Table 2.

The impact of provider use of Joint Longitudinal Viewer in the prior 6 months of outpatient primary care visits on provider ordering of duplicate images.

Characteristics					Odds ratio (95% CI)a			P value
Provider characteristics
	Joint Longitudinal Viewer provider			0.44 (0.24-0.78)			.005
	Provider history of ordering imaging studies (quartiles)
		1	Refb			Ref
		2	0.87 (0.33-2.32)			.78
		3	2.73 (1.23-6.06)			.01
		4	4.15 (1.86-9.25)			.001
	Provider type
		Physician	Ref			Ref
		Physician assistant/Nurse practitioner	1.32 (0.75-2.32)			.34
Patient characteristics
	Gender
		Female	Ref			Ref
		Male	1.68 (0.89-3.17)			.11
	Age (years)
		<30	Ref			Ref
		30-39	2.19 (1.17-4.11)			.01
		40-49	0.78 (0.35-1.73)			.54
		≥50	1.27 (0.46-3.56)			.65
	Race
		White	Ref			Ref
		Black or African American	1.09 (0.56-2.13)			.80
		Other	1.63 (0.82-3.21)			.16
	Elixhauser comorbidity score
		0	Ref			Ref
		1	0.44 (0.23-0.83)			.01
		2	0.40 (0.18-0.90)			.03
		3 and above	0.15 (0.04-0.52)			.003

^aThe odds ratio and 95% CIs are estimated from the logistic regression model controlling for all variables shown in the table as well as facility (random effects) and fiscal month.

^bRef indicates baseline in the analysis.