Improvement Insights Blog

Well, the unseasonably warm winter weather was bound to end some time. Sure enough, last week Mother Nature shook her head and remembered what she was supposed to be doing during the winter. I found refuge from the cold and the snow at my favorite table at my favorite coffee shop. It’s my favorite coffee shop because Jason, the owner, makes the best café au lait and beignets this side of Café Du Monde in New Orleans. The table against the back wall offers the geographic benefits of both having a great view of the snow out the window and being just the right distance from the pot belly stove that Jason keeps stoked on cold days.

I had just finished my café au lait and was staring at the empty cup wondering if I should order another when the question was answered for me. A tray was set down on my table with two full cups and a plate of Jason’s delicious beignets, covered in powdered sugar that was as white and plentiful as the snow outside. I looked up and saw my friend Annie smiling at me.

“Hi! Mind if I join you?” she asked.

“Hey, Annie! Of course! Pull up a chair.” I replied, motioning for her to sit down.

“Thanks! It just so happens that I was looking for you, and I knew that you could never refuse Jason’s café au lait and beignets,” she said, taking her seat and grabbing her cup and one of the pastries.

“Looking for me?” I asked. “Why? Are things still going smoothly at the hospital?” I had known Annie for a long time, but we had the opportunity to work together a year ago when I was hired to help establish a Quality Improvement strategy at the hospital where Annie worked as a nurse.

“Oh, things are fine… but they can always be better, right?” she replied. “I put together some new efforts, and I wanted to make sure I was going about it the right way.” She pulled a USB drive out of her pocket. “I just happen to have some data right here…”

“Well, you’re lucky I happen to have my laptop with me,” I said, placing my computer on the table.

“Oh, please,” she said, jokingly dismissing my comment. “When do you NOT have your laptop with you? In all the time I’ve known you, I’ve never seen you without it. I suspect you sleep with it under your pillow.”

“Nah, I had to stop doing that,” I replied. “It overheats when you cover it up like that.” I opened a spreadsheet of her data on patient falls when she motioned for the mouse. I passed it to her.

“Okay, so I’ve got patient fall data here, with Total Patient Falls and Total Patient Days,” she said, indicating columns B and C. “I’ve divided them together to create a percentage here in column D. What I do is highlight the labels in column A and the percentage in column D and create an XmR Individuals chart:

“Once I get the chart,” she continued, “I indicate the date that we implemented the process change by selecting that point and clicking ‘Show Process Change.'”

She did so on the chart, and the XmR chart updated to show this:

“So,” she concluded, handing me the mouse and taking a sip of coffee, “This is what I came up with. What could I be doing better?”

I looked at Annie puzzled. “What do you mean? This shows that you’re doing an amazing job! You’ve not only lowered the center line significantly, you’ve tightened the control limits. That means you’ve created a more consistent process that has resulted in fewer patient falls. You took what I taught you a year ago and put it to use. I’m thrilled!”

“Well, thank you,” she replied. “You were a good teacher. But what I meant to ask was is there a different tool or chart I should be using?”

“Ah… I see. Well, now that you mention it, there is a different chart that might suit you a little better: the u chart.”

“Mmm-hmm…” she mumbled as she bit into a beignet. “How is that different?”

“Well, instead of just using the percentage, the u chart takes into account the numerator and denominator.”

Annie’s face scrunched up in puzzlement. “I don’t understand. Percentages are universal. 25% is 25%, no matter what, right?”

“Well, let me ask you this: If you looked at two months of patient falls, and the first month was exceptionally quiet and you only saw 4 patients, but one of them stumbled and fell, that’s 25%, right?” I asked.

“Right…” she answered, hesitantly.

“And then the next month, you saw 4,000 patients, but 1,000 of those patients fell, that’s still 25%, right?”

“Ohhhh…” she said, nodding slowly. “One patient falling out of four could just be a fluke from such a small sample, while 1,000 patients falling out of 4,000 indicates a systemic problem, right?”

“Exactly,” I agreed. “The u chart uses the raw data of Total Falls and Total Patient Days, kind of like you’re always urging me to eat more raw fruits and vegetables in my diet.”

“And you keep ignoring me and eating lunch at Joe the Hot Dog Guy’s cart,” she muttered, smiling.

“Anyway…” I continued, smiling. “You would highlight the labels in column A as well as the raw data in columns B and C, then select the u chart.”

“And then after selecting the point where you made the process change and clicking ‘Show Process Change,’ you get this chart.”

“Why are the lines for the UCL and LCL all jaggedy?” Annie asked.

“Because your sample size isn’t the same each month, so it’s taking that into consideration.” I said. “If the total patient days is a smaller number, the UCL and LCL will be farther apart; if it’s larger, they’ll be closer to the center line.” Annie nodded, and I continued. “You’ll see a few more points in red here than you do on the XmR chart.”

“I see that. Why is that?” Annie inquired.

“Think about it like the difference between putting your ear to a patient’s chest to hear a heartbeat vs. listening through your stethoscope. The stethoscope blocks out more outside noise and is a more sensitive device than your ear alone. In the same sense, the u chart is more sensitive because it takes into account both the numerator and denominator.” I replied, leaning back and taking a sip of my coffee.

“I understand,” Annie said. “It looks like the data points themselves on the u chart are the same as the X chart I was using, so the data points on the chart are still the Total Patient Falls divided by the Total Patient Days, right?” I nodded. “But the UCL and LCL vary because the denominator is different each month.”

I nodded again. “It sounds like you’ve got it.”

Annie smiled. “This will help tremendously. Thank you so much!” She stood up and put on her coat. “You know, my sister is visiting this weekend, and wanted me to find out if you wanted to meet at Pareto’s Big Bar and play some darts again.”

“Your sister April?” I asked, and Annie smiled. “Tell her that my interest level is WAY above the UCL.”

“I will,” said Annie, smiling as she wrapped her scarf around her neck. “And then after that I’ll explain to her what a UCL is.”

If you’re interested in learning more about the u chart in QI Macros (and its close relative, the p chart), click HERE and HERE and HERE and HERE to read about it on the QI Macros website.