Antimicrobial susceptibility of Streptococcus pneumoniae, Haemophilus influenzae and MoraxeUa (Branhamella) catarrhaus isolated in the UK from sputa

Transcription

1 Journal of Antimicrobial Chemotherapy (1991) 28, Antimicrobial susceptibility of Streptococcus pneumoniae, Haemophilus influenzae and MoraxeUa (Branhamella) catarrhaus isolated in the UK from sputa M. PoweO', D. McVey*, M. H. Kassbn*, H. Y. Chen' and J. D. Williams' 'Department of Medical Microbiology, The London Hospital Medical College, Turner Street, London El 2AD; b Lederle Laboratories, Fareham Road, Gosport, Hampshire PO13 OAS, UK Four hundred and thirty-one Streptococcus pneumoniae, 1272 Haemophilus influenzae and 305 Moraxella (Branhamella) catarrhalis were isolated from sputa and identified in 28 UK laboratories during a ten week period in Disc diffusion susceptibility testing was performed in each centre using identical methods. Species-specific susceptibility breakpoints applied to data for six antimicrobial agents were determined from the distribution of isolates according to zone diameters of inhibition measured in participating laboratories and were correlated with minimum inhibitory concentration data obtained with 302 isolates sent to the coordinating centre. Inter-laboratory reproducibility was estimated by comparing peripheral and coordinating centre results for these 302 isolates and by distributing five reference strains to all laboratories for testing. Reduced susceptibility to ampicillin and amoxycillin-clavulanate was detected in < 3% of S. pneumoniae, but 81% were resistant to tetracycline and 6-5% to erythromycin. Resistance to ampicillin due to production of /Hactamase occurred in 94% of H. influenzae; another 5-2% were resistant to ampicillin and amoxycillin-clavulanatc but were /J-lactamase-negative. 4-5% were resistant to tetracycline and most (86-6%) had MICs > 1 mg/l of erythromycin. Zone diameters around ampicillin discs were > 10 mm smaller than those around amoxycillin-clavulanate discs for 241 (79%) of M. catarrhalis. Although only 193/241 had been reported to be 0-lactamase positive by participating laboratories, data obtained at the coordinating centre confirmed that 5> 10 mm and < 3 mm zone size differences correlated with 0-lactamasepositive and -negative isolates respectively. No M. catarrhalis were resistant to amoxycillin-clavulanate and < 4% were resistant to either tetracycline or erythromycin. The prevalence of resistance to cefaclor was highest among H. influenzae (5-2%) and lowest among S. pneumoniae (0-9%). Only seven of 2008 isolates (two to three per species) were resistant to cefixime. The data suggest that the prevalence of resistance to ampicillin, tetracycline and erythromycin must be taken into consideration when treating respiratory infections. Introduction Streptococcus pneumoniae, Haemophilus influenzae and Moraxella (Branhamella) catarrhalis are species which colonize the respiratory tract and may be isolated from sputa obtained from patients with respiratory symptoms indicative of infection. While it is not always clear as to whether the clinical picture is entirely due to the presence of these organisms in expectorated sputum, there is no doubt that in some cases they can /91/ S0Z00/ The Britiih Society for Antimicrobial Chemotherapy

2 250 M. PoweD et al be entirely responsible for the respiratory illness and in others can contribute to the pathogenic process (Gransden, Eykyn & Phillips, 1985; Devalia et al., 1988; Makela, 1988). Many of these patients are treated by oral administration of antimicrobial agents but there are now problems due to the advent of drug resistance in each of these three species. During a UK survey in 1977 resistance to tetracycline was detected in 6-8% of S. pneumoniae, but none of the 866 isolates collected were resistant to erythromycin; and only in one isolate was the minimum inhibitory concentration (MIC) of penicillin > 006 mg/l (Howard, Hince & Williams, 1978). Sporadic reports from the UK of penicillin-resistant (MIC >012 mg/l) and erythromycin-resistant S. pneumoniae followed this study (George et al., 1988; Gransden et al., 1988). Antimicrobial resistance in this species and the occurrence of multi-resistant organisms are now worldwide problems (Appelbaum, 1987). There is considerable data available regarding the UK prevalence of antimicrobial resistance in non-capsulate H. influenzae, which are the predominant types associated with the respiratory tract. Among 2371 such isolates collected from 24 centres in England and Scotland in 1986, 6% were found to produce /Mactamase. Another 4% showed non-/mactamasc-mediated resistance to ampicillin (MIC > 1 mg/l) while 2-7% were resistant to tetracycline, 3% to cefaclor and 02% to cefixime (Powell et al., 1987). These figures are similar to those for Wales in 1986 and for Ireland in 1988 except that prevalences of /Mactamase-positive isolates were higher at 8-4% and 102% respectively (Howard & Williams, 1988, 1989). There are no previous surveys of the prevalence