Previous Article | Next Article ![]()
Antimicrobial Agents and Chemotherapy, July 2008, p. 2480-2485, Vol. 52, No. 7
0066-4804/08/$08.00+0 doi:10.1128/AAC.01118-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Päijät-Häme Social and Health Care Group, Clinical Microbiology, Lahti, Finland,1 National Public Health Institute, Department of Human Microbial Ecology and Inflammation, Turku, Finland,2 National Public Health Institute, Department of Bacterial and Inflammatory Diseases, Antimicrobial Research Laboratory, Turku, Finland,3 National Agency for Medicines, Helsinki, Finland4
Received 24 August 2007/ Returned for modification 23 October 2007/ Accepted 18 April 2008
|
|
|---|
|
|
|---|
Hence, the most important fields of use of SXT combinations are in the treatment of urinary tract infections and upper respiratory tract infections. According to a point prevalence study conducted in the Finnish primary health care system, by indication, 81% of all SXT prescriptions were for respiratory tract infections and 15% were for urinary tract infections (21). Since it is a relatively inexpensive drug, this combination has been used widely in developing countries. There are also a few special indications for SXT use, such as the prophylaxis of Pneumocystis carinii infections among AIDS patients and the treatment of infections caused by Stenotrophomonas maltophilia (13).
Emerging resistance among the major respiratory tract pathogens S. pneumoniae, H. influenzae, and M. catarrhalis has undoubtedly decreased the use of SXT. This resistance, however, varies worldwide. In the late 1990s, 31.9 to 88.6% of European and 24.2 to 89.4% of Asian S. pneumoniae isolates were susceptible to SXT. Among H. influenzae strains, a similar variation was noticed (22). According to a recent report from the United States, resistance among pneumococci is decreasing after peaking in 1999-2000 (9). A very small proportion of M. catarrhalis isolates, 0.1 to 2.6%, in the SENTRY surveillance program in 1997-1999 (12) showed resistance to SXT.
The relationship between antimicrobial use and resistance has been evaluated in several studies, using different approaches and analytical methods. Many studies have shown the connection between increased consumption and increased resistance of certain antimicrobial agents (4, 6, 11). As Baquero et al. point out in their review (2), an increase in use has been linked to an increase in resistance at all ecological levels, including patients, small communities, different areas of the same country, the national level, and also the international level. The effect of reduced antimicrobial consumption on the prevalence of resistance has not always been as obviously demonstrated. Seppälä and coworkers (23) showed a significant decline in erythromycin resistance among Streptococcus pyogenes isolates after a reduction in macrolide use in Finland. In Iceland, the prevalence of non-penicillin-susceptible pneumococci even increased in certain areas after significant reductions in total antimicrobial use owing to the spread of a resistant clone (1). On the other hand, resistance can be extremely persistent; in England, sulfonamide resistance persisted in urinary Escherichia coli isolates as long as 9 years after prescription of SXT was formally restricted (3). This had, however, already been predicted earlier (14).
The present study was undertaken to investigate whether the outpatient consumption of SXT is associated with resistance in Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis. We analyzed this relationship by comparing regional resistances of the three pathogens separately to local SXT consumption in 21 central hospital districts in Finland.
(Preliminary results of this work were presented as a poster at the 15th European Congress of Clinical Microbiology and Infectious Diseases, Copenhagen, Denmark [P1456].)
|
|
|---|
Antimicrobial resistance. The regional SXT resistance data between 1998 and 2004 for S. pneumoniae, H. influenzae, and M. catarrhalis were obtained from 26 clinical microbiology laboratories belonging to the FiRe network (Finnish Study Group for Antimicrobial Resistance). The reporting activity varied annually. For two central hospital districts (Kainuu and Ahvenanmaa), no resistance data for this period were available. When several laboratories reported results from a particular central hospital district, the data were combined. Resistance rates based on at least 30 tested strains per laboratory per year were included in the study. The resistance data consist of data for all consecutive (isolated from both inpatients and outpatients) clinical isolates in each laboratory over the whole year.
The national resistance data presented for the years 1988 to 2004 were obtained from the FINRES report (http://www.ktl.fi/portal/english/projects/fire/finres).
The resistance data are based on routine susceptibility results from the participating laboratories. The methods were based on the CLSI (formerly NCCLS) guidelines (8), with some exceptions: in 1998 and 1999, 20/26 laboratories performed these tests according to CLSI methods. Five laboratories used modified disc diffusion methods for H. influenzae (Isosensitest agar or PDM agar with supplements or chocolate agar), and one laboratory used Etest and chocolate agar for testing. Since 2000, all laboratories have basically followed the CLSI standard. The quality control program includes susceptibility testing of S. pneumoniae ATCC 49619 and H. influenzae ATCC49247 on a weekly basis as internal quality controls and proficiency testing programs by Labquality (www.labquality.fi) and UKNEQAS (www.ukneqasmicro.org).
