% of patients with NDC vs Procedure Code

Calculate their estimate weight from first claim date we observe. That should be closest to most normal dose

Add weight statistics and estimated dosing

Look at this population.

look at gaps between dosing that these patients are receiving alt. Do we have the time between dosing getting lower.

Can we add in Dosing information and Day supply

There is interest into looking at the dosing for these patients. Let's look further into this data. @schuldtr

PDC Thresholds:

90+, (Elmor developing rest)

Highest cost

Let's adjust this to two months.

This seems to be the sweet spot. - Robert

Add a table describing these patients.

Look at who is and is not switching.

Crush Medicaid into Other Pay\er

Add notation here that If 2l then LOT2Linename == LOTENDS

Add any median

Median IQR

WIthin 45 Days N=

Updated to reflect 13% not N=13

Arliene and Sophia: We only have 273 of our patients with a record in the Medication orders table. Not everyone is represented.

I am aware of typo I will update when I finish Sankey plot. As you can see this report takes 58 minutes to generate.

No need to break eyes.

Only keep visits that match from Eye table. Check this.

Robert confirm if this is visits or patients

David, here is the context. We have 83% of patients being treated for both eyes. That will mean that the N of injections probably appears higher since we aren't breaking them down into injections by specific eye.

Possible project for Sophia? D&I - looking at disparities in ttt

Removed the 1:1 match - consolidated tables to be based on each matching pair.

Overall Cohort - Moved everything descriptive of overall cohort to this first section

Final sample size for our study

Is this an artifact of how they are recording the DX.

Add in a count of the N of Study Patients and N of Sample Patients

Missing added as a future proof option for future data uploads.

When we filter out records for condition == ALL we lose all Not Recorded and OU records in the Eye table.

Patients with a race value, but it is not useful for analysis. I am considering this missing if we have unspecified. Do you want me to keep it like that?

Slightly higher Medicare Cost and Risk patients in the Race/Ethnicity grouping. Possibly due to the older age

Fairly significant difference here. Not missing is older in comparison.

Hmm this seems strange. Can you explain how this was calculated?

Makes sense to me. Looks to be similar across cohorts

Can you clarify?