drug allergy reaction is always temporally related to drug exposure. The timing of the reaction is dependent on the mechanism of the reaction. For example, an IgE-mediated reaction usually occurs within 15 minutes to one hour of exposure.
Physical examination should focus on the skin since drug allergies frequently involve the skin. It is important to distinguish between a morbilliform rash and an urticarial rash; the former is usually benign whereas the latter indicates the presence of IgE antibodies, a risk factor for anaphylaxis. Mucosal involvement may indicate StevensJohnson syndrome or toxic epidermal necrolysis, both serious conditions. Chest examination may reveal wheezing secondary to bronchospasm, or stridor caused by laryngeal oedema.
As β-lactam antibiotics are small molecules incapable of sensitising the immune system on their own, they must act as haptens by binding to serum proteins, thereby 3
Laboratory tests may be helpful in diagnosing organspecific allergic reactions such as haemolytic anaemia, hepatitis, nephritis etc. In situations where anaphylaxis is suspected, a serum tryptase level drawn within 3 to 4 hours of the reaction may help diagnose anaphylaxis but sensitivity is low and the assay is expensive. Blood tests for penicillin-specific IgE is not reliable because of insensitivity and lack of an appropriate minor determinant reagent.
Table 1. Skin-test reagents for penicillin allergy •
Major determinant: - Penicilloyl polylysine (Prepen)
Minor determinant mix (MDM) - Potassium penicillin G - Sodium benzylpenicilloate - Benzylpenicilloyl-n-propylamine
1 mg/ml 0.01 M 0.01 M
Management of β -lactam Allergies These drugs can cause many different types of allergic reactions, from minor drug rashes to fatal anaphylaxis. Patients with a vague history of allergy to penicillin are frequently seen in clinical practice. Mendelson et al. skin tested 240 children and adolescents with a history of penicillin or amoxicillin allergy and found only 8.75% to have positive skin tests to these antibiotics. Similarly, Macy et al. skin tested 348 patients referred to their allergy clinic with a history of penicillin allergy, and only 60 (17.2%) were found to have positive skin tests. These results suggest that without skin testing, the majority of patients labelled as penicillin allergic would be avoiding these drugs unnecessarily.
Table 2. Type and specificity of positive penicillin skin test reactions in 384 patients Reagent PrePen Penicilloate Penilloate Penicilloate + penilloate Penicillin + PrePen Penicillin Amoxicillin All other combinations*
Before starting a patient on β-lactam antibiotics, the physician must ascertain whether there is a history of any reaction to the drug, and the severity of any reaction. In general, trivial drug rashes are not contraindications to the use of these drugs if they are absolutely needed, since the symptoms of the reaction can often be controlled to allow the patient to complete the course of treatment. However, these drugs must be absolutely avoided if the patient has a history of Stevens-Johnson syndrome, toxic epidermal necrolysis or anaphylaxis. The first two entities, both serious conditions, should be easily recognisable. However, the patient may not be able to distinguish a morbilliform or maculopapular rash from an urticarial rash. Pichichero reported that 22% of patients with a history suggestive of a non-IgE mediated rash nevertheless had positive skin tests to penicillin.
Puncture (no. of patients)
Intradermal (no. of patients)
24 2 6 1 6 2 1 12
*All of these subjects had positive responses to at least one of the commercially available skin test materials: amoxicillin (3 of 15), PrePen (11 of 15), and penicillin (1 of 15). From: Macy E, Richter PK, Falkoff R, et al. J Allergy Clin Immunol 1997; 100:586-91.
