For years in Sweden the rate of penicillin-nonsusceptible pneumococci (PNSP) with penicillin MIC ≥0.12 mg/l has been 3 to 5%. 1, 2 However, in Malmöhus County, in the southern part of Sweden, the rate of PNSP increased in the late 1980s and reached ∼10% in 1992. 3, 4 In this county a spread of several clones of PNSP, especially among preschool children, was observed. 5 Approximately 50% of the registered PNSP strains were of serotype 9V. 6 In 1995 the South Swedish Pneumococcal Intervention Project was initiated, aiming at reducing the spread of PNSP. 7–9 This is an ongoing community-based project with special emphasis on children in day care, including repeated cultures, contact tracing and exclusion from day care whenever a child carrying PNSP is identified. The project also includes measures to reduce antibiotic utilization in the county. The initial epidemic spread seems to have abated in the county, the incidence of PNSP being curtailed at 8 to 10% since 1992. 10
In Sweden reliable data on antimicrobial sales have been available for years at a national and county level. 11 Since 1998 antibiotic sales data are also registered by the address of the consumer, rendering possible reliable statistics at a municipality level. The aim of this study was to examine the correlation between antibiotic utilization and the frequency of PNSP in single municipalities in a county where PNSP have been present for 10 years. Using the comprehensive data available, we tested the hypothesis that the frequency of PNSP in children ages 0 to 6 years is positively correlated to the utilization of antibiotics at a population level.
MATERIALS AND METHODS
The former Malmöhus County (since January 1998 part of the larger Skåne County) is situated in the southernmost part of Sweden. The county comprises 20 municipalities (Fig. 1). Eight of the municipalities are mainly urban, and 12 are rural. On December 31, 1998, the total population was 829 353, of which 68 030 were children 0 to 6 years old. The largest municipality, Malmö, had 254 904 inhabitants, and of those 20 289 were children ages 0 to 6 years. The smallest municipality, Svalöv, had 12 707 inhabitants, of which 1150 were children age 0 to 6 years.
Utilization of antibiotics
All antibiotic prescriptions for outpatient care, served at Swedish pharmacies between January 1 and June 30, 1998, to 0- to 6-year-old children with residency in Malmöhus County were included, and the municipality of the patient registered. The prescribed antibiotics were given as defined daily doses (DDD) per 1000 inhabitants per day, and all prescriptions of phenoxymethylpenicillin (PcV), ampicillin/amoxicillin (including amoxicillin/clavulanic acid), cephalosporins, trimethoprim-sulfamethoxazole, macrolides and other antibiotics (grouped together) were registered. DDDs are based on adult dosages. When children are considered, the lower dosages for children must be taken into account, and the DDDs should be multiplied by a factor of 2 or 3, depending on age/weight and type of infection.
All pneumococcal strains recovered from nasopharyngeal cultures from 0- to 6-year-old children, analyzed at the departments of clinical microbiology in Lund, Malmö and Helsingborg between July 1, 1997, and June 30, 1998, were considered in this study. The three laboratories served the entire population of the former Malmöhus County until the June 30, 1998. From the laboratory in Helsingborg (serving the populations in Helsingborg, Bjuv and Höganäs), data on the number of PNSP but not data on the total number of pneumococci were available. The nasopharyngeal specimen were cultured on blood agar plates, and the isolates were identified as Streptococcus pneumoniae on the basis of colony morphology and susceptibility to optochin. 12 The strains were screened for penicillin resistance by the disc diffusion method, using 1-μg oxacillin discs. The MICs of penicillin were determined by E-test (AB Biodisk, Solna, Sweden). 13 The total number of cultures positive for pneumococci and among them the number of PNSP with penicillin MIC ≥0.12 mg/l were registered by place of residency. In this study only strains from 0- to 6-year-old children were accounted for, and strains from each patient were registered only once, even if multiple cultures positive for pneumococci were found. Strains known to be recovered at contact tracing or screening were excluded from analysis.
Correlation analyses were performed with Pearson’s correlation coefficient with two tailed significance to evaluate the linear association between the variables antibiotics utilization and frequency of PNSP. All calculations were retested with Spearman’s correlation coefficient to avoid impact of any outliers.
Pneumococci with reduced susceptibility to penicillin
The frequency of PNSP varied largely among the communities. In the 17 municipalities with data on both PNSP and total number of isolated pneumococci, the proportion of PNSP varied between 0 and 49.5%. The registered number of PNSP (MIC ≥ 0.12 mg/l) per 1000 children and municipality varied between 0 and 10.3, the highest number being found in the densely populated southwestern part of the county (Table 1).
