Determining Drug Exposure Based on Medication Dispensing Data: A Validation Study of Vitamin K Antagonist Treatment Episodes Against INR Records

Figures

• 

  FIGURE 1. Illustration of fixed and dynamic exposure times used to construct VKA treatment episodes. (A) This figure illustrates an example of applying a fixed exposure time of 100 days between subsequent dispensed VKA prescriptions (numbered 1–8) and the addition of 100 days to the date of last VKA dispensing within the treatment episode. (B) This figure illustrates the use of a dynamic method. The dynamic exposure time is calculated as the average period between the preceding two dispensed VKA prescriptions, with a fixed exposure time of 100 days for the first two dispensings. The dynamic exposure time was added to the date of last VKA dispensing within the treatment episode to construct the end date. This figure was created in BioRender. Kruip (2024). https://BioRender.com/z70g565.

• 

  FIGURE 2. Results of fixed exposure times. (A) The percentage of INR measurements outside dispensing‐based VKA treatment episodes was calculated by dividing the number of INR measurements that occurred outside the VKA treatment episode by the total number of INR measurements. (B) The VKA‐exposed person‐time ratio was calculated by dividing the person‐time exposed to VKA according to dispensing‐based treatment episodes vs. INR‐based treatment episodes. In both panels, the x‐axis displays the fixed exposure time, that is, the number of days allowed between subsequent dispensed VKA prescriptions within the same treatment episode. The different colors represent the number of days added to the date of last VKA dispensing within the treatment episode.

• 

  FIGURE 3. Results of dynamic exposure times. (A) The percentage of INR measurements outside dispensing‐based VKA treatment episodes was calculated by dividing the number of INR measurements that occurred outside the VKA treatment episode by the total number of INR measurements. (B) The VKA‐exposed person‐time ratio was calculated by dividing the person‐time exposed to VKA according to dispensing‐based treatment episodes vs. INR‐based treatment episodes. In both panels, the x‐axis displays whether the dynamic exposure time was calculated based on the preceding two or three dispensed prescriptions. The different colors represent the fixed exposure time that was applied for the first two or three dispensings of the treatment episode. The upper panels display the percentage by which the dynamic exposure time was increased.

• 

  FIGURE 4. Median number of dispensing‐based VKA treatment episodes. Median number of dispensing‐based VKA treatment episodes with first and third quartiles are displayed for both fixed (A) and dynamic (B) methods. (A) The y‐axis displays the fixed exposure time, that is, the number of days allowed between subsequent dispensed VKA prescriptions within the same treatment episode. The different colors represent the number of days added to the date of last VKA dispensing within the treatment episode. (B) The two colors indicate whether the dynamic exposure time was calculated based on the preceding two or three dispensings. The x‐axis displays the fixed exposure time that was applied for the first two or three dispensed prescriptions of the treatment episode. The upper panels display the percentage by which the dynamic exposure time was increased.