Scandinavian Journal of Infectious Diseases, 2010; 42: 254–259 ORIGINAL ARTICLE I nvasive disease and paediatric carriage of Streptococcus pneumoniae in Ghana ERIC S. DONKOR1 ,2, M ERCY J. N EWMAN 1 , JOSEPH O LIVER-COMMEY3 , ELIZABETH B ANNERMAN1 , NICHOLAS T. K. D. D AYIE 1 & EBENEZER V. B ADOE3 F rom the 1 Department of Microbiology, University of Ghana Medical School, Accra, Ghana, 2 Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, University of London, London, UK, and 3 Department of Child Health, University of Ghana Medical School, Accra, Ghana Abstract This study was carried out primarily to evaluate the public health burden related to Streptococcus pneumoniae in Ghana and to provide related preliminary molecular epidemiological data on the organism. Invasive and nasopharyngeal specimens were screened for S. pneumoniae, and isolates were subjected to serotyping, multilocus sequence typing (MLST) and antibiotic susceptibility testing. Overall, the prevalence of S. pneumoniae in cerebrospinal fl uid (CSF) was 1.7%, in blood was 0.2%, and in nasopharyngeal specimens was 15.3%. The prevalence of multiple drug resistance among the isolates was 48.6%, while the percentage resistance to various drugs was in the range of 11.1–84.0%. Serotyping of the S. pneumoniae isolates showed 7 different serotypes (3, 6B, 9, 10, 14, 16 and 23F). The extent of coverage of serotypes by the 7-valent pneumococcal conjugate vaccine was 57.1%, for the 10-valent vaccine was 57.1%, and for the 13-valent vaccine was 71.4%. MLST of 7 housekeeping genes of the organism showed a high level of genetic diversity among the isolates. S. pneumoniae appears to be an important organism in invasive infections in Ghana, being the most prevalent organism in CSF in this study. The high multiple drug resistance of the organism observed heightens the public health burden, which may be controlled by pneumococcal conjugate vaccines to a large extent. Introduction Resistant strains of the organism have been reported Streptococcus pneumoniae is part of the normal in both developing and developed countries, and bacterial fl ora of the upper respiratory tract, but is contribute to the high mortality of its diseases. also associated with several invasive infections, includ- The enormous burden of S. pneumoniae, especially ing meningitis, pneumonia, and sepsis. Worldwide, as a result of its increasing and widespread drug the annual mortality of meningitis and pneumonia resistance, underscores the importance of control due to the organism is over 1.6 million [1,2]. Condi- through vaccination. Quite recently, conjugate vac- tions including HIV infection, sickle cell anaemia cines for S. pneumoniae have been developed to pre- and a variety of organ failures are associated with an vent infections in children, and the vaccines have increased risk of serious S. pneumoniae diseases. been shown to be superior to the previous polyvalent Though a global problem, the public health impact polysaccharide vaccine of the organism [4,5]. of S. pneumoniae is higher in the developing world, In Ghana the isolation rate of S. pneumoniae is where children 5 years of age are most affected, low, partly because of a lack of laboratory facilities and the organism is responsible for 10–20% of all and technical expertise required for the isolation of deaths in this age group [1,3]. this organism. As a result there is inadequate infor- T he public health burden related to S. pneumo- mation on its public health impact, though studies niae is heightened by the increasing resistance of the in some parts of the country have reported a high organism to essential antimicrobial drugs, parti- incidence of S. pneumoniae in meningitis [6,7]. cularly penicillin, cephalosporins and macrolides. Additionally, little is known epidemiologically about C orrespondence: E. Sampane-Donkor, Room 204b, Pathogen Molecular Biology Unit, Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK. Tel: +44 20 7927 2032. E-mail: eric.sampane-donkor@lshtm.ac.uk. (Received 15 July 