The total utilization of antibiotics by 0- to 6-year-old children varied among the 20 municipalities, from 8.50 to 19.73 DDD per 1000 children per day (Table 2). The areas with higher than average utilization were most often situated in the southwestern part of the county (Fig. 1). The municipalities with high total utilization also had a higher utilization of macrolides and trimethoprim-sulfamethoxazole. In all municipalities PcV was the most commonly used drug (range, 5.68 to 10.97 DDD per 1000 children per day).
Correlation between antibiotic utilization and PNSP
There was a significant correlation between the utilization of antibiotics (DDD per 1000 children per day) and the frequency of PNSP (proportion of PNSP of all isolates of pneumococci from nasopharyngeal cultures) in the 17 municipalities (correlation coefficient, 0.69;P = 0.002) (Fig. 2;Table 3). Correlation coefficients significant at the 0.001 level were noted for trimethoprim-sulfamethoxazole, amoxicillins, macrolides and cephalosporins. There was no significant correlation between the use of PcV and the frequency of PNSP (Table 3).
Bacterial resistance to antimicrobial agents has become a worldwide problem both in hospital and community settings. 14 Antibiotic multiresistance has reduced the therapeutic arsenal for common respiratory tract infections as well as for invasive infections. For years the highest incidences of PNSP have been reported from South Africa, Spain and Hungary. 15 Lately several studies from South America have reported a frequency between 15 to 40% of PNSP in invasive infections, 16, 17 and increasing incidences are reported from the US, France and other countries. 15, 18, 19 Most studies are based on the incidence of PNSP in cultures from hospitalized patients. Low incidences of PNSP have been reported from Italy, Germany, Holland, Sweden, Norway and Finland. 15
In Sweden the frequency of PNSP in national laboratory surveys has been 3 to 5%. 1 In southern Sweden reports from Malmö University Hospital showed a frequency of 10% in the late 1980s 4 and at Lund University Hospital 8% in 1992. 3 The incidence of cases of bacteremia caused by PNSP has been <5% each year in the area. 20 More than 80% of the tested cultures in the laboratories have been from the nasopharyngeal cultures of children ages 0 to 6 years with respiratory tract infections. 10 A specimen for culture is rarely obtained at the first visit but is recommended if given empiric therapy fails. In the present study all children age 0 to 6 years with a nasopharyngeal culture positive for PNSP at the three microbiologic laboratories during 12 months were registered in the respective municipality where the child lived. This enabled us to calculate the number of recovered PNSP per 1000 children ages 0 to 6 years in each municipality. Some rural municipalities are small, but all have >1000 children ages 0 to 6 years. The total number of PNSP found during 1 year is not impressive but is illustrative of a steady state 10 years after the first findings of PNSP in the county. The different frequencies of PNSP, comparing results from the national surveys (3 to 5%), Malmö University Hospital (10%) and in this study comparing municipalities (0 to 49.5%), illustrate the difficulties in interpreting resistance data even if reported from the same country. In the present study antibiotic utilization in the area, geographic area and number of PNSP per total number of pneumococci, age and microbiologic method are accounted for. Despite the identical national recommendations for sampling, it can be discussed whether the indications for sampling were equally complied with by the physicians in the different municipalities.
Respiratory tract infections are the major reason for an antibiotic prescription in western countries, especially for children. In Sweden PcV is the recommended drug for empiric treatment of most cases of bacterial respiratory tract infections. Amoxicillin is recommended as a second line antibiotic in cases of acute otitis media after therapeutic failure of PcV. Macrolides and tetracyclines are mainly recommended when suspecting an atypical etiology, e.g. Mycoplasma or Chlamydia. Because of a steadily increasing utilization of antibiotics in Sweden in the 1980s and the beginning of the 1990s, the Swedish Strategic Program for the Rational Use of Antimicrobial Agents and Surveillance of Resistance was launched in 1993. 10 Between 1993 and 1997 the national utilization of antibiotics was reduced by 20%. 11, 21 In Malmöhus county the reduction of antibiotic use in children during 1993 through 1997 was 40%. 11 Despite this the children in Malmöhus County still use more antibiotics than children in other Swedish counties.
Since 1998 not only age, gender and data on the served pharmaceutical agent is registered at the pharmacy but also the patient’s address. The use of a drug by the inhabitants in a specific geographic area can therefore be obtained. Earlier, data were registered on the area in which the pharmacy was located, which gave errors at a municipality level because of commuting, open hours, etc. Therefore we chose to include only the antibiotics statistics for the last 6 months of the study period, in DDD per 1000 children per day per municipality. Compared with earlier statistics, based on the location of the pharmacies, the variation between the areas in the utilization of antibiotics was similar to the present data based on the residency of the consumer. There were surprisingly large differences between neighboring municipalities, both in total usage and in the use of different antibiotics. These large differences cannot likely be explained by differences in the incidence of bacterial infections.