Therefore, if these minor determinants were not available, 20% of the penicillin allergic patients would have been missed. Interestingly, just one out of 60 patients reacted to amoxycillin skin test only, indicating that side-chain sensitivity to amoxycillin is an uncommon occurrence. When skin testing is carried out using major and minor determinants and amoxicillin, a very high negative predictive value is achieved. In Macy's study, only 5.2% of patients with negative skin tests developed a reaction on oral challenge, and all were mild. In another study of 247 children referred for allergic reactions to penicillin, amoxycillin or cephalosporins, 34% of them had a positive skin test. 2.5% of all skin test negative patients had a reaction on drug challenge, all mild. 1.7% of all skin test and challenge negative patients had a reaction to a subsequent course of antibiotics, all mild. These studies showed that skin tests have excellent specificity and negative predictive value. Skin testing is also very safe; in the various published reports, none of the patients with a history of severe anaphylaxis to penicillin suffered any adverse reactions to skin testing, although such patients have a higher incidence of skin test positivity. In a study of 58 patients with a history of severe anaphylaxis to
Skin testing has proven to be very reliable in predicting the risk of anaphylactic reactions in patients with a history of penicillin reactions. Penicillin is metabolised to major and minor determinants. The major determinant is penicilloyl and the minor determinants include penicilloate, penilloate and benzylpenicillin. Skin test reagent for the major determinant penicilloyl polylysine is available (Prepen). Although the minor determinants are responsible for a substantial proportion of serious anaphylactic reactions, no skin test reagent for penicilloate and penilloate is commercially available. Sodium amoxicillin is usually included in the panel since some patients react to its side-chain instead of the β-lactam group (see Table 1). Other investigators believe that ampicillin should also be included. In Macy's study, 12 out of 60 (20%) penicillin allergic patients reacted to penilloate and/or penicilloate only (see Table 2).
Vol. 8 No. 4
penicillin, the authors concluded that skin testing with 500U of penicillin and major determinant is relatively reliable, safe and practical even in individuals extremely sensitive to penicillin. Mendelson recommended that the initial starting dose for skin testing should be reduced by a factor of 100 for patients highly sensitive to penicillin.
In summary, if a patient with a history of penicillin allergy requires antibiotics, substituting a chemically unrelated antibiotic should be the first course of action. If penicillin is absolutely required, skin testing with major and minor determinants and amoxicillin will reliably assess the risk of a serious reaction. If the patient is skin test negative, a test dose of oral amoxicillin should be given under medical supervision and the patient observed for one hour before starting the course of treatment. The patient must understand that skin testing will only predict anaphylactic reactions and minor skin rashes may still occur. If the patient is skin test positive, an alternative antibiotic can usually be found. Under exceptional circumstances where penicillin is absolutely necessary, e.g. tertiary syphilis, the patient can be desensitised. Desensitisation is achieved by either oral or intravenous methods. In both methods, the patient is given minute doses of the β-lactam drug starting at 0.05 mg orally or 0.01 mg intravenously. The next higher dose is given after a 15-minute interval if there is no reaction. The dose is gradually increased over the next 4 hours until the patient can tolerate a full therapeutic dose of the drug. This should be carried out by an allergist experienced with the procedure under strict medical observation. After desensitisation, treatment must begin right away and must not be interrupted. After a successful course of treatment, the patient must be desensitised again just before subsequent courses of penicillin in the future.
Skin tests for penicillin should include intradermal injections, since standard skin prick tests alone are less sensitive. Skin prick testing should however be performed first, especially in suspected highly sensitised patients, with intradermal testing to follow only if the skin prick test is negative. Appropriate starting solutions for skin testing include penicilloyl polylysine (Schwarz Pharma), penicilloate (0.01 M), penilloate (0.01 M), benzylpenicillin (1 mg/ml) and amoxycillin (1 mg/ml). Since penicilloate and penilloate are not commercially available, they must be self-prepared and stored in the freezer (Ref: Macy E et al. J Allergy Clin Immunol 1997;100:586-91). 0.02 ml of the solutions are injected intradermally and the wheal and flare reaction is read after 15 minutes. Histamine and saline are used as positive and negative controls respectively. Positive responses consist of a wheal of 5 mm or more in diameter with surrounding erythema greater than the wheal, a negative response to control solution and a positive response to histamine. The incidence of skin test positivity declines with time following the initial reaction, and is down to less than 10% after 10 years. This is because of waning penicillinspecific IgE levels in the absence of antigenic stimulation. A new course of penicillin in this situation may restimulate IgE production by memory B cells. Many of the history positive but skin test negative patients are therefore at risk of resensitization when given therapeutic courses of penicillin in the future. The reported rates of resensitization vary from 1% to 16%.
Prevention of Drug Allergies and the Multiple Drug Allergy Syndrome Patients allergic to penicillin are 10 times more likely to become allergic to other drugs. This susceptibility not only applies to anaphylactic type reactions, but also extends to drug rashes, exfoliative dermatitis, toxic epidermal necrolysis and Stevens-Johnson syndrome. In a study of 120 patients with antibiotic allergies, 19% were allergic to at least one other antibiotic; 15.8% reacted to three or more antibiotics. 42% of those patients allergic to two or more antibiotics were allergic to NSAIDs, in contrast to only 18% of those patients allergic to one antibiotic. It is therefore clear that some patients are highly susceptible to developing allergies to a large number of drugs.
