Prediction of the potential benefit of different pneumococcal conjugate vaccines on invasive pneumococcal disease in German children

Transcription

1 Pediatr Infect Dis J, 2002;21: Vol. 21, No. 11 Copyright 2002 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A. Prediction of the potential benefit of different pneumococcal conjugate vaccines on invasive pneumococcal disease in German children RÜDIGER VON KRIES, MD, MONIKA HERMANN, MSC, ALEXANDRA HACHMEISTER, ANETTE SIEDLER, PHD, HEINZ J. SCHMITT, MD, ADNAN AL-LAHHAM, PHD AND RALF RENÉ REINERT, MD Background. In the US a pneumococcal conjugate vaccination program with a 7-valent conjugate vaccine was successfully implemented in How much invasive pneumococcal disease can potentially be prevented by the 7-valent (or 11-valent) vaccine in Europe? Methods. Prospective, active surveillance of invasive pneumococcal disease in German children age <16 years was performed between 1997 and Age- and disease-specific coverage and incidence rates were assessed in children old enough to benefit from complete vaccination to estimate the annual number of cases potentially preventable. Results. A total of 1,743 cases were reported; 667 isolates were serotyped. Coverage of 7-valent (11-valent) conjugate vaccines in children age 6 months and older was age- and diagnosisdependent, ranging from 10.5% (15.8%) to 78.3% (82.6%) for meningitis and from 13.6% (68.2%) to 75.0% (89.3%) for nonmeningitis invasive pneumococcal disease cases. Of an estimated annual number of 176 children with pneumococcal meningitis age 6 months or older, 112 (122) cases had serotypes included in the 7-valent (11-valent) conjugate vaccine compared with 181 (254) of 324 nonmeningitis invasive pneumococcal disease cases, with 37 of the 73 cases covered by the Accepted for publication July 8, From Institute for Social Pediatrics, Ludwig-Maximilians- Universität, Munich (RVK, MH, AH); Robert-Koch-Institute, Berlin (AS); Center for Preventive Paediatrics, Johannes Gutenberg- Universität, Mainz (HJS); and National Reference Centre for Streptococci, Institute of Medical Microbiology, Rheinisch- Westfälische Technische Hochschule, Aachen (AA, RRR), Germany. Key words: Pneumococcal conjugate vaccine, invasive pneumococcal disease, surveillance, serotype coverage, capturerecapture. Address for reprints: Rüdiger von Kries, M.D., M.Sc., Department of Pediatric Epidemiology, Institute of Social Pediatrics and Adolescent Medicine, Ludwig-Maximilians-University Munich, Heiglhofstrasse 63, D München, Germany. Fax /315; ag.epi@lrz.uni-muenchen.de. DOI: /01.inf a valent vaccine only in children older than 5 years. Regarding meningitis in this age group the potential benefit was equally poor for both the 7-valent (12 of 37 cases) and the 11-valent vaccine (15 of 37 cases). Conclusion. Coverage of the 7- and 11-valent conjugate vaccines depends markedly on age and disease. The additional potential benefit of the 11-valent compared with the 7-valent vaccine for pneumococcal meningitis was marginal. INTRODUCTION Streptococcus pneumoniae is a significant cause of meningitis and septicemia in infants and young children and is associated with high case fatality rates and proportions of serious sequelae. 1 5 In the United States a pneumococcal conjugate vaccination program with a 7-valent conjugate vaccine was successfully implemented in This prompted a decline in the rates of invasive pneumococcal diseases (IPD) that was greater than predicted from the vaccination rates, suggesting herd immunity. 6 Replacement of nonvaccine serotypes in IPD cases was not reported. 6 The registration of the 7-valent pneumococcal conjugate vaccine in Europe was not followed by similar pneumococcal conjugate vaccination programs, in spite of the high rates of both macrolide and penicillin resistance reported from many European countries Furthermore the rates for pneumococcal meningitis in children in Europe are similar to those in the US. 11 One reason for the reluctance to follow the US example in Europe may be the 10 to 15% lower estimated coverage for childhood IPD cases with the 7-valent pneumococcal conjugate vaccine in Europe. 12 Does the 7-valent vaccine really cover 10 to 15% fewer IPD cases in Europe than in the US? Because only children older than 5 months have a chance to be fully vaccinated and because most invasive pneumococcal disease affects children younger than 5 years, the more relevant question is: What proportion of invasive disease in European children older than 5 months and 6 years of age are infected with strains represented in the 7-, 9- and 11-valent vaccine? The 7-valent vaccine

