Table 3.

Multivariable Regression: Factors Associated with Physicians' Numbers of Prescriptions and Potentially Inappropriate Medication Orders

Variables	Prescriptions	PIM Orders
Lack of benefit/risk information about deprescribing^*	−0.40 (0.16)^{^†}	−0.06 (0.03)^{^†}
Number of medications^*	−0.67 (0.24)^{^†}	−0.07 (0.04)
Proportion of minority patients	−3.72 (0.81)^{^†}	−0.34 (0.14)^{^†}
Proportion of patients aged ≥75	−2.92 (1.13)^{^†}	−0.45 (0.20)^{^†}
Use of the Beers List^{^‡}		−0.17 (0.08)
Intercept	16.31 (1.32)^{^†}	1.40 (0.24)^{^†}
Adjusted R² (AIC)	0.3939 (228)	0.2062 (12.4)

Data are β (standard error). All variables included in the fitted models are reported in this table.
↵* The importance of triggers/barriers were assessed using 5-point Likert scales ranging from 1(Not important) to 5(very important).
↵† P < .05.
↵‡ Reference group includes those physicians who did not use the Beers List.
AIC, Akaike information criterion; PIM, potentially inappropriate medication.