The issue of cross-reactivity between penicillins and cephalosporins is less clear. Both classes of drugs contain a β-lactam ring structure. Increased hypersensitivity to first-generation cephalosporins exists in those patients who have histories of penicillin allergy, but cross-reactivity between penicillin and second- and third- generation cephalosporins appears to be low. Cross-reactivity between penicillins and cephalosporins with similar side chains appear to be more frequent. If a cephalosporin is to be used in a penicillin allergic patient, inclusion of the drug in the skin test panel is advisable with the understanding that the negative predictive value of such a test is unknown. In a patient with known allergy to a cephalosporin, substituting another cephalosporin with a different side chain structure is usually safe.
The responsibility for preventing drug allergies rests squarely on the shoulders of medical practitioners. Frequent, short courses of antibiotics are responsible for sensitising many patients to these drugs. We must educate our patients and ourselves when antibiotics are truly indicated. By avoiding inappropriate use, the incidence of drug allergies will decrease, which will ultimately benefit our patients and ourselves.
How Atypical is Atypical Pneumonia? Patrick C. Y. WOO Associate Professor, Division of Infectious Diseases Department of Microbiology, The University of Hong Kong, Queen Mary Hospital
What is Atypical Pneumonia?
the term "walking pneumonia" is coined. Of course, some cases could be very severe. Since the causative agents are not the typical pyogenic bacteria, the pneumonia does not respond to beta-lactams, such as amoxicillin, amoxicillin-clavulanate, penicillin G, or even the often used "big-guns" such as the carbapenems (e.g. imipenem and meropenem), anti-pseudomonal penicillin (e.g. piperacillin-tazobactam) third and fourth generation cephalosporins (e.g. cefepime), of which one of the major mechanisms of bacterial killing is through inhibition of cell wall murein synthesis, which somehow trigger the autolytic activities of murein hydrolases, hence disrupting the delicate balance between murein synthetic and degradative activities.
Atypical pneumonia is pneumonia caused by infectious agents other than the "typical" pyogenic bacteria (Streptococcus pneumoniae, Haemophilus influenzae, Staphylococcus aureus, Klebsiella pneumoniae, Moraxella catarrhalis). It is caused by "atypical" agents such as Mycoplasma pneumoniae, Chlamydia species, Legionella species, and respiratory viruses (including mainly influenza viruses A and B, parainfluenza viruses 1, 2, and 3, adenoviruses, and respiratory syncytial virus). In Hong Kong, Mycoplasma pneumoniae causes the majority of the cases, followed by Chlamydia pneumoniae and more importantly in Hong Kong, Mycobacterium tuberculosis. Clinically, the episode starts as a mild respiratory illness with persistent symptoms for more than 10 days before presentation. The clinical symptoms and signs are often disproportional to the chest radiograph changes. Thus
A summary on the at-risk groups, associated symptoms and signs, and complications of the better known causes of atypical pneumonia are shown in Table 1.
Table 1. Summary of the major causes of atypical pneumonia and the corresponding at-risk groups, associated symptoms and signs, and complications Atypical agents
At risk group
Other associated signs and symptoms
Young children and adults
Prodrome 3-4 week
Headache, chills, myalgia, arthralgia, skin rash, bullous mucosal eruptions
Haemolytic anaemia, idiopathic thrombocytopenic purpura, disseminated intravascular coagulation, hepatitis, pancreatitis, myocarditis
Often no fever
Hoarseness and sinusitis
Myocarditis, erythema nodosum
Neonate, immuno- Cough and tachypnea suppressed host
Diffuse pulmonary infiltrate
Eosimphilia, increase IgM and IgG
Chlamydia psittaci Contact with birds Headache and bradycardia
Horder's spots, splenomegaly, deranged LFT
Contact with animal parturient tissue
Endocarditis, granulomatous hepatitis
Exposed to High fever and rigor contaminated water and cooling system
Abrupt onset of chills and headache
Diarrhoea, vomiting, Persistent infiltrate for weeks to abdominal pain, confusion, months. Bacteraemia up to 40% of impaired renal function test, all cases white cell count up to 30
Vol. 8 No. 4
When to Suspect Atypical Pneumonia or Include Atypical Pneumonia as one of the Differential Diagnosis?
