Table 1.
Criteria for the construction of adherence and drug exposure measures.
| Prescription-dispensation design with a fixed 12-month time-window (PD Cohort) | |
| Prescription fill requirements | No prescription filled is required for inclusion |
| Index date | Date of first prescription (filled or not) |
| Numerator (days’ supply) | Days covered by filled medication from the index date to the end of follow-up, provided they are covered by a medical prescription |
| Denominator (follow-up days) | The 365 days immediately after the index date provided they are covered by a medical prescription |
| Censoring previous to the end of study date | At the date of death, loss of continuous health plan coverage or periods not covered by a doctor’s prescription during the 12 months of follow-up. No other causes of censoring (including switching or patient discontinuation) are allowed |
| Immeasurable time | Periods under hospitalization (both acute and long-term care) are not accounted for |
| Stockpiling | A maximum of 2 months of stockpiling is allowed |
| Dispensation-only design with a fixed 12-month time-window (D1 Cohort) | |
| Prescription fill requirements | At least one prescription filled is required for inclusion |
| Index date | Date of first prescription filled |
| Numerator (days’ supply) | Days covered by filled medication from the index date to the end of follow-up |
| Denominator (follow days) | The 365 days immediately after the index date |
| Censoring previous to the end of study date | At the date of death, loss of continuous health plan coverage. No other causes of censoring (including doctors or patient discontinuation or switching) are allowed |
| Immeasurable time | Periods under hospitalization (both acute and long-term care) are not accounted for |
| Stockpiling | A maximum 2 months of stockpiling is allowed |
| Dispensation-only design with a fixed 12-month time-window, requiring at least two prescriptions filled separated by 6 months (D2 Cohort) | |
| Prescription fill requirements | At least two prescription filled separated by 6 months are required for inclusion |
| Index date | Date of first prescription filled |
| Numerator (days’ supply) | Days covered by filled medication from the index date to the end of follow-up |
| Denominator (follow days) | The 365 days immediately after the index date |
| Censoring previous to the end of study date | At the date of death, lost of continuous health plan coverage. No other causes of censoring (including doctors or patient discontinuation or switching) are allowed |
| Immeasurable time | Periods under hospitalization (both acute and long-term care) are not accounted for |
| Stockpiling | A maximum of 2 months stockpiling period is allowed |
| Rationale | Measure used by the Pharmacy Quality Alliance (PQA). Due to restricted inclusion criteria, the PQA measure excludes most part of non-adherent patients |
| Days covered by medication in the 12-month time-window (drug exposure) | |
| Fill-up requirements | No prescription filled is required for inclusion |
| Index date | Date of first prescription (filled or not) for the PD Cohort and date of first prescription filled for D1 and D2 Cohorts. |
| Numerator (days’ supply) | Days covered by filled medication from the index date to the end of follow-up |
| Denominator (follow days) | The 365 days immediately after the index date |
| Censoring previous to the end of study date | No censoring was applied. |
| Immeasurable time | Periods under hospitalization (both acute and long-term care) are not accounted for |
| Stockpiling | A maximum 2 months of stockpiling is allowed |
| Rationale | Rough measure of drug exposure during the follow-up period |