of antimicrobial resistance among M. catarrhalts throughout the UK although two local studies have reported that 35% and more than 70% were /Mactamase-positive (McLeod et al., 1986; Winstanley & Spencer, 1986). Recent studies have estimated that 84% of M. catarrhalis in the USA and 83% in France are /Mactamase-positive (Chardon, Bellon & Lagier, 1990; Jorgensen et al., 1990). Although all M. catarrhalis are inherently resistant to trimethoprim (MIC > 8 mg/l), resistance to either tetracycline or erythromycin appears to be uncommon. In one study in the USA, 1 % of 457 strains were resistant to tetracycline and 3-5% were not susceptible to erythromycin (Brown et al., 1989). This study involved identification and disc-diffusion susceptibility testing of S. pneumoniae, H. influenzae and M. catarrhalis isolated from, and thought to be significant in, sputum cultures (i.e. would normally be reported to the clinician) by 28 UK laboratories. The prevalence of resistance in each species to six antimicrobial agents which can be administered by the oral route was then determined by analysis of the diameters of zones of inhibition recorded by the participating centres. Collection and identification of isolates Materials and methods Twenty-eight geographically scattered UK laboratories contributed data regarding S. pneumoniae, H. influenzae and M. catarrhalis which were considered to be clinically significant when isolated from sputa obtained from hospital in-patients (excluding those in intensive care units) and community patients. The total number (whatever the species) of isolates examined was limited to a maximum of 100 per laboratory within a ten week collection period. All laboratories identified S. pneumoniae on the basis of susceptibility to optochin and H. influenzae by X and V disc satellitism. Methods

3 SosceptibOity of patbogensfromsputa 251 routinely used for identification of M. catarrhalis varied. However 26/28 centres applied a minimum of DNase and/or tributyrin tests as well as sugar utilization and oxidase reactions, to Gram-negative diplococci. Susceptibility testing Disc diffusion susceptibility testing was performed in each participating laboratory by a standardized method using materials provided by a single supply distribution centre (Lederle Laboratories, Fareham, Hants, UK). From fresh overnight cultures of speciated organisms, suspensions were prepared in peptone water in order to produce semi-confluent growth following swab inoculation of plates containing Iso-Sensitest agar (Oxoid) supplemented with 5% (v/v) lysed horse blood and 10 mg/l NAD. The following discs were then applied: ampicillin (2 fig), amoxyciuin-davulanate (2+1 fig), erythromycin (5 ng), tetracycline (10 fig), cefaclor (30 fig) and ccfixime (5 ng). Plates were incubated for 18 hours in an atmosphere of 5-10% CO 2 and air at 37 C C. Laboratories were requested to use their usual control strains in these tests. Zone diameters were measured and recorded to the nearest whole millimetre. Tests for production of /J-lactamase were performed according to the usual practice of each laboratory (four used an acidometric method while the rest used commerical methods based on nitrocefin). Inter-laboratory reproducibility During the study, laboratories were asked to send at least two isolates of each species for which zone diameters around cefixime discs were > 22 mm to the coordinating laboratory (London Hospital Medical College) for susceptibility testing. In addition, any isolates for which zones around cefixime discs were ^ 22 mm were also requested for re-testing, regardless of any other drug resistance detected. Identity of the isolates was confirmed by methods similar to those used in the participating centres, except that the criteria of Doern & Morse (1980) were applied to presumptive M. catarrhalis. As an additional check on inter-laboratory reproducibility, five reference strains were sent to each participating centre at the start of the study. These were tested on two separate occasions in all laboratories by study protocol methods. The reference strains were S. pneumoniae ATCC , H. influenzae ATCC , two ampicillin-resistant H. influenzae (one /J-lactamase-positive and one negative) and a /J-lactamase-negative Neisseria spp. for which zones of inhibition for ampicillin and amoxycillin-clavulanate were mm. Susceptibility testing at coordinating centre Disc diffusion testing was repeated for all isolates sent to the centre and MICs of the same six antimicrobials were determined by an agar dilution method (supplemented Iso-Sensitest agar as before). The inoculum used was 10 4 cfu/spot for all three species. All H. influenzae and M. catarrhalisreceivedwere tested for /Mactamase production by the starch hydrolysis method (Catlin, 1975). MICs of penicillin and diameters of zones of inhibition around 1 fig oxacillin discs were determined for all S. pneumoniae received. The ATCC 5. pneumoniae and H. influenzae reference strains, as well as M. catarrhalis ATCC 25238, were used as controls.