Investigations and Treatment for In-patients with Suspected Atypical Pneumonia Most "atypical pneumonia" is empirically treated on the basis of clinical and/or epidemiological grounds. Investigations of in-patients suspected to have atypical pneumonia could include: 1. Nasopharyngeal aspirate for direct detection of viral antigens and virus culture. 2. Sputum for Legionella culture or urine for Leginoella antigen detection. 3. Serum for detection of antibody against Mycoplasma, Chlamydia, and Legionella. 4. If tuberculosis is suspected, early morning sputum for acid fast bacilli on three consecutive days.
1. A n y c a s e s o f s e v e r e c o m m u n i t y - a c q u i r e d pneumonia. Since atypical pneumonia can be severe, it should be treated empirically in any cases of severe community-acquired pneumonia. 2. Patients with respiratory tract symptoms that lasted for more than ten days. 3. Patients with pneumonia that do not respond to beta-lactam treatment. 4. Gram smear of sputum is positive for white blood cells, but no pyogenic bacteria that cause the typical pneumonia can be recovered from the sputum, or the culture results show "commensals only".
Treatment of patients with suspected atypical pneumonia should include a macrolide or doxycycline. In patients with confirmed influenza pneumonia, amantidine/ rimantidine or a neuraminidase inhibitor could be given.
ABCs of Antimicrobial Resistance Margaret IP Associate Professor, Department of Microbiology The Chinese University of Hong Kong, Prince of Wales Hospital Background
Table 1. Emergence of antibiotic resistance
Shortly after the discovery of penicillin by Alexander Fleming and its widespread use as an antimicrobial agent in the 1940s, bacteria resistant to penicillin began to appear. Pathogenic bacteria are living organisms and yearn to survive. Under the pressure of antibiotic usage, bacteria have to find ways to resist these antibiotics. This was first exemplified by penicillinase-producing Staphylococcus aureus. Penicillinase is an enzyme that breaks the penicillin compound and inhibits its action on the bacterial cell wall. Often, these enzymes are naturally present in some bacteria and are part of their defense mechanism against other organisms. These enzymes are encoded by DNA in genes found in certain bacteria. Pathogenic organisms acquire these genes and begin to express the resistant mechanisms. As new classes of antibiotics were introduced, various resistance mechanisms to specific target sites on different bacteria appeared successively.
1950s 1965 1967 1976 1983 1988 1998 2001
Penicillin-resistant Staphylococcus aureus Ampicillin-resistant E. coli Penicillin-resistant Streptococcus pneumoniae Penicillin-resistant Neisseria gonorrhoea Extended-spectrum β-lactamase Klebsiella pneumoniae Vancomycin-resistant enterococci (VRE) Vancomycin-resistant Staphylococcus aureus (VRSA) Carbapenam-resistant K. pneumoniae
enzymes are present in Gram-positive and Gram-negative organisms and particularly play an important role in drug resistance among Gram-negative bacteria. Alteration of Target Site Penicillin binding proteins (PBP) are binding sites for the β-lactam antibiotics. Changes or absence of these binding proteins thus leads to drug resistance. A well-described example is in the case of penicillin-resistant Streptococcus pneumoniae whereby the PBPs have altered and have reduced affinity to penicillins and cephalosporins.
In this millenium, we are facing the challenge of a battle against bacterial infections caused by multidrug-resistant bacteria. For example, Streptococcus pneumoniae is now resistant to a number of commonly used orally administered antibiotics and Klebsiella pneumoniae may be resistant to multiple antibiotics including second and third generation cephalosporins, aminoglycosides and even carbapenams. Some of these bacteria are listed in Table 1 and the list is by no means exhaustive.
Impaired Permeability In Gram-positive bacteria, the outer membrane is penetrated by passive diffusion whereas in Gram-negative bacteria, entry of antibiotics is facilitated by protein channels (porins). Changes in the porin channels affect drug accumulation into the cell. An example is the loss of the outer membrane protein (OmpF) in E. coli leading to resistance to fluoroquinolones.
Mechanisms of Antimicrobial Resistance Antibiotics act at various sites in the bacterium to inhibit growth. These sites are often enzymes or essential proteins and most antibiotics have to pass through the cell wall or outer membranes to reach these targets. A number of mechanisms are involved (Table 2).