2 1018 THE PEDIATRIC INFECTIOUS DISEASE JOURNAL Vol. 21, No. 11, Nov includes serotypes 4, 6B, 9V, 14, 18C, 19F and 23F. The 9-valent conjugate additionally contains serotypes 1 and 5, whereas the 11-valent conjugate vaccine additionally contains serotypes 1, 5, 3 and 7F. In a comprehensive 4-year surveillance study for childhood IPD in Germany, the coverage of serotypes included in 7-, 9- and 11-valent pneumococcal conjugate vaccines for different age groups and diagnoses was examined. Moreover the number of potentially preventable IPD cases per year was assessed. Because the pneumococcal serotype distribution in Germany is similar to that observed in other European countries, 1, 3 5, 13, 14 it is possible that these data can serve as a model for the potential impact of current and future pneumococcal vaccines in Europe. MATERIALS AND METHODS Study design. This study was based on prospective, active surveillance of IPD in the German population age 16 years. As of December 31, 1997, the size of the population at risk was (Federal Statistics Office, Wiesbaden, Germany). The study period was from January 1, 1997, to December 31, The first 2 years of surveillance have been reported previously. 5 Patients were enrolled in the study if they had been admitted to a pediatric hospital and if S. pneumoniae had been isolated from at least one culture of blood, cerebrospinal fluid or a sample from any other normally sterile body site. Cases were identified through two surveillance systems, one hospital-based and the other one laboratorybased. They were independently reported by the pediatric hospitals and by the microbiology laboratories on a monthly basis by returning postcards/questionnaires. A nothing to report option was included to estimate participation in the surveillance system. Physicians reporting cases of IPD received a questionnaire and were asked to give additional information about the site of the infection, clinical diagnosis and outcome. On their monthly postcards laboratories gave information on the body site from which S. pneumoniae had been isolated and on the age and gender of the child concerned. Cases with S. pneumoniae isolation from cerebrospinal fluid were categorized as meningitis cases regardless of the presence of other positive cultures (e.g. blood cultures). Cases where S. pneumoniae was isolated from other body sites were categorized as nonmeningitis cases unless there was both a positive blood culture and a physician s diagnosis of meningitis (based on the clinical picture and the cell count in the cerebrospinal fluid). Case reports from hospital surveillance and laboratory surveillance were linked on the basis of the second letters of the patient s first and surname, month and year of birth, gender, date of sampling in the hospital/date of culturing of the pathogen in the laboratory, place of hospital/laboratory and laboratory identification number. To assure accurate linkage of data, an automated linkage procedure was supplemented by additional linking by an independent person. Microbiologic investigations. All microbiology laboratories were asked to send pneumococcal strains isolated from a normally sterile body site to the German National Reference Centre for Streptococci for confirmation and serotyping. Each isolate was confirmed as S. pneumoniae by optochin sensitivity and by bile solubility. Capsular typing was conducted by the quellung reaction with group and factor sera provided by Statens Seruminstitut (Copenhagen, Denmark). Statistics. Vaccine coverage was calculated as the proportion of serotypes that would be included in current and forthcoming conjugate vaccines among cases typed between 1997 and Pearson-Clopper confidence intervals were used for vaccine coverage to allow for variability of coverage. The number of (age-specific) IPD cases observed in children in Germany was derived from the cases ascertained by either source (aggregated registry). The true number of cases (ascertainment corrected estimate) was estimated using the two source capturerecapture method. 