4 252 M. PoweOrt al Table L Antimicrobial resistance in S. pneumoniae, H. tnfluenzae and M. catarrhalis as determined from the zone diametei breakpoints indicated Organism S. pneumoniae H. tnfluenzae M. catarrhalis Antimicrobial agent ampicillin amoxycillin-clavulanate tetracycline erythromycin cefaclor cefixime ampicillin ( *) ampicillin (fi~) amoxycillin-clavulanate tetracycline erythromycin cefaclor cefixime ampicillin amoxycillin-clavulanate tetracycline erythromycin cefaclor cefixime Zone diameter breakpoint (mm) (MIC [mg/l] correlating with zone) < 25 (>0-12) < 25(3*0-12) < 24 (5*4) <21(Ssl) <21 (Ss 16) <22(>2) not applicable < 18 (5=1) < 18(^1) =S 18 (5=4) < 17 (5=1) < 18 (> 16) <19( 2) not applicable < 19(5 2) < 18 (Ss4) < 17 (5*1) <18(5= 16) <19(5=2) Number (%) resistant 9 (21) 11 (2-6) 35 (81) 28(6-5) 4(09) 3(0-7) 119 (9-4) 67 (5-2r 67 (5-2) 57 (4-5) 1101 (86-6)* 66 (5-2) 2(0-2) 241 (79)* none 9(3) 11 (3-6) 4 (1-3) 2(0-7) 'See text for explanation of these figures. * 193 reported to beft-lnctnrrune-poativeand 48 in which 0-lactamase production was inferred because zones for amoxycillin-ciavulanate were > 10 mm larger than zones for ampicillin (see text). Results The participating laboratories contributed data on a total of 2008 isolates, consisting of 431 S. pneumoniae, 1272 H. influenzae and 305 M. catarrhalis. The contribution from each centre varied from isolates with an average of 72 per centre and a mode of between 90 and 100 per laboratory. At least two isolates of each species per laboratory were sent to coordinating centre of which 11 (presumptively 3 S. pneumoniae, 6 H. influenzae and 2 M. catarrhalis) were found to have been misidentified, therefore the data derived from these isolates were deleted from the database. Zone diameter and MIC data were then compiled for 302 isolates, consisting of 80 S. pneumoniae, 133 H. influenzae and 89 M. catarrhalis. Susceptibility breakpoints Zones of inhibition recorded in participating centres were plotted on separate histograms for each drug versus each species. Susceptibility breakpoints were determined from these graphs and from zone diameter and MIC data relating to the 302 isolates sent to the coordinating centre (Table I). Zone diameter breakpoints were situated where delineations between susceptible and resistant populations were clearly apparent on histograms and were found to correlate well with accepted MIC breakpoints when

5 Susceptibility of pathogensfromsputa 253 data from the coordinating centre were analysed. The exception to the rule was erythromycin and H. influenzaer, in this instance, the modal zone diameter of inhibition was 6 mm with 54% of isolates producing zones < 13 mm around erythromycin discs. The breakpoint (<17mm for resistance) applied was the zone diameter which correlated with separation of the small fully-susceptible population (MIC < 0-5 mg/l erythromycin) from those isolates with intermediate or definite resistance to the drug (MIC Js 1 mg/l). S. pneumoniae The 11 S. pneumoniae resistant to amoxycillin-clavulanate included eight of the nine isolates that were resistant to ampicillin. The zone diameter for amoxycillin-clavulanate for the ninth ampicillin-resistant isolate and the ampicillin zone diameters for the other two amoxycillin-clavulanate-resistant isolates were all 26 mm. These 11 isolates included eight which were resistant to erythromycin, the four resistant to cefaclor and two of the three resistant to cefixime (Table I, columns 2 and 3). These latter two isolates were also resistant to tetracycline, but both were susceptible to erythromycin. H. influenzae Not all participating laboratories performed a /Mactamase test on every H. influenzae and M. catarrhalis isolated. There were 16 H. influenzae for which zones of inhibition around ampicillin discs were < 18 mm but no /Mactamase test result was available. However, zones of inhibition around amoxycillin-clavulanate discs were also < 18 mm (and within ±2 mm of the data for ampicillin) for all 16 isolates. These isolates were therefore added to the total number which were reported to