Table 1.

Summary of sites included in the study, number of patients, and local cluster and quiet period definitions.

Location	Patients	Years	Local cluster definition (# of KD cases in 7 days)	Number of clusters	# of patients in clusters	Fraction of patients in clusters	Local quiet period definition (# days with no KD cases)	Number of quiet periods
Bay Area, CA USA	822	16	4	33	180	0.21	9	146
Boston, MA USA*	1017	18	4	43	223	0.22	15	96
DC, USA	193	5	4	8	44	0.23	9	41
Denver CO, USA	629	13	4	30	179	0.28	11	99
Emilia-Romagna, IT	502	18	3	36	139	0.28	16	121
Hawaii, USA	403	11	4	10	48	0.12	14	72
Los Angeles, CA USA	1603	16	6	53	458	0.29	5	229
New Zealand	1008	19	4	54	274	0.27	14	110
Seoul, Korea	1205	13	6	31	257	0.21	7	123
San Diego, CA USA	1332	18	5	47	315	0.24	8	177
Seattle, WA USA	646	17	4	22	112	0.17	18	80
Wakayama Prefecture, Japan	2255	20	6	76	619	0.27	6	209

Bold indicates use of the 99.5th percentile as a definition for either cluster or quiet period; black indicates use of 97.5th percentile. We used the more conservative cluster definition in lower incidence areas because an integer value (e.g., 2 cases in 7 days) might span a large percentile range of the distribution of densities in a low-incidence area. We used the higher threshold to define quiet periods in higher incidence regions by a similar logic, to ensure we had used tail values when constrained to use integer values (*The full record from Boston included 1596 patients over 35 years, with 293 patients in 58 clusters, or 18%, and 213 quiet periods. For the analysis here we restricted analysis to after 2000).