The association between antibiotic utilization and resistance in the community has not been well-examined. An increasing frequency of PNSP in hospitalized patients subsequent to increasing sales of aminopenicillins in the community has been demonstrated in Spain. 22 A study from Iceland indicated a correlation between the total use of antibiotics at a municipality level and the prevalence of PNSP. 23 The relationship between trimethoprim-sulfamethoxazole and PNSP in the present study is in line with findings in previous studies on individual antibiotics use in Iceland, 23 as well as in a day-care study in southern Sweden. 8 In the present study also the utilization of other antibiotics (amoxicillin, macrolides and cephalosporins) was correlated with the frequency of PNSP in a municipality. Both trimethoprim-sulfamethoxazole and macrolides are effective in eradicating the normal flora in the nasopharynx, which might explain why they may enhance the risk of acquiring a PNSP. Knowledge of the effect of different beta-lactams on the nasopharyngeal flora is limited. In a study from Spain use of oral cephalosporins has been suggested to select PNSP. 24 Amoxicillin eradicates pneumococci more effectively than PcV in the nasopharynx 25 and could thus leave an ecologic niche enhancing the establishment of PNSP. Because PcV is used only in the Nordic countries, a comparison with amoxicillin has not been possible in other countries. Even though PcV was the most used antibiotic, there was a lack of correlation between the use of PcV and the frequency of PNSP. Most probably this is because PcV affects the normal flora in the nasopharynx less than the other antibiotic groups. However, the results in this study must be interpreted with precaution, because the municipalities with a high utilization of macrolides and broad spectrum antibiotics also had a high total utilization of antibiotics. Hence the explanation for the higher frequencies of PNSP in these municipalities might rather be the high total antibiotics utilization.
Because the municipality with the highest frequency of PNSP could be considered as an outlier, calculations were also made with Spearman’s correlation coefficient. The respective results did not particularly differ; hence only the results for which Pearson’s correlation coefficient was used are mentioned in the text/tables.
It was not possible to bring forth retrospectively the total number of pneumococcal isolates in 3 of the 20 municipalities. Hence we could not calculate the frequency of PNSP in these municipalities. Further studies are needed with prospective data from a larger number of municipalities to establish the correlation between antibiotic utilization and frequency of PNSP.
In conclusion we found a significant correlation between the utilization of antibiotics in children ages 0 to 6 years and the frequency of PNSP in the community. Trimethoprim-sulfamethoxazole seems to be the most selective antimicrobial drug, followed by amoxicillins, macrolides and cephalosporins. PcV seemed to have the least ecologic impact of all antibiotic groups. The present study supports the importance of proper use of antibiotics, especially in children. This study also indicates that one explanation of the relatively low frequencies of PNSP in Sweden may be the use of PcV as the drug of choice in respiratory tract infections.
We thank the personnel at the participating microbiology laboratories in Lund, Malmö and Helsingborg. The study was supported by grants from the Swedish Society of Medicine, Elsa Lundberg’s and Greta Fleron’s Foundation, Tore Nilsson’s Foundation, Skåne County Council’s Research and Development Foundation and the Network for Pharmacoepidemiology Foundation.
1. Olsson-Liljequist B, Burman LG, Kallings I. Antibiotic susceptibility of upper respiratory tract pathogens in Sweden: a seven year follow-up study including loracarbef. Swedish Respiratory Tract Study Group. Scand J Infect Dis 1992; 24: 485–93.
2. Kihlström E, Normann B. Occurrence of pneumococci with resistance or decreased susceptibility to penicillin in southeast Sweden. Scand J Infect Dis 1995; 27: 489–94.
3. Ekdahl K, Kamme C. Increasing resistance to penicillin in Streptococcus pneumoniae
in southern Sweden. Scand J Infect Dis 1994; 26: 301–5.
4. Forsgren A, Walder M. Antimicrobial susceptibility of bacterial isolates in south Sweden including a 13-year follow-up study of some respiratory tract pathogens. APMIS 1994; 102: 227–35.
5. Ekdahl K, Ahlinder I, Hansson HB, et al. Duration of nasopharyngeal carriage of penicillin-resistant Streptococcus pneumoniae
: experiences from the South Swedish Pneumococcal Intervention Project. Clin Infect Dis 1997; 25: 1113–7.