2009; accepted 11 November 2009) ISSN 0036-5548 print/ISSN 1651-1980 online © 2010 Informa UK Ltd. (Informa Healthcare, Taylor & Francis AS) DOI: 10.3109/00365540903490000 Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only. S. pneumoniae in disease and carriage, Ghana 255 the organism, especially in relation to vaccination, isolates were serotyped using the latex agglutination which could be a major constraint to the effective reaction with specifi c antisera [9]. Multilocus control of the disease burden in Ghana. This sequence typing (MLST) was carried out on 7 research paper reports primarily on the public bur- housekeeping genes of the organism: a roE , gdh , gki, den related to S. pneumoniae in terms of its invasive r ecp , s pi , xpt and d dl . Standard protocols were fol- diseases and drug resistance, and also provides lowed for polymerase chain reaction (PCR) and preliminary data on the molecular epidemiology of direct sequencing of PCR products with the aid of the organism in Ghana. an ABI Prism 310 genetic analysis system [11]. Allelic profi les were analysed with the aid of applica- tions on the MLST homepage (http://spneumoniae. M aterials and methods mlst.net). The ATCC 49619 strain of S. pneumoniae was used as a control in the laboratory tests. Study sites and subjects T he study was carried out at 3 major hospitals in Ghana from July 2006 to December 2007. During Data analysis this period, specimens of cerebrospinal fl uid (CSF) T he data collected were entered in MS-Excel and were collected from meningitis patients, while blood analysed to address the objectives of the study. The specimens were taken from pneumonia and septicae- strategies taken to analyse the data involved descrip- mia patients. The routine at the study hospitals was tive statistics, including geometric means, frequen- to collect specimens from all patients who met the cies, ranges and prevalence rates of the study diagnosis inclusion criteria. Demographic character- variables. Signifi cant differences, associations and istics of the patients, as well as conditions predispos- interrelationships of the variables were also assessed ing to increased risk of pneumococcal infection were at a level of p  0.05. Specifi c analyses were carried also collected. A survey was also carried out to deter- out to: (1) determine prevalence rates of S. pneumo- mine the nasopharyngeal carriage of S. pneumoniae niae from specimens including blood, CSF and among children aged 13 years. This was carried out nasopharyngeal swabs; (2) compare prevalence rates at the Child Health Department of one of the study of S. pneumoniae with other isolated organisms from hospitals, namely Korle-Bu Teaching Hospital. This blood and CSF; (3) determine prevalence rates of hospital is one of the largest in Ghana, and also a S. pneumoniae resistance to various antimicrobial major referral centre. Because patients visit Korle-Bu drugs; (4) compare S. pneumoniae penicillin resis- Teaching Hospital from all over the country, the tance in invasive disease and carriage; (5) estimate hospital was selected for the carriage study, in order the burden of S. pneumoniae infections; (6) compare to obtain a more representative sample. Based on the burden of S. pneumoniae in different age groups; 95% confi dence limits with an allowable error of (7) assess the serotypes of S. pneumoniae isolated 10%, we randomly sampled at least 124 healthy from study subjects; (8) evaluate the genetic diversity children visiting Korle-Bu Teaching Hospital for a among the S. pneumoniae isolates; (9) assess the review; nasopharyngeal specimens were collected extent of coverage of S. pneumoniae serotypes by from the children for laboratory analysis. pneumococcal conjugate vaccines. L aboratory analysis of specimens R esults T he specimens were cultured for S. pneumoniae on Numbers of study subjects, clinical specimens blood agar supplemented with 5% sheep blood agar, and isolates and suspected colonies of S. pneumoniae were confi rmed by the optochin inhibition test [8]. CSF Over the study period, a total of 4566 patients at the specimens were also screened for S. pneumoniae by study hospitals were clinically diagnosed with men- the latex agglutination test [9]. The isolated S. pneu- ingitis, pneumonia or sepsis. Their age ranged from moniae strains were tested for susceptibility to a 0.003 to 78 years, and a total of 1551 patients were range of antimicrobial agents by the Kirby–Bauer aged 5 years. Clinical specimens collected from the method [10]. In addition, minimum inhibitory con- patients for laboratory analyses included 3779 blood centration (MIC) values were determined for the and 787 CSF specimens; 1220 blood and 331 CSF drugs using the Etest [9]. NCCLS breakpoints were specimens were from patients aged 5 years, while used to determine the susceptibility of S. pneumoniae 2559 blood and 456 CSF specimens were from isolates. For penicillin, the most commonly used drug patients aged 5 years. A total of 124 nasopharyn- against S. pneumoniae, the MIC defi nition of resis- geal specimens were also collected for investigation. tance was a MIC value 1.0 mg/ml. S. pneumoniae Overall, a total of 16 S. pneumoniae isolates were Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only. 256 E. S. Donkor et al. obtained from invasive specimens (blood and CSF) among the study subjects. S. pneumoniae preva- and 19 from nasopharyngeal specimens. lence was 0.19%. A group of 5 Gram-negative organisms comprising Salmonella typhi (0.21%) and non-typhoidal salmonella (0.4%), Enteroba- P athogens isolated from CSF cter (0.58%), Acinetobacter (0.56%), Klebsiella Out of the 787 CSF specimens cultured, 23 (2.9%) (0.45%), and E. coli (0.26%), were isolated from produced bacterial growth and a total of 11 different the blood specimens. pathogenic organisms were isolated from the speci- Table II also shows prevalence rates of the isolated mens. Table I shows the prevalence rates of the organisms for study subjects aged 5 years and those various organisms isolated from the CSF specimens. aged 5 years. Overall, prevalence rates in subjects The most frequently isolated organism was S. pneu- 5 years of age were 0.26–1.90% and in subjects moniae with a prevalence rate of 1.1% among the aged 5 years were 0–0.18%. Signifi cant differences study subjects. A total of 5 different Gram-negative in prevalence rates of the organisms between the 2 organisms were isolated with a prevalence rate of age groups were more commonly observed for the 0.89%, comprising Escherichia coli (0.25%), Gram-negative bacteria. For both age groups, S. Pseudomonas aeruginosa (0.25%), Salmonella spp. aureus was the most prevalent organism. (0.13%), Proteus spp. (0.13%) and Flavobacterium meningosepticum (0.13%). Neisseria meningitidis S. pneumoniae prevalence in invasive and was isolated with a prevalence rate of 0.38%, while non-invasive specimens Haemophilus infl uenzae, Staphylococcus aureus and viridans streptococci were each isolated with a prev- B ased on positive cultures, the prevalence of alence rate of 0.13%. The only fungus isolated from S. pneumoniae in invasive specimens was 0.35% the CSF was Cryptococcus neoformans with a prev- (16/4556), comprising 0.19% (7/3779) for blood alence rate of 0.13%. With the exception of S. pneu- and 1.1% (9/787) for CSF. Non-culture detection of moniae, all the organisms isolated from the CSF S. pneumoniae in CSF specimens showed a preva- specimens were isolated from patients aged 5 years. lence rate of 18.2% (2/11). Thus the overall S. pneu- Out of the 9 isolates of S. pneumoniae from CSF, moniae prevalence in CSF specimens based on 4 were from children 5 years while 5 were from culture and non-culture detection methods was 1.72% subjects aged 5 years; this represents prevalence (11/798). The overall S. pneumoniae prevalence in rates of 1.2% (4/331) and 1.1% (5/456) in the 2 age invasive specimens in subjects 5 years was 0.58% groups, respectively, with no signifi cant difference. (9/1220) and in those aged 5 years was 0.23% (7/3015). An overall