Table 2.

Associations between Project Lazarus implementation and opioid analgesic prescribing, by strategy, North Carolina, 2009–2014

				Univariate Models^a			Multivariable Adjusted Model^b
Strategy	Implementation Level	County-months	Rx OAs dispensed	IRR	95% CI	CLR	IRR	95% CI	CLR
Diversion Control	None	4971	23,512,894	1 (ref.)			1 (ref.)
Diversion Control	Any	2229	17,848,868	1.15	1.12, 1.17	1.044	1.06	1.03, 1.09	1.056
Naloxone Policies	None	6216	34,103,605	1 (ref.)			1 (ref.)
Naloxone Policies	Any	984	7,258,157	1.08	1.06, 1.11	1.038	0.97	0.95, 0.99	1.048
Community Education	None	5969	30,353,079	1 (ref.)			1 (ref.)
Community Education	Any	1231	11,008,683	1.11	1.09, 1.13	1.042	1.00	0.97, 1.03	1.058
Provider Education	None	4962	25,608,213	1 (ref.)			1 (ref.)
Provider Education	Any	2238	15,753,549	1.13	1.10, 1.15	1.042	1.00	0.97, 1.03	1.060
Support for Patients with Pain	None	6684	34,023,002	1 (ref.)			1 (ref.)
Support for Patients with Pain	Any	516	7,338,760	1.08	1.06, 1.10	1.029	0.96	0.93, 1.00	1.069
Hospital ED Policy	None	5485	29,009,970	1 (ref.)			1 (ref.)
Hospital ED Policy	Any	1715	12,351,792	1.11	1.08, 1.13	1.047	0.99	0.96, 1.01	1.055
Addiction Treatment	None	1559	13,442,436	1 (ref.)			1 (ref.)
Addiction Treatment	Any	5641	27,919,326	1.16	1.11, 1.20	1.085	1.04	0.99, 1.08	1.090

IRR incidence rate ratio, CI confidence interval, CLR confidence limit ratio, ED Emergency Department

Results from Poisson GEE with county-month population offset

^aUnivariate unadjusted models for each strategy

^bAdjusted multivariable model adjusts for the other six strategies, county health status, year and season