• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jcmPermissionsJournals.ASM.orgJournalJCM ArticleJournal InfoAuthorsReviewers
J Clin Microbiol. Oct 2007; 45(10): 3408–3410.
Published online Aug 15, 2007. doi:  10.1128/JCM.01393-07
PMCID: PMC2045369

Validation of Nasopharyngeal Sampling and Culture Techniques for Detection of Streptococcus pneumoniae in Children in Kenya[down-pointing small open triangle]


We compared nasopharyngeal swabs against nasal wash cultures for detecting colonizing pneumococci and examined the effect of frozen storage in skim milk-tryptone-glucose-glycerin on culture. Among the 55 children with positive nasal wash cultures, swab cultures were positive for 47 (85%). Of the 96 swabs positive on direct plating, 94 (98%) were positive when recultured after freezing.

Before beginning a large, community-based study of nasopharyngeal carriage of Streptococcus pneumoniae, we conducted three studies to validate our sampling methodology. The objectives of the studies were (i) to determine the sensitivity of a nasopharyngeal swab compared to that of a nasal wash culture, (ii) to determine whether colonization is symmetrically distributed in the nasopharynx and estimate the loss of sensitivity inherent in sampling only a single side, and (iii) to determine whether freezing of nasopharyngeal swabs has a detrimental effect upon the sensitivity of the culture.

We conducted two field studies (study 1 and study 2) and a laboratory study (study 3) of nasopharyngeal carriage sampling and culture techniques, respectively, with children at the start of the rainy season in the Kilifi District between 11 April and 10 May 2006. Study 1 compared nasopharyngeal swabs with nasal wash samples taken 10 to 15 min apart for the same 62 children. Study 2 compared separate nasopharyngeal swabs of the right and left nostrils taken a few minutes apart for a different sample of 62 children. In study 1, the nasopharyngeal swab was taken first; in study 2, the side studied first was determined systematically before the study began to represent both sides equally. The populations were convenience samples of children less than 5 years of age who were presented to the pediatric outpatient department of Kilifi District Hospital, Kenya, with minor illnesses not requiring hospitalization. The sizes of the samples were chosen to define sensitivity at 90% with a precision of ±10%. All of the samples in studies 1 and 2 were cultured on the day of collection. For study 3, the 62 nasopharyngeal swabs from study 1 and the 62 swabs of the right nostril from study 2 were recultured after freezing in skim milk-tryptone-glucose-glycerin (STGG) (2) at −80°C for 2 months. The study was approved by the Kenya Medical Research Institute/National Ethical Review Committee and The Oxford Tropical Research Ethics Committee, and written informed consent was obtained for all participants.

Nasopharyngeal swab samples were taken with rayon-tipped, flexible, aluminum-shaft swabs (Medical Wire and Equipment Company, Town, United Kingdom) according to World Health Organization guidelines (4) by trained fieldworkers and transported in STGG to the KEMRI/Wellcome Trust laboratory within 4 h. To obtain a nasal wash sample, 5 ml of sterile physiological saline was instilled into one nostril and sample liquid was gently aspirated out through the other nostril into a sterile container. The swabs and the pellets obtained from the wash samples after centrifugation for 1 min were processed according to the WHO guidelines (4). S. pneumoniae was identified by colony morphology, [proportional, variant]-hemolysis, optochin sensitivity, and bile solubility. Four morphologically distinct colonies were subcultured for typing from each primary plate. Pneumococci were serogrouped by latex agglutination and serotyped by the Quellung reaction using polyclonal rabbit antisera (Statens Serum institute, Copenhagen, Denmark). The sensitivities of the paired swabs and/or wash cultures were determined from matched-pair two-by-two tables, and performance differences were tested using McNemar's chi-square test.

The median age of the 62 participants in study 1 was 12 months (interquartile range, 8 to 24 months), and that in study 2 was 17.5 months (interquartile range, 9 to 26 months). In study 1, pneumococci were detected in the nasopharynxes of 55 (90%) children by either swab or nasal wash cultures (Fig. (Fig.1).1). Of the 61 nasal wash cultures analyzed, 55 (90%) were culture positive for S. pneumoniae, and in 5 (9%) of these, two different serotypes were identified. Of the 62 nasopharyngeal swabs cultured in study 1, 47 (76%) were positive, and in 9 (19%) of these, two or more serotypes were identified; in one swab, three serotypes were cultured simultaneously. Compared against that of nasal wash cultures, the sensitivity of nasopharyngeal swabs for detecting any pneumococcal colonization was 85% (95% confidence interval [CI], 73 to 94%). In study 2, pneumococci were detected in the nasopharynxes of 46 (74%) children by either a left or a right-sided swab (Fig. (Fig.1).1). Of the 124 swabs taken, 87 were culture positive, and in 15 (17%) of these, two different serotypes were identified. There was no systematic evidence of bias in colonization for either the right or the left swab (P = 0.18), so each swab set was combined to derive a sensitivity estimate of 95% (95% CI, 88 to 98%) against the gold standard of a positive result on either swab.