be /Mactamase negative but resistant \o both ampicillin and amoxycillin-clavulanate, so increasing the number of isolates to 67 (5-2%) (Table I, columns 3 and 4), of which 29 were resistant to cefaclor. Of the remaining 37 cefaclor-resistant isolates 30 were ampicillin-susceptible while the other seven produced /Mactamase. Two isolates were resistant to cefixime, one produced /Mactamase and the other gave zones ^ 22 mm for the other /Mactams. Only ten H. influenzae were resistant to > four antimicrobials. M. catarrhalis /J-Lactamase production was reported for 193 (63%) M. catarrhalis. The modal zone diameter of inhibition for ampicillin in this group was 15 mm but the range of zone sizes was 6-33 mm and zones of > 30 mm were recorded for three isolates. Of the remaining 112 isolates, 58 were reported to be /Mactamase negative. Zone diameter data for ampicillin and amoxycillin-clavulanate for 45/58 M. catarrhalis were within 3 mm of each other and ranged from mm. In contrast, zone diameters around amoxycillin-clavulanate discs were > 10 mm larger than those around ampicillin discs for the other 13 isolates, suggesting that they were /Mactamase-positive. No /Mactamase data was available for the other 54 M. catarrhalis. A similar comparison (< 3 and > 10 mm difference) between zones of inhibition gave two distinct groups of isolates, suggesting that at least 35/54 were /Mactamase-positive, 31 of which produced zones < 18 mm for ampicillin. Hence, the final number of isolates which were known or were very likely to be /Mactamase-positive was 241, increasing the total percentage from 63% to 79%. MICs of ampicillin for 60 /Mactamase-positive M. catarrhalis tested

6 I Table II. Susceptibility testing data for five reference strains (total = 54 reports for each strain) in participating laboratories Reference strain H. influenzae 518 (fi-f S. pneumoniae (ATCC ) H. influenzae (ATCC ) OH Neisseria spp. (fi ) H. influenzae 186 (/?+)* */7-Lactaraase negative; */J-lactamaje positive; 'R, resistant; 'S, susceptible. % agreement for correct result amoxycillinampicillin clavulanate tetracycline crythromycin cefaclor 41 (sy 100 (S) 98 (R) 100 (S) (S) (S) (S) (S) 93 (R) 98 (S) 81 (R) 91 (R) 83 (S) (65) cefixime S(100) S(100) S(IOO) S(100) S(100)

7 Susceptibility of pathogens from sputa 255 Table HI. Comparisons of susceptibility testing data in participating and coordinating centres for 302 isolates Percentage of results in agreement with coordinating centre finhino* fnr rji tiwjes findings for each species Antimicrobial agent 5. pneumonias H. influenzae M. catarrhalis (n = 80) (n-133) (n = 89) all all all S R results S R results S R results Ampicillin Amoxycillin-clavulanate Tetracycline 97 25* * 97 Erythromycin 99 50* 98 0* Cefaclor * * 99 Ccfixime 100 5C 'Based on < 8 isolates. in the coordinating centre were between 0-03 and 1 mg/l (MIC M of 0-12 mg/l). The range of zone diameters observed for these 60 isolates was 6-31 mm but a zone size difference > 10 mm as described above correlated consistently with production of /2-lactamase. No M. catarrhalis were resistant to > three antimicrobials. Inter-laboratory reproducibility Although there was considerable variation between zone diameters of inhibition recorded in participating laboratories for the five reference strains, the number of times these strains would have been misreported as susceptible or resistant, according to the breakpoints adopted, was generally very small (Table II). The exceptions were ampicillin and amoxycillin-clavulanate for the ampicillin-resistant, /Mactamase-negative H. influenzae 518 and for erythromycin with all five reference strains. A comparison between the data obtained for the 302 isolates which were retested in the coordinating laboratory showed that total agreement (i.e. all results) was ^ 93% for M. catarrhalis and > 98% for S. pneumoniae for all six agents (Table III) but varied from % for H. influenzae. Discrepancies involved reports of susceptibility by participating laboratories for isolates considered resistant at the coordinating centre in 41 of a total of 80 