Skip to main content
. 2015 Jan 9;71(3):363–368. doi: 10.1007/s00228-014-1792-9

Table 2.

Concurrent validity of indicators of prescribing quality based on cut-offs of number of drugs in the medication list (polypharmacy indicators)

Number of drugs (cut-off for indicator) Patients identified by the indicator
n
According to gold standard Sensitivity (95 % CI) Specificity (95 % CI)
Suboptimal drug treatment Appropriate drug treatment
Confirmed
n (PPV)
Not identified
n (%)
Identified
n (NPV)
≥1 192 141 (0.73) 0 (0) 8 (1.00) 1.00 (0.97; 1.00) 0.14 (0.07; 0.23)
≥2 189 141 (0.75) 0 (0) 11 (1.00) 1.00 (0.97; 1.00) 0.19 (0.10; 0.29)
≥3 176 136 (0.77) 5 (3.5) 19 (0.79) 0.96 (0.93; 0.99) 0.32 (0.22; 0.45)
≥4 163 129 (0.79) 12 (8.5) 25 (0.68) 0.91 (0.86; 0.95) 0.42 (0.30; 0.54)
≥5 149 121 (0.81) 20 (14) 31 (0.61) 0.86 (0.80; 0.92) 0.53 (0.41; 0.65)
≥6 137 111 (0.81) 30 (21) 33 (0.52) 0.79 (0.72; 0.86) 0.56 (0.44; 0.68)
≥7 114 96 (0.84) 45 (32) 41 (0.48) 0.68 (0.61; 0.76) 0.69 (0.57; 0.79)
≥8 84 71 (0.85) 70 (50) 46 (0.40) 0.50 (0.42; 0.60) 0.78 (0.66; 0.86)
≥9 71 61 (0.86) 80 (57) 49 (0.38) 0.43 (0.35; 0.52) 0.83 (0.71; 0.90)
≥10 49 45 (0.92) 96 (68) 55 (0.36) 0.32 (0.25; 0.40) 0.93 (0.82; 0.97)

CI confidence interval, NPV negative predictive value, PPV positive predictive value