Table 3.

Regression models display the effect of the health care breach type on individuals affected^a.

Dependent variable=ln(Individuals affected)	Model 1		Model 2		Model 3
	β	P value	β	P value	β	P value
Hacking or IT incident	Included	—b	Included	—	Included	—
Improper disposal	–1.162c	<.001	–1.176c	<.001	–1.176c	<.001
Loss	–1.342c	<.001	–1.339c	<.001	–1.339c	<.001
Other	–1.014c	<.001	–1.085c	<.001	–1.085c	<.001
Theft	–1.168c	<.001	–1.218c	<.001	–1.218c	<.001
Unauthorized access or disclosure	–1.336c	<.001	–1.362c	<.001	–1.362c	<.001
Unknown	–1.249d	.02	–1.291d	.02	–1.291d	.02
Business associate	Included	—	Included	—	Included	—
Health plan	—	—	–0.382c	<.001	–0.382c	<.001
Health care clearing house	—	—	–0.393	.49	–0.393	.49
Health care provider	—	—	–0.411c	<.001	–0.411c	<.001
Stagewise fixed effect	Included	—	Included	—	Included	—
StHIPAAe	—	—		—	1.271d	.03
Constant	9.144c	<.001	8.743c	<.001	8.743c	<.001
Observations	5158	—	5140	—	5140	—
R 2	0.12	—	0.14	—	0.14	—

^aBreach report is present in Multimedia Appendix 2.

^bNot applicable.

^cP<.001.

^dP<.05.

^eStHIPAA: stronger than Health Insurance Portability and Accountability Act.