instances of disagreement. The data suggesting lowfigures (< 50%) for agreement on resistance which appear in Table III are based on < 8 strains tested except in the case of amoxycillin-clavulanate with H. influenzae where 9/16 isolates considered resistant onre-testingwere reported as susceptible by participating laboratories. These 16 H. influenzae were all /Mactamase-negative. It is also notable that 6/29 isolates which were found to be ampicillin-resistant and /Mactamase-negative were considered susceptible on first testing. Agreement on the data of tests for /Mactamase was 96% for M. catarrhalis and 98% for H. influenzae. In any multicentre study in which identification and susceptibility testing of isolates is performed in each participating laboratory questions must be raised regarding contri- Discussion

8 256 M. Powell et al butions from individual centres. The numbers of resistant isolates obtained from each of the participating centres were small. Although the total numbers of isolates of each species examined varied between centres, local percentages of resistance obtained suggested that overall figures were not distorted by disproportionate contributions from participating laboratories. With the exception of H. influenzae, re-examination of isolates in the coordinating centre indicated that misidentification of species did not appear to be a major source of error in this study. It is notable that 26 of the 28 laboratories applied one or both of DNase and tributyrin tests to presumptive M. catarrhalis, either of which have been put forward as suitable sole criteria for identification by Doern (1990), provided that the organism is known to be a Gram-negative diplococcus and is oxidase-positive. This study sought to reduce inter-laboratory discrepancies of disc diffusion susceptibility testing by employing a standardized method. In addition, attempts were made to estimate the frequency of errors by the use of reference strains and by repeat testing of a sample of isolates in a coordinating centre. Isolates were not chosen randomly for repeat testing although the criteria laid down ensured that the selection process resulted in submission of organisms of variable drug resistance patterns. The data generated from these comparisons suggested that thefinalfigure (5-2%) for the prevalence of non- /7-lactamase-mediated resistance to ampicillin among H. influenzae might be an underestimate. In addition, susceptibility testing to erythromycin was less reproducible than for other agents but this might have been partly due to the use of a 5 fig disc, instead of the 15 fig disc which is more suitable for H. influenzae, and to the effects of a CO 2 -supplemented atmosphere. The zone diameter breakpoints adopted were chosen after completion of the study period so that all data could be taken into consideration. While 2 fig amoxycillin discs are not ideal for the detection of low level resistance to penicillin among S. pneumoniae, correlation with MICs of penicillin > 0-12 mg/l supported the use of a > 25 mm breakpoint for susceptibility under these tests conditions and correlated well with results using oxacillin (1 fig) discs in the coordinating centre. Other zone and MIC breakpoints were similar to those used in previous UK surveys (Howard et al., 1978; Powell et al., 1987) and to MIC breakpoints adopted by the NCCLS (1985) in the USA. For the new-orally-administered cephalosporin, cefixime, the MIC breakpoint for the fully-susceptible population is < 1 mg/l (Fuchs, Barry & Jones, 1986). From zone diameter and MIC data it was quite clear that ^ 19 mm correlated with reduced susceptibility among H. influenzae and M. catarrhalis but that < 22 mm was appropriate for S. pneumoniae. Antimicrobial resistance in S. pneumoniae was generally low, although resistance to tetracycline and erythromycin was increased compared with UK figures of 6-8% and nil respectively in 1977 (Howard et al., 1978). Reduced susceptibility to ampicillin was uncommon, which is in keeping with previous reports of 0-1% and 1-3% for penicillinresistant isolates (Howard et