S. pneumoniae prevalence rate of 15.3% (19/124) was observed for the nasopharyn- Pathogens isolated from blood geal specimens. The prevalence rates in subjects 5 A total of 206 (5.5%) of the 3779 blood cultures years and those aged 5 years were 27.2% (9/33) produced growth. A wide range of organisms and 10.3% (10/97), respectively. was isolated and the prominent ones are reported in Table II. The most frequently isolated organism A ntimicrobial drug resistance of was S. aureus with a prevalence rate of 1.32% S. pneumoniae isolates The prevalence of multiple drug resistance among S. T able I. Microbial pathogens isolated from cerebrospinal fl uid pneumoniae isolated from the study subjects specimens. was 48.6% (17/35); the rates among invasive and non- Organism na % invasive isolates were 50% (8/16) and 47.4% (9/19), respectively, while the rates among isolates from sub- Staphylococcus aureus 1 0.13 jects 5 years and those aged 5 years were 56.3% Streptococcus pneumoniae 9 1.1 (9/16) and 52.6% (10/19), respectively. Resistance to Viridans streptococci 1 0.13 Escherichia coli 2 0.25 the various antimicrobial drugs tested were cefotaxime Pseudomonas aeruginosa 2 0.25 11.1%, chloramphenicol 11.1%, ciprofl oxacin 12.5%, Salmonella spp. 1 0.13 cefuroxime 14%, penicillin 19.4%, erythromycin 28%, Proteus spp. 1 0.13 cloxacillin 43.6%, ampicillin 48%, tetracycline 62%, Flavobacterium meningosepticum 1 0.13 co-trimoxazole 84%, and gentamicin 84%. The MIC Haemophilus infl uenzae 1 0.13 Neisseria meningitidis 3 0.38 value ranges (mg/l) were as follows: chloramphenicol Cryptococcus neoformans 1 0.13 0.75–8.0; cefotaxime 0.125–1.5; ampicillin 0.064– 3.00; erythromycin 0.016–0.032; tetracycline a Number of isolates. 0.094–64.0; and penicillin 0.047–2.0. Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only. S. pneumoniae in disease and carriage, Ghana 257 T able II. Microbial pathogens isolated from blood specimens. Organism NT % NC % NA % Staphylococcus aureus 50 1.32 30 1.9 20 0.9 Streptococcus pneumoniae 7 0.19 4 0.26 3 0.13 Escherichia coli 10 0.26 6 0.39 4 0.18 Non-typhoidal salmonella 15 0.4 11 0.71 4 0.18 Salmonella typhi 8 0.21 5 0.32 3 0.13 Acinetobacter spp. 21 0.56 19 1.23 2 0.09 Enterobacter spp. 22 0.58 22 1.4 0 0 Klebsiella spp. 17 0.45 17 1.1 0 0 NT, total number of isolates from all subjects; NC, total number of isolates from subjects aged 5 y; NA, total number of isolates from subjects aged 5 y ears Serotype distribution and genetic diversity of rates of 4.9% in Tanzania, 2% in Kenya, 1.8% in S. pneumoniae isolates Ethiopia, 1.4% in Uganda for CSF; and 1.4% in Kenya, 0.8% in Ethiopia, 0.2% in Uganda for blood Thirteen S. pneumoniae isolates were serotyped; [2]. In this study, the detection of S. pneumoniae 61.5% (8) were typeable while 38.5% (5) were not. antigens in CSF by latex agglutination showed an The typeable isolates were of 7 different serotypes: overall prevalence rate of 18.2%. Though this involved 3, 6B, 9, 10, 14, 16 and 23F. few specimens, it shows the prevalence of the organ- Table III shows the MLST results of some S. ism is actually higher than that reported by culture pneumoniae isolates. Genetic analyses of the 7 house- (1.1%). Though not as reliable as culture, the latex keeping genes (a roE , gdh, g ki , r ecp , spi , xpt, d dl ) of the agglutination test has been demonstrated to be quite isolates showed wide variability as shown by the reliable, yielding a specifi city of 91% and sensitivity allele numbers corresponding to the various genes of of 86.6% [12]. The relatively low rate of detection of the isolates; while the s pi and xpt gene sequences S. pneumoniae and other organisms by culture is were the same in some isolates, the other 5 house- probably due to the treatment of patients with anti- keeping genes were entirely different. Ten (35.7%) of microbial agents before samples were taken. This was the 28 housekeeping genes sequenced showed novel confi rmed in interviews and discussions with both sequences, and this included 1 gene each of a roE patients and