Efflux Once the antibiotic entered into the bacterial cell, it is possible for the antibiotic to be extruded out of the cell by specific pumps. These are complex, energy-driven mechanisms and involve channels that drive 'toxic' ions out of the cell. An example is the MexA-MexB-OprK efflux system in Pseudomonas aeruginosa that leads to multidrug resistance to tetracycline, chloramphenicol, and fluoroquinolones.
Enzymatic Inactivation β-lactamase enzymes hydrolyze the β-lactam rings in penicillins and cephalosporins to inactivate their action on bacterial cell wall. These Table 2. Mechanisms of antimicrobial resistance Antimicrobial resistance mechanisms
Destruction or inactivation of the antibiotic
β-lactamases e.g. Klebsiella sp.
Alteration of target site to reduce binding of antibiotic to the target
Penicillin binding protein (PBP) e.g. Streptococcus pneumoniae
Impermeability – blockage of mechanism by which antibiotic enters cell
Outer membrane protein (OmpF) in E. coli
Efflux – mechanism by which antibiotic cell
MexA-MexB-OprK efflux system in is driven out of Pseudomonas aeruginosa
Alternative metabolic pathway to bypass agents
Dihydrofolate reductase overproduction in E. coli
Vol. 8 No. 4
Bypass of Metabolic Pathway Resistant organisms may synthesize new enzymes, e.g. trimethoprim-resistant dihydrofolate reductase, as well as the normal drug sensitive enzyme to bypass the metabolic pathway in the production of tetrahydrofolate for DNA synthesis.
organisms' susceptibility patterns are essential to guide clinicians to the best choice of treatment. In a study performed recently by Ling J et al., 2003, her group examined 4741 specimens obtained from 3977 patients by 89 general practitioners during a sixteenmonths' period from July 2000 in Hong Kong. Specimens were collected from patients suspected of infection. The specimens included swabs from various sites, midstream urine, sputum and stool. The commonly isolated organisms from each category of sites and the frequency with which these organisms were isolated are listed (Table 3). The study, in general, supported the data produced
Incidence of Pathogens from the Community Antibiotics are frequently prescribed by general practitioners. These are often administered before the knowledge of the bacterial culture and antibiotic susceptibility results. Recent local data on both the distribution of causative organisms in different sites of infection commonly seen in the community and the
Table 3. Common bacterial pathogens from various sites (based on specimens from general practitioners in Hong Kong, data adapted from Ling J, et al. JAC2003) Specimen type
Positive culture / No. of specimen
No. of Isolates (% +ve culture)
β-haemolytic streptococci (Group C and G) Group A Streptococcus Staphylococcus aureus Streptococcus pneumoniae
136 (39%) 51 (15%) 38 (11%) 12 (3%)
Haemophilus influenzae Streptococcus pneumoniae Pseudomonas aeruginosa
67 (33%) 28 (14%) 23 (11%)
Staphylococcus aureus Escherichia coli Anaerobes Proteus/Morganella spp. Pseudomonas aeruginosa
50 (27%) 10 (5%) 10 (5%) 9 (5%) 8 (4%)
Escherichia coli Proteus/Morganella spp. Enterococcus spp. Coagulase negative Staphylococcus Klebsiella spp.
207 (65%) 23 (7%) 20 (6%) 19 (6%) 13 (4%)
Salmonella sp. Vibrio parahaemolyticus
22 (55%) 15 (38%)
Candida albicans* Neisseria gonorrhoea Staphylococcus aureus
109 (19%) 47 (8%) 12 (2%)
Antimicrobial prescribing – Checklist: • Clinical diagnosis : indication for antibiotic • Appropriate clinical workup • Likely pathogens • Choice of antibiotics - Any contraindications and allergies - Dosage and route of administration - Duration of treatment • Modifications according to clinical response and culture results Choice of antibiotics: Site of infection Pharyngitis
Empirical choice (adult) Penicillin V 250-500 mg qid 10 days if Group A Streptococcus suspected/confirmed.
Skin/soft tissue infection
Cloxacillin 500 mg qid (+/- amoxicillin 500 mg tds) or amoxicillin/clavulanate 375 mg tds 7 days
Nitrofurantoin 50 mg qid 7 days or fluoroquinolone e.g. ofloxacin 200-400 mg bd 3 days
Vaginal candidiasis common. Rule out STD as indicated and evaluate sexual partner. For presumed gonococcal urethritis, spectinomycin 2 g IM x 1 or ceftriaxone 125 mg IM x 1.
