Prevalence of antibiotic resistance and serotypes in pneumococci in England and Wales: results of observational surveys in 1990 and 1995BMJ 1996; 312 doi: https://doi.org/10.1136/bmj.312.7044.1454 (Published 08 June 1996) Cite this as: BMJ 1996;312:1454
- Alan P Johnson, clinical scientista,
- David C E Speller, consultant microbiologista,
- Robert C George, consultant microbiologistb,
- Marina Warner, medical laboratory scientific officera,
- Gil Domingue, medical laboratory scientific officera,
- Androulla Efstratiou, clinical scientistb
- a Antibiotic Reference Unit, Laboratory of Hospital Infection, Central Public Health Laboratory, Colindale, London NW9 5HT
- b Streptococcus and Diphtheria Reference Unit, Respiratory and Systemic Infection Laboratory, Central Public Health Laboratory, London NW9 5HT
- Correspondence to: Dr Johnson.
- Accepted 3 April 1996
Objective: To assess the prevalence of antibiotic resistance and serotype distribution among pneumococci in England and Wales in 1990 and 1995.
Design: Observational surveys in March 1990 and March 1995. During two weeks in each survey period all pneumococci isolated in public health laboratories in England and Wales were collected and assessed for sensitivity to antibiotics and the distribution of serogroups or serotypes.
Setting: The network of public health laboratories throughout England and Wales.
Subjects: 1127 individual patient isolates of Streptococcus pneumoniae obtained during the two surveys.
Main outcome measures: Sensitivity or resistance to a range of antibiotics; serogroup or serotype.
Results: The prevalence of intermediate or full resistance to penicillin increased from 1.5% in 1990 to 3.9% in 1995 and resistance to erythromycin increased from 2.8% to 8.6%. About 92% of isolates belonged to serogroups or serotypes included in the currently available pneumococcal vaccine.
Conclusion: Resistance to penicillin and erythromycin has increased among pneumococci in England and Wales. Continued surveillance to assess further increases in the prevalence of pneumococcal resistance to antibiotics is essential.
Resistance to cefotaxime or ceftriaxone, which may be the drugs of choice for pneumococcal meningitis caused by penicillin resistant pneumo- cocci, is still fairly rare
Roughly 92% of pneumococcal isolates in England and Wales belong to serogroups or serotypes included in the currently available 23 valent vaccine
Between 65.1% and 73.9% of pneumococcal isolates from children aged 2 or less belong to sero- groups or serotypes included in the five, seven, or nine valent conjugate vaccines under development for use in children of this age group
Continued surveillance of pneumococcal resist- ance to antibiotics is essential if clinicians are to make rational decisions concerning the treatment of pneumococcal infections
During the past three decades antibiotic resistant pneumococci have been reported world wide.1 2 To assess trends in the prevalence of resistance in England and Wales the Public Health Laboratory Service assessed antimicrobial resistance and serogroup or serotype distribution of all pneumococci isolated in the nationwide network of public health laboratories during defined periods in 1990 and 1995.
During two weeks in March 1990 and March 1995, 52 public health laboratories submitted all isolates of Streptococcus pneumoniae to the Central Public Health Laboratory. After receipt isolates were confirmed as pneumococci, serogrouped or serotyped, and tested for antibiotic sensitivity as described.3 Sensitivity or resistance to antibiotics was determined according to published criteria.4 5 Data were analysed and compared by the χ2 test.
A total of 1127 single patient isolates were studied (544 in 1990, 583 in 1995). The age distribution of patients in the two surveys was similar (P=0.82), roughly one third being under age 10 and one third being aged 60 or over. In 1995 there was a significant increase in the proportion of pneumococci isolated from the ear (P<0.05) but no significant change in the proportions of isolates from other sites.
In 1990 eight isolates (1.5%) submitted by six different laboratories exhibited either intermediate or full resistance to penicillin (table 1). Six of the isolates showed intermediate resistance to cefotaxime. In the 1995 survey 23 isolates (representing 3.9% of the study population) showed intermediate or full resistance to penicillin (table 1). This increase was significant (P<0.02). Full resistance to cefotaxime and ceftriaxone was noted in only 0.7% (four) and 0.3% (two) of the isolates respectively. The 23 penicillin resistant isolates were from 18 laboratories, 14 laboratories submitting one such isolate.
