Table 3. Mixed effects regression¹ of opioid outcomes on between-person and within-person patient PageRank percentile and controls among high-risk patients with more than one prescriber per quarter (n = 68,401; n obs = 273,604).

	1: Num Rx		2: MME>90		3: Overdose		4: OUD
	IRR	(CI)	OR	(CI)	OR	(CI)	OR	(CI)
Female	1.00	(0.99–1.01)	0.42***	(0.39–0.46)	1.00	(0.88–1.14)	0.78***	(0.71–0.85)
Age(10 years)	0.99***	(0.99–1.00)	0.62***	(0.60–0.64)	0.92**	(0.87–0.97)	0.69***	(0.66–0.71)
Insurance type2
POS	0.85***	(0.84–0.87)	0.31***	(0.28–0.35)	0.66***	(0.54–0.79)	0.74***	(0.65–0.83)
HMO	0.88***	(0.87–0.89)	0.47***	(0.43–0.52)	0.81*	(0.69–0.96)	1.12*	(1.00–1.25)
State3
Kentucky	0.93	(0.86–1.01)	1.51	(0.79–2.88)	0.87	(0.54–1.41)	1.16	(0.53–2.56)
Virginia	0.91**	(0.84–0.98)	3.51	(2.76–4.47)	1.04	(0.80–1.36)	0.48	(0.22–1.04)
Tennessee	0.92*	(0.85–0.99)	1.40	(1.10–1.78)	0.97	(0.77–1.22)	3.92***	(1.85–8.29)
Ohio	0.98	(0.91–1.05)	2.14	(1.67–2.73)	1.33**	(1.09–1.60)	0.54	(0.26–1.15)
North Carolina	1.11**	(1.03–1.19)	1.23	(0.96–1.58)	1.08	(0.93–1.26)	0.75	(0.36–1.59)
Georgia	0.91*	(0.85–0.98)	0.74	(0.39–1.41)	[omitted]	1.26	(0.60–2.65)
Cancer diagnosis	1.03***	(1.02–1.03)	1.51***	(1.43–1.59)	2.50***	(2.20–2.84)	1.05	(0.97–1.14)
# opioid prescribers	1.35***	(1.35–1.35)	2.89***	(2.83–2.96)	1.49***	(1.41–1.56)	1.29***	(1.26–1.33)
Network prominence
BP PageRank (10%ile)	1.04***	(1.04–1.04)	1.68***	(1.64–1.73)	1.08**	(1.03–1.13)	1.45***	(1.40–1.50)
WP PageRank (10%ile)	1.04***	(1.04–1.05)	1.20***	(1.18–1.21)	1.09**	(1.03–1.14)	1.14***	(1.11–1.16)
ICC	0.71		0.85		0.54		0.73
BIC	1,115,970		218,590		17,629		79,097

¹Random intercept models adjusted for state fixed effects; incidence rate ratios or odds ratios and confidence intervals are presented

²Omitted category = Other

³Omitted category = West Virginia

* p<0.05

** p<0.01

*** p<0.001