...
The "Medications" feature is often used with patient registries and/or to collect concomitant medications for clinical trials. It combines simple entry for data entry staff with sophisticated analysis of medication usage for investigators. From the perspective of data entry staff, medications are entered by merely selecting from a list and then entering in the applicable the begin date (, end-date (if applicable), dose, times pre day, route of administration, indication, and any notes.
Data Entry Fields
From a data analysis perspective, medication usage in datasets is automatically looked-up and connected with visit dates (e.g., date of a clinic visit , planned protocol encounter dates such as baseline, week 1, week 4or study encounter (e.g., "Baseline", "Week 1", etc.))! In datasets, investigators can choose to include 1) the dose of a specific medication, 2) the dose equivalence across multiple medications, and/or 3) whether a patient has ever been exposed to a single or a whole list of medications (see image below).
...
- Should I use the medication feature or create a "Concomitant Medication" CRF?
- The time saving analytical routines of the Medications feature nearly always trumps building a separate Concomitant Medication case report form (CRF). However, because the fields of the Medication feature are hard coded, in some cases a separate CRF is created because Because fields can not be added or modified, the medication feature is typically not used when where there is an interest in medication usages in specific use characteristics (e.g., was it titrated, were there any Adverse Effects in the first 24 hours, etc.), although in some cases it works to use the "Notes" or "Indication" fields to document such characteristics.
- What if a medication is not taken every day (e.g., PRN)?
- What to do depends on the aims of the study/registry. For instance, if total prescribed dose is what is of interest, then it is entered as if taken every day. If the "mean daily dose" is what is of interest, then that is what is entered for dose (i.e., an estimated average over a set number of days. Once calculated, the mean dose is entered as "daily"). If documentation of usage is what is of interest, then enter the total dose and put PRN in the notes. The main thing to consider is StudyTRAX pulls whatever is entered into the dose field into data sets, thus analytical concerns mainly drive the decision.
- Missing information?
- Start and End date.
- What to do depends of the aims of the study. Some examples:
- Subject on medication, but doesn't know start date: In this case, if the research question of interest is length of time on the medication, then adopting a conservative estimation rule is recommended. The meaning of "conservative" will varying depending on the aims of the study, but generally it would be defined to work AGAINST the hypothesis so that the investigator is confident the results minimize type I errors.
- Subject knows the year, but not month: Same as above in terms of adopting a conservative rule. That is, work against the hypothesis. Typically one of the following is adopted: "1/1/xxxx", "6/1/xxxx", or "12/31/xxxx".
- What to do depends of the aims of the study. Some examples:
- Medication Dose
- If the dose is not known, documentation vs. analytical aims will drive the decision. That is, if there is no need to document the use of the medication, the medication would NOT be entered, or if the medication does need to be documented 0 is entered for the dose. On the analytical side of things, typically a conservative rule is adopted (e.g., in a retrospective study, the medication is mentioned in the EMR, but not dose, use the smallest known dose).
- Number of times per day
- Handled the same as missing "Dose". A "0" is entered for number of times per day if the medication is only needed for documentation. For analytically purposes, the conservative rule is often set to "enter 1".
- What if a medication is taken multiple times on the same day?
- Since data analysis is at the "Day" level, whether or not this is one or multiple entries depends on whether it is consider important to separate the information. For example, patient registry designs may prefer multiple entries because the information is used to generate clinic notes within StudyTRAX. Conversely, a research design may only have an the total daily dose and thus combines this into one total dose entry.
...