15 Incidence of IPD cases was determined by dividing these estimators by the sum of person-years (within the respective age group) using official population estimates from the German Federal Statistical Office. The annual amount of meningitis/nonmeningitis IPD cases caused by serotypes covered by the pneumococcal conjugate vaccines was assessed for different age groups by multiplying age-specific capture-recapture (ascertainment corrected estimates) incidences, coverage in our data and the annual number of children at risk in the respective age stratum from the German Federal Bureau of Statistics. The confidence intervals consisted of the lower/upper boundaries of the confidence interval of the vaccine coverage estimator multiplied with the expected annual number of cases per age stratum. Summarized annual numbers of cases (i.e. age categories 0 to 15 years and 0.5 to 5 years) in Tables 2 and 3 were calculated as the sum of all age-specific cases. Accordingly the respective proportions of cases covered by the respective vaccine in these summarized age categories were calculated as the number of expected annual cases covered by the respective vaccine divided by the total number of expected annual cases in these tables. RESULTS In the hospital surveillance system the average return rates of the monthly reporting cards was 95.1%. From the questionnaires that had been prompted by

3 Vol. 21, No. 11, Nov THE PEDIATRIC INFECTIOUS DISEASE JOURNAL 1019 TABLE 1. Cumulative incidence rates per 10 5 person-years of IPD cases in Germany during the surveillance period 1997 to 2000 Incidence Rates/10 5 Person-Years 2 yr 5 yr 16 yr Cases recorded by either source All IPD Meningitis Nonmeningitis IPD Population estimates (capture-recapture) All IPD Meningitis Nonmeningitis IPD the initial reports on the monthly cards, 97.2% were sent back to the study center. In the laboratory surveillance system the respective return rates for the monthly reports were 95.5%. In total 841 valid cases could be ascertained by the hospital surveillance system and 1472 valid cases could be ascertained by the laboratory surveillance system. After eliminating 570 cases reported to both systems, an overall number of 1743 cases were collected by both sources together (meningitis, 741; nonmeningitis IPD, 1002). The estimated total number of meningitis cases was 837 and that for nonmeningitis cases was 1409, bringing the total number to 2246 cases with adjustment for underreporting by the capture-recapture method and suggesting that 88.5% of the meningitis and 71.1% of the other cases were reported by either source in the surveillance system. The incidence estimates for the conventionally reported age categories of pneumococcal meningitis, nonmeningitis cases and all IPD cases combined are shown in Table 1. The highest incidence was observed in the first 2 years of life with a rapid decline thereafter; 60% of all patients were male. Details concerning clinical diagnoses and outcome were available in the 841 cases reported by the hospital surveillance system. The 841 cases consisted of 494 (58.7%) meningitis cases, 314 (37.3%) cases with a positive blood culture and no evidence of meningitis, 17 (2.0%) cases of pneumonia with isolates from pleural empyema only and 16 cases (1.9%) with pneumococcal isolates from other sterile sites (joint, bones, peritoneum, or abscess material). The proportion of children with missing information or outcome not clear yet in the hospital surveillance system was 12.3% (61 cases) in the meningitis cases and 7.8% (27 cases) in the nonmeningitis IPD cases. According to the assessment of the treating physician in the 494 meningitis cases, 283 children (57.3%) recovered completely, 41 children (8.3%) died and 109 (22.1%) children recovered with sequelae. These sequelae included 57 cases with central nervous system (CNS) damage, 40 cases with hearing deficit or deafness and 12 cases with multiple or other defects, which comprised 8 cases with hearing loss (deafness) plus CNS damage, 2 cases with skin defects resulting from disseminated intravascular coagulation, 1 case with CNS damage plus persistent renal failure and 1 case requiring revision of a ventriculoperitoneal shunt. In the 347 nonmeningitis IPD cases 295 children (85.0%) recovered completely, 7 children (2.0%) died and 18 (5.2%) children recovered with sequelae, namely a case with a postischemic infarction occurring during the course of septicemia, 1 case with hearing deficit or deafness, 11 cases with pleural thickening and 5 cases with multiple (pleural thickening plus sacroileitis, neurologic plus cardiac symptoms) or other defects (splenomegaly, myocarditis or skin defects). Serotyping was achieved for 667 of the 1743 reported cases. In total 43 different serotypes belonging to 28 serogroups could be identified. In 3 cultures the serogroup was known, but no information about serotype was available. The 10 most common serotypes were 14 (23.5%), 18C (7.9%), 1 (7.6%), 19F (6.7%), 23F (6.7%), 6B (5.4%), 4 (4.9%), 7F (4.6%), 9V (4.5%) and 19A (4.5%). During the surveillance period the proportion of serotype 14 in children age 16 years increased from 15.6% in 1997 to 32.5% in Coverage by the 7-, 9- and 11-valent vaccines for all childhood IPD cases was 59.8, 68.5 and 76.6%, respectively. Serotype 1 accounted for 7.6% of the increase in coverage from the 7-valent to the 9-valent vaccine, whereas only 1.0% belonged to serotype 5. The increase in coverage from the 9-valent to the 11-valent vaccine consisted of 4.6% for serotype 7F and 3.4% for serotype 3. The differences in coverage of the 7-, 9- and 11-valent vaccines were age-dependent as shown in Figure 1. In meningitis cases (Fig. 1A) the coverage of the 7-valent vaccine increased from 22% in the age group 0 to 2 months to 70% in the age group 6 to 11 months, remained above 60% until the age of 5 years and then decreased to 11% in the 10- to 15-year age group. Fourteen different serotypes could be identified in the 19 typed cases of meningitis in children age 10 to 15 years. Coverage of the 9- and 11-valent vaccines in meningitis cases was within the 95% confidence interval of the age-specific coverage estimates of the 7-valent vaccine with exception of the young infants age 0 to 5 months. In this age group the 11-valent vaccine covered 83.0%, 34.0% above that of the 7-valent vaccine, of all cases with point estimates outside the 95% confidence interval of the 7-valent conjugate vaccine. This substantial difference was mainly because serotype 7F accounted for 17.0% of all meningitis cases in infants age 0 to 5 months. In nonmeningitis cases (Fig. 1B) the coverage of the 7-, 9- and 11-valent vaccine gave a picture similar to that for the meningitis cases until the age of 5 years

4 1020 THE PEDIATRIC INFECTIOUS DISEASE JOURNAL Vol. 21, No. 11, Nov FIG. 1.A, coverage rates of age-specific typed cases of pneumococcal meningitis cases in Germany during the surveillance period 1997 to 2000; B, coverage rates of age-specific typed cases of nonmeningitis IPD cases in Germany during the surveillance period 1997 to OO, 7-valent serotypes; top, upper 95% confidence interval for 7-valent serotypes; bottom, lower 95% confidence interval for 7-valent serotypes;, 9-valent serotypes;, 11-valent serotypes. except for the 9-valent serotype till age 1 year. In contrast to meningitis the difference of 36.4% in coverage of the 11-valent vaccine (79.5%) to the 7-valent vaccine (43.2%) of the nonmeningitis IPD in infants age 0 to 5 months is mainly caused by serotype 3, accounting for 22.7% of these cases. In children age 6 to 15 years 50.0% of all nonmeningitis cases were caused by serotype 1, which is included in the 9-valent vaccine and accounted for most of the difference in coverage between 7-valent and the 11-valent vaccines in this age group. The impact of the age-specific coverage rates on the expected annual number of meningitis cases is presented in Table 2. The expected annual number of pneumococcal meningitis cases was 210, with 13 cases occurring in those 2 months of age or below (i.e. before the first vaccine inoculation is recommended) and a further 21 meningitis cases in children age 3 to 5 months who may be only partially vaccinated. The respective number of pneumococcal meningitis cases that would have benefited from complete vaccination is 176. Of these cases 112 were covered by the 7-valent vaccine, 116 by the 9-valent vaccine and 122 by the 11-valent vaccine. Adding cases that might have at least partially benefited from vaccination (age 3 to 5 months), the figure increased to 197 cases, for which