al., 1978; Wall, Emmerson & Lamport, 1981). In addition, Lafong et al. (1988) reported that 1% of 488 S. pneumoniae collected in Belfast were of intermediate susceptibility to penicillin whereas 12% were resistant to tetracycline and all were susceptible to erythromycin. More recent figures from Belgium are slightly higher than those of the present study in that penicillin, erythromycin and tetracycline resistance were reported in 2-7%, 11-5% and 10-4% of isolates respectively (Verhaegen et al., 1990). Nevertheless, isolates not susceptible to penicillin (MIC > 0-12 mg/l) and multi-resistant S. pneumoniae have already caused localized outbreaks in the UK

9 Susceptibility of pathogen from spott 257 (Gould, Magee & Ingham, 1987; Moore & Williams, 1988). Vigilance is clearly necessaary even though the prevalence of such organisms appears to be much lower than in many other countries (Appelbaum, 1987). Philpott-Howard & Williams (1982) reported that 5-3% of 1841 H. influenzae were /Mactamase-positive while 2-5% were /Mactamase-negative but required > 1 mg/l ampicillin for inhibition. The numbers of ampicillin-resistant /Mactamase-positive and -negative H. influenzae found in this study suggested that there has been a substantial increase in the prevalence of these isolates in the UK during the last decade. Resistance to tetracycline (4-5%) was higher in 1990 than the 2-7% recorded in 1986 (Powell el al., 1987). Previous UK surveys have not examined resistance to erythromycin among H. influenzae because the modal MIC of the drug for this species (1-2 mg/l) is above the breakpoint of < 0-5 mg/l erythromycin for fully susceptible isolates (equivalent to a zone diameter around a 5 fig disc of > 18 mm in this study). Of the 1272 H. influenzae examined 87% were considered to show intermediate (MIC > 1 mg/l but < 4 mg/l) or definite resistance (MIC > 8 mg/l) to erythromycin. Using the same NCCLS-recommended MIC breakpoint, Doern et al. (1988) reported that only 1% of isolates from the USA were fully susceptible to erythromycin following determination of MICs of the drug for 2811 H. influenzae. It is well-recognized that /Mactamase-positive M. catarrhalis are susceptible to very low concentrations of ampicillin and that zone diameters may be very large (Luman et al., 1986 Wallace, Nash & Steingrube, 1990). This latter fact may explain why so many M. catarrhalis were not tested for /Mactamase production in the participating centres even when there was a ^ 10 mm difference between zone diameters around ampicillin and amoxycillin-clavulanatc discs. Several other studies have suggested that all M. catarrhalis should be tested for production of /Mactamase and that all found to be positive should be regarded as resistant to ampicillin (Wallace et al., 1990). Some authors have recommended that nitrocefin is used for detection of /Mactamases expressed by this species (Wallace et al., 1990), however, the starch hydrolysis method used in the coordinating centre correlated perfectly with ^ 10 mm and < 3 mm differences between zone sizes for ampicillin and amoxycillin-clavulanate. The 79% /Mactamase-positive figure derived closely resembles the reports of > 80% from France and USA. Antimicrobial resistance to other agents (apart from trimethoprim) does not appear to be a problem in this species at present in the UK. Resistance to cephalosporins was low except for 5-2% of H. influenzae and cefaclor which is higher than the 3% of isolates collected in 1986 which required > 16 mg/l cefaclor for inhibition. More than half (36/66) H. influenzae were also resistant to ampicillin (seven were /Mactamase positive and 29 were negative). Of the 74 cefaclorresistant H. influenzae collected in 1986, three were /Mactamase-positive and 47 showed non-/mactamase-mediated resistance to ampicillin, while a smaller proportion of isolates (24/74) were ampicillin-susceptible when compared with this study. In contrast, only 7/2008 isolates were resistant to cefixime. The MIC*, of cefixime determined in the coordinating centre was 0-25 mg/l for S. pneumoniae and M. catarrhalis and 0-06 mg/l for H. influenzae. Cefixime is very stable in the presence of /Mactamases produced by the latter two species and generally retains activity for ampicillin-resistant, /Mactamase-negative isolates in vitro (Sanders, 1989). This study has demonstrated that resistance to antimicrobial agents must be taken into consideration when treating respiratory infections associated with the presence of one or more of these three species in sputa. On current evidence, ampicillin is no longer

