data&donuts
  • Data & Donuts (thinky thoughts)
  • COLLABORATor
  • Data talks, people mumble
  • Cancer: The Brand
  • Time to make the donuts...
  • donuts (quick nibbles)
  • Tools for writers and soon-to-be writers
  • datamonger.health
  • The "How" of Data Fluency

data & donuts

"Maybe stories are just data with a soul." -- Brene Brown

Opioids by the numbers...

9/12/2017

 
I am a visual person. Before I dig into the data or start a granular analytics project I need to be visual. I have the csv files saved neatly in a file but my first thought is to ask a few general questions that might show me what sort of data I need.
For example, Medicare Part D Opioid Prescriber Summary Files are a great place to begin asking questions. The Excel file was pulled into Tableau and this is a first pass of data exploration. What jumps out first is the opioid prescribing behavior of Family Practice in the 4 selected states. Definitely demands a little more granularity. The bottom graphic sorts along an average prescribing rate and we can compare across states to see specialist prescribing behavior by state. This tells a slightly different story suggesting a deeper look will be required.
Picture
Apologies for the blurry image but I did a quick exploration within US Census Bureau Data using their Topologically Integrated Geographic Encoding and Referencing (TIGER) tool. The tool is great if you don't have a data visualization program to use and it works quite well inside the platform. Use the link if you want to have a look around or try out Tableau Public--a free version of the popular desktop application .

West Virginia is top of mind as the highlighted county below has been reporting the highest number of opioid deaths. I was curious to find out more about certain demographics. This is part of a large client project so my purpose is to share the resources--not necessarily the outcomes. But I also explored education, employment, military service, race, and several other variables to see what might be unique about this county. On a hunch I looked at workman's comp and disability rates as well.
Picture
Not as scalable as analytics but the West Virginia 2016 Mental Health National Outcome Measures (NOMS): SAMHSA Uniform Reporting System is a great example of a resource to either scrape for data or read for context when looking at state level or county level patterns in data.
westvirginia-2016.pdf
File Size: 1072 kb
File Type: pdf
Download File

There are multiple data sources helpful in questioning the data in opioid abuse. The CDC Wonder Mortality Data reports underlying cause of death data as well as additional sites examining industry payments to physicians and drug overdose treatment and rehabilitation facilities. Be sure to take the time to read the data files to find out how data was collected and what the variables are measuring.

"Statistics are like swimwear - what they reveal is suggestive but what they conceal is vital." 
-Ashish Mahajan, Lancet 2007

​Ask any data questions either on twitter or LinkedIn...

Comments are closed.

    Telling stories...

    Finding, curating, tidying, analyzing, and communicating your data creates many opportunities for discussion and collaboration...

    Take a look around...
    Follow @datamongerbonny

    Categories

    All

    twitter...

    Tweets by datamongerbonny
Proudly powered by Weebly
  • Data & Donuts (thinky thoughts)
  • COLLABORATor
  • Data talks, people mumble
  • Cancer: The Brand
  • Time to make the donuts...
  • donuts (quick nibbles)
  • Tools for writers and soon-to-be writers
  • datamonger.health
  • The "How" of Data Fluency