Table 1.

Sample Sizes for Intervention and Control Groups in the Analyses Associated with the Four Interventions

	Intervention Group		Control Group
Intervention/Analysis	Source	Sample Size		Source	Sample Size
	Member mailings and advertising*		Members living in Michigan	Physicians	31,576		Members living outside of Michigan‡	Physicians	106,732
		Patients	188,216		Patients	194,171
		Claims	4,637,462		Claims	5,152,602
Generic sampling†	Patients of physicians selected for generic sample program	Physicians	289§	Patients of physicians not selected for generic sampling program	31,287
		Patients	22,163		Patients	166,053
		Claims	422,449		Claims	4,215,013
Physician incentive	Patients of group practices selected for physician incentive program	Physicians	1,127	Patients of matched groups not selected for physician incentive program	Physicians	980
		Patients	39,474		Patients	31,934
		Claims	773,414		Claims	569,508

^*The member-to-physician ratio is higher for the intervention group because the density of BCBSM members is higher in Michigan than in any other state

^†The intervention group counts members and claims of members if they were associated with a physician that participated in the generic sampling intervention at any point during the study even if the member was mostly seen by nonparticipating physicians (thus the ratios of members and claims to physicians is higher than for the control group)

^‡These are almost all employees of nationally based employers

^§Physicians were counted in the intervention group for the generic sampling intervention if they received generic samples in any year