10 258 M. Powell et al a reliable agent for such infections. Amoxycillin-clavulanate, however, is active against the three species studied with the exception of those S. pneumoniae and H. influenzae which have non-enzymic mechanisms of resistance to /J-lactams. The prevalences of resistance to tetracycline and erythromycin appear to be increasing slowly and the latter agent cannot be considered to be active against the majority of H. influenzae. Cefaclor is generally very active against S. pneumoniae and M. catarrhalis but less so against H. influenzae. Cefixime has been shown to be very active for these respiratory pathogens with few resistant isolates in circulation. Along with some of the 4-quinolone agents and a few of the more recently developed macrolides, the new cephalosporins, such as cefixime, are among the agents which will be increasingly relied upon to treat respiratory infections in the future. Acknowledgements The authors wish to thank the medical microbiologists and laboratory staff in the following centres for their participation and cooperation in the study: Aberdeen Royal Infirmary; Royal Victoria Hospital, Belfast; Queen Elizabeth Hospital, Birmingham; Bristol Royal Infirmary and Frenchay Hospital, Bristol; Royal Sussex County Hospital, Brighton; PHL Chelmsford; City Hospital, Edinburgh; West Park Hospital, Epsom; Southern General Hospital, Glasgow; Kettering District General Hospital; Seacroft Hospital, Leeds; Leicester Royal Infirmary; Royal Liverpool Hospital; Whipps Cross Hospital, London; Preston Hall Hospital, Maidstone; Wythenshaw Hospital, Manchester; Borders General Hospital, Melrose; Royal Victoria Hospital, Newcastle; John Radcliffe Hospital, Oxford; Llandough Hospital, Penarth; Poole General Hospital; St Mary's Hospital, Portsmouth; Royal Berkshire Hospital, Reading; Hope Hospital, Salford; Northern General Hospital, Sheffield; South Warwickshire Hospital; Warwick; Clatterbridge Hospital, Wirral. References Appelbaum, P. C. (1987). World-wide development of antibiotic resistance in pneumococci. European Journal of Clinical Microbiology and Infectious Diseases 6, Brown, B. A., Wallace, R. J., Flanagan, C. W., Wilson, R. W., Luman, J. I. & Redditt, S. D. (1989). Tetracycline and erythromycin resistance among clinical isolates of Branhamella catarrhalis. Antimicrobial Agents and Chemotherapy 33, Cattin, B. W. (1975). Iodometric detection of Haemophilia influenzae /Mactamase: rapid presumptive test for ampicillin resistance. Antimicrobial Agents and Chemotherapy 7, Chardon, H., Bellon, O. & Lagier, E. (1990). French multicenter trial on the isolation of Branhamella catarrhalis in general hospitals. In Abstracts of the International Congress for Infectious Diseases, Montreal, Canada. Abstract 606. Devalia, J. L., Hannanycri, Y., Cundell, D. R., Davies, R. J., Grady, D. & Tabaqchali, S. (1988). Variation in histamine synthesis by Gram-negative and Gram-positive respiratory tract bacteria and the effect of cefaclor. In Update on Haemophilus influenzae: How Virulence, Incidence and Resistance Affects Treatment (Ringelmann, R., Ed.), pp (International Congress and Symposium Series, No. 128). Royal Society of Medicine Services, London. Doem, G. V. (1990). Branhamella catarrhalis: phenotypic characteristics. American Journal of Medicine 88, Suppl. 5A, 33S-5S. Doern, G. V., Jorgensen, J. H., Thornsberry, C, Presnot, D. A., Tubert, T., Redding, J. S. et al. (1988). National collaborative study of the prevalence of antimicrobial resistance among clinical isolates of Haemophilus influenzae. Antimicrobial Agents and Chemotherapy 32, Doern, G. V. & Morse, S. A. (1980). Branhamella (Neisseria) catairhalis: criteria for laboratory identification. Journal of Clinical Microbiology 11,

