Skip to main content
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: J Am Pharm Assoc (2003). 2021 Apr 15;61(5):555–564. doi: 10.1016/j.japh.2021.04.007

Table 4.

Examining Types of Misuse Using Regression Models with Propensity-Score Matching: Pre-Implementation Compared to Post-Implementation (n=87)

Models with or without PS p-value 95% Confidence Interval
Coef. Stand. Error Lower Bound Upper Bound
Drug/Age Misuse
 (post vs. pre) −0.003 0.071 0.969 −0.142 0.136
 (PS, post vs. pre) −0.049 0.047 0.404 −0.142 0.043
Drug/Drug Misuse
 (post vs. pre) −0.597 0.438 0.369 −1.455 0.260
 (PS, post vs. pre) −0.457 0.372 0.401 −1.186 0.273
Drug/Drug Misuse_(avoid combination)
 (post vs. pre) −0.556 0.043 0.369 −0.141 0.030
 (PS, post vs. pre) −0.049 0.043 0.401 −0.134 0.036
Drug/Drug Misuse (consider therapy modification)
 (post vs. pre) 0.022 0.212 0.969 −0.392 0.437
 (PS, post vs. pre) 0.049 0.151 0.817 −0.246 0.345
Drug/Drug Misuse_(monitor therapy)
 (post vs. pre) −0.564 0.291 0.286 −1.133 0.006
 (PS, post vs. pre) −0.457 0.267 0.191 −0.980 0.066
Drug/Disease Misuse
 (post vs. pre) −0.333 0.202 0.359 −0.729 0.620
 (PS, post vs. pre) −0.407 0.195 0.102 −0.790 −0.248
Drug/Label Misuse (Exceeds Daily-Dosage)
 (post vs. pre) −0.100 0.080 0.369 −0.253 0.053
 (PS, post vs. pre) −0.160 0.066 0.015 −0.291 −0.030
Drug/Label Misuse (Exceeds Single-Dosage)
 (post vs. pre) −0.361 0.057 0.011 −0.472 −0.250
 (PS, post vs. pre) −0.358 0.062 0.011 −0.479 −0.273
Drug/Label Misuse (Exceeds Timing/Frequency)
 (post vs. pre) 0.072 0.123 0.877 −0.170 0.314
 (PS, post vs. pre) 0.284 0.132 0.102 0.026 0.542
Drug/Label Misuse (Exceeds Use Duration)
 (post vs. pre) 0.011 0.070 0.969 −0.126 0.148
 (PS, post vs. pre) 0 0.061 0.999 −0.120 0.120
Drug/Label Misuse (Inappropriate Indication)
 (post vs. pre) 0.025 0.069 0.969 −0.109 0.159
 (PS, post vs. pre) −0.012 0.025 0.759 −0.061 0.037

Note: PS = Propensity Score Matching; both the raw and the propensity-adjusted p-values were adjusted for multiple comparisons using the pFDR