6. Melander E, Ekdahl K, Hansson HB, et al. Introduction and clonal spread of penicillin and trimethoprim/sulfamethoxazole-resistant Streptococcus pneumoniae
, serotype 9V, in southern Sweden. Microb Drug Resist 1998; 4: 71–8.
7. Ekdahl K, Hansson HB, Mölstad S, Söderström M, Walder M, Persson K. Limiting the spread of penicillin-resistant Streptococcus pneumoniae
: experiences from the South Swedish Pneumococcal Intervention Project. Microb Drug Resist 1998; 4: 99–105.
8. Melander E, Mölstad S, Persson K, Hansson HB, Söderström M, Ekdahl K. Previous antibiotics consumption and other risk factors for carriage of penicillin-resistant Streptococcus pneumoniae
in children. Eur J Clin Microbiol Infect Dis 1998; 17: 834–8.
9. Gunnarsson O, Ekdahl K. Previous respiratory tract infections and antibiotics consumption in children with long- and short-term carriage of penicillin-resistant Streptococcus pneumoniae.
Epidemiol Infect 1998; 121: 523–8.
10. Ekdahl K, Cars O. Role of communicable disease control measures in affecting the spread of resistant pneumococci: the Swedish model. Clin Microbiol Infect 1999; 5: 4S48–4S54.
11. National Corporation of Pharmacies. The prescription survey 1993–1997. Stockholm: 1998.
12. Lund E, Henrichsen J. Laboratory diagnosis, serology and epidemiology of Streptococcus pneumoniae
. Methods Microbiol 1978; 12: 241–62.
13. Skulnick M, Small GW, Lo P, et al. Evaluation of accuracy and reproducibility of E-test for susceptibility testing of Streptococcus pneumoniae
to penicillin, cefotaxime, and ceftriaxone. J Clin Microbiol 1995; 33: 2334–7.
14. Caputo GM, Appelbaum PC, Liu HL. Infections due to penicillin-resistant pneumococci. Arch Intern Med 1993; 153: 1301–10.
15. Baquero F. Pneumococcal resistance to beta-lactam antibiotics: a global geographic overview. Microb Drug Resist 1995; 1: 115–20.
16. Berezin EN, Carvalho ES, Casagrande S, Brandileone MC, Mimica IM, Farhat CK. Streptococcus pneumoniae
penicillin non-susceptible strains in invasive infections in Sao Paulo, Brazil. Pediatr Infect Dis J 1996; 15: 1051–3.
17. Hortal M, Algorta G, Bianchi I, et al. Capsular type distribution and susceptibility to antibiotics of Streptococcus pneumoniae
clinical strains isolated from Uruguayan children with systemic infections. Microb Drug Resist 1997; 3: 159–63.
18. Kaplan SL, Mason EO Jr, Barson WJ, et al. Three-year multicenter surveillance of systemic pneumococcal infections in children. Pediatrics 1998; 102: 538–45.
19. Geslin P, Fremaux A, Sissia G, Spicq C. Streptococcus pneumoniae
: serotypes, souches invasives et resistantes aux antibiotiques—situation actuelle en France. Presse Med 1998; 27 (Suppl 1): 21–7.
20. Ekdahl K, Mårtensson A, Kamme C. Bacteraemic pneumococcal infections in Southern Sweden 1981–96: trends in incidence, mortality, age-distribution, serogroups and penicillin-resistance. Scand J Infect Dis 1998; 30: 257–62.
21. Mölstad S, Cars O. Major change in antibiotics use following a national program. Scand J Infect Dis 1999; 31: 191–5.
22. Baquero F, Martinez Beltran J, Loza E. A review of antibiotic resistance patterns of Streptococcus pneumoniae
in Europe. J Antimicrob Chemother 1991; 28 (Suppl C): 31–8.
23. Arason VA, Kristinsson KG, Sigurdsson JA, Stefánsdóttir G, Mölstad S, Gudmundsson S. Do antimicrobials increase the carriage rate of penicillin resistant pneumococci in children? Cross sectional prevalence study. BMJ 1996; 313: 387–91.
24. Baquero F. Trends in antibiotic resistance of respiratory pathogens: an analysis and commentary on a collaborative surveillance study. J Antimicrob Chemother 1996; 38 (Suppl A): 117–32.
25. Eliasson I, Holst E, Molstad S, Kamme C. Emergence and persistence of beta-lactamase-producing bacteria in the respiratory tract in children treated with beta-lactam antibiotics. Am J Med 1990; 88 (5A):51S–55S.