RESUMO
PURPOSE: The objective of this study was to evaluate sampling methods for recovery of mutans streptococci (MS) and lactobacilli (LB) in children 9 to 36 months of age. METHODS: Tongue and plaque specimens collected on cotton swabs and stimulated saliva were diluted and plated on selective and nonselective media. Tongue specimens on a swab and mouth mirror were inoculated directly on selective agar media (MS only). Sampling methods were compared by frequency of recovery of MS or LB, correlation of microbial counts with dmfs scores, and potential of specific microbial counts to predict caries presence or absence. RESULTS: The mean dmfs score of 87 subjects was 6.3; 48 subjects were caries free. Levels of MS and LB were consistently higher in plaque than in other sampling techniques (P<.001), and frequencies of recovery of MS were highest in plaque (P<.041) and tongue (P<.006). Frequency of LB recovery did not differ significantly between sampling methods. Counts of MS or LB in total subjects and subjects aged 9 to 24 months correlated positively with dmfs scores (P<.028). Threshold levels of MS which were predictive of presence of caries were: (1) plaque=>2x10(5); (2) tongue=>10(4); (3) saliva=>10(5); (4) mirror=>50; and (5) swab=>50. Comparable levels of LB were: plaque, >10(3); tongue, >10(2) and saliva, >10(3). Specificities associated with these predictions were higher than sensitivities for all sampling methods. CONCLUSIONS: (1) All sampling methods were adequate for microbial risk assessment tests in children under 3 years of age; (2) MS was a stronger indicator of caries status than LB.