None, often self-limiting
*Yeast and not bacterium
by the Antibiotic Resistance Surveillance initiated in July 1999 by the Department of Health and published on its website http://www.info.gov.hk/dh/publicat/index.htm.
sensitive to ampicillin in the presence of clavulanic acid or sulbactam. In Haemophilus influenzae, sensitivity to ampicllin and cefaclor were 61% and 53% respectively. Sensitivity rates remained high (98%) for ampicillin/ clavulanate, cefuroxime and fluoroquinolone.
Antimicrobial Susceptibility Patterns The percentage of antimicrobial susceptibility for these organisms are listed in Table 4. For Gram-positive bacteria, 3 isolates of MRSA were obtained and the remainder Staphylococcus aureus isolates were all susceptible to cloxacillin and the susceptible rates of trimethoprim (97%) and ofloxacin (98%) were high. 90% of β-haemolytic streptococci were sensitive to penicillin. The 10% of β-haemolytic streptococci noted to be of reduced susceptibility to penicillin were found in Group C and G streptococci only and their role in pharyngitis is doubtful and controversial. All Group A streptococci remained sensitive to penicillin. For Streptococcus pneumoniae, 19% of the isolates were sensitive to penicillin. The remainder 81% of the isolates were all of reduced susceptibility to penicillin with minimum inhibitory concentration in the range of 0.12-0.25 ug/ml. In addition, the susceptible rates to tetracycline, cotrimoxazole, and clarithromycin were exceedingly low, of 13%, 13% and 10% respectively. All enterococci were sensitive to ampicillin and 96% were sensitive to fluoroquinolones.
Antimicrobial Prescribing – Choice of Antibiotics Antibiotics should only be prescribed when there is a high suspicion of bacterial infection. Once the indication is made, appropriate workup with microbiological investigations should be performed. The choice of antibiotics depends on the site of infection, the likely pathogen, and the antimicrobial susceptibilities. The rationale for the empirical therapy for a number of community acquired bacterial infections are produced, as follows, and are based on the local data obtained from Ling et. al., 2003 and the Department of Health. The appropriate use of these agents requires correct dosage and duration of therapy and should be modified accordingly when the culture and sensitivities become available. Acute Pharyngitis As Group A streptococcus is the commonest bacterial pathogen, penicillin V for 10 days would be the drug of choice. Erythromycin may be used in penicillin allergic patient but the sensitivity rate is relatively low, of 67%, among these community strains. However, acute pharyngitis is also often of viral in origin, particularly in children with other associated URI symptoms.
Among Gram-negative bacilli, fluoroquinolone sensitivity ranged between 78-95%. Only 45% of E. coli is susceptible to oral cefuroxime (axetil), but 84% was
Table 4. Antimicrobial susceptibility patterns of common bacterial pathogens (based on specimens obtained from general practitioners in Hong Kong, data adapted from Ling J, et al. JAC2003) Organism (No. of isolates) Mt
Percentage of Antibiotic Susceptible (%) Amp Sxt Tetra Oflox/Levo
Gram positive cocci Staphylococcus aureus (159) β-haemolytic streptococcus (200) (Group A, C and G) Streptococcus pneumoniae (50) Enterococcus spp. (31)
19 (81)+ -
Gram negative cocci Neisseria gonorrhoea (47)
34 3 38 61 91 -
84 97 100 98 -
96 100 100(Caz)
70/98 98/98 79/100 100/100
50 90 82 91 91
78 95 88 98 91 100
54/44 90/90 77/100 91/68 -
Gram negative bacilli Escherichia coli (245) Klebsiella spp. (57) Proteus/Morganella spp. (34) Haemophilus influenzae (64) Salmonella sp. (22) Pseudomonas aeruginosa (61)
45 90 97 98 -
Abbreviations: Mt, methicillin; Pen, penicillin; Amp, ampicillin; Sxt, cotrimoxazole; Tet, tetracycline; Oflox, ofloxacin; Levo, levofloxacin; C, chloramphenicol; Clarithro, clarithromycin; Spect, spectinomycin; Cxm, cefuroxime; Ctx, cefotaxime; Gent, gentamicin; Netil, netilmicin; Na, nalidixic acid; Cip, ciprofloxacin, Tm, trimethoprim; Sulpha, sulphamethoxazole, Cec, cefaclor; Caz, ceftazidime * Minimum inhibitory conc.