Statistical analysis. The resistance rates of S. pneumoniae, H. influenzae, and M. catarrhalis in the 21 central hospital districts were compared to local consumption figures for SXT in the previous year.
A linear mixed model for repeated measures was used to model the association between regional resistance and regional antimicrobial consumption of the previous year. The percentage of resistant strains was taken as the dependent variable, while antimicrobial consumption and time were the explanatory variables. A random coefficient model with random intercept and slope was fitted. Mixed models were fitted using the Proc Mixed in SAS system for Windows, version 9.1. The level of statistical significance was set to 0.05.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. SXT consumption for outpatient care in 21 central hospital districts in Finland (1997-2003)
|
![]() View larger version (43K): [in a new window] |
FIG. 1. SXT and TMP consumption in Finland from 1987 to 2003 (DDD/1,000 inhabitants/day), including sales figures for pharmacies and hospitals. J01EE, all SXT combinations; J01EA, TMP. Figures for SXT for the period 1987-1989 were not available.
|
![]() View larger version (12K): [in a new window] |
FIG. 2. SXT resistance (%) in S. pneumoniae (Spn), H. influenzae (Hi), and M. catarrhalis (Mc) in Finland from 1988 to 2004.
|
|
View this table: [in a new window] |
TABLE 2. Regional SXT resistance in Streptococcus pneumoniae (1998-2004)
|
|
View this table: [in a new window] |
TABLE 3. Regional SXT resistance in Haemophilus influenzae (1998-2004)
|
|
View this table: [in a new window] |
TABLE 4. Regional SXT resistance in Moraxella catarrhalis (1998-2004)
|
|
View this table: [in a new window] |
TABLE 5. Connection between SXT use and resistance among S. pneumoniae, H. influenzae, and M. catarrhalis isolates
|
|
|
|---|
SXT resistance in Finland increased among pneumococci from the end of the 1980s until 1999 and after that decreased slightly, only to rise again. The same pattern was seen for H. influenzae. For M. catarrhalis, resistance increased in the 1990s but has since decreased to a very low level (4.3% in 2004). At the regional level, time on its own did not explain changes in resistance among S. pneumoniae and M. catarrhalis isolates, and among H. influenzae isolates, the change in resistance was borderline. When the resistance levels are estimated at the national level, it seems that a decreasing trend exists for M. catarrhalis. For H. influenzae or S. pneumoniae, such conclusions cannot be drawn.
There are several studies showing significant correlations between antimicrobial consumption and resistance in S. pneumoniae. Most often, the use of macrolides has been linked to erythromycin resistance. The coselection of penicillin-resistant pneumococci has been linked to the use of several drugs or drug classes (7). In Finland (20), the regional resistance rates of consecutive pneumococcal isolates were compared to rates of use of several antimicrobials for the two previous years, and a statistically significant correlation was found between SXT consumption and resistance (P = 0.043). The connection between the use of SXT and resistance in pneumococci has also been shown at the individual level (16). In contrast, no link between macrolide use and azithromycin resistance was observed among invasive or noninvasive H. influenzae isolates in Slovenia (6). In Finland (19), the increase in use of cephalosporins was linked to increased prevalence of β-lactamases in M. catarrhalis but not in H. influenzae. In Sweden, the production of β-lactamase was compared to the total use of antibiotics in areas of high consumption and low consumption (18). Among M. catarrhalis isolates, the number of β-lactamase producers was significantly higher in the area of highest mean consumption of all antibiotics. Among H. influenzae isolates, no significant difference was shown, although larger numbers of β-lactamase-positive isolates were detected in the areas of high antimicrobial consumption.
In the 1970s and 1980s, TMP resistance was linked to increased consumption of TMP alone and of SXT for E. coli (15), but not later, for E. coli isolates from urinary tract infections (17). Thus, it seems that restricting the use of SXT may result in resistance declining, but only in pneumococci. On the other hand, M. catarrhalis may be subject to selection pressure, but not with SXT. In the case of H. influenzae, the present study supports the results of the few previous studies, showing that resistance in H. influenzae is not highly susceptible to selective pressure.