11 SosceptibiBty of pathogens from spau 259 Fuchs, P. C, Barry, A. L. & Jones, R. N. (1986). Cefixime disk susceptibility test criteria. Journal of Clinical Microbiology 24, George, R. C, Snell, J. J. S., Cooper, P. G. & Erdman, Y. J. (1988). Penicillin-resistant pneumococci. Lancet i, Gould, F. K., Magee, J. G. & Ingham, H. R. (1987). A hospital outbreak of antibiotic-resistant Streptococcus pneumoniae. Journal of Infection 15, Gransdcn, W. R., Eykyn, S. J. & Phillips, I. (1985). Pneumococcal bacteraemia: 325 episodes diagnosed at St Thomas's Hospital. British Medical Journal 290, Howard, A. J., Hince, C. J. & Williams, J. D. (1978). Antibiotic resistance in Streptococcus pneumoniae and Haemophihts influenzae. Report of a study group on bacterial resistance. British Medical Journal i, Howard, A. J. & Williams, H. M. (1988). The prevalence of antibiotic resistance in Haemophihts influenzae in Wales. Report from the Standing Specialist Advisory Group for Microbiology in Wales. Journal of Antimicrobial Chemotherapy 21, Howard, A. J. & Williams, H. M. (1989). The prevalence of antibiotic resistance in Haemophihts influenzae in Ireland. Journal of Antimicrobial Chemotherapy 24, Jorgensen, J. H., Doera, G. V., Maher, L. A., HowelL A. W. & Redding, J. S. (1990). Antimicrobial resistance among respiratory isolates of Haemophilus influenzae, Moraxella catarrhalis, and Streptococcus pneumoniae in the United States. Antimicrobial Agents and Chemotherapy 34, Lafong, A. C, Crothers, E., Bamford, K. B. & Rooney, P. J. (1988). Distribution of serotypes and antibiotic resistance among pneumonococci in Northern Ireland. Journal of Infection 16, 235^2. Luman, I., Wilson, R. W., Wallace, R. J. & Nash, D. R. (1986). Disk diffusion susceptibility of Branhamella catarrhalis and relationship of /?-lactam zone size to /J-lactamase production. Antimicrobial Agents and Chemotherapy 30, Makela, P. H. (1988). Unencapsulated Haemophilus influenzae what kind of pathogen? European Journal of Clinical Microbiology and Infectious Diseases 7, McLeod, D. T., Ahmad, F., Capewcll, S., Croughan, M. J., Calder, M. A. & Seaton, A. (1986). Increase in bronchopulmonary infection due to Branhamella catarrhalis. British Medical Journal 191, Moore, E. P. & Williams, E. W. (1988). Hospital transmission of multiply antibiotic-resistant Streptococcus pneumoniae. Journal of Infection 16, National Committee for Clinical Laboratory Standards (1984). Performance Standards for Antimicrobial Disk Susceptibility Tests. 3rd edn; Approved Standard M2-A3. NCCLS, Villanova, PA. Philpott-Howard, J. & Williams, J. D. (1982). Increase in antibiotic resistance in Haemophihts influenzae in the United Kingdom since 1977: report of study group. British Medical Journal 284, Powell, M., Koutsia-Carouzou, C, Voutsinas, D., Seymour, A. & Williams, J. D. (1987). Resistance of clinical isolates of Haemophilus influenzae in United Kingdom British Medical Journal 295, Sanders, C. C. (1989). /7-Lactamase stability and in vitro activity of oral cephalosporins against strains possessing well-characterized mechanisms of resistance. Antimicrobial Agents and Chemotherapy 33, Verhaegen, J., Glupczynski, Y., Verbist, L., Blogic, M., Vandeven, J., Yourassowsky, E. et al. (1990). Capsular types and antibiotic sensitivity of pneumococci isolated from patients with serious infections in Belgium 1980 to European Journal of Clinical Microbiology and Infectious Diseases 9, Wall, R. A., Emmerson, A. M. & Lamport, P. (1981). Penicillin-resistant pneumococci in London. Lancet ii Wallace, R. J., Nash, D. R. & Steingrube, V. A. (1990). Antibiotic susceptibilities and drug resistance in Moraxella (Branhamella) catarrhalis. American Journal of Medicine 88, Suppl. 5A, Winstanley, T. G. & Spencer, R. C. (1986). Moraxella catarrhalis; antibiotic susceptibility with special reference to trimethoprim. Journal of Antimicrobial Chemotherapy 18, (Received 22 October 1990; accepted 19 April 1991)