clinicians. It has been demonstrated and spi , and 2 genes each of g dh , gki , recp and ddl. that after the initiation of parenteral antibiotics, CSF sterilization may occur after 4 hfor S. pneumoniae, and pre-treatment of patients with antimicrobial Discussion agents before specimen collection could consider- S . pneumoniae is an agent of immense public health ably reduce the isolation rate of bacteria in CSF, importance and has attracted more attention and to a higher extent in blood [13,14]. recently, due to the development of pneumococcal S. pneumoniae was the most prevalent organism conjugate vaccines. In this study, we evaluated the in CSF and is probably the most important cause of public health burden of S. pneumoniae and collected meningitis in Ghana. In 2003, the organism was epidemiological data relevant for the utilization of implicated in an outbreak of meningitis in Ghana pneumococcal conjugate vaccines in Ghana. Based [15]. In blood specimens, the prevalence of S. pneu- on culture, the prevalence of S. pneumoniae in CSF moniae was relatively low for both age groups and was 1.1% and in blood was 0.2%. Recent studies S. aureus and Gram-negative organisms were more based on blood cultures have reported prevalence important. The prevalence of S. pneumoniae in T able III. Multilocus sequence typing (MLST) of Streptococcus pneumoniae isolates. Isolate aroE g dh gki recp s pi xpt ddl G36 5 4 2 4 4 1 1 G18 51 a 2 a 12 a 41a 55a 3 244a G1222 1 1 8 18 6 3 8 G2508 12 184a 42a 107 a 6 239 179 a aroE , g dh, g ki , r ecp, spi , xpt , and ddl represent the housekeeping genes sequenced. a Indicates a novel allele and its sequence is very similar to that of the attached allele number as defi ned by the MLST database. Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only. 258 E. S. Donkor et al. nasopharyngeal specimens was higher for children proportion of disease caused by S. pneumoniae aged 5 years (27.2%) than those 5 years; preva- may be preventable. Though a limited number of lence rates of up to 87% have been reported in this isolates was used in this evaluation, similar fi ndings age group [16]. S. pneumoniae occur as part of the have been reported in other parts of Ghana and normal fl ora in the nasopharynx. However, nasopha- several other countries [11,21]. Owing to the ryngeal carriage of the organism is considered a risk limited number of isolates that were serotyped, it factor for invasive disease, and this may partly account may be diffi cult to comment adequately on the for the relatively high morbidity of S. pneumoniae in relationship between serotype and drug resistance. children aged 5 years [16,17], which was also However, the high multiple resistance among the observed in this study. S. pneumoniae isolates suggests signifi cant resis- M ultiple drug resistance, a major public health tance among the non-vaccine serotypes. Owing to problem, occurred at a high prevalence rate of 48.6% serotype replacement, it may be necessary to mon- among the S. pneumoniae isolates in our study, itor drug resistance of non-vaccine serotypes in this which is consistent with a recent report [18]. This era of pneumococcal vaccination. discourages empirical antibiotic treatment of the M LST is the defi nitive, internationally accepted organism and underscores the need for susceptibility method for monitoring the spread of clones through testing. Disk diffusion test results were generally con- the S. pneumoniae population [23,24]. MLST of the sistent with MIC values and showed high resistance 28 S. pneumoniae genes showed wide genetic varia- to co-trimoxazole, gentamicin, tetracycline and tions, which is quite common with the organism [25]. ampicillin. Many microorganisms have been reported The high proportion of novel sequences identifi ed is as having a high percentage resistance to most of interesting and shows that the population structure these drugs for several years, and the rates of resis- of the organism in Ghana may be quite different from tance have been rising over the