In 1990 five isolates showing intermediate penicillin resistance belonged to serotype 23F whereas the three fully resistant isolates belonged to serotypes 6B, 9V, and 23F. Of the 23 isolates showing intermediate or full resistance to penicillin in 1995, nine belonged to serotype 9V and the others to serotypes 6B, 19A, 19C, 19F, and 23F.
The prevalence of erythromycin resistance increased from 2.8% (15 isolates referred from 10 laboratories) in 1990 to 8.6% (50 isolates from 29 laboratories) in 1995 (P<0.001) (table 1). Of the 29 laboratories submitting erythromycin resistant isolates in the second survey, 16 submitted one such isolate and seven submitted two. The erythromycin resistant isolates in 1990 belonged to serogroups or serotypes 3, 5, 6, 9, 14, 19, and 22 whereas those in 1995 belonged to serogroups or serotypes 3, 6, 14, 15, 19, and 23.
There were no significant changes in the prevalence of resistance to other antibiotics (table 1). Resistance to vancomycin or rifampicin was not detected in either survey. In 1990 nine isolates (1.7%) were resistant to three or more antimicrobial agents whereas in 1995, 11 isolates (1.9%) were multiresistant. In both surveys multiresistant isolates belonged to serogroups 6, 9, 19, or 23.
In the two surveys 37 serogroups or serotypes were noted, 27 occurring in both surveys. The seven most common in both were 3, 6, 9, 11, 14, 19, and 23 (table 2). In each survey about 92% of the isolates belonged to serogroups or serotypes included in the currently available 23 valent pneumococcal vaccine.6 Similarly, in each survey roughly 73% of isolates from children aged 2 or less belonged to serogroups or serotypes included in a nine valent conjugate vaccine currently undergoing development for use in children.7 Conjugate seven valent and five valent vaccines are also being evaluated,7 and roughly 66% of isolates from children aged 2 or less belonged to serogroups or serotypes proposed for inclusion in the five valent vaccine.
The two surveys showed a significant increase in the prevalence of intermediate or full resistance to penicillin. As noted previously,8 however, resistance to cefotaxime or ceftriaxone—which may be the drugs of choice for meningitis caused by pneumococci with reduced sensitivity to penicillin—is still fairly rare. Of potentially greater relevance was the increase in the prevalence of resistance to erythromycin to 8.6% in 1995. Erythromycin is widely used for pneumococcal infections in patients who are allergic to penicillin or who are infected with penicillin resistant organisms. In addition, erythromycin resistant pneumococci show cross resistance to the related macrolides clarithromycin and azithromycin,9 10 which are extensively used for empirical treatment of community acquired respiratory infection.
The trend of increasing resistance to β lactams and macrolides is consistent with reported data on sensitivity testing of blood and cerebrospinal fluid isolates of pneumococci in England and Wales. In these isolates the prevalence of penicillin resistance increased from 0.3% of 2667 isolates in 1989 to 2.5% of 2751 isolates in 1994 and the prevalence of erythromycin resistance increased from 3.3% of 2522 isolates in 1989 to 11.2% of 2689 isolates in 1994.11 12 Increasing rates of antibiotic resistance among pneumococci have also been noted in individual hospitals in the United Kingdom.13 14 Interestingly, there was no significant change in the prevalence of resistance to a range of other antimicrobial agents, including tetracycline, chloramphenicol, and trimethoprim. Furthermore, these surveys and other studies15 show that vancomycin resistance has not emerged in pneumococci, which is important given that rifampicin or vancomycin may be useful for treating cephalosporin resistant pneumococcal meningitis in children.16
Continued surveillance of antimicrobial resistance and serotype distributions in S pneumoniae is essential if clinicians are to make rational decisions about the management and prevention of pneumococcal infection.
We are grateful to colleagues throughout the Public Health Laboratory Service, whose help made these surveys possible. We also acknowledge the contributions of Dr G Colman, Dr B Cookson, Professor E M Cooke, Dr P G Cooper, Mrs L Ball, Mrs T Parsons, Mrs K Broughton, and Dr A V Swan.
Funding The Public Health Laboratory Service.
Conflict of interest None.