5 Vol. 21, No. 11, Nov THE PEDIATRIC INFECTIOUS DISEASE JOURNAL 1021 TABLE 2. Model analysis of annual amount of pneumococcal meningitis cases in Germany Age Group Yearly Incidence ( 10 5 )of Meningitis Meningitis Estimation of No. of Cases/yr Cases covered by x-valent vaccine 11-valent 9-valent 7-valent 0 2 mo (6.7, 11.9)* (2.2, 8.2) (0.8, 6.0) 77.8% 38.9% 22.2% 3 5 mo (14.3, 20.2) (11.9, 18.8) (9.6, 17.2) 86.2% 75.9% 65.5% 6 11 mo (33.1, 45.1) (31.3, 43.8) (30.4, 43.1) 74.6% 71.4% 69.8% 1 yr (25.6, 34.3) (25.6, 34.3) (24.2, 33.2) 78.3% 78.3% 75.0% 2 yr (11.8, 18.3) (11.8, 18.3) (10.9, 17.8) 82.6% 82.6% 78.3% 3 yr (4.8, 9.9) (4.8, 9.9) (4.8, 9.9) 73.3% 73.3% 73.3% 4 5 yr (8.7, 16.1) (6.4, 14.8) (6.4, 14.8) 80.0% 66.7% 66.7% 6 9 yr (7.3, 15.8) (6.3, 15.1) (5.5, 14.3) 63.2% 57.9% 52.6% yr (0.6, 7.1) (0.2, 6.0) (0.2, 6.0) 15.8% 10.5% 10.5% 0 15 yr % 65.5% 61.5% 6 71 mo (0.5 5 yr) % 74.5% 72.3% * Numbers in parentheses, 95% confidence interval around number of expected yearly cases based on variation of vaccine coverage. Percentage are the proportion of cases covered by respective vaccine. 126 cases were covered by the 7-valent vaccine, 132 cases were covered by the 9-valent vaccine and 140 cases were covered by the 11-valent conjugate vaccine. For all conjugate vaccines the coverage in meningitis cases decreased substantially in children age 10 to 15 years. In conclusion there were no age strata with extreme differences in the number of meningitis cases that would be covered by the different vaccines, with the exception of infants (age 0 to 2 months), who are too young for any pneumococcal conjugate vaccine given according to the German vaccination schedule (3 doses at 2, 3 and 4 months plus booster dose in the second year of life). Table 3 shows the expected annual number of nonmeningitis IPD cases, which was 358 in total: 16 of these were observed in children age 0 to 2 months (i.e. before the first vaccine shot of pneumococcal conjugate vaccines is recommended). An additional 18 nonmeningitis IPD cases occurred in children age 3 to 5 months, who would have been only partially vaccinated. The respective number of nonmeningitis IPD cases that would have benefited from complete vaccination is 324. Of these cases 181 were covered by the 7-valent vaccine, 231 by the 9-valent vaccine and 254 by the 11-valent vaccine. Compared with the 7-valent vaccine, the additional benefit in coverage would be 50 cases for the 9-valent vaccine ( 15%) and 73 cases for the 11-valent vaccine ( 23%). DISCUSSION With a comprehensive surveillance system that used two independent sources for reporting IPD cases in children and the application of the capture-recapture method to adjust for reporting bias, we found pneumococcal meningitis rates that were similar to other 1, 3, 4, 14, 16 European countries. As pointed out previously by ourselves 5 and by others 11 underascertainment of nonmeningitis cases is likely because blood culturing in hospitalized children with fever of unknown origin appears to be less common in Germany than in the US; furthermore it is very rarely performed in German outpatients, which accounted for about two-thirds of the nonmeningitis IPD cases in the US. 17 The case fatality rate for pneumococcal meningitis in Germany (8.3%) was somewhat lower than in other European countries. 1, 3, 4 Typed and untyped cases did not differ markedly with respect to age (median), gender, outcome (death or sequelae) or diagnosis (meningitis/other IPD) (data not shown). Serotyped cases therefore are likely to be representative of all childhood IPD cases in Germany. The most important finding of our study was the

