Ask Jay – What Kind of Chart Should I Use?
People still seem to be confused about what kind of Quality Improvement chart to use. Here’s the short and simple answer:
Continue Reading "Ask Jay – What Kind of Chart Should I Use?"
Improvement Insights Blog
Latest "Jay Arthur Blog" Posts
NAHQ NEXT 2020 HQ Showcase – Jay Arthur
If you’re a registered NAHQ Next 2020 attendee, you can view Jay Arthur’s showcase until October 31, 2020.
1. Click on the link below.
https://www.webcastregister.live/nahq2020/welcome
2. Click on the NAHQ Next top banner to log in. It will take you to the page below. Log in with your name and email address:
3. Once you’re logged in, visit the On-Demand Sessions and scroll to the very bottom to view Jay’s session, “Zero Harm and the Trillion Dollar Prescription.”
COVID-19 Cases by Type of Facility
The New York Times published a list of cases by facility. I tweaked it to get Pareto Charts of the biggest problems. Nursing homes and rehab facilities were 64% of the total, then prisons and food processing (e.g., beef, pork, poultry, etc.).
Let’s Work Together
This is a fun thing we did last week. While Watching the movie “Forrest Gump,” Jay heard this song and had the idea that this song is particularly applicable to our current situation. Several of the people that work for QI Macros are musicians, so we tried to do something fun while we’re working remotely. This was recorded in 4 separate homes and then assembled and arranged by Nicholas. We hope you enjoy it:
U.S. Deaths by Age Group as of 4/24/2020
People over age 55 account for 92% of COVID-19 Deaths (data from CDC). Sweden is using similar data to leave the country open for most citizens while asking seniors to stay at home. How do we reopen the economy? Self-quarantine seniors; let everyone else get back to work.
Here’s the 2019-2020 Influenza (i.e., Flu) deaths. Again, seniors are 83.5% of deaths. The flu death rate is about 1 per 1,000. COVID-19 death rate is 1-2 per 100, perhaps lower in people under the age of 55 and higher for those over age 55.
Countermeasure: Self-quarantining seniors will help flatten the curve and prevent overwhelming healthcare.
Colorado COVID-19 Cases and Deaths in Control Charts
While most COVID-19 charts show cumulative cases and deaths, I wondered what would happen if we turned the cumulative data into a daily data and plotted it as a control chart. It is possible to use c or XmR charts for this purpose. Process changes chosen based on runs and trends in the data.
Daily new cases still unstable, but hovering around 339/day. Cases began to rise 3/18. Potentially leveling off 3/26. the date of state “stay-at-home” order.
Daily deaths still unstable, but hovering around 12/day. Deaths began rising 3/21, but appear to be stabilizing. Significant spike on 4/2, root cause unknown.
Continue Reading "Colorado COVID-19 Cases and Deaths in Control Charts"
COVID-19 Paramedic Dashboard 2020
One of our QI Macros users offered to share his dashboard of paramedic response during the Seattle area response to COVID-19. His team transported the first COVID patient in America. As you can see, turnaround times (TAT) at the hospital averaged 30 minutes and temperatures spiked in transported patients.
How Crisis Affects U.S. Healthcare Workers
Back in 2002, thousands of miles from New York City, I worked with a hospital in Oregon. I was amazed to find that they had staffing data from 9/11 about the up and down for that month. As you can see in this X Chart, absenteeism (understaffing) was lower for each of the four following days, 9/12-15, and then recovered. I have noticed this pattern with COVID-19 as well: a week of paralysis followed by a return to normal. Unfortunately, healthcare workers haven’t had the luxury of downtime when dealing with the new crisis.
So don’t be surprised if crisis and uncertainty cause you or someone you know to hit the pause button.
Continue Reading "How Crisis Affects U.S. Healthcare Workers"
Reducing Blood Sample Over-Collection
At the ASQ Lean Six Sigma Conference in Phoenix this week, Katie Castree with Accumen presented an excellent improvement story about reducing over-collection of blood tubes in a hospital. Here’s the story:
Baseline: 317 extra tubes of blood collected every day (115,705/year)
93% of tubes were not used (clinicians thought it was much higher and resisted changing)
Most of the unused tubes were collected in the Emergency Department (focus on the ED, not the entire hospital).
After the countermeasure (not collecting tubes unnecessarily), extra tubes dropped from 317 per day to 118 per day saving $12,335/year and 0.27 FTEs. Over time, extra tubes dropped to only 84 per day, a 74% reduction.
When are bar charts dangerous?
Answer: Almost always.
- When everything is presented in a bar chart, everything looks the same. And before long, everyone forgets what they’re looking at. Dan Roam
Bar charts always remind me of a Jack-O-Lantern with bad teeth.
Bar charts are one of the most misused charts. The most common mistake is using bar charts for time-series data (i.e., dates):
Bar charts are best used for categories—types of defects, objects, or whatever. Even so, the bars are often in a random order which makes distinguishing the important from the unimportant pretty difficult. Simply sorting the bars into a descending order makes them easier to analyze.
