Health Quality & Statistics – Industry News of the Week

By | January 20, 2014

Greetings, friends. I hope you had a safe and productive weekend.  

Here’s what’s new and exciting in the wide, wooly world of health care quality and statistics … 

Health Care Quality

  • One of my earliest jobs in the hospital was “weekend intake” for our case-management team: I’d take FTP’d admission reports and manually re-key them into the case-management software. Yay. In those days, West Michigan had a bit of a peak of teen pregnancy — it wasn’t unusual for me to transfer OB records for females as young as 14. Times change, and for the better; teen pregnancy here has declined markedly over the last decade. But does MTV’s show 16 and Pregnant have anything to do with it? Reason magazine thinks so — and if the conclusions are right, we might have some insight into non-standard ways of enculturating healthier behaviors among traditionally difficult-to-reach populations.
  • Medscape’s Business of Medicine section has a special series right now on “The Noncompliance Epidemic.” Good stuff — especially Neil Chsanow’s “Why Are So Many Patients Noncompliant?”
  • Last week’s Health Reform Watch in the Washington Post’s Wonkblog has generated a fair amount of controversy in some circles. Sarah Kliff has done some good work reporting on the Affordable Care Act, but the last two articles — about the happyspeak by some health insurers — suggest that the rift between insurers and HHS are deeper than they appear. When senior execs go on the record speaking in double- or triple-negatives ….
  • A new report by the Robert Wood Johnson Foundation includes “10 sweeping recommendations aimed at improving the health of all Americans.” Central to this is shifting from a prevention to a wellness mindset.
  • In “What Patients Don’t Tell Their Doctors,” in the Jan. 13 New York Times, Dr. Abigail Zuger outlines some problems clinicians encounter when they speak with their patients. Read the article, then ask yourself: How much of the stuff we need to document with ICD-10 are we NOT going to see in the medical record, because providers don’t want to ask the question?
  • Rural health IT adoption — in particular, Meaningful Use — popped on the radar of Becker’s Hospital Review.  Good read for an overview of the challenges our ambulatory and rural-health centers may face as IT demands grow.
  • HHS announced that it would soon start answering FOIA requests for Medicare physician reimbursement — at an individual level. The American Medical Association is not pleased. 

Statistics & Methodology

  • You know the drill: You get a question, you research and perform analysis, you obtain findings with a clear conclusion, and you publish the results to inform policy. But what happens when the statistical evidence is inconclusive? Blogger Andrew Gelman addressed the subject informally last week. Good gut check for people who tend to think that every complex system can be reduced to a conclusive number.
  • A brief post from Simply Statistics reminds us to question whether missing data is missing at random or not, and if not — then what can you infer by what’s not present?
  • The Data Scientist Salary Survey notes that among tools used by data scientists, SQL tops the list, followed (in order) by R, Python, Excel, Hadoop, Java, Network/Graph, JavaScript, Tableau, D3, Mahout, Ruby … and finally, SAS/SPSS. Interesting perspective on what tools/technologies may be ascendant at the moment. Statisticians and quality-improvement folks can sometimes overlook the quality of a dataset … a problem minimized when the statistician is also doing his own data pulls directly from a server using SQL.
  • Like playing with R? Catch David Smith’s R-blogger’s December 2013 roundup of R-focused articles from last month.
  • … and to conclude the Trifecta of R … if your organization needs to do sophisticated analysis but budget is tight and Excel just can’t quite cut it, look into R. Isaac Peterson recently did a nice post titled “Why R is Better Than Excel for Fantasy Football (and most other) Data Analysis” — Peterson outlines the basics. I’m thinking “quality analytics in a small-practice setting” here. Bottom line is that a solid analytics platform need not require capital spending, it just requires a simple software installation and someone who knows both stats and R programming.
  • Matt Asher’s “The week in stats (Jan. 13th edition)” offers a few morsels of statistics-related fun.
  • For a longer-form review of how data-visualization can communicate and explore, but not identically, check out Nina Zumel’s article, “The Extra Step: Graphs for Communication versus Exploration” from the Win-Vector Blog.
  • People who know me know that I love Scotch. Mmm, Scotch. The spirit is delightful because it resists easy flavor classification. A light, floral Speyside malt tastes nothing like a peaty Islay malt. For some insight into how statistics can help us classify Scotch flavors, check out “Where the whisky flavor profile data come from.” And reach for a fresh tumbler. 

Notes from Jason

  • I’m organizing and chairing a session titled “Identifying Causal Factors Leading to Optimal Chronic Disease Self-Management Outcomes” at the VI European Congress of Methodology. The conference, to be held at Universiteit Utrecht (Utrecht, Netherlands) July 23-25, 2014, is now accepting abstracts for all sessions. Please review the list of approved sessions. If you’re interested in submitting an abstract, follow the instructions or let me know and I can connect you to the conference planners.

 Have a great week!

Be Sociable, Share!