Random colony selection versus colony morphology for detection of multiple pneumococcal serotypes in nasopharyngeal swabs.

Detection of multiple serotype carriage is important in understanding pneumococcal epidemiology. We compared random selection of 4 pneumococcal colonies per swab with selection by colony morphology. Multiple serotypes were detected in 20% of 98 swabs; 14% by morphology and 17% by random selection. Selection by morphology was more efficient per colony, but underestimated the true rate of multiple carriage.