Table 1. Estimation of the switch frequency (weekly probability for a persistently carried pneumococcal strain to switch its capsule).

Study	Number of reported switches	Cumulated persistent carriage duration (weeks, estimated)	Estimated switch frequency (weeks−1)
(Barnes et al) [27]	1	121	8.2×10−3
(Sluijter et al)* [28]	1	169	5.9×10−3
(Meats et al) [19]	0	3852	0
(Bogaert et al) [29]	3	3068	9.7×10−4
Random effects pooled estimate from the 4 studies [30]			1.5×10−3 [4.6×10−5–4.8×10−3]

^*Regarding the second study [28], it is to be noted that although the article provided the results of all positive samplings in chronological order for all children, it did not specify the corresponding sampling times. We therefore had to approximate the overall persistent colonization duration by supposing that all successive positive samplings occurred consecutively at a 1 month interval.