Despite the decrease in consumption at the national level, increases in resistance among pneumococci and also among H. influenzae isolates from 2002 to 2004 were evident (Fig. 2). In Iceland, the spread of an international multiresistant clone explained the increased carriage of non-penicillin-susceptible pneumococci in an area of low and declining use of antimicrobials (1). Resistance against erythromycin, tetracycline, chloramphenicol, and SXT was very common among these isolates. Similarly, in Spain, dissemination of several multiresistant pneumococcal clones was associated with a high prevalence of macrolide resistance in pneumococci (5). We are not able to analyze multiresistance among single S. pneumoniae isolates from our resistance data, and the isolates have not been serotyped. Consequently, we cannot confirm the occurrence of multiresistant clones in Finland, which in our opinion may explain this increase in resistance. Our view is supported indirectly by the increase in penicillin nonsusceptibility (from 8.5% to 10.8%) and erythromycin resistance (from 15.4% to 20.2%) between 2002 and 2004 in all S. pneumoniae isolates (http://www.ktl.fi/portal/english/projects/fire/finres). The clonality among multidrug-resistant H. influenzae strains is not as obvious as that among pneumococci (10). Therefore, the reason for the increase in SXT resistance remains open at this point.
Our method differs from that of many studies in that we compared consumption and resistance over a relatively long period (7 years), and also geographically. The consumption data used in our analysis (sales from wholesalers) are estimated to represent at least 99% of the actual sales. Resistance data were not available from two (2/21 districts) central hospital districts. The remaining 19 districts cover
98% of the population in Finland. Thus, our data have an extensive coverage. In Finland, resistance data have been collected annually from all health care districts since 1997. As shown in Tables 2, 3, and 4, the reporting activity for SXT resistance has varied, especially in the case of pneumococci. The large amount of missing information may have influenced the analysis of pneumococci. Also, there are other factors possibly biasing the results. The resistance data are not based on a one-per-patient principle; a common practice in Finnish laboratories, however, is to perform susceptibility testing only once if the same pathogen is cultured from several samples during an infection episode. On the other hand, the regional consumption figures may include prescriptions purchased in one health care district but used elsewhere. Parts of prescriptions may even remain unused. S. pneumoniae, H. influenzae, and M. catarrhalis cause infections in and are carried mainly by young children; therefore, antimicrobial consumption by this age group can be considered most important for the development and spread of resistance. Our consumption data and resistance data were, however, collected during the study period as total figures, and we are not able to analyze the connection by age groups. From the FINRES data of 2005 (http://www.ktl.fi/portal/english/projects/fire/finres), we know that SXT resistance was higher (23.1%) among young children (<5 years) than in the rest of the population (16.2%). It is also known that 60% of SXT has been prescribed for otitis media, an infection presenting mostly in children of <5 years old (21). However, we do not know how these figures have altered during the study period. Therefore, the effect of age on our results cannot reliably be evaluated.
In conclusion, the use of SXT explains changes in resistance in pneumococci but not in H. influenzae and M. catarrhalis. The selection of coresistance and persistence of resistance mechanisms may increase and maintain resistance against SXT in respiratory tract pathogens despite declining consumption of the drug.
Members of the FiRe network in 2004 were Petteri Carlson and Merja Rautio (Jorvi Hospital), Risto Renkonen and Anna Muotiala (Medix-Diacor Laboratories), Martti Vaara and Eveliina Tarkka (HUSLAB), Riitta Karttunen and Tarja Ojanen (Kanta-Häme Central Hospital), Jari Karhukorpi (Pohjois-Karjala Central Hospital), Antti Nissinen (Keski-Suomi Central Hospital), Pekka Ruuska (Kainuu Central Hospital), Henrik Jägerroos (Lappi Central Hospital), Martti Larikka (Länsi-Pohja Central Hospital), Simo Räisänen (Keski-Pohjanmaa Central Hospital), Ulla Larinkari and Benita Forsblom (Kymenlaakso Central Hospital), Ulla Kärkkäinen (Kuopio University Hospital), Hannu Sarkkinen and Pauliina Kärpänoja (Päijät-Häme Central Hospital), Maritta Kauppinen and Seppo Paltemaa (Etelä-Karjala Central Hospital), Päivi Kärkkäinen (Mikkeli Central Hospital and Savonlinna Central Hospital), Markku Koskela (Oulu University Hospital), Sini Pajarre and Raija Manninen (Satakunta Central Hospital), Sinikka Oinonen and Virpi Ratia (Seinäjoki Central Hospital), Paul Grönroos (Health Center Koskiklinikka), Risto Vuento and Oili Liimatainen (Tampere University Hospital), Maj-Rita Siro (Health Center Pulssi); Erkki Eerola (Turku University), Olli Meurman and Kaisu Rantakokko-Jalava (Turku University Hospital), Suvi-Sirkku Kaukoranta (Vaasa Central Hospital), and Pentti Huovinen and Katrina Lager (National Public Health Institute).
Published ahead of print on 28 April 2008. ![]()
Members of the FiRe Network are listed in Acknowledgments. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»