years not only for what is known. A similar observation has also been clinical isolates but also for the normal fl ora of the made in S. pneumoniae isolates in the Gambia [26]. healthy population [19,20]. Lower prevalence rates There is a need for further typing of S. pneumoniae of resistance were observed for ciprofl oxacin, from this region to provide more information on chloramphenicol, cefotaxime and cefuroxime. With molecular clones particular to the region. the exception of chloramphenicol, these drugs have O ur study showed that S. pneumoniae plays an been on the Ghanaian market for a relatively short important role in invasive infections in Ghanaian period of time and therefore may not have been used children, especially meningitis. The high multiple or misused extensively. In addition, some of these resistance of the organism, which limits treatment drugs are expensive and are used only for serious options, heightens the public health burden of the infections, thus limiting their usage. The resistance organism. This underscores the need for vaccina- rate against penicillin, which has been an important tion, and conjugate vaccines that were shown to have drug in the treatment of S. pneumoniae infections in serotype coverage of up to 70% constitute a major Ghana, was 19.4%. A study in 2005 in Ghana intervention in reducing the public health burden reported a rate of 12% of intermediate penicillin of this organism in the country. resistance [21]. While S. pneumoniae is still suscep- tible to the drug, high resistance is predicted in the next couple of years due to poor enforcement of Acknowledgements regulations restricting access of the general public to T his study was funded by a grant from the antimicrobial agents. Global Alliance for Vaccines and Immunisation S erotyping of S. pneumoniae identifi ed sero- (Pneumococcal vaccines Accelerated Development types 3, 6B, 9, 10, 14, 16 and 23F. Most of these and Introduction Plan; PneumoADIP) based at serotypes have been implicated in invasive infec- Johns Hopkins University, Bloomberg School of tions worldwide. We did not encounter serotypes 1 Public Health. The funding received is gratefully and 5, which have been commonly reported in inva- acknowledged. The technical assistance offered by sive infections in West Africa [21,22]. This may be clinicians and laboratory personnel of the Korle-Bu due to the relatively small number of isolates sero- Teaching Hospital, Tamale Teaching Hospital, and typed following loss of viability of some of the S. Sunyani Regional Hospital is also acknowledged. pneumoniae isolates. The extent of coverage of sero- types by the 7-valent pneumococcal conjugate vac- cine was 57.1%, for the 10-valent vaccine was E thical approval: The protocol of the study was 57.1%, and for the 13-valent vaccine was 71.4%.It approved by the Ethical and Protocol Committee of is encouraging that about 57–70% of the serotypes the University of Ghana Medical School. Informed are covered by conjugate vaccines and therefore this consent was obtained from study subjects before Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only. S. pneumoniae in disease and carriage, Ghana 259 enrolment, and in the case of children, consent was parenteral antibiotic pretreatment. Paediatrics 2001;108: obtained from parents. 1169–74. [14] F eldman WE. Effect of prior antibiotic therapy on concentra- tions of bacteria in CSF. Am J Dis Child 1978;132:672–4. Declaration of interest: No confl ict of interest to [15] Leimkugel J, Forgor AA, Gagneux S, Pfl uger V, Fliel C, Awine E, et al. An outbreak of serotype 1 Streptococcus declare. pneumoniae meningitis in northern Ghana with features that are characteristic of Neisseria meningitidis meningitis epidemics. J Infect Dis 2005;192:192–9. R eferences [16] C ardozo DM, Nascimento-Carvalho CM, Souza FR, Silva NM. Nasopharyngeal colonization and penicillin [1] W orld Health Organization. Pneumococcal vaccines: resistance among pneumococcal strains: a worldwide 2004 World Health Organization position