Skip to main content
. Author manuscript; available in PMC: 2015 Jul 1.
Published in final edited form as: Pharmacoepidemiol Drug Saf. 2014 Mar 24;23(6):628–635. doi: 10.1002/pds.3609

Table 2.

Comparing the prevalence, sensitivity, positive predictive value (PPV), and specificity of nine alternative claims-based definitions of antipsychotic polypharmacy

Source Parameters Identified with polypharmacy Prevalence Sensitivity PPV Specificity
>14 days of polypharmacy N (%)
    Chart 51 (30.0%)
Medicaid Allowable gap: 32 days 50 (29.4%) 92.16% 94.0% 98.47%
Allowable gap: 14 days 46 (27.1%) 82.35% 91.3% 96.63%
Allowable gap: none 45 (26.5%) 82.35% 93.3% 96.63%
>60 days of polypharmacy
    Chart 48 (28.9%)
    Medicaid Allowable gap: 32 days 43 (25.9%) 83.33% 93.0% 97.03%
Allowable gap: 14 days 36 (21.7%) 68.75% 91.7% 94.57%
Allowable gap: none 24 (14.5%) 50.00% 96.0% 91.64%
>90 days of polypharmacy
    Chart 44 (27.5%)
    Medicaid Allowable gap: 32 days 39 (24.4%) 81.82% 92.3% 99.25%
Allowable gap: 14 days 32 (20.0%) 68.18% 93.8% 95.00%
Allowable gap: none 18 (11.3%) 40.91% 100.0% 91.16%

Note: The Medical chart was the criterion standard (n = 312 medical records). The chart-based assessment did not allow for medication gaps in order to capture the duration of antipsychotic medications and polypharmacy as prescribed. Prevalence of the measures is calculated based on nine alternative definitions, using the medical chart-based measure to establish the denominator for individuals on antipsychotics for >14 (n = 170), >60 (n = 166), and >90 days (n = 160).