FIG. 1.
Matched-pair tables of the three studies. As there was no evidence of laterality (P = 0.18), the left and right swabs were combined to estimate the sensitivity of a single swab against the gold standard of a positive result on either swab. pos, ...

In study 3, among the 124 nasopharyngeal swab samples cultured immediately, 96 were positive for S. pneumoniae. When these samples were recultured after 2 months of frozen storage, 94 of them were positive for S. pneumoniae (sensitivity, 98%; 95% CI, 93 to 100%). Among the samples that were initially negative on culture, two were positive for pneumococci when recultured after freezing, so the total number of children who were classified as carriers was the same for each culture exercise. In the immediate cultures, 77 samples grew one pneumococcal serotype, 15 grew two different serotypes, and 1 grew three serotypes. In frozen cultures, these numbers were 77, 17, and 0, respectively. Of the 93 swab specimens for which both the fresh and frozen cultures yielded at least one pneumococcus, the serotypes isolated from the same specimen at the two time points were different in 7 instances (Table (Table11).

Serotype distribution of 221 S. pneumoniae isolates from 124 nasopharyngeal swabs cultured both immediately (fresh) and after freezing for 2 months

We found nasopharyngeal wash cultures to be more sensitive for pneumococcal detection than nasopharyngeal swabs. Sampling the nasopharynx via both nostrils was also superior to sampling by one nostril alone. It is not possible to say whether this is because pneumococcal colonization can be confined to one side of the nasopharynx or because the inherent insensitivity of a single swab is improved by repetition. Subject compliance in longitudinal studies is inversely related to the number and discomfort of the study techniques involved. Nasal washing combined with two swab samplings is considerably less acceptable to study subjects than a single nasopharyngeal swab sampling; hence, the loss of follow up that is likely to accompany this procedure is unlikely to be justified by the additional sensitivity observed. The differences in serotype distributions between STGG samples cultured immediately and those cultured after 2 months of frozen storage might suggest that serotypes have differential capacities to survive freezing and thawing, and if this is true, freezing could represent one way to select for certain serotypes. However, the difference is equally likely to be due to the limited sensitivity of a single culture of an STGG sample that contains multiple serotypes.

One of the shortcomings of the study was that the sampling was done during the rainy season, when carriage prevalence is higher in this community (1). In the less conducive conditions that prevail in the dry season, differences between alternative sampling and laboratory methods might have been more apparent. Second, the swabs were frozen for only 2 months and it is not known whether a longer duration of freezing might lead to a fall in sensitivity. The main strength of the study is that all of the comparisons were paired, either within the individual or within the swab.

Among the earliest methodological improvements in laboratory processing of nasopharyngeal swabs was the use of mouse inoculation (3), which was found to be similar in sensitivity to the use of selective media (with gentamicin). The two techniques together significantly increased the yield of S. pneumoniae (3). Each technique alone was imperfect, but combining the techniques is impractical in large epidemiological studies involving thousands of swabs. One conclusion that seems to apply to all present sampling and culturing techniques is that repetition noticeably increases the yield of pneumococci, a sign of imperfect sensitivity.

A single nasopharyngeal swab performs moderately well in the detection of S. pneumoniae carriage, and the additional gains associated with two swabs or a nasal wash are not sufficiently large to offset the disadvantages of discomfort and unacceptability to subjects. Culturing swab specimens that have been frozen in STGG for 2 months does not appreciably reduce the sensitivity for detecting S. pneumoniae but does yield a slightly different spectrum of serotypes. The imperfect sensitivity of nasopharyngeal swab cultures needs to be factored into the interpretation of carriage studies. Additionally, there is a clear need to develop more-sensitive methods for detecting colonization by multiple serotypes of S. pneumoniae in nasopharyngeal cultures.


This paper was published with the permission of the Director, Kenya Medical Research Institute.

This work was funded by the Wellcome Trust of Great Britain by a fellowship award to J. A. G. Scott (061089).


[down-pointing small open triangle]Published ahead of print on 15 August 2007.


1. Abdullahi, O., et al. The descriptive epidemiology of Streptococcus pneumoniae and Haemophilus influenzae nasopharyngeal carriage in children and adults in Kilifi District, Kenya. Pediatr. Infect. Dis., in press. [PMC free article] [PubMed]
2. Gibson, L. F., and J. T. Khoury. 1986. Storage and survival of bacteria by ultra-freeze. Lett. Appl. Microbiol. 3:127-129.
3. Hendley, J. O., M. A. Sande, P. M. Stewart, and J. M. Gwaltney, Jr. 1975. Spread of Streptococcus pneumoniae in Families. I. Carriage rates and distribution of types. J. Infect. Dis. 132:55-61. [PubMed]
4. O'Brien, K., H. Nohynek, et al. 2003. The pneumococcal vaccine trials carriage working group. Report from a WHO working group: standard method for detecting upper respiratory carriage of Streptococcus pneumoniae. Pediatr. Infect. Dis. J. 22:133-140. [PubMed]

Articles from Journal of Clinical Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...