paper. Can Commun update. Braz J Infect Dis 2006;10:294–306. Dis Rep 1999;25:150–1. [17] Obaro SK. Prospects for pneumococcal vaccination in [2] GAVI’s PneumoADIP. Surveillance and research report. African children. Acta Trop 2000;75:141–53. Baltimore: John Hopkins Bloomberg School of Public [18] Van Bambeke FR, Reinert R, Appelbaum PC, Tulkens PM, Health; 2007. Peetermans WE. Multidrug-resistant Streptococcus pneu- [3] S cott JA. The preventable burden of pneumococcal disease moniae infections: current and future therapeutic options. in the developing world. Vaccine 2007;25:2398–405. Drugs 2007;67:2355–82. [4] Centers for Disease Control and Prevention. Direct and indi- [19] N ewman MJ, Seidu A. Carriage of antibiotic resistant E. coli rect effects of routine vaccination of children with 7-valent in adult intestinal fl ora. West Afr J Med 2002;21:48–50. pneumococcal conjugate vaccine on incidence of invasive [20] O keke IN, Fayinka ST, Laminkanra A. Antibiotic resistance pneumococcal disease—United States, 1998–2003. MMWR in Escherichia coli from Nigerian students,1986–1998. Morb Mortal Wkly Rep 2005;54:893–7. Emerg Infect Dis 2000;6:393–6. [5] Selman S, Hayes D, Perin LA, Hayes WS. Pneumococcal [21] Holliman R, Liddy H, Johnson J, Adjei O. Epidemiology of conjugate vaccine for young children. Manag Care 2000; invasive pneumococcal disease in Kumasi, Ghana. Trans R 9:49–52. Soc Trop Med Hyg 2007;101:405–13. [6] Newman MJ. Bacterial meningitis in southern Ghana: a [22] Adegbola R, Hill PC, Secka O, Ikumapayi UN, Lahai G, 3-year review of bacterial isolates. Ghana Med J 1991;25: Greenwood BM. Serotype and antimicrobial susceptibility 349–53. patterns of isolates of Streptococcus pneumoniae causing [7] Acquaye JK. Pyogenic meningitis. Ghana Med J 1981;5: invasive disease in The Gambia 1996–2003. Trop Med Int 115–7. Health 2006;11:1128–35. [8] B owers EF, Jeffries LR. Optochin in the identifi cation of [23] Coffey TJ, Enright MC, Daniels M, Morona JK, Morona R, Streptococcus pneumoniae. J Clin Pathol 1955;8:58–60. Hryniewicz W, et al. Recombinational exchanges at the [9] Bauer AW, Kirby WM, Sherris JC, Turck M. Antibiotic capsular polysaccharide biosynthetic locus lead to frequent susceptibility testing by a standardized single disk method. serotype changes among natural isolates of Streptococcus Am J Clin Pathol 1966;45:493–6. pneumoniae. Mol Microbiol 1998;27:73–83. [10] Baron JE, Peterson LR, Finegold SM. Bailey and Scott diag- [24] Brueggemann A, Peto T, Crook D, Butler J, Kristinsson K, nostic microbiology. 9t h ed. Missouri: C. V. Mosby Co.; 1994. Spratt B. Temporal and geographic stability of the sero-group [11] Enright MC, Spratt BG. A multilocus sequencing typing specifi c invasive disease potential of Streptococcus pneumo- scheme for Streptococcus pneumoniae: identifi cation of niae in children. J Infect Dis 2004;190:1203–11. clones associated with serious invasive disease. Microbiology [25] Hiller NL, Janto B, Hogg JS, Boissy R, Yu S, Powell E, et al. 1998;144:3049–60. Comparative genomic analyses of seventeen Streptococcus [12] R ai GP, Zachariah K, Sharma R, Phadke S, Belapurkar KM. pneumoniae strains: insights into the pneumococcal suprage- Pneumococcal antigen detection in cerebrospinal fl uid: a nome. J Bacteriol 2007;189:8186–95. comparative study on counter immunoelectrophoresis, latex [26] A ntonio M, Dada-Adegbola H, Biney E, Awine T, agglutination and coagglutination. Comp Immunol Micro- O’Callaghan J, Pfl uger V, et al. Molecular epidemiology of biol Infect Dis 2003;26:261–7. pneumococci obtained from Gambian children aged 2–29 [13] Kanegaye JT, Soliemanzadeh P, Bradley JS. Lumbar punc- months with invasive pneumococcal disease during a trial ture in paediatric bacterial meningitis: defi ning the time of a 9-valent pneumococcal conjugate vaccine. BMC Infect interval for recovery of cerebrospinal fl uid pathogens after Dis 2008;8:81. Scand J Infect Dis Downloaded from informahealthcare.com by Memorial University of Newfoundland on 09/20/13 For personal use only.