Opioid Prescriber Map Discussion

jgulotta

The Data Services team has been working on some visuals to address the Opioid Prescriber Summary File data set provided by CMS.gov (https://us.v-cdn.net/6029860/uploads/editor/5q/v5fftvrxuyeh.png). The Map and charts created using this data display the Opioid Prescription Rate and can be grouped by state, county, or zip code. It also reports the most prevalent prescribers in the area, with anesthesiologists excluded from the report. The goal is to create an interesting visual that can be used to inform other data sets and potentially address the prevalence of opioid prescriptions within the Healthcare sector.

If there are any thoughts on how we could expand on the data set, what we can join it to, or how we can enhance the visuals, please feel free to let us know.

Thanks!

AMyrka

This looks great! I will probably have some questions for you once I think about it. Can I share with some opioid experts?

dtriller

Thanks for sharing this. My initial thoughts are that the presentation has some strong initial biases that should be avoided (unless the source of the interpretation is cited). For example, terms like "bad" and "over-prescriber" make determinations about the results that may not be well founded. I think the dataset itself could be useful to research and public initiatives, but additional variables might need to be added to make it more useful. Percent of claims is likely insufficient. Can other details of prescriber specialty, avg patient acuity (e.g. comorbidity score), comorbidities (cancer, e.g.), be added to give more insights? Also, presenting data in form of days supply might also be helpful. A doctor may have many claims for small increments (e.g. 3 day supplies) and be doing a thoughtful job, while looking "bad" based on # claims. Conversion to morphine equivalents might also be useful, as it would account for differences in potency of drugs and quantities being prescribed.

jgulotta

Anne, Please feel free to share with people and let us know what kind of feedback they give you.

dtriller: thank you for the notes! I will go back to the team with your comments on the word choice. Unfortunately, the data set published does not include much more than the specialty and the prescribing rate/claim count, but reading your comment has given us a lot of insight regarding the type of data it may be useful to try to join this data set with.

Thanks for your thoughts and input!