Examining the role of firearm involvement in repeat intimate partner violence assaults

Table 2 Time to recurrent IPV events

	Conditional gap-time frailty model		Conditional gap-time model
	Hazard ratio	Confidence interval	Hazard ratio	Confidence interval
Panel A: all police reported offenses
Firearm threats at index	1.22	0.76–1.98	1.23	0.92–1.65
Firearm involvement in past IPV	1.40	0.96–2.02	1.25	1.02–1.54
Firearm involvement in recidivism	0.99	0.70–1.39	1.06	0.78–1.44
N	1004		1004
Number of failures	660		660
Log likelihood	− 3117.51		− 3317.98
Wald χ²	96.47***		90.48***
I-likelihood	− 3302.63		–
Θ	0.68***		–
Likelihood ratio for Θ	400.93***		–
Panel B: police-reported physical assaults
Firearm threats at index	1.07	0.55–2.08	1.23	0.89–1.70
Firearm involvement in past IPV	1.90	1.13–3.21	1.45	1.15–1.82
Firearm involvement in recidivism	1.03	0.69–1.53	1.07	0.76–1.51
N	886		886
Number of failures	540		540
Log likelihood	− 2351.12		− 2621.08
Wald χ²	194.90***		154.67***
I-likelihood	− 2590.41		–
Θ	1.36***		–
Likelihood ratio for Θ	539.93***		–

Models control for perpetrator characteristics including age, substance use, prior incarceration, history of non-IPV assaults, history of strangulation, violation of conditional release order, relationship type, economic abuse, prior IPV history, and arrest at index assault
*** denote p<0.001

ISSN: 2197-1714