6 1022 THE PEDIATRIC INFECTIOUS DISEASE JOURNAL Vol. 21, No. 11, Nov TABLE 3. Model analysis of annual amount of pneumococcal invasive nonmeningitis IPD in Germany Age Group Yearly incidence ( 10 5 )of Nonmeningitis IPD Nonmeningitis IPD Estimation of No. of Cases/yr Cases covered by x-valent vaccine 11 -valent 9-valent 7-valent 0 2 mo (9.5, 15.3) (5.4, 12.3) (3.1, 9.8) 79.2% 54.2% 37.5% 3 5 mo (10.1, 17.0) (4.9, 13.1) (4.9, 13.1) 80.0% 50.0% 50.0% 6 11 mo (45.4, 59.4) (38.0, 53.6) (33.9, 50.2) 78.3% 68.1% 62.3% 1 yr (63.0, 76.3) (56.6, 71.0) (54.5, 69.2) 77.8% 71.1% 68.9% 2 yr (25.8, 33.8) (24.7, 33.2) (20.7, 30.8) 87.5% 85.0% 75.0% 3 yr (18.2, 25.9) (17.4, 25.4) (15.8, 24.3) 79.5% 76.9% 71.8% 4 5 yr (18.7, 31.7) (14.8, 28.7) (8.9, 22.9) 71.4% 60.7% 42.9% 6 9 yr (21.5, 29.2) (17.6, 27.4) (3.9, 14.6) 89.3% 78.6% 28.6% yr (15.9, 30.4) (14.4, 29.2) (1.0, 12.3) 68.2% 63.6% 13.6% 0-15 yr % 69.4% 54.6% 6 71 mo (0.5 5 yr) % 71.4% 64.6% * Numbers in parentheses, 95% confidence interval around number of expected yearly cases based on variation of vaccine coverage. Percentages are the proportion of cases covered by respective vaccine. relationship of the serotypes to age and diagnosis. A poor coverage of the 7-valent vaccine compared with the 9- and 11-valent vaccines was mainly observed in children too young to have been immunized by any pneumococcal conjugate vaccine and in children older than 5 years of age for meningitis and older than 3 years for nonmeningitis cases. Although poor coverage in older children has also been reported in other studies, 1, 13 the brisk decline in coverage after the age of five was clearly illustrated in our experience. More importantly the poor coverage after the age of 5 years in pneumococcal meningitis cases poses a problem for the 9- and 11-valent vaccines as well. The additional benefit of these vaccine in the prevention of potentially vaccine-preventable pneumococcal meningitis in children older than 5 years was only 3 and 8%, respectively. The overall better performance of the 9- and 11-valent vaccines in children older than 5 years was mainly related to the nonmeningitis IPD cases that accounted for a 50 and 57% better coverage, respectively, in potentially vaccine preventable nonmeningitis IPD cases. Death and serious sequelae were mainly a consequence of pneumococcal meningitis. The additional benefit of the 11-valent vaccine for pneumococcal meningitis in children age 6 months or older was only 5.7%. If prevention of death and sequelae is the focus of a pneumococcal vaccination program, then there is no reason to wait for licensing of the higher valent pneumococcal vaccines. Because serotype coverage in meningitis in older children even for the 11-valent vaccine was poor, giving children an additional dose of the 23-valent polysaccharide vaccine at school entry (at 6 years of age) might be an option in programs for prevention IPD in children. ACKNOWLEDGMENTS We thank the Erhebungseinheit für seltene pädiatrische Erkrankungen in Deutschland for continuous support for this study and all reporting physicians for their time and efforts. This study was financially supported in part by a grant from Deutsches Zentrum für Luft-und Raumfahrt, Projektträger des Bundesministeriums für Bildung und Forschung (PID-ARI-Net, Förderkennzeichen 01KI9913/0) and by Wyeth-Lederle, Münster, Germany. REFERENCES 1. Miller E, Waight P, Efstratiou A, Brisson M, Johnson A, George R. Epidemiology of invasive and other pneumococcal disease in children in England and Wales Acta Paediatr Suppl 2000;89: Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States in 1995: Active Surveillance Team. N Engl J Med 1997;337: Spanjaard L, van der Ende A, Rumke H, Dankert J, van Alphen L. Epidemiology of meningitis and bacteraemia due to

7 Vol. 21, No. 11, Nov THE PEDIATRIC INFECTIOUS DISEASE JOURNAL 1023 Streptococcus pneumoniae in the Netherlands. Acta Paediatr Suppl 2000;89: Venetz I, Schopfer K, Muhlemann K. Paediatric, invasive pneumococcal disease in Switzerland, Swiss Pneumococcal Study Group. Int J Epidemiol 1998;27: von Kries R, Siedler A, Schmitt HJ, Reinert RR. Proportion of invasive pneumococcal infections in German children preventable by pneumococcal conjugate vaccines. Clin Infect Dis 2000;31: Black SB, Shinefield HR, Hansen J, Elvin L, Laufer D, Malinoski F. Postlicensure evaluation of the effectiveness of seven valent pneumococcal conjugate vaccine. Pediatr Infect Dis J 2001;20: Fenoll A, Jado I, Vicioso D, Berron S, Yuste JE, Casal J. Streptococcus pneumoniae in children in Spain: Acta Paediatr Suppl 2000;89: Geslin P, Fremaux A, Sissia G, Spicq C. Streptococcus pneumoniae: serotypes, invasive and antibiotic resistant strains: current situation in France (in French). Presse Med 1998; 27(Suppl 1): Marton A, Meszner Z. Epidemiologic studies on drug resistance in Streptococcus pneumoniae in Hungary: an update for the 1990s. Microb Drug Resist 1999;5: Overweg K, Hermans PW, Trzcinski K, Sluijter M, de Groot R, Hryniewicz W. Multidrug-resistant Streptococcus pneumoniae in Poland: identification of emerging clones. J Clin Microbiol 1999;37: Hausdorff WP, Siber G, Paradiso PR. Geographic differences in invasive pneumococcal disease rates and serotype frequency in young children. Lancet 2001;357: Hausdorff WP, Bryant J, Paradiso PR, Siber GR. Which pneumococcal serogroups cause the most invasive disease: implications for conjugate vaccine formulation and use: part I. Clin Infect Dis 2000;30: Bundesamt für Gesundheit. Invasive Pneumokokken- Infektionen in der Schweiz. Epidemiol Bull 2000;20: Kaltoft, Zeuthen N, Konradsen HB. Epidemiology of invasive pneumococcal infections in children aged 0 6 years in Denmark: a 19-year nationwide surveillance study. Acta Paediatr Suppl 2000;89: McCarty DJ, Tull ES, Moy CS, Kwoh CK, LaPorte RE. Ascertainment corrected rates: applications of capturerecapture methods. Int J Epidemiol 1993;22: Eskola J, Takala AK, Kela E, Pekkanen E, Kalliokoski R, Leinonen M. Epidemiology of invasive pneumococcal infections in children in Finland. JAMA 1992;268: Zangwill KM, Vadheim CM, Vannier AM, Hemenway LS, Greenberg DP, Ward JI. Epidemiology of invasive pneumococcal disease in southern California: implications for the design and conduct of a pneumococcal conjugate vaccine efficacy trial. J Infect Dis 1996;174: Pediatr Infect Dis J, 2002;21: Vol. 21, No. 11 Copyright 2002 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A. Decreased number of antibiotic prescriptions in office-based settings from 1993 to 1999 in children less than five years of age NATASHA B. HALASA, MD, MARIE R. GRIFFIN, MD, MPH, YUWEI ZHU, MD, MS AND KATHRYN M. EDWARDS, MD Objective. Increasing rates of antibiotic resistance have stimulated efforts to decrease antibiotic use. To assess the success of these efforts, we analyzed antibiotic prescribing trends in children younger than 5 years old, the group with the highest use, from 1993 to Methods. Data from the National Ambulatory Medical Care Survey were analyzed to determine antibiotic prescribing patterns for office-based visits from 1993 to 1999 for children <5 years old. Accepted for publication July 1, From the Departments of Pediatrics (NBH, YZ, KEM) and Preventive Medicine and Medicine (MRG), Vanderbilt University School of Medicine, Nashville, TN. Key words: Antibiotic use, antibiotic resistance, Streptococcus pneumoniae, children, National Ambulatory Medical Care Survey. Reprints not available. DOI: /01.inf e1 Data were stratified by US regions, patient s race and gender. Antibiotic prescription rates per 1000 population were calculated with population data from the US Census Bureau as the denominator. Specific prescribing of penicillins, cephalosporins, macrolides and sulfas was also assessed. Results. Overall antibiotic prescribing in the office-based setting peaked in 1995 at 1191 antibiotic courses per 1000 children, then declined to 698 per 1000 in 1999, a decrease of 41%. Antibiotic prescribing was consistently higher in whites than blacks; however, declines in prescribing over time were observed in both groups. Although there was wide regional variation in antibiotic prescribing in the early 1990s, by the late 1990s prescribing rates were similar in all regions. Prescriptions for penicillins and